Relationship of amplifier sound to transformer quality?


Is this significant?
ptss
Two tube power amp designers/builders who say yes are Tim de Paravicini (EAR-Yoshino) and Roger Modjeski (Music Reference). Roger will personally hand wind you one for $1,000. Tim says he can tell you how an amp reproduces bass by looking at the transformer.
Thanks bdp24. Do you know which transformer manufacturers are most highly regarded?

Jap manufacturer Hashimito transformers compete just as great as Audio Note hand-wounded transformers.
One of the reasons Jadis electronics sound so good is their hand-wound transformers. 
If the transformer is not wrapped with low frequency high permeability mu metal sound quality will suffer. No two ways about it. The toxic effects of the induced magnetic field extends to everything within what, about a foot of the transformer? At least.

ptss---The DIY guys who build their own amps are the ones to ask about transformers. They all have their own personal favorite(s), but you see the name Magnequest a lot. rcprince is right, the Jadis transformers are really nice, but proprietary---you can't get one (without buying the whole amp!). Give Modjeski a call---he's pretty open about recommending stuff like that.
If the transformer is not wrapped with low frequency high permeability mu metal sound quality will suffer. No two ways about it. The toxic effects of the induced magnetic field extends to everything within what, about a foot of the transformer? At least.


Magnetic field outside of toroidal transformer is theoretically zero.  It is completely contained inside of the core.  Toroidal transformer has no gaps making it more sensitive to presence of DC voltage.
EveAnna Manley is so convinced of this that they make their own trannies in house.
"Magnetic field outside of toroidal transformer is theoretically zero. It is completely contained inside of the core. Toroidal transformer has no gaps making it more sensitive to presence of DC voltage."

Theory doesn’t seem to hold for a lot of things in audio. I wrapped a toroidal transformer, the aftermarket transformer in my uber modded Oppo 103, and heard good results. Maybe it doesn’t take much leakage to be audible, who knows?
Does anyone know who supplies Levinson,Spectral or Boulder?
The heavier the better!
True for bass Don, but that compromises the highs. As in most things Hi-Fi, it's a balancing act. That's the art of transformer design.
if the transformer is hand wounded using high quality silver wires with iron-nickel alloy in Double-C configuration, will this give deep bass without compromising the highs and still sound quick & agile? 
"if the transformer is hand wounded using high quality silver wires with iron-nickel alloy in Double-C configuration, will this give deep bass without compromising the highs and still sound quick & agile?"

I imagine that combination of metals would work rather well since, you know, iron-nickel alloy is mu metal. I suspect it would sound like a million dollars.

I had several encounters with the late great James Bongiorno (Hadley, S. A. E., Great American Sound, SUMO and Spectra AmpZilla fame) who would agree that transformer design and quality has a significant effect on the sound of an amp.  His very own amp designs testify to that. Look at the old Ampzilla, Son of and Grandson amps, all having huge high quality transformers. 

He told me that the transformers were one reason that old Sansui receivers and amps sounded so good. At the time I didn't agree with him as I wasn't a huge Sansui, GAS, or SAE fan. My how things change.

I came to realize and appreciate what genius he was, and that more often then not everything he said was true.  All of his designs are musical and engaging as was the man.





Two tube power amp designers/builders who say yes are Tim de Paravicini (EAR-Yoshino) and Roger Modjeski (Music Reference). Roger will personally hand wind you one for $1,000. Tim says he can tell you how an amp reproduces bass by looking at the transformer.
Another couple designers (one already mentioned) who think alike are Kondo Audio Note Japan & Convergent Audio (CAT). 
For tube amps the bigger the xformer the better for its bass. In the CAT JL2 the output xformers are 55lbs each (which contributes to the total 180 lbs) & CAT also winds his own xformers.
In s.s. amps the quality & size of the power xformer is important. Shielding the power xformer to attenuate the noise has a big effect on the amp's micro-details. Plitron makes some very good low-noise toroidal xformers. The size of the power xformer has much to do with the max output current of the amp's output stage which in turn has an effect on how difficult a load (speaker) that amp can drive.  
Monolith Magnetics makes some very nice (and expensive) transformers. The OPTs I use in my 833C SE amps weigh 62lbs each, with amorphous double-C cores and virgin teflon interwinding insulation. Here's a picture of one during its construction.

http://i1243.photobucket.com/albums/gg559/Quagmire22/833OPTcoilandcore_zps9fb41a95.jpg

They're not cheap but high quality very rarely is.
And yes, the transformers have a huge effect on the sound of tube-based amps, preamps, DACs, whatever.
It is a direct correlation.

...quality transformer is about 80% of final sound...it is essential...
Another factor in transformer selection....
  • The transformer acts like a reservoir for electrical energy.
  • When loud, dynamic music (very fast, like drums) is encountered, the energy required to realistically reproduce that instrument is provided initially from the transformer
  • A small trasformer is unable to provide all the energy and as a result there is a tiny drop in the circuits internal voltage
  • This is music, the the internal voltage is always changing - not good for an amp
  • The voltage fluctuations degrade the amps output signal
  • The Bigger the transformer, the better the amp is able to handle the demand for electrical energy, fluctuations are minimal and the result is better quality music
Naim is one company that has always followed this approach. They even build seperate power supplies for most of their gear and it pays dividends. Many other high end amps adopt the same philosophy.

I’ve made seperate power supplies with large transformers for a few components and each time the improvements were significant - well worth the investment. And that was simply increasing the size of the transformer.

Factor in the other advantages listed above and you can see why good designers consider transformers crucial to good design.

Regards
williewonka
1,019 posts
08-04-2016 9:45am
Another factor in transformer selection....
The transformer acts like a reservoir for electrical energy.
When loud, dynamic music (very fast, like drums) is encountered, the energy required to realistically reproduce that instrument is provided initially from the transformer
A small trasformer is unable to provide all the energy and as a result there is a tiny drop in the circuits internal voltage
This is music, the the internal voltage is always changing - not good for an amp
The voltage fluctuations degrade the amps output signal
The Bigger the transformer, the better the amp is able to handle the demand for electrical energy, fluctuations are minimal and the result is better quality music
williewonka, what you have written is simply bad information i.e. not correct. Please don't write incorrect material & misguide the public. it does nobody any good. Your conclusion to use a large xformer is correct  tho'.

The transformer does *not* act as a reservoir of energy. That is the job of the power supply capacitors. Capacitors store charge & if there is a transient in the music, it is these power supply caps that provide the burst of current needed by the amp. The transformer is a conduit for providing current to re-charge these power supply caps. The larger the power xformer, the more current it can handle & the faster the power supply caps re-charge ready very quickly for that next transient in the music.

A small trasformer is unable to provide all the energy and as a result there is a tiny drop in the circuits internal voltage
not correct - a very small transformer cannot handle large currents hence cannot re-charge the power supply cap quick enough. So, the power supply caps suffer a droop in voltage for much longer a time than if a bigger, higher current xformer would have been used.

The voltage fluctuations degrade the amps output signal
correct. it is called amplitude modulation (AM) of the music signal which is a form of distortion due to a badly filtered power supply. 

The Bigger the transformer, the better the amp is able to handle the demand for electrical energy, fluctuations are minimal and the result is better quality music
the conclusion is correct but the bigger xformer is *not* why the fluctuations are minimal. The fluctuations are minimal due to the power supply caps filtering/suppressing power supply ripple. IOW, you could use a very large power xformer & have no power supply caps & you would have huge power supply ripple which would destroy the quality of the music.
Post removed 
Post removed 
Bombaywalla - my apologies - the post was poorly worded.

I was attempting to use metaphors that people without detailed electrical backgrounds could relate too - I obviously failed.

My intention of using the "reservoir" metaphor was to communicate the fact that every transformer has a finite power delivery capability and not that it actually "stores" electrical energy like a battery or capacitor

But, just like a battery - If you exceed the delivery capability of a transformer, then things will happen at the output terminals of the transformer that then leads to degraded performance of the rest of the power supply and ultimately the entire component. 

Designers of quality components understand the transient nature of audio  and design sufficient "capability" into their transformers to deal with them, taking into account...
  • winding materials
  • insulations materials
  • core materials
  • Laminated vs. toroid 
  • etc..

Components built to a "price point" often have power supplies that use a transformer conforming to a more "standard design" which are often less able to match the transient demand that the rest of the circuitry is actually capable of achieving and therefore results in the component operating at a reduced level of performance.

E.G. - I 've had a couple of components that used an AC Wal-Wart "power supply" and simply replacing the Wal-Wart supply with a "more capable" transformer elevated the performance of both components significantly.

For me - Good component design starts with the transformer.

I just hope I've redeemed myself - a Little at least :-)


williewonka,
thanks much for your clarifying post. Yes i believe that you have done well to redeem yourself. My apologies too for coming down on you so hard. I could have phrased that better myself.
It's hard to know how technical the audience is when trying to explain some this material. I've found that watering it down to suit a non-technical audience can add more ambiguity & confusion rather than stating it in a technical way & later on, based on the questions from the audience, explaning certain sections using more layman analogies. More often than not the audience here at Audiogon is pretty technical & they can stomach quite a bit before barfing but one never really knows...
Your last post shows you clearly understand the virtues of a good transformer in the design of a component....thanks.
Good info guys. There are many parts in audio designs that can impact the sound. Ps caps trannies resistors etc. polite on makes good tram
nsformers as I repair counterpoint products. Significant change in sound not always but yes.  If the power supply is poor then I would not waste my time. The dad I build has a better powers supply that most amps I see.  Happy listening.
First of all, hope you are all well. Old thread but a question came up I know you guys can answer. Jim Stuart of Meridian made a comment years ago when asked about the 'lean' nature of the old Meridian 105 monos. He said what they need is a bigger transformer. The amp and psu are in separate chassis. I just happen to have an extra pair of psu. Could I connect them in parallel to double their rating? It would be a simple matter of splicing the umbilicals of the psu's and then into the amp as usual. The caps are a pair of 6800uf in each supply so I would up one set to 10,000uf and delete the other? TIA
Transformers  are the heart of any amp should be the best.
Jim Stuart of Meridian made a comment years ago when asked about the 'lean' nature of the old Meridian 105 monos. He said what they need is a bigger transformer. The amp and psu are in separate chassis. I just happen to have an extra pair of psu. Could I connect them in parallel to double their rating?
Yes, it's a good idea!
It would be a simple matter of splicing the umbilicals of the psu's and then into the amp as usual.
If you can find the plug and socket, my suggestion is make a "Y" connector so no modification to the original power supplies.
The caps are a pair of 6800uf in each supply so I would up one set to 10,000uf and delete the other?
My suggestion is add a power diode in series to each + and - outputs. Again, no need to modify the original power supplies.

https://www.kepcopower.com/support/diodes.htm
Do you mean at the outputs of the amp/speaker connections? What value would be suitable?
I should point out I'm not interested in redundant operation. I want increased power/current/va rating
Do you mean at the outputs of the amp/speaker connections? What value would be suitable?
No, the diode is in series to each +55V and -55V of the power supplies outputs to Amplifier. The reason of adding those diode in series is to isolate both power supplies and let each power supply only charging its own capacitor.
120V/30A diode is adequate.
I should point out I'm not interested in redundant operation. I want increased power/current/va rating
Since one power supply can provide enough power, parallel two power supply will definitely increase output current capacity and also can serve the function of redundant operation.
a] a quality output transformer is the heart of any good tube amplifier.b] do not try to combine the outputs of two separate transformers (AC power nor audio output). they will both be unhappy to say the least.c] do not try to combine the outputs of two separate AC or DC supplies. They will be more than unhappy.
@speedskatere,this is for ss psu.

@imhififan...So in this configuration, would it be beneficial to increase C to 20,000uf total for each psu, total 40,000uf per pair of psu? Or keep each psu at 13,600? TIA
would it be beneficial to increase C to 20,000uf total for each psu, total 40,000uf per pair of psu?
Yes, check dimensions if the capacitors can fit in before ordering. 
One function of the capacitor is to smooth AC ripple voltage on the supply rails, higher capacitance is better, provided its not over-stressing the transformer and rectifier, 10,000uF should be fine but do not go any higher.
I will probably need to install a soft start?
Maybe, but I don’t think a soft start on 150VA transformer with 2x10,000uf cap is necessary, IMO.

https://neurochrome.com/pages/the-ultimate-guide-to-soft-start-design
csontos-
Could I connect them in parallel to double their rating?
No. And anyone who understands what a transformer is and how it works should be able to see at an instant why this will not work. Since no one does then I guess everyone can benefit from this.

Any electrical current in any wire produces an electromagnetic field around that wire. Also any electromagnetic field that crosses any wire induces a current in that wire. This is basic electronics everyone needs to understand. This is fundamentally the reason why RFI and EMI are problems, why designers space power supplies away from the signal, etc etc.

Motion is critical to this effect. The motion can be either physical motion, as in an electric generator like a MC cartridge, or it can be field motion as with an alternating current.

This is why you do not coil excess wire. The fields around one wire induce currents in adjacent wires and this messes with the sound.

Okay so that is the principle. Now what is a transformer? What is all this hand winding BS? It is BS, right??

Not even. A transformer winding is a thin wire wrapped round and round real tight. This results in a powerful field being generated. This mess of windings where the power or signal (whatever you want to call it, it is all the same) comes in is called the primary.

Next to this is another mess of windings called the secondary. What happens, the alternating fields of the primary cross the coils of the secondary inducing a current in the secondary.

Got it? There’s lots more to it. The ratio of windings P:S determines whether it is a step up or step down transformer. But this is enough to understand why two transformers will not work and why you need a bigger transformer. The current the transformer can handle or transform depends almost entirely on the physical size of the transformer. Because of two things- more current calls for both larger gauge wire and more windings. Both of which increase size.

So now everyone should understand it will do no good to run them one after the other. One will not "hurt" the other. Transformers do not have feelings so there will be no problem with one being "unhappy" or "not getting along" or anything like that. You can connect one to another to another all day long. Won’t give you any more capacity, which is what you asked about. For that you need a bigger transformer.

Got it?
Right, but now they will be doubled up on each channel. Or that will not change inrush?
but now they will be doubled up on each channel. Or that will not change inrush?
As I mentioned before, those diode in series is to isolate both power supplies and let each power supply only charging its own capacitor.
Thanks so much for your help. Much appreciated.
If this is not just transformers, if it is two complete power supplies you are talking about combining together in parallel, then what you are really talking about is doubling the capacitance. With twice the power supply storage capacity you should notice a profound, or at least decent, improvement in bass slam, with more authority well up into the midrange. The top end may even improve. Hard to say how much. The existing power supplies being mono and separate units are already probably very overbuilt. But I have yet to see the power supply that wasn't improved by bigger/better/more caps. It is the kind of extravagant expense no sane manufacturer would ever do. But as they say in racing, run what you brung!
The transformers are not running daisy-chained. They are in parallel which means combined together working simultaneously. If I were to hook them up in series, it would double voltage instead. So yes, It will double the va rating or current capability with a slight loss due to leakage and bring C from 13,600uf per channel to 40,000uf per channel since I will put in bigger caps than oem. An interesting experiment for this kit. This is similar to paralleling an amp for increased current capability vs bridging for increased voltage or output as in watts.
I meant to say similar to paralleling two channels of a stereo amp or in this case a pair of mono amps. However just the PSUs paralled accomplishes the same thing.
They are in parallel which means combined together working simultaneously.
Actually, both PSU will not output exactly same voltages (e.g. one PSU output 55.1V and the other PSU output 54.9V), the one which output higher voltage will do all the work until the current demand from the load is high enough to make the voltage sag to the same voltage as the lower output voltage PSU, then the lower output voltage PSU will start supplying the current to meet the requirements.
I suppose voltage differences must be very slight. Will the power diodes prevent circulating current in the event the voltages don't match well? I suppose I will check to see how close they are before I proceed.
Voltage differences is highly expected, that is the reason insert power diode in series to isolate both PSU.
It's perfectly normal one PSU supply current to the load all the time and the other PSU kick-in when higher current is required. That is how this configuration works.