Does anyone care to ask an amplifier designer a technical question? My door is open.

Roger, normally things go pretty easy on this forum unless certain individuals get involved as you've already sorted out. When they jump in, that's usually a signal that the usefulness of the thread is over.

I am puzzled why the idle current goes down rather than up as it warms. Usually its the other way round because of the negative temp coefficient of transistors.

Actually this is pretty common with bipolars. They can get into a phenomena known as 'thermal runaway' if this is not well controlled.

However, my latest amps are Class A, and I suspect that a more realistic match is obtained by culling outliers by HFE, then match from VBE using the bias at constant potential and sufficient to generate the operating current. Finally, instead of using matched emitter resistors, I use emitter resistors tailored to the output devices, so that each emitter resistor sees the same potential drop.
Your thoughts? Any advice appreciated.

There's a difference between idle and dynamic operation. Matching the hfe over the range in which the device operates is going to be a better method. Otherwise what will happen is a particular transistor can 'hog' current at higher operating points and it will thus be the thing the sets the distortion of the amplifier at power, especially full power.

with Class D switching freq over 450/500 KHz these days do you still think output filtering is an issue?. How do I educate more about this ?. If the issue of filtering not a function of switching freq please correct me. Thanks again.

There's lots of information on the web! You might want to read the papers of Bruno Putzey who has done a lot to further the art. The filter is there to filter out the switching frequency, which it cannot do completely. What is left is a sine wave called 'the residual'. Usually the filter is set to about 60-80KHz, so as to avoid phase shift within the most sensitive area of the audio passband, but otherwise get the residual down as low as possible. The higher the switching frequency the easier this is to do- the speaker inductance itself starts to play a huge role at higher frequencies.

Mac Turner brought me in to consult on his. What disturbes me about them is that they all get their sound from the IC in the front end. Perhaps some are discrete front ends but I have not seen any.

Actually you don't need any ICs in the front end prior to the encoding scheme, if you set things up right. The encoding scheme of course is going to use some sort of chip- our amp for example uses a high speed comparator chip. But the audio signal is applied directly to it.

Ralph, did you perhaps mis-read this. He said the idle goes down not up as the amp warms. This is the opposite of thermal runaway.

Yes- you are correct- totally misread that one. It is pretty weird. I’m wondering how the thermal feedback is accomplished. If its done with an active device, that device might be seeing its hfe drift up as it heats up- thus causing the outputs to shut down more than expected. A schematic would be interesting.

Nelson Pass has designed an interesting amplifier based on the old VFET/SIT devices. As solid state goes one of the most musical I’ve heard.
https://www.diyaudio.com/forums/pass-labs/276711-sony-vfet-amplifier-2-a.html
It also employs low voltage rails, about 25-28 volts.

I'm still in the Polypropylene camp. Best thing for the money. Remember any part I buy is 5 x cost to the final buyer. So I cant buy too many $50 caps expensive resistors.

For a long time polystyrene was an alternative that performed better and was no more expensive. Things fell apart though after the German company that supplied polystyrene got out of the business. It was easy to hear how the polystyrene was better- and this seemed to verify the specs on paper.

The whole argument about 2nd harmonic distortion being "benign" totally ignores Intermodulation distortion which is far worse and always higher.

While I agree with your point here (and that of the rest of the post from which this is taken) this particular statement is false. It is possible to have low IM while THD is considerably higher, although clearly that isn't the case with the example you cited.

"Well, it’s the sound that matters". If a poor design, showing obvious performance weaknesses, sounds "good", something is very wrong somewhere.

The problem here is that the bench specifications used by the industry do not reflect the physiology behind how the ear/brain system perceives sound. The most egregious example is how we perceive sound pressure, which is done via the higher ordered harmonics- 5th and above. The reason this is a problem is that to obtain really 'good' specs on paper, a fair amount of loop negative feedback has to be employed to suppress distortion. As Norman Crowhurst pointed out 60 years ago, feedback introduces distortion of its own and its entirely higher ordered harmonics. Because our ears convert all distortions into a tonality, this causes circuits employing feedback to be brighter and harsher than the original signal.

None of the above is controversial- we've know this stuff for decades. But the industry continues to do nothing about it, which means that there has been little progress. IMO/IME the harmonics produced by a given circuit should be weighted so that higher orders are given a greater weighting than lower orders (for example small amounts of a 7th are considerably more audible than a 2nd at the twice the level). IMD of course should be kept low.

Question: Tube Phono preamp. Would replacing the 1N4007 diodes (in bridge config) in power supply with HEXFREDs or Schottkys improve it? The bridge feeds a CRC filter which then feeds a tube regulator. Thanks.

If the power transformer is not properly snubbed, then you will get less noise. But if its snubbed properly there won't be any difference.

I read a report by JA measuring a Prima Luna tube amp (some years ago). He found an output impedance of 8 ohms! This means a DF of 1 ohm or less! Combined with the typical varying impedance of most speakers this is way too high! How can supposedly competent engineers get away with something like this? Because the result is far from neutral, accurate SQ! No matter how pleasing the "golden ear" crowd claims!

Because the ear converts distortion into tonality, just because you have flat frequency response does not mean it will sound flat. A small amount of higher ordered harmonic distortion can introduce brightness. This is an additional reason of why two amps can have the same frequency response yet sound different. IOW, its not just the output impedance, its also the distortion signature.

All OTL amps like high impedance because they have lots of voltage but limited current. Since current is the limit use the forumla

Power = current squared x impedance. The amplifier max current is the same for both speakers but 16 ohms gives you twice the power of 8.

@ramtubes
Roger, I think if you revisit the above comments you will find them to be incorrect. An OTL has to be able to drive real world loudspeakers and so can produce the same currents at the output as any other amplifier. FWIW, the output tubes in most OTLs can easily blow a 10 amp fuse in certain situations without damage to the tubes.

Your use of the formula is not accurate. As an example (I think you have clio9's M-60s on hand) look at the output power at clipping into 4, 8 and 16 ohms. Now this is a smaller OTL, but I think you can see that its output power does not behave as you stated above. The half power of 16 ohms occurs at 4 ohms, not 8.

Two versions of the Futterman circuit. East coast/West coast. Harvey was a hoot. I visited him and his gang around this time. Sadly or not, he folded too. Is there a Futterman curse?

There most certainly is! Every manufacturer that has made a Futterman amplifier has had to go out of business. This is because there is more to the circuit than meets the eye, and no-one was able to do the execution such that a reliable amplifier resulted. The reliability problem is what drove them out. Of course, there are notable exceptions- the Fourier company simply under-rated parts and used sloppy construction, to the point where it would not have mattered what they made- they would have failed and gone out of business anyway. Harvey used surplus and unreliable capacitors in his amps (unreliable because they were not used in the right application).

Since I dont believe wire has a sound I prefer colors. I use all 9 colors in my amplifiers. With colors you can actually start to see the circuit without a schematic.

If you look carefully you will find some long wires repeated so you could simplify things there.

Our  wire is custom-built. Because of that we have to buy a lot of wire at any one time. The long wires repeated is so that all the power tubes have exactly the same series resistance involved (although the cathode resistors dominate that aspect) and transmission line effects are minimized (the output section has bandwidth to several MHz so stability is important). Once you understand how the dielectric behaves you find no need for Teflon.

If my amps are being run at levels that do not bring on distortion, why do they still have that classic "tube-amp" character even at those low listening levels? If it’s not the clipping characteristics that are coming in to play...what is it that produces that classic tube sound as I described it?

@prof
The fact of the matter is that the amps make audible distortion, which is the coloration you hear. Below a certain low power level they can often be making more distortion than at higher levels!

I have a 3 disc LP set of Theodorakis’s “Canto General” that Atmosphere produced, and I found the engineering to be disappointing. It has been muddy, and there has been no bass to speak of.

@unreceivedogma
There is plenty of bass on that recording! We had the biggest bass drum in the state at the time. But it is very deep, and some systems don't play it very well. Since I recorded it, I can use it as a reference and I can tell you that many tone arms don't play that bass right either.

Preamps are much easier and to me not so interesting

Roger, IME this statement is false. Many good amplifier designers think that a good preamp is no big deal and then go right ahead and design a poor preamp as a result. This is totally because they really in fact for real don't know what a preamp does! - which is to say, a lot more than just the gain and bandwidth, that sort of thing. If a preamp isn't right, it makes no difference how good the amp or speakers are, the missing information can't be recovered downstream.

I wanted the RM-200 to have good CMRR (hum rejection in simple terms). One cannot do that with a tube at the input.

We get pretty high figures and we do it with a tube.... As a hint, look into 2-stage CCS circuits. You aren't going to get good numbers without a decent CCS, a resistor or a single-stage CCS won't hack it.

It also has something few amps do not have which is the abilty to drive a dipping load with increased power rather than decreased power. Neither CJ, Rogue, ARC or anyone else I can think of has done that.

A good number of 300b SETs can do that. The Wolcott did as well.

Horn speakers tend to have peaks in the response.

Some but not all. CAD has done a lot to improve horn response- if there are problems in the throat where it couples to the horn, all is lost. But I've heard several horn setups where this common problem is overcome, and the result is that they sound very much like ESLs.

I have 6,000 LPs. A lot of them have a good bottom, some have an excellent bottom.

I love the Theodorakis performance, but alas it’s like listening through cotton.

Let’s see what it sounds like in my new room. I’ll let you know.

I have a similar number of titles. The bass really is there. There were only 1000 pressed, so its not likely to be a worn stamper. 

I still find that OTLs at low impedance are current limited. As to 10 amps without damage. When I put a 6AS7 on the curver tracer and go just a bit above the peak rated cathode current I see flakes of cathode coating coming off like sparks from a sparkler at much less that one amp. WIth the grid being so close they can easily fall into the grid wire and POOF. Horizontal output tubes that Futterman and I use are specified for high peak current about 1 amp.

I guess this really depends on what is meant by 'current' (since the word has become a charged term in audio)! And a lot depends on the 6AS7 in question too- the GAs don't hold up; most American tubes have problems in our circuit as well since they really aren't intended for fixed bias operation. We prefer the Russian variant; they hold up the best of any we've seen.

BTW, I like your approach to the whole 'damping' thing and I also appreciate your use the the phrase 'output regulation' (which I see as opposed to 'output impedance') which I see as a more accurate term. You are spot on that far too much attention has been placed on damping factor- have you seen this article by the former head engineer of EV? http://www.dissident-audio.com/Loudspeakers/CriticalLSDamping.pdfIn is we see that no speaker made needs a 'damping factor' of over 20:1 and some need quite a bit less! FWIW the original AR-1 was designed for a 1:1 damping factor.
Also kudos for the comments about too much power. Many amplifiers make excess distortion at lower power levels and when too much power is available in the amp, 90% of the listening will be in this higher distortion region- so not really taking advantage of the amp's capabilities.

BTW I first met Bill Johnson and Robert Fulton at Bob Fredere's house in Minneapolis where the two would meet for listening sessions (at the time the D-150 was ARC's SOTA amp). I wound up running a set of Fulton J's and then Premiers for several years.

The 6AS7 is a pass tube in a DC power supply. Heres a link, one has to scroll down a bit and read the application paragraph at the beginning. I dont think brand is going to make a lot of difference as they are all made for the same application as stated clearly here. .

I've certainly seen the page for the 6AS7 :)  What I am telling you is the the 6H13C is a different tube (the Russian variant). Once preconditioned, they can hold up better than an American tube. Far less likely to see the cathode coating falling apart, at least until the tube gets weak.

Regarding bias stability, the amp has no need for a servo. If the DC Offset is unstable, its likely noise in the driver tube as the output tubes obtain their bias from the driver tube. Normally the DC Offset is the sort of thing that you set or at least check once every 6 months or so. IOW, its **very** stable!

Current is well defined by science. I didnt know there was a special audio definition.???

There is! I regard the audio versions as common myth, often bandied about inappropriately:
http://www.atma-sphere.com/Resources/Common_Amplifier_Myths.php

Although we disagree on many things I do appreciate your Gentlemanly approach, unlike that other fellow who left us.

Thanks - let's hope he stays away. His approach produces so much noise, its impossible to have an actual conversation, not to mention his creation of an entirely new wing of physics (or at least alternate meanings to words to which no-one was previously aware)...

Just to be clear, I have a lot of respect for you as I do Nelson Pass, John Curl, David Berning and a number of others. There's a lot of snake oil in this business so its refreshing when we don't have to deal with that. Like you, I've been at this a long time but went down a different path a long time ago:http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.phpand while I understand completely how the Voltage Paradigm works (IOW I don't make amps to be 'tone controls'), I'm not at all convinced that the Voltage Paradigm is the only way to achieve the most neutral presentation. For me, the reason was best expressed by Norman Crowhurst, who pointed out a good 60 years ago that while feedback of course suppresses distortion, it also introduces some of its own, which tends to be entirely higher ordered harmonics. Its not that I'm against feedback, but its inappropriate or inexpert application does bother me, and for that I'll use the current ARC amps as an example.

The problem is that the ear converts all forms of distortion into tonality (and the ear/brain system has tipping points where that tonality can be favored over actual frequency response), and the the most egregious problem in audio IMO/IME is brightness (and its twin brother, harshness), which in transistors is entirely caused by distortion; also in many tube amps that use feedback. This is because the ear uses higher ordered harmonics to sense sound pressure, so is really sensitive to them as a result (moreso than most test equipment due to the range that the ear has to cover)! The line I draw in the sand is I want it to sound like real music as opposed to a just a good stereo. To this end, I do my best to vet every customer's system and expectations in order to make a sale. This limits my sales for sure, but it also results in really excellent results both in sound and customer loyalty when everything is set up right. 

Regarding the different gain in the M-60s, its entirely possible that the CCS is damaged. Clio9 was an early adopter of Mk3.3 and I suspect he has earlier CCS boards in his amps. IIRC that is... If you can send a photo of the CCS board to my email (found on the atma-sphere.com website) that will tell me a lot.

Ralph, I measured your preamp's RIAA and it was +4 db in the bass. That is not accurate EQ. My RIAA accuracy is +/- 0.2 dB.
.
What interests me is how it came to be thought that preamps are harder. What designer is going around saying that? Im not.

The preamp's math conforms to Stanley Lipshitz's math and the parts are within 1% or better, FWIW.
The thing about a good preamp is that there are so few! In that regard I would say there are a lot more competent amps than there are preamps.

Well, I lied. I have to respond. Ralph-I can't believe that you said that about me after I have done nothing on this Board (in multiple threads) but praise you and after you agreed that preamps are not easier to design (well) than amps.

@fsonicsmith
Sorry- my comment was not aimed at you- I had missed that you had left. I was in fact referring to someone else, who (thankfully) has not posted on this thread at all. I see now that in my haste to get thru all the new posts that I missed several posts somehow and yours was one of them. So I misinterpreted Roger's comment. I apologize- I've no reason to drag your name through the mud!

I’m sure you misspoke when you referred to pi-squared. The 6.28 factor you referred to is of course correct, but that is 2 x pi, not pi-squared

Yup- thanks for that. My medication is a bit too powerful today...

It adds better isolation on high frequencies between power supplies.But what do you think does it have any real effect?

What do you think about separate power  supplies between left and right channels for 300B SET amplifier? Does it do a significant difference?

A choke is helpful as it can be used to help filter out the 60Hz sawtooth waveform that is part of the rectification process. They can be quite helpful in that regard, as to do it with resistors and filter caps essentially requires more filter caps and also means a bigger voltage drop in the power supply, where voltage might be at a premium. IOW you can have a fairly low voltage drop across the choke due to a low DC resistance, yet still filter out noise in the supply.

Separate supplies is not a bad idea at all, although in a class A1 SET you shouldn't have a lot of noise in the supply. If you have separate supplies though, you could also have separate chassis for left and right. This would give you the ability to run shorter speaker cables, which is really helpful when working with amps with a higher output impedance such as an SET.

The trick is that when things stereo share a common power supply its possible for them to talk to each other (the term is 'crosstalk') and essentially interact unless great care is taken to deal with the power connections and ground connections correctly.  Otherwise you can increase IM distortion and IM needs to always be kept as low as possible. 

Hi Ralph. Im curious if you use an inverse RIAA network to check your EQ. If so where did you get the values? I built mine from the advice of Peter Moncrieff, Mitch Cotter and Dick Sequerra who were the first to discover that many RIAA EQ curves were off. Of course I used precision parts.

Of course! Its been a really long time and I do not remember where I got the values, but when used I got flat response. FWIW it also plays flat when I cut a lacquer on my lathe and play it back on the preamp. The Westerex has pretty tight curves in this regard; the electronics are matched to the cutter head.

5. What is the best and cost effective way to separate power supply between output tube B+ and driver tube B+?

@alexberger

You can of course make sure that the timing constants in the power supply are low enough that the driver and output section can’t talk to each other. Here’s the math:

F = 1/RC times Pi squared.

That formula results in some hard to work with numbers, since F is frequency, R is resistance and C is capacitance in Farads, which is really inconvenient. So I usually useF=1,000,000/RC * 6.28
(6.28 being Pi squared).  This formula results in F in cycles per second (Hertz), R in ohms and C in uf (microfarads), which is more real-world.


In your power supply, C is the power supply bypass capacitor, which is probably an electrolytic device. R is the resistor between the power tube B+ power supply filter cap and the filter cap for the driver. The thing is, there will be a certain amount of current that the 6SN7 needs, so you have to make sure the resistor is large enough wattage to survive that.

For that you need Ohm’s Law which is R=V/I


R is resistance in ohms, V is volts, I is current (C was taken already so I is the convention for current).


To calculate wattage (of the resistor) W=V x I


So for example, if you use a 40uf filter cap, and the driver tube is drawing 9mA (0.009amps) between the two sections, then for the power supply to have a cutoff at 0.5Hz the resistor value will be 8,000 ohms. There will be a 72 volt drop across the resistor and it should be a 2 watt device.


Now the trick here is to make sure that the coupling capacitor has a cutoff frequency higher than the power supply cutoff! Otherwise, the circuit can become unstable, prone to ’motorboating’ (a low frequency oscillation) and IMD will be higher. Use the same formula to calculate the value of the coupling capacitor; R will the resistor in the grid circuit of the power tube that the capacitor is driving. To get 0 phase shift at 20Hz, the coupling cap should go 10 octaves lower, or to 2 Hz; our power supply cutoff is safely below that (although it would not hurt to go an octave lower by doubling the value of the filter capacitance). The problem here is that SET output transformers often don’t have good low frequency bandwidth, and the 300b is right on the cusp of where 20-20Hz is actually sort of possible with a good output transformer. Even though your speaker may not go that low, its a good idea to get as much bandwidth as you can to preserve phase relationships that the ear uses to detect soundstage width and depth. So get a good output transformer.


If the manufacturer has issues with low frequencies saturating the transformer, you can reduce the value of the coupling capacitor, but I would be hesitant to do much of that as phase shift in the bass robs the amp of perceived bass impact, even though its flat on the bench.


A less cost effective way to do the power supply for the driver is to use a separate power transformer and power supply for it. In this way no matter what signal condition exists in the output section, no amount of noise in the power supply of the output tube (such as a general voltage sag) can talk to the driver section. However in SETs, this usually isn’t a problem unless you run the power tube in class A2 or class A3. The latter classes of operation can draw more power from the power supply as output power is increased; class A1 does not do that.

What is the role and importance of significantly upgraded power cords in component design? It seems to me that the PC is a fundamental part of the electronic device and that the manufacturer should optimize the cord for the device in building it. What are your thoughts on the matter as someone who actually builds electronics?

There is a voltage drop across any power cord that can influence the output power of the amplifier. What is less well understood is that it can affect distortion too. Much depends on the amplifier design of course, but these effects are measurable and no surprise in some cases that they can be heard too.

Question about the effects of different impedance settings and phono cartridges.

The loading is there for the preamp, not the cartridge! I know this sounds counterintuitive but here's how it works:
The cartridge has an inductance and the tone arm cable has a capacitance (along with the input capacitance at the preamp). Together they form a tuned circuit. With a Low Output Moving Coil cartridge, this resonance can be at several MHz. If the preamp is unhappy about Radio Frequency Interference (RFI) at its input, it will affect the sound- the additional distortion will likely make it bright. The loading resistor detunes the resonant circuit and thus prevents the RFI. Problem solved?
The downside is usually two things: 1st, by loading the cartridge in this manner you are causing it to do more work. The energy has to come from somewhere and that means the cantilever will be more stiff- less able to track higher frequencies! The second problem is that (generally speaking) a preamp that requires loading to sound right probably also has stability issues, usually because the designer hasn't sorted out the facts in the 2nd paragraph below your quote. This instability often results in excess ticks and pops that **actually would not be audible were the preamp actually stable**!

This latter fact comes as a surprise to many people, but when you are exposed to a stable phono circuit, it is very relaxing to not be hearing so many ticks and pops. This has nothing to do with bandwidth BTW. For more on this seehttp://www.hagtech.com/loading.htmlandhttps://www.whatsbestforum.com/threads/cartridge-loading-a-misnomer.15077/

Thanks.
I didn't totally follow all that, but that's due to my level of ignorance.

@prof

Don't feel bad. A lot of designers don't get it either!

Good question. Cartridges are very different in how they respond to loading. My Denon 103 is a 14 ohm cartridge (as I recall) and likes 100 -200ohms load. More load drops signal level, dont ever go that far, and makes the sound rather dead. No load is rather bright.

On the other hand the Lyra cartridges are so low in impedance that they dont respond to loading so we, in the SF audio society did some tests and found the Lyra best unloaded.

There is not any relation to speakers and amps that I would care to make. A cartridge is a source, the load is a resistor. Not much else going on.

Roger, you might want to do some reading at the links I posted in my prior post. Most of this post (except perhaps the comments about subjective listening) is incorrect. As I pointed out earlier, the load is not for the cartridge's benefit- its about the preamp.

The 6SN7 was made for black and white TV and was never, to my knowledge, used in the audio chain. 

There's plenty of old school audio electronics that used the 6SN7. As you know, its geometry is similar to that of the 6CG7/12BH7 and 12AU7 (the latter being the same as the 12BH7 but with the entire structure sawed in half). If you need examples, Google:
https://www.google.com/search?client=ubuntu&channel=fs&biw=1440&bih=729&tbm=isch&...

Now if the 6SN7 is not really a good audio tube, why does VAC use it on the entire line of amps as input tubes?  He could have used a 6SL7 or 6CG7 or variant.

Its because it *is* a good audio tube. That is why so many manufacturers used it on the old days, prior to octal based tubes being replaced by miniature tubes. Its considerably less microphonic than 6DJ8s, although almost any signal tube has to be hand picked for low microphonics.


Folks, you can't just pull a tube out of a box and expect that it going to do the job. It really should be tested, but testing for microphonics means listening to the tube to see if its acceptable.


I don't agree with George's comments about microphonics equating to euphonic character! As anyone who has played with tubes knows, microphonics is *not* musical or euphonic. If you hear that quality in a microphonic tube, its doing that **in spite** of the microphonics!

I think microphonics can sound euphonic. I built a Bottlehead be pre-and while it was fantastic the tubes in that application were very microphonic. It lends everything a rounder more robust sound as it’s like a delayed Feedback.

That's a new one on me- usually microphonics causes harshness!

Please explain. Im all ears.

Actually Roger, I did that earlier. You must not have read the post?
The loading is for the benefit of the preamp, if its sensitive to RFI. If not, no loading is needed. IOW if you need loading to deal with brightness, the preamp has a problem with RFI. The loading resistor detunes the tank circuit caused by the cartridge and tone arm cable and thus knocks out the RFI caused by the tank circuit when driven into excitation by the energy of the cartridge. Here’s a couple of links that address this in greater detail; the link to the What’s Best forum includes posts by Jonathan Carr, a noted designer of LOMC cartridges. The one to Jim Hagerman’s website has some of the math and some charts that show whats going on:
http://www.hagtech.com/loading.html
https://www.whatsbestforum.com/threads/cartridge-loading-a-misnomer.15077/

Why, because it was made for audio, not B&W television.

There are some great books on the history of Heath written my people who were there.

This other stuff about sawing tubes in half make no sense. Ive taked to tube designers and they dont talk that way... at all.

With regards to the 6SN7:

From the RCA Receiving tube manual, RC-19, published 1959, page 229:
2nd sentence (after the part about being used in television circuits):
"Each unit may also be used in multivibrator or resistance-coupled amplifier circuits in radio equipment"
The same phrase about ’resistance-coupled amplifier’ is used to describe the 6SL7, 12AX7, 12AT7 and so on.

Roger, when you say that you don’t know of anyone that used the 6SN7 in audio applications, as in this post:

The 6SN7 was made for black and white TV and was never, to my knowledge, used in the audio chain.

And then turn around and say this:

My dad built the WM-2 in 1956 so I am very aware of 6SN7, as was Heathikit who used them everywhere they could. Are you aware that most of their early products were built largely from WWII surplus of which there was tons. They would buy tons of surpus and then figure out what to do with it. 6SN7 were in great abundance.

--- Could you clarify what you meant here? These statements appear contradictory.

Regarding the sawed in half comment, sure, tube manufacturers don’t talk that way and for the record, I don’t manufacture tubes. But I do use them. Look at a 12AU7 and compare it to a 6SN7 (GE 6SN7s are the best example for this) or a 6CG7; the 12AU7 plate structure is half the height. And when you look at the specs, extremely similar to the 6CG7/6SN7; the geometry and spacing was maintained. So ’sawed in half’ is a good layman’s description. It makes plenty of sense.

To those who write up pages of pseudo science and create paradigms to make excuse for bad specs I no longer care to see here. Skilled people in this industry have come up with some minumum standards for noise, distortion and output impedance.

"Skilled people"? Do you mean marketing?
Uh, Roger, in a way this seems aimed at me (the use of the word ’paradigm’; I’ve not seen anyone else here use that word). If so, you’re way off base (and I regard the attack as un-called for). Don’t think for a minute that we built our amps without feedback because we couldn’t apply it! In case you don’t know what is meant by the Power Paradigm, as opposed to the Voltage Paradigm, I did explain it at this link:
http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php

It is in no way an excuse for bad specs- and in fact our amps have some pretty good specs (if proper measurement technique is used, which means **don’t ground a speaker terminal during testing**, which is the mistake that almost everyone except Charles Hanson made/makes). I suspect you didn’t read the paper at the link very carefully, since you claimed that you read it, yet still with the remonstrations!

Briefly:The Power Paradigm is what was around prior to the Voltage Paradigm; the ground work for the latter being laid down by MacIntosh and EV in the late 1950s (it was not until the early 1970s that it had fully taken root). At the time, the only way to build an amplifier that acted as a voltage source was to use enough negative feedback to get proper output regulation. But it was well known at the time that this didn’t work for all speakers made- the speakers had to be built to work with the concept as well. That is why I use the term ’Paradigm’ (and I *also* use it because any thought outside of that platform is often regarded as heresy).


Here is an interesting example that at which you should take a look so you know I’m not making this up:
https://www.google.com/search?client=ubuntu&channel=fs&q=fisher+a-55&ie=utf-8&oe=utf...
The first hit of this Google search for a Fisher A-55 is an image from a YouTube video. In the image we see the damping control of the amplifier. It is labeled at full counterclockwise " Constant Voltage", at noon "Constant Power" and fully clockwise "Constant Current". And yes, we’ve had this conversation before.
Speakers built in the old days that were under the Power Paradigm were usually equipped with midrange and tweeter level controls. These controls were not there to adjust the speaker to the room, they were there to adjust the speaker to the voltage response of the amplifier, which was an unknown (examples: vintage horn systems, Acoustic Research AR-1, 1960s KLH loudspeakers, the large Advent...). The Voltage Paradigm was an attempt to get away from having to do that- more ’plug and play’ so to speak.

Like any problem, the solution introduced its own problem- that of brightness as a coloration. This is a problem in every amplifier that employs loop feedback. How bad the problem is depends on the skill of the designer. The brightness is not caused by a frequency response error of course, it is due to residual levels of higher ordered harmonic distortion caused by the feedback application itself. See Norman Crowhurst.

The traditional argument is that the residual distortion is negligible. It is not!! The ear *uses* higher ordered harmonics to sense sound pressure; adding them in even trace amounts results in brightness that does not show up as a FR error because the ear converts the distortion into tonality.

(In their book "Control Design And Simulation", Jack Golten and Andy Verwer discuss (in chapter two) with regard to applying mathematical models to the real world: "...mathematical models invariably involve simplification. Assumptions concerning operation are made, small effects are neglected and idealized relationships are assumed."

It is the mark of a good engineer to know when and which things should be assumed, neglected or idealized. I maintain that violating one of the human ear/brain most fundamental perceptual rules is not good engineering.)

That the ear converts distortion into tonality is well-known. That the ear uses higher ordered harmonics to sense sound pressure is also well understood and non-controversial. It can be demonstrated by a test using simple test equipment.

Many high end designers recognize the brightness-as-coloration problem; the issue is what to do about it? That is why you see so many zero feedback amps that have "bad specs" but for some reason sound quite good. As I’ve pointed out many times, the "specs" don’t recognize the human hearing perceptual rules so what looks good on paper does not always translate to what sounds good as well. This fact is also non-controversial- we see it all the time in the pages of Stereophile, where a reviewer liked an amplifier quite a lot, and yet John Aitkinson is perplexed because the equipment measured poorly. The simple explanation, offered by HH Scott’s head engineer, is that the wrong thing is being measured.

BTW, these comments are not to be construed that every amp that measures poorly must sound better than it measures! There are some bad products out there and you and I are likely in agreement on many of them.

What can I expect to pay for a recap (plus any normal servicing)? I can actually drop off the amp at Krell myself (they are an hour or so away).

While it might be old enough that some of the filter capacitors could be getting near failure, that is not the problem here. Filter caps don't make static noise. Is this problem in both channels or just one? If one, I would suspect the input transistors.

My pre amp is a Conrad Johnson Classic II. Output is 200ohms.
The Krell input 47K--despite this, I am using an attenuator before the amp inputs.

The input impedance has nothing to do with whether you need attenuation or not, FWIW. I'd get an estimate from Krell before deciding your course.

When an amp specifies a maximum current what exactly does that mean and how is it relevant? On another thread there's discussion of an amp that has max current of 29 amps and puts 100 watts into 8 ohms. Since 29 amps would be a whole lot more than the current into 8 ohms at 100 watts or into any normal load this must have some other relevance. 

Power is equal to volts times amps and is also equal (thru Ohm's Law) to current (squared) times Ohms.
In this case, to make 100 watts into 8 ohms, you divide 100 by 8 and take the square root, which is about 3.4 amps. That's all thats needed to make 100 watts if the load is 8 ohms and it makes no difference what kind of amplifier it is. So what is the 29 amps??

Let's do the math the other way- 29 squared is 841 watts if into a 1 ohm load, but if this amp can double power as impedance is halved, clear down to 1 ohm, the wattage would 800 watts, not 841! The math can't lie about stuff like this, so clearly if the 29 amps is real, it has to be something else. Often it is- its the amount of current that can flow if the power supply is shorted for 10 milliseconds. Mostly that's a measure of storage that the capacitors in the power supply have and probably does not say much about how the amp measures or sounds. Also FWIW, there are tube amps that have that much 'current'. IMO, the 'current' (please note quotes) is often a misleading figure as current can't exist without voltage no way no how.

I am likely one of a very few designers that actually will repair others equipment. I do this because I like to see others work, how they make things, how their things perform. I have tested over 400 amps and preamps and can tell you all about what is going on in them.

 Well there's another thing we share. I put myself through college doing consumer electronics repair. I got the job right out of high school in 1974. I enjoy seeing how things are built and how problems were solved. We don't see much gear shipped to us, but we do a lot of local repairs on competitors equipment, as well as guitar amps, semi-pro and pro audio.

(although I am unclear if the phone company developed the standard which I have heard referred to as the 600 ohm standard).

The 600 ohm aspect has to do with the spacing of the lines on telephone poles. You've seen them many times - that spacing causes the resulting transmission line to have a characteristic impedance of 600 ohms. This means that if terminated by a 600 ohm load, there will be no reflections at the termination in the transmission line- its not that the impedance of the transmission line is 600 ohms.
So output transformers driving a balanced line were set up to drive a 600 ohm load. When the balanced line system was brought into studio and radio station applications, the 600 ohm became a standard which stood for many years and many pro audio components still support it.

Well, it turns out that it was my friend Ralph (we have never met, I simply call him my friend) who misspelled Charley's name.  

He did the measurements of our old M-60 kit for Glass Audio. I met him and hung out back when he was still Avalon Loudspeakers. Different world... nice guy. He is missed.

Also, while I don’t recall the exact numbers, Ralph has stated in some past threads that he has observed remarkably high levels of energy emanating from LOMC cartridges at ultrasonic or RF frequencies.

@almarg
The Hagerman site shows a 30db(!) peak. I can confirm that.

About 35 years ago I used to subscribe to the cartridge-needing-a-load theory. I set out to build a device that would come up with the correct loading so as to eliminate guesswork on the part of the user, which I was hoping to build and sell.

What I found when ringing the cartridges was that they really didn't ring until you got to some very high frequencies well past audio! This is easy to understand- if you pass a squarewave through one, there isn't much inductance to mess with the squareware at audio frequencies.

So if the brightness of an unloaded cartridge isn't due to ringing, what is doing it? I was lucky- a serendipitous event caused me to realize that the preamp played the bigger role.  At that point it became a simple engineering task to insure that a phono preamp design would be resistant to this sort of problem.
In a way, I did get what I set out to achieve- being able to sort out the correct load, which for all LOMC cartridges is 47K. What is needed is a preamp that is unperturbed by a 30 db peak 100KHz or above into the low MHz.... Not that hard once you know what's afoot.

Some products allow the music to communicate to you, some don’t, and a combination of components may speak to me, but not to you. I know this isn’t a technical question, but I felt compelled to write it nevertheless.

Dr Herbert Melcher showed that there are tipping points in the brain. Normally music is processed by the limbic system. But when things audio go amiss, the processing is unconsciously transferred to the cerebral cortex. At that point the music 'subjectively' loses its 'soul'. For this reason, its really important to know how the ear/brain system perceives sound, so as not to violate its perceptual rules (if you want the gear to be emotionally involving). Once you know that, it becomes and engineering task that is fairly mundane.

He made excellent equipment and I can see from JAs measurements he did a good job on the three major characteristics of a good amplifier, one that would drive a wide variety of speakers well. On this I am in complete agreement with Charley. Im not sure why I need to listen to his amps, I am confident they sound find. Charley and I are on the same page with what is important.

Charles made **zero feedback** amps and IMO his are some of the best solid state amps made. What I like about his approach is that he solved the issue of an amp that acts as a voltage source while also lacking the usual coloration of brightness caused by the distortion of added feedback. The industry needs this sort of diversity and he is missed.

I feel one of the great benefits of tube products is the limited bandwidth eliminates susceptibility to much of the higher frequency ac noise. Comments?

You can have plenty of bandwidth and not have noise problems. Noise is best handled by proper power supply design and good grounding technique. I find also that fully balanced differential circuits are handy for noise rejection as well. This means that you can have a wide bandwidth tube amplifier and have it be very quiet, even though it may not be shielded by a chassis.

I read both links. The first is common knowledge and nothing about RF. The second is long so if there is something in particular you want to me to read please quote it here. Are we playing "Wack A Mole’ here?

Im really tired of RF being the devil for everything. Usually if there is RF sensitivity you will hear an AM radio station. If there aint no radio coming in there aint no RF. Lets get real please.

For the What’s Best forum, just look at JCarr’s posts. No whack a mole- not even sure what that comment is even supposed to mean.

No-one has said RF is the devil for everything except you, just right now. But these two links did point at RFI (or ultrasonic or near ultrasonics, in the case of MM cartridges) being a problem with LOMC cartridge operation... I don’t see how that can be construed as ’everything’.

Ralph has not honored the stated purpose and rather come here to once again to hawk his paradigms and unusual ideas about cartridge loading and RFI. You are into vinyl I see. Do you agree with his loading suggestons? Dont we load a cartridge to change its sound? That last one about loading a cartridge for the sake of the preamp was so out of the world. Most of just a juse a ferrite bead to stop RF from coming in.

This statement is outlandish and false. Please note though that I am not attacking you, just the veracity of this statement, unlike you who sees fit to attack me personally. I have honored the stated purpose, as I am a designer of amps and my door is open.

If you say you looked at both links and came away with the idea that RFI has nothing to do with it, the only conclusion that can be drawn is that you didn’t in fact read the links or failed to comprehend their significance. When we are talking about noise in the MHz range, that’s RFI, especially when it is the result of excitation of a tank circuit.

I do agree, Ralph should have never entered this thread nor should you. I dont see a question and I dont see a contribution. You left, said you wouldnt come back, but you did.

Roger, you seem to appreciate my being courteous when I offer corrections (which you seem unwilling to accept), but you seem unwilling to be civil in return.


Contrary to yours and others impressions, this is not your thread- its a public forum.


As a bit of a hint, I find that keeping decorum in spite of personal attacks (including outright untruths) is a simple tool to maintain credibility. FWIW I have an EE just like you and I don’t engage in pseudo science any more than you. I can prove everything I say. I suspect that you are not *as* acquainted with equipment from the old days- 1960s and before, as I might be. You have said that my talk about the Power Paradigm and the Voltage Paradigm is the pseudo science to which you refer; to me that just suggests that you don’t have a grounding in history. This despite my showing an example of what I was talking about (one that we have discussed prior); in spite of that evidence (the Fisher A-55 I linked, and any amplifier with a ’damping’ control) right in front of your face, yet you still accuse me of pseudo science!


But:
You are getting out of hand here. This entire thread is in violation of Audiogon posting rules, for the simple fact that you primarily use it for promotion.
On occasion I get accused of the very same thing, but those that do so will note that I have yet to start a single thread on this site (despite being active on it since its inception in the 1990s), let alone one that focuses primarily how great I am in tandem with how great my products are, all the while demeaning (in some cases, falsely) others. I do mention my products when asked and also in the context of related questions (such as balanced operation; quite often because there is much misunderstanding about how balanced lines are supposed to work).

This is not your thread! - if that were the case it would be on your website. Instead you are posting on a forum that is open to the public. My presence here has mostly been to keep the record straight when things have gone off the rails. The 6SN7 thing and the cartridge thing are two examples.

I can see that it bothers you as much as it does me when you see things that are untrue. If I can offer a bit of advice:We often go thru live with made up stories. In fact we live as if those stories are true. When life and the stories don’t agree we suffer. At that point, to end the suffering the best thing to do is drop the made up story. You will note that I do not attack the voltage rules as being made up. I regard them as an engineering solution, but one that has some bad applications. I’ve pointed out why earlier and don’t need to repeat myself. The old power rules are one way out of the resulting colorations of the voltage rules, but they come with their own price (limited market, for one, more fiddling and more careful equipment matching for another).


IOW, if you think I made up the idea of the power rules, you are simply not grounded in history.


I am perfectly happy to have that conversation on a different thread. I’ve been having this conversation for the better part of 30 years and its not been debunked by any engineer that listens- only those who don’t and are further unaware of how the voltage rules were developed.

I dont mind any designer coming here if we both understand electrical science in its accepted current form and speak the proper language. If you want to read made up science there are pleanty of other threads on here for that.

I have a a question for Ralph, Why have you not sent an amplifier to Stereophile for review? They have reviewed all of mine and preamps too.

I got my degree at the UofM. I don’t subscribe to anything made up. Enough with the innuendo already- if you don’t get what I’m talking about, just say so, otherwise lay off. John Curl got it, Nelson Pass got it, other engineers on this site got it...

Regarding Stereophile:
We sent a preamp to TAS for review. It did quite well- the reviewer bought it and said so in the review. Then he moved over to Stereophile prior to the publication of the review; it got published there. At that point we knew we were in trouble when we got the news that he had jumped ship to Stereophile, as we had been warned by people from several companies (ARC among them) that if you can’t afford advertising with them, they will treat you as a sacrificial lamb to show how hard hitting their reporting is. Sure enough, some months after the initial review was printed (in which I had to challenge JA’s measurement comments, as the Audio Precision at that time didn’t properly support a balanced input, despite it appearing so- a different conversation), and after being asked if we would advertise in the magazine, the reviewer was later instructed by Stereophile editors to print a followup. Apparently when he was installing defecting tubes in the preamp, it didn’t work right. We tested several of those tubes; one was a Telefunken that was so dead that I thought my tester was at fault as it got no reading at all when the Telefunken was lit up. I had to test some known good tubes to know that I was not going nuts! To date I’ve yet to see another 12AT7 that bad that still had a vacuum. We got blamed in the followup for that failing on the part of the preamp. Do you have amps and preamps that function correctly on bad tubes?

A second product, the MA-2, was destined for a nice review in 2004 by Paul Bolin when he was with Stereophile, but a company that was trying to take over my company threatened Stereophile with a lawsuit if they didn’t give the amp back to them. Paul let me know this was happening, and since we owned that amp I went and picked it up. While the lawsuit never went down, JA was sufficiently rattled by that event that he removed Paul from the Stereophile staff.

why do some amplifier manufactures say to use a preamp between their amplifiers and the streamer when you are only using a streamer and power amp, no vinyl and no other sources. I understand the need for a volume control and to have the impedance's low for the streamer/DAC and high for the power amp. My streamer/Dac is <100 ohms and the power amp is 47K. There is an implication that the preamp does other things that affect the SQ.

The answer is highly system dependent of course, but one reason is that a preamp is or should be designed to minimize artifacts occurring from the interconnect cable between the preamp and amp. Sometimes digital sources are not very good at this. A second reason is that a preamp can provide a fairly high impedance load for the source, meaning that the latter does not have to work as hard to make its signal. This could result in lower distortion- and no place better in the system to get things right!

Source output 22R, amp inputs 23k5 or 100K. Adding a preamp with attendant switching, circuitry, noise and additional cables is little more than a subtle tone control. It maybe different, but it may not be better.

Some volume controls in DACs are actually far more than subtle in their operation. Stripping bits isn't the best way to control volume. For that reason the buffered volume control that can be in an active line section could be more neutral.


One thing is certain: when it comes to preamps, passive volume controls or running direct, because there are so many different products and approaches to this problem no generalization about the matter could be seen as true.

Apparently some people just cannot restrain themselves or have difficulty with reading comprehension

-Or there might be a third explanation.

minimize artifacts occurring from the interconnect cable between the preamp and amp.

A second reason is that a preamp can provide a fairly high impedance load for the source

Source output 22R, amp inputs 23k5 or 100K. Adding a preamp with attendant switching, circuitry, noise and additional cables is little more than a subtle tone control. It maybe different, but it may not be better.

I agree with ieales and not the answer given above his, but then you all know that. What are these artifacts, are they Egyptian?

If you've heard cables sound different from one to another, that is an artifact. Cables have capacitance, inductance and characteristic impedance (hence the 600 ohm standard used with balanced lines for several decades).

The thing that intersts me is that most complain that the bass of passives is usually lacking. However the bass of passives goes to DC with no phase shift. Tube preamps do not go to DC and at 40 Hz will start to have some phase shift. Is the phase shift what they like?

It would be interesting to see if it is indeed phase shift. Many preamps don't have any phase shift right to 20Hz but I suspect that passives often do- not because the passive itself does of course, but because it in tandem with certain sources might.

A preamp need not have DC response to have immeasurable phase shift at 20Hz. For that it only need to go 2Hz.

Technical question: Do certain electronic components, e.g. capacitors, need to "form" before they operate at their best?

Electrolytic capacitors do need to be formed. Most of this is done at the factory, simply by charging the cap to the forming voltage (and is actually the difference *in some cases* between two voltage ratings). This is to prevent the cap from being damaged at installation! However the caps are shipped without charge and can sit unused. They do need to form up properly again when installed.  In some of our larger amps, we often see the amp blow a fuse during turn-on (due to current inrush) that it won't ever blow later after the caps have formed up properly (which seems to take a week or so of intermittent use). 

More than once I've rejuvenated a system by re-plugging ALL connectors, both internal and external.

+1

If the capacitance of a cable causes rolloff, is that an artifact?

Of course.

The answer is that the industry cares more about power than nuance. As well many speaker designers don't realize the relationship between amp and speakers so they don't realize that by making the speaker lower impedance, the amp will make more distortion. In high end audio, its more about nuance than power, and higher impedance loads do that with greater ease.

Or there might be a third explanation.
Attainment of the tipping point of your cortex?

Simpler than that.


This is a public forum. Anyone who wishes can post anywhere on the forum, including this thread as long as they are a member.


I've been here since its inception.


I'm also an engineer, an amplifier designer and I answer questions.


But here's the kicker: This thread was put up as a promotional device. You can tell because the manufacturer is talking about his products, often more than once in many of his posts. This is against the Audiogon rules, but because I and others are posting here as well, the entire thread doesn't have to be taken down.

I notice on this account that Georgehifi isn't being taken to task; although while he is a manufacturer, he's not an amp designer. But this thread has drifted to the topic of preamps on occasion and on that account he's been posting too.

Does that give you the right to be a self proclaimed forum cop.

Of course not- neither am I attempting to enforce any rules.

That's the pot calling the kettle black

Oh- can you point me to a thread I've created on this site? Meanwhile, irony is still a thing, huh?

Wow, you've really got something up you this morning, must be the bagging Class-D's getting on the other thread, and you keep saying (promoting) your is coming soon and it will fix all the problems with Class-D.

?? Did I promote such a thing on that thread? I did provide info but IIRC no information whatsoever about what we're up to.

correct me if I’m wrong.

That has been my attempt in the past. I try to do it without personal attacks- the principle being: attack the argument, not the poster.

A promotional device in what manner? Do you think Roger is using this thread to try and sell more products? Since you called out Roger on this I will say you are walking a fine line here Ralph, especially given what you once told me in a conversation we had about marketing. The thread was started so that Roger can share his knowledge and experience with the community. He has requested that other designers not respond to the questions asked of him, and from my perspective it’s not that you don’t bring value to the discussion (and I get it’s a public forum), but when you bring up the self promotion stuff, I have to agree with George that it is the pot calling the kettle black.
You consistently promote balanced differential designs, not always mentioning products specifically, although correct me if I am wrong, but you have mentioned the MP-1 was one of the first balanced preamps in high end audio on more than one occasion. I have read many threads where you reference how your preamps solve the issue of cable artifacts coloring the sound, that you make them unity gain with a buffered output, and you have mentioned or alluded to your amplifier designs many a time as well. Granted you do so in a gentlemanly manner and as a means to educate. However, isn’t that what Roger is doing as well?

No. And I agree, its a fine line!

Tony, Roger is a friend of yours so this might be sensitive, but here’s the difference. I just try to present the facts and nothing else. I don’t see Roger doing that here; take a look at the attacks in his post just prior. I’ve got no problem with his answering questions; I’ve been doing that here for over 20 years. The difference is in that context he also sees fit to attack others- Cary, ARC, Atma-Sphere (and also me personally) and so on. And lots of mention of his amps and preamps that goes beyond just the facts. In my case when I mention our gear I make no claims about the sound, just statements about what it is- for example that our preamps are balanced. Take a look at the examples you cite. When I’ve made that statement you quoted (and others that Roger quoted), it was simply fact with zero comment about how the equipment sounds or performs, with the exception of the fact that if the equipment supports the balanced standard, then the cables used will be transparent. Go take a look. The reason for this is the audiogon rules- I can’t (or thought I can’t) just get up and say how great my stuff is while demeaning others. ’I was so amazed how this detergent cleaned the stains that that other detergent couldn’t.’ That’s *advertising*.

BTW I kept my mouth shut about this until specifically asked. FWIW, I’m a moderator on another site, and I know better than to attack others personally as we see in Roger’s posts below (and previous). I have a thick skin, which is why I know better than to return in kind:

I never imagined other amplifier designers would want to answer questions directed to me. If you think one of my answers is wrong you are wellcome to chime in but not with your paradigm or unsupported theories. I dont welcome any unsupported theories, poorly vetter answers from flawed articles.

....

Roger:


The problem here is simple: you cannot disprove anything that I’ve posted, while I can prove that its real, and have already done so on this thread. This might be the 4th time I’ve pointed this out. I am simply pragmatic; which to the best of my ability you seem to equate with ’pseudo science’. How much proof do you need? The problem here is not that I am wrong, nor is it that you don’t have engineering talent- yours are some of the better transformer coupled amps I’ve seen. But right now it seems that when presented with something that you don’t know about, it appears that you’d rather dismiss it than cause your hand to move and investigate (re.: cartridge loading, power rules). Good engineering practice is good science. In my case, I see if I can measure it; that’s how I found out that power cords can affect equipment performance both measurably and audibly.

You have a paradigm to promote with which I totally disagree. When I bring up that a widely varying impedance speaker will not sound as the designer intended, you bring your paradigm.. Perhaps your amp provides a tone control some like. I have 2 m-60s in my shop right now in my A/B test rack, anyone is welcome to come listen. One is stock one is my mod with feedback. It appears you have abandoned feeback on some psychological level rather than listening. We are just listening.

So please dont bring your, not vettet, paradigm to a scientific discusstion which is in general disagreement with it. That is self promoting to a high degree. Why post here when you have known for years we disagree about damping, distortion, current, tube applications and a host of other things? Ive read your paper over and over again and it makes little sense.

You are the only outside designer who has entered this thread.

When you answer questions from your point of view I have to deal with that and it makes more work for me. We already know what you are going to say you have said it 6,798 times.

If you really believe this then you missed the boat about what this is about.


Put it another way: Flat frequency response from any speaker in any room is flat out (if you will pardon the expression) **impossible**. You can’t name a single speaker measured by anyone that is really in fact actually flat. Plus, you can’t fix it with an equalizer- they don’t have the resolution.


We like to think speakers are flat, but such thinking is engaging in made up stories. Its fantasy.


And it turns out that for the last 80(!) years we’ve known that distortion is interpreted by the ear as tonality! See the Radiotron Designer’s Handbook, 3rd edition (page 67 IIRC). This fact is indisputable.


So what can we conclude? Certainly that the ear places an extreme emphasis on certain distortions (again, see the Radiotron), while not caring nearly so much about others.


Since feedback causes those distortions out of its application (see Norman Crowhurst), its a **guarantee** that any amp that employs it will have some coloration (brightness) due to the reasons stated above.


Brightness is the single biggest objection that people raise about audio reproduction. Women tend to have more intact hearing than men and its common for them to raise objections to brightness more than men. Anyone here with a GF or wife can attest to this. So maybe feedback to achieve a voltage source isn’t the way to go; women are after all part of the marketplace (WAF).

As I pointed out in my email to you a few days ago, in the old days before the voltage rules were introduced (1950s), speakers had to deal with the issue of unknown voltage response in amps because a lot of amps (SETs for example) didn’t employ feedback. This is the Power Paradigm, and if it irks you that I put a name to it, I’m sorry - you weren’t around when I did that- I also use ’power rules’ and ’power source’ as alternates. These speakers had controls on them to adapt the speaker to the voltage response of the amp. JBL, EV, Altec, Acoustic Research, KLH... you’ve seen these controls! I have to imagine that you must have thought they were to adjust the speaker to the room?

Some modern speakers have these controls too - Sound Lab ESLs, Classic Audio Loudspeakers and plenty more, if they are designed for amps with a high output impedance (power source). The Sound Labs aren’t, but because ESLs aren’t Voltage Paradigm devices; they need the controls in order to work with voltage source amps.


The idea of the modern Power Paradigm is simple: just don’t make the distortions to which the ear is keenly sensitive, and after that do your best to get flat response from the speaker. The Voltage Paradigm has it the other way ’round.
This is why I advocate the ZERO autoformer, as it allows you to adjust the voltage response of the amp without using feedback, which I regard as the bigger sin for reasons stated above.

Quite simply I (and other designers like those that make SETs) am not trying in my designs to do what you are trying to do. I’m trying to do what SETs do, but without so much distortion (coloration) and with wider bandwidth. Its not woo voodoo- I don’t go in for that anymore than you do; its all just engineering once you know what the problem is, which is stated above, but ad nauseum: the ear converts distortion into tonality. Get rid of the distortion, and the presentation **can** be more neutral.

But the only other relatively minor issue I’ve ever perceived in his paradigm paper is that as worded it might lead **some** readers to believe that the high output impedance and other characteristics of his amplifiers (and various other tube amplifiers) would result in precisely constant power delivery into varying load impedances. (In fact I’ve seen one or two posts by members here who do not have significant technical backgrounds in which that belief has been stated). Whereas the reality is simply that they will come considerably closer to accomplishing that than an amp which acts as a voltage source. To a greater or lesser degree depending on the amp’s output impedance and on how the speaker’s impedance varies as a function of frequency.

Thanks Al, and spot on.

No amp is a true power source. But if the speaker load (which can be quite variable and that's OK) is high enough then the amp can be **relatively** constant power on that load.

When you graph the amp's power vs load impedance curve, it looks very much like an airfoil profile in cross-section. There is a maximum power output, and at lower impedances the output power falls off rapidly as more of the power is simply dissipated by the output section itself. Above that maximum, power falls off very slowly as load impedance is increased. For example our M-60 is only a few watts less at 30 ohms as compared to 16.
While this is not constant power, its pretty close as the difference is less than 1/2 db.

With transformer coupled amps (VAC comes to mind) you have the ability to select the most ideal winding to push the amp more towards a voltage source or more towards a power source- depending on how the power tubes are thus loaded and the resulting output impedance. We use the ZEROs for that same purpose- years ago we used to use our Z-Music autoformer for that same reason, before the ZEROs existed.

Let’s look at these links as an example which are a couple you frequently drop into your posts, including on this thread which is where I pulled them from:

http://www.atma-sphere.com/Resources/Common_Amplifier_Myths.php

http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php

@Clio9
I put these papers up for the simple reason that I got tired of writing the same thing over and over (for example, the article about common amplifier myths gets linked a lot, as that is a very common myth around these parts). I recently added another article about balanced operation; its a lot easier to drop a link than write 200-800 word explanations all the time. That is the only place I know of where I can host something like that.
Other than that, I *do* walk my talk. Beyond that if there is a statement of error, I offer the correction as I have done here. Normally that isn’t a problem, but in this case I’m getting personally attacked over stuff that is normally pretty easy to understand. While I didn’t intend it that way, apparently my presence here is part of why this thread hasn’t been removed.

I dont like the high output impedance or high distorton. I do like Ralph’s amplifiers better than his paper. I wish he would just drop the paper. Its embarassing.

Can’t he just say what his amps do without having to invent this story?

Loudspeakers that operate under Power Paradigm rules are speakers that expect constant power, regardless of their impedance. Examples include nearly all horns (currently the Avantgarde Trio is the only known exception), ESLs, magnetic planers, a good number of bass reflex and acoustic suspension designs. Horns, ESLs and magnetic planers do not get their impedance curve from system resonance and so benefit from a constant power characteristic and indeed, many of these speaker technologies are well-known as good matches with Power Paradigm amplifier designs.

This paragraph is particularly disturbing. Is he saying ESL speakers are constant impedance?

@ramtubes

I’m not saying the ESLs have a flat impedance curve- far from it. Many vary from bottom to top by about 10:1.

Here is the meat of it right here:

Speakers are not power driven, they are voltage driven. Lets look at the low end resonance. Most speaker climb to 40-60 ohms at resonance (35 hz lets say). The speaker designer does not want the voltage to rise into that peak. He already accounted for that peak with the mechanical damping of his speaker.

I am really curious right now how many times I’ll be explaining this!
While the above quote is mostly true these days, its not actually 100% correct due to exceptions. First of all of course, voltage doesn’t exist without current and current times voltage is power, so obviously all speakers are actually driven by power. The bit about them being ’voltage driven’ is a **charged term or phrase**, similar to the idea of ’RMS power’; see:https://forum.audiogon.com/discussions/rms-power
On that thread, we see that the term ’RMS power’ is really something that was created to satisfy FTC rules for testing amplifiers back in the 1970s. Outside of that the phrase is ambiguous. I pointed this out at the inception of that thread. There are other charged terms in audio as well- ’damping factor’ and ’output impedance’ are two that come to mind- they have *very* specific meanings in audio, not so much elsewhere in the electronics industry (IMO/IME, Roger, your use of the phrase ’output regulation’ is more accurate).

This is the case with the phrase ’voltage driven’ as well. In this case, what it means is that the speaker is is intended to be driven by an amplifier that acts as a voltage source.

The reason is that the speaker has a resonance caused by a driver in a cabinet, as we see in the quote from you just above. Since the resonance expresses an impedance peak, if you don’t throttle back the power into that peak (IOW, maintain constant voltage), you get a tonal coloration. In effect, the impedance curve of such a speaker is also a map of its efficiency vs frequency: the higher impedances needs less power, lower impedances needs more. A voltage source satisfies this.


Now an ESL is not a driver in a box. In a nutshell its impedance curve is essentially based on a capacitor. Thus its impedance curve does not also represent its efficiency vs frequency; its not following the same rules. Many solid state amps act as voltage sources; if you put one on a Quad, the result is a loss of bass and way too bright.


Now as I mentioned in my email to you, some speaker designers build their speakers to work with Power Paradigm amps, and so apply a different technique. The driver in the box has a peak, but if you do the math and place the peak at a point in the box such that the box is starting to roll off before the peak, what happens is you can extend the LF response of the speaker down another half octave for a given box size. The Acoustic Research AR-1, the world’s first acoustic suspension loudspeaker, was designed for an amplifier with a 7 ohm output impedance.

You claim that this flies in the face of electrical theory, but you then have to explain a lot! Where did the voltage rules come from? Who introduced them? When did that happen? What was out there before that? I’ve of course told you... multiple times at this point. Its audio history.

Regarding distortion:

About high order harmonics I will ask you. Would you rather have an amplifier with 3% 3rd and 0.01 7th-11th or one with less than 1% 3rd and 0.02 7th -11th? My point is that the high ordered harmonics are rather small. Intermodulation will follow a similar path.

The one with the lessor higher ordered harmonics will sound smoother/more like music to the human ear. Most would prefer that, given a speaker on which both amps make good FR. The ear treats the 3rd very much like the 2nd and in that regard its relatively insensitive. We really should be weighting the various harmonics according to the ear/brain’s sensitivity to them. But our current test and measurement regime pretty well ignores human hearing rules that have been discovered since the 1960s. But if we did that sort of rating, a 0.02% of the 7th would look only fair because its pretty audible.

It is a popular myth that if an amp is high in THD it will also be high in IMD.


This is the inconvenient truth about human hearing; feedback violates arguably the most fundamental rule of human hearing perceptual rules: how we sense sound pressure. This is why SETs have made such a huge comeback in the last 28 years.


But only in high end audio.


High end is fundamentally different from regular audio, the difference being in regular audio its all about the money. In high end its all about seeing how far you can push it, making it sound real instead of like a good hifi. That is where I draw the line- IMO/IME if the speaker requires that the amp employ feedback, it will never sound real, at best it will sound like a good stereo.

I am comparing a 5ch and 7ch amp. Same brand and specs exactly the same except for one difference -- one is a 5ch and the other a 7ch. Same rating of Tor power supply ( 2 total valued at 1,230VA and 1,025VA)

All things being equal - will sound quality be relatively the same? Dynamics? ... or is one better than the other?

Also -- ideally how big of a power supply is best suited for an amp rated at 200wpc at 8ohms driving speakers that dip to low 3ohms?

It *might* be that the amp with 5 channels could sound a little better, solely based on your comments above, due to 5 channels sharing a common power supply instead of 7. IME probably not a very big difference in any event!
With regards to the latter question, the power supply for one channel should be capable of sustaining 14 amps continuous to sustain the output section of the amplifier (assuming that the amp is doubling power as the load is halved); a bit dependent on the class of operation (for example, a class D amp won't need any more than that, while a class A amp will need a bit more to sustain its driver section.   Generally though its a good idea to have some reserve available in the power supply so that it runs more reliably. I prefer to have the power supply to have considerably more energy than the amp is going to need to make so that it won't limit current when things get complex.

Isn't the presumption with most speakers that they will be driven by a voltage source? And by assuming that, isn't it consequently presumed that the amplifier will provide the current according to the impedance while driving the voltage regardless?

Yes and yes. But emphasis on **most** speakers; not all speakers expect the amp to be a voltage source. Examples: Coincident Technology, Lowther, Audiokinesis, Pure Audio Project, Spatial Audio, Classic Audio Loudspeakers, pretty much any speaker that is used with an SET (so most horns); high end audio is a diverse community.

Aren't the output stages of amplifiers typically followers? If so, that would make them voltage sources, wouldn't it? If you operated the output devices in a mode that provided both current and voltage gain than I suppose you could call the amp a power source, but that's rare and any reactance in the speaker will exacerbate nonlinearity, wouldn't it?

Not all amps have followers for output sections. Most transformer-coupled tube amps for example do not. Most OTLs however do. Yet both transformer coupled and OTL tube amps can behave as a voltage source if sufficient feedback is applied. IOW its all about the design.

@clio09
In one post Roger says it doesn't make much difference, OTOH in the very same post describes the power cords that he made up special and how anyone can do it. I have to assume that he felt it was worth doing.

JA measured a low 84dB sensitivity but with a higher/smoother impedance he deemed them "a very easy load for the partnering amplifier to drive."

@prof
Tube amplifier power has traditionally been expensive. Inefficient loudspeakers as mentioned above only became possible in the era of solid state since that power is so much cheaper. Unless you are in a very small room, I'd consider a different speaker as 140 watts isn't a lot of power when dealing with a sensitivity that low, regardless of the 'difficulty' of the speaker.

@clio09

this is the power cord he was referring to that does make a difference under certain circumstances:

http://tubeaudiostore.com/pocoyoucanac.html

Otherwise as Roger noted in his last post just get a cord with good connectors.

Apparently though Roger sees or saw fit to build his own power cord setup using some pretty heavy gauge wire (see page 12); this suggests (if actions speak louder) that he does think they make a difference, else why go through the bother?

(FWIW a power cord with light gauge but good connectors will not bring home the bacon, generally speaking. I've seen power cords, not just the connectors, heat up.)


I first heard how dramatic a power cord can be at the 1990 CES in George Cardas' room. The impact and nature of the bass was easily heard by all present to be quite different and with more impact when stock power cords were replaced by a pair that George offered. I found it disturbing; at the time I just didn't understand what was happening (and it clearly was not snake oil as it was repeatable) but I also knew enough to not buy the explanations that were offered (which clearly **were** snake oil).


But in time I found out that all you have to do is to simply measure the effects a power cord has on an amplifier. Generally speaking, the more power the amplifier draws, the more you will hear differences between power cables. 

As for the reference to light gauge cords not bringing home the bacon, what gauges are we talking about?

18ga for sure- not suitable for larger current draws. The wire gets warm along its entire length with a big enough amp- that suggests its limiting current, and certainly the voltage at the amp will not be that of the wall.

Connect two diodes back to back in parallel and insert in hot lead of power cord.

The diodes are bypassed by electrolytic capacitors, which conduct only when neither diode is conducting.

Ralph, what is OTOH mean? Im a newbee.The cords with diodes make a difference when needed. I think I have been clear and consistant here.

OTOH: On The Other Hand; other slang acronyms you may find handy:
IMO: In My Opinion
IME: In My Experience
IIRC: If I Recall Correctly
YMMV: Your Mileage May Vary
If there are only diodes in the cord, there will be no conduction near the AC crossover point. Electrolytics are thus used as I mentioned prior so that the AC waveform is undistorted. However, it appears you are conflating two different subjects in your comment above.

Your amp draws heaps of current when played hard. Something heavy ga might help a little at full power. At idle Im not so sure. I note the current draw of the M-60 goes up quite a lot at full power testing.

How long can I run an M-60 sinewave at 60 watts into 8 ohms before the tubes get unhappy?

Not very long, but long enough so that organ pedal tones and the like can be reproduced without damage. The tube is linear with grid current.

Is it pure Class D or something else?

It is a good example of class D.

No damage, and from what I can surmise, due to luck- one of your connections was apparently loose, otherwise neither speaker would have played. Essentially on the non-playing side, it sounds like the output of the amp was shorted. Fortunately you caught it and sorted it out, so no damage to the amp.

This is assuming that the biwire connection on the loudspeaker employs a common connection between the woofer and HF section.

I've found over the years that its not wise to make too many assumptions. If you have a Digital Voltmeter, check the continuity between the woofer and HF sections (with nothing connected to it) and see if they really are completely isolated.

I swap between amps with some that only have single-ended inputs and others with only balanced inputs and others with both.  My preamp has outputs for both.  I recently acquired a very nice 4m balanced IC.  For the cases when I want to use the single-ended input only amp, the CAT JL-3 monos, I will need to use an XLR to RCA adapter into the amp.  The adapter I bought has the neg line tied to ground.  But this results in the preamp's neg output line being shorted to ground which does not seem to be a good thing.

Its probably not- check with the manufacturer of the preamp. If the output is transformer coupled, then this is required. So it really depends on the preamp!

But if I were you, I would simply use the RCA output if using an amplifier with a single-ended input. There is no advantage of using the really nice balanced cable if you are going single-ended- the single-ended requirement actually results in the **entire** connection being single-ended so the cable is not operating as designed and intended.

Ralph, couldn't he use a Jensen transformer to make he XLR to RCA change? Would it still offer the proper grounding?

Yes.

Okay, does atmasphere believe that power cables can make huge differences in an amplifier’s sonic performance and why?

I don't use the word 'believe' because I made the measurements which showed that the power cord **without question** can affect the amplifier's performance. The measurement is easy- the voltage drop from one end of the power cord to the other. Its not rocket science and its not mumbo jumbo woo. Anyone with a 3 1/2 digit DVM can do it. After you know what the voltage drop across the cord is, then measure the resulting differences in the amplifier: output power, distortion and output impedance. You will find that they all change. So by this simple means you have a tool to correlate with what you hear.
So in a nutshell, I don't believe it, I **know** it.

Would a "normal" power cord supplied with a piece of gear have enough excess power passing capacity so that any drop in voltage due to the cord, is from one satisfactory voltage to a lesser, yet still satisfactory voltage to power the unit? Or are you saying that the power cords supplied with a piece of gear could be starving the unit for power and maximum performance?

Much is depends on the equipment used, so the answer as a generalization is 'both' although I'm uncomfortable with the phrase 'excess power passing capacity'. Power cords like anything else are subject to Ohm's Law.

Sorry, I meant amps, not volts.

@bwguy

 That amp can draw over 10 amps at full power. Probably not a good idea.
Does the amp hum if nothing is connected to it- just speakers?

my M-60s are supplied straight from the wall.  They are not happy running through the P-10!  From the point of view of distortion from the M-60's as a function of voltage, would you expect a measurable difference in distortion over the range of 115-125 Volts?

You would certainly be able to measure a difference, not just distortion but also power.

It seems that conditioners like the P10 are not happy with higher constant loads on them- they work fine with higher power solid state amps that have otherwise low quiescent current draw, but tube amps with larger draws (for filament circuits, perhaps also class A) seem to cause them trouble. For that reason we usually recommend running the amps straight out of the wall.
If you want a really nice conditioner, a company called Elgar used to make some that are quite impressive.

What Al said- The issue here is something called 'characteristic impedance'. The BNCs are designed for cables with a characteristic impedance of either 50 or 75 ohms. Without getting into the designer's head, it uncertain what CI the RCA connection is actually working with, since RCAs have no termination standard and are used with all sorts of impedances. Hence Al's comment about waveform quality; reflections of the signal (noise) may be the result. 

could you explain what shorting plugs are?

@bwguy
They are connectors that have all their connections shorted together. Installed at the input of an amp or preamp, in this way one can be sure that no noise is present at the input of that piece. 

Just because they make a difference it does not follow that a power cord does. In power cables there is not much to measure and a lot of other wire to consider.

I would not base an assessment of a power cord on the effects of interconnects or speaker cables. Ohm's Law works much better :)
You can measure the voltage drop of the power cord with a simple DVM.
As to its effects, its easier to see how it affects the equipment to which its connected. Some gear is more sensitive to AC input voltage than others, but in general, measure power output, output impedance and distortion. These will vary according to power cords, but not according to the cost of those cables  :)

I am sold on tubes for analogue audio but am confused by all of the information on power. I see from many posts that tube power need not be very high or as high as the speaker manufacturer claims as a requirement, i.e. a 200 WPC SS amp is needed to drive a speaker with 85db sensitivity (the manufacturer requires a minimum of 75 WPC, but likes at least 100 WPC), yet I have used a tube power amp with 40 WPC on the speakers and it sounds terrific. I have read that it is in the output transformers and SS amps are generally direct coupled.
Will you please explain this phenomenon?

@rollintubes
With all amplifiers its all about frequency response and distortion as to how they are going to sound. Solid state amps tend to have the flattest frequency response, but the difference between that and a good tube amp is slight and on many speaker loads may not even be measurable. Yet the solid state amp often sounds brighter and harsher, even though its got flat frequency response.

The reason is distortion. Our ears detect sound pressure by detecting the presence of higher ordered harmonics in any sound (likely because pure sine waves are extremely rare in nature). Solid state amps typically make more of these higher ordered harmonic distortions than tube amps do.

So when you are trying to make power, this really comes into play in a number of ways. As you push an amplifier towards full power, tubes and transistors behave differently- at clipping (anything over full power) transistor amps produce a large amount of higher ordered harmonics. Tubes don't, until they are over-driven really hard (and thus their distortion becomes audible as break-up).

So a smaller tube amp can act like its a lot bigger than it seems compared to a transistor amp. This is because music has lots of transients, and the distortion may only be showing up on the transients. The ear interacts with these distortions, telling you that the sound is louder. But the tube amp tends to do this in a way that is less noticeable, so essentially its giving you false loudness cues (this is particularly true of SET amplifiers, which is why they are often cited for being so 'dynamic' for how little power they have).


The other thing that comes into play is decibels. A 3db (decibel) increase is not very noticeable to the ear, but it takes twice as much power to do it. Your solid state example of 200 watts is just a little over 6 db more powerful than your 40 watt tube amp. That's not a huge increase volume-wise, so the additional distortion the tube amp makes at or near full power is able to fool your ear.


In a nutshell, tube amps have a more pleasant overload character so you can push them into overload and not even know it. That is why tube power **seems** more powerful than solid state power. Its not, but you might need a sound pressure level meter to see what is happening.

Why shouldn't decent tone controls, and maybe even loudness controls come back in fashion?

@4krow

A really nice tube preamp that had excellent tone controls was the Harmon Kardon Citation 1. If you can find one and have it properly refurbished, problem solved (so long as you don't need low output moving coil capability, although you could solve that with a stepup transformer). It employed switches to execute the tone controls, so when set to flat it really was flat. Its tricky to build a good tone control circuit (which requires an additional gain stage) that does not color the signal beyond the effect of the tone control itself, which is why tone controls went the way in the quest for greater transparency.

But other than possible tube noise floor raising...is the actual sound likely to alter in any way as tubes age?

Yes. If their transconductance falls below the minimum values, the performance will suffer and can affect things like bandwidth, distortion and dynamic qualities. In a nutshell, can sound 'sleepy'.

Much better than a 33 band EQ in my opinion, and in some cases even a parametric EQ.

@4krow
Its **very** tricky to build a 33 band EQ that doesn't mess up the sound! Even a 10-band is challenging. Parametrics have a lot in common with the filters used in analog synthesizers- they are pretty complex circuit-wise and so they are going to have a sonic footprint even when set to flat. 

BTW the RIAA curve we are familiar with that is 40 dB of EQ from top to bottom is not the RIAA curve. It is the RIAA plus MM/MC curve.

Try telling the manufacturers of cutter electronics that.

Does anyone find this interesting or alarming that we call the RIAA curve the big one and not make much mention that it is for velocity transducers only?

No. Its well understood that is the case. There has been some controversy over other types of transducers as you point out, won't be properly equalized. IMO/IME the producers of such cartridges should offer their products with an equalizer to set things right.