Maybe Leaving SET for Solid State because I want bass

@faustuss 

A lot to unpack here. I'll start with the SETs and 1% issue. Then cover the false statements that appear based either vindictively or just out of not really knowing what you were talking about. 

So cracking on:

The first thing is that with SETs most of the distortion is the 2nd harmonic which is mostly innocuous to the ear. I don't think there even is an SET that can do only 1% THD at full power, but even with much higher levels its pretty obvious people don't find that disturbing. 

So  the 1% thing really is obituary. Its a value chosen to reflect more modern amps. The Knight KA95 I had when I was a kid made about that much at full power. A good number of the TI class D chips in use today make about 10% at full power.

WRT your comments about the M-60 kit; it was an early one that he put together in about 1997. All of our Mk2 (which span 1996 to 2004) and Mk3 (2004 to present Mk3.3) power specs are rated at 120V not 117; that little difference in line Voltage has a large effect on a zero feedback tube amp, let alone one that has a large number of power tubes. 

As the AC line Voltage goes down, not only does the B+ go down but so does the filament Voltage, thus reducing the transconductance of the tubes. This happens across the entire circuit. The early Mk2s like the kit in question didn't have a very good CCS in the Voltage amplifier circuit so its performance changed with line Voltage in addition to the effects on the power tubes. When we introduced the cascode version of the CCS about 1998, the performance of the Voltage amplifier was maintained quite well over an AC input Voltage range from 107 to 130VAC. So I hope you are able to understand we have made improvements in the nearly 30 years since that kit was made! The Mk3.3 took that another step in that direct by going to a solid state CCS.

During testing, if one speaker terminal is held at ground, which can happen if you use a grounded signal generator and analyzer, the distortion goes up considerably as the drive to the power tubes is unbalanced. You have to make special arrangements when testing the amps on this account since the output terminals are not grounded. 

All the gain and most of the distortion in our OTLs occurs in the Voltage amplifier. It has to swing all the Voltage that is seen at the output, which is a pretty big ask. The driver and output tubes have no gain and as a result only contribute between 3 to 5% of the total distortion of the amplifier; IOW the Voltage amplifier is where roughly 95% of the distortion of the amp occurs, quite unlike most transformer coupled tube amps.

On this account if you want the distortion down the Voltage amplifier tubes must have matched sections. If that is done, and if the amp has 120V as its IEC connector, our current OTLs measure at about 0.5% THD at rated power, which is not the same as clipping power. The M-60 will clip at about 70 Watts rather than its rated 60 Watts into 8 Ohms. OTOH if no attention is paid to the input tubes the THD will likely be about 3% THD. Matching the power tubes seems to have little effect, unsurprising has they have no gain. It is important for them to be good though!

Your comments about damage to speakers, oscillation,  DC Offset and tube life are simply false. If they were true we'd not have been around for over 49 years 

Our OTLs are unconditionally stable in that the input or output signal condition or load cannot cause oscillation.

If a tube shorts, it blows the fuse by the power cord. This very simple mechanism plus making sure the amps have large enough resistors in the output section to survive tube failure is part of why the amps are so reliable. 

The DC offset generated by the amps if the amps are not set up right is quite low and any low power speaker can handle it. It is easily monitored by a VU meter on the front of the amp where the DC Offset control is as well; its easy to set. If you see it change on the VU meter, its an indication something is up with a tube since the settings tend to be very stable over months of use.

The power tubes often last well over 10,000 hours. 

 All this said, I happen to be of the opinion that tube amps are on borrowed time, which is different from saying I don't like them. But there are class D amps that are just as smooth and detailed in the mids and highs now so I wonder, if sound quality is the goal, why anyone would bother with tube amps unless they just like playing with them for fun (which is something I do).

Hegel. My H200 integrated is nice and warm. I wanted even more so I use a tube preamp and run through the Hegel's Home Theater bypass. Tube preamp SS amp. The best combo. 

@atmasphere 

Interesting your initial response is "Then cover the false statements that appear based either vindictively or just out of not really knowing what you were talking about." instead of "that a boy, nice job doing your homework". Then what is a consumer supposed to know and what do I have to be vindictive about by simply posing the question to someone in the industry who avails himself to these forums.?!

"The first thing is that with SETs most of the distortion is the 2nd harmonic which is mostly innocuous to the ear. I don't think there even is an SET that can do only 1% THD at full power, but even with much higher levels its pretty obvious people don't find that disturbing."

You find me a hobbyist who isn't familiar with this old diatribe which wasn't even the premise of my post but a much broader question regarding the validity of what the industry wants us to know. Sorry if I quoted a thirty something year old review in which your rationalization of why your product didn't meet spec but why wasn't it designed for the AC powerline voltage available at the time? Still, there is my statement - "Interesting though I wasn't able to uncover any objective test bench evaluations of your OTL amps published since 1998, Only the subjective ones from publications that don't routinely do them." when I'm sure the few remaining publications that provide test bench data are always requesting manufacturer's samples for review and I couldn't find any regarding subsequent products included in your catalog except as I mentioned the subjective reviews found in the rest of the popular print and online publications.

To further address some of the salient questions I had and your response to them, sag in AC mains voltage is a common occurrence when evaluating power amp power output measurements but seems to usually occur with high power, high current amplifiers when the tester is trying to measure output into very low impedances. Usually, the manufactures claims of full output into those loads can't be verified because of the limited amount of current available from the wall outlet (real world). Also, my concern about the DC offset in which you responded about a meter on the front panel and a control knob so it can be readily monitored and adjusted and the setting may not require tweaking for months. I don't have to adjust bias on my tube amps for years and if it's a set bias amp I don't have to think about it at all unless I notice a degradation in sound in which case I would simply replace the existing quad with a new matched one and of course I routinely have my equipment cleaned and evaluated by an experienced tech. Then there's the question of whether if loading an OTL's outputs into resistors has any proven sonic merit over having the outputs of a conventional tube amps into an output transformer which I believe is one of the aspects to the widespread attraction to the way tube amps sound. As for their days being numbered, they said that about vinyl in the 1980s.

Please don't be offended but I consider this valid exchange regarding a very pertinent issue.

Another vote for a couple of powered subs.

That's what subs are for.

That's what I added to my B&W 803 D3s.

Transformed my system.

Certainly more than any amp change.

Then what is a consumer supposed to know and what do I have to be vindictive about by simply posing the question to someone in the industry who avails himself to these forums.?!

@faustuss You could start by doing better research before making untrue statements. Just say’n. Here’s another exampe:

why wasn’t it designed for the AC powerline voltage available at the time

It was. The longer answer is when it was designed we often saw +120VAC in many parts of the country. 117V has been on the low side for a very long time.

Usually, the manufactures claims of full output into those loads can’t be verified because of the limited amount of current available from the wall outlet (real world).

IMO, anyone doing measurements that doesn’t have a way of dealing with that isn’t playing with a full deck. Get a variac for heaven’s sake! Then you can insure the line Voltage is correct by doing measurements. 

We ran into that problem with our MA-3 which makes 500 Watts. It has AC line correction to deal with brownouts and can make full power even if the line has dropped below 100VAC. 

I don’t have to adjust bias on my tube amps for years and if it’s a set bias amp I don’t have to think about it at all unless I notice a degradation in sound 

That’s usually the case with our stuff too. 

Then there’s the question of whether if loading an OTL’s outputs into resistors has any proven sonic merit over having the outputs of a conventional tube amps into an output transformer which I believe is one of the aspects to the widespread attraction to the way tube amps sound.

I hope you really didn’t mean this. Who loads OTL outputs into resistors?? Resistors don’t make sound and aren’t loudspeakers... so don’t expect to hear anything when you do that. If you meant something else, either your sentence is poorly constructed or you have some sort of misconception about how OTLs work. I can’t tell but you mentioned this twice...

WRT to output transformers you have a variety of challenges and they are not a source of widespread attraction. First is they have limited bandwidth- the higher the power the more limited they get. Somewhere in the 60 to 120Watt range is about the upper limit for what might be considered ’hifi’; IOW full power from 20Hz to 20KHz.

Its an excellent transformer that has bandwidth to 5Hz; most are specced to 10Hz. Distortion will rise in the amplifier as the frequency is decreased towards the cutoff. This is very measurable and interferes with bass impact, messes with midrange clarity as does the phase shift that accompanies. Of course if you are after high power, the ability to play bass properly may mean that the highs will be limited. 

The next problem is the sigmoid curve of hysteresis. It distorts the signal in its own right and also causes the load line of the power tubes to be somewhat elliptical. If you don’t know the significance of that, look it up! Its gets much worse for series feed SETs where the output transformer has a gapped core! If the load line is elliptical the power tubes will be making more distortion and all output transformers result in some elliptical behavior. 

If there is a mismatch in the power tubes, that will appear to the transformer as a DC component. This can cause the transformer to make greater distortion.

You can compensate for a lot of this with feedback of course. However, there is an upper limit where the poles and odd resonances of the output transformer limit how much feedback you can use before you have stability problems. To correct for some of that of course there is lead and lag compensation (the former is that little cap in parallel with the feedback resistor... the latter a network usually seen at the output of the Voltage amplifier). 

A very skilled designer can deal with that though but that will mean a very expensive carefully designed output transformer, both of which can be counted on one hand. The best example of that is likely the KrohnHite UF101 which employed up to 80dB of feedback which was quite a feat!

OTLs allow for much wider bandwidth with high power and since there are less poles and zeros (since there’s no output transformer), much higher feedback can be used if desired. 

OTLs do have their downsides, but you didn’t mention them other than the number of tubes used. 

FWIW, if you have no need to check the bias on your amps over years of use, either you really don’t use them all that much,  you are doing yourself a disservice or they are auto bias, which is something we’ve employed for about 21 years. We didn’t automate the DC Offset since it is trivial compared to the correct bias setting, and since we have a VU meter on the amps it can serve multiple duty as it allows you to see the relative output power of the amp, see if the tube condition has changed and can even be used to test the power tubes with the right procedure. 

The reason tubes are on borrowed time is several factors all happening at the same time.

The first is that class D is slowly invading the musical instrument amplifier market. We have several in our studio right now. Guitar amps are a much larger tube market than high end audio! About 80% of tube production goes to guitar amps. When a 100 Watt guitar amp weighs only 15 pounds as opposed to 85 pounds, people notice that when they are moving out of a venue at 3:AM. These days an awful lot of guitar players get their ’sound’ from effects pedals rather than the ’sound’ of the amp as was common in the 60s and 70s. 

So as that market fades it will be harder to maintain production at the same levels! The market is going to look a lot different in 10 years. 

The next problem is the effects of Covid and war in Ukraine. That has driven tube prices up all over  regardless of the country of origin. Russian tubes are only allowed into the US from Russia because Mike Matthews of New Sensor in NYC owns some tube factories in Russia. Otherwise its hard to get Russian stuff and they were one of the world’s larger tube sources. So when a tube factory in China burned down a few years ago, suddenly JJ was on the hook and could not keep up with sales. Prices skyrocketed. We’re paying 3x what we used to for output tubes. 6SN7s, 12AT7s, 12AU7s and so on are all a lot more expensive. 

Tariffs have not helped matters at all. They will not be the reason anyone starts making tubes in the US again. Forget about RCA or GE bringing tubes back  

Finally, class D amps have evolved to such a state that they now can be as smooth and revealing as the best tube amps, without the issues of heat, reliability and size. Why would you want to run tubes if you can get the same or better sound quality for less money and bother? The only real answers are not for sound quality; just for fun and nostalgia (which is perfectly legit), seeing how someone was able to do the art of tubes by making an amp look cool. If those things are not important but sound quality is, then there are class D amps that are a better deal and a whole lot easier to live with.

I should point out though that class D amps vary in sound quality more than tube amps do, so if you’ve not heard the right one or ones you may well be wondering what I’m talking about!

So yes, tubes are on the way out. I realize back in the 1960s they were declared ’obsolete’ and have now been around longer as such then when they were the only game in town. But technology does march along, so even though a declaration in the past was proven false does not mean that will always be so.