Pioneer M-22 Power Amp - Epic Japanese Gear From the Past

It's not overshoot (note how the first oscillation in the 20kHz plot is under the average, not over). It's also a really long ring (ie: barely damped at all). And note how the frequency of the ring is about 500kHz in both the 20kHz test and the 100kHz test. That seems odd, although maybe it's just because your square-wave generator has the same pulse slope for both?

I'm afraid you're over my pay-grade.

 

No the square generator is ok. What bother me is the difference between cold and hot. It is the same on both channels.

 

Perhaps repeat the experiment? I think JeffYoung is suggesting an external influence.

 

I did repeat it last night. I turned off the amp to let it cool. The results are the same as the first time. When cold, the square waves are perfect. When warm they're not. Anyway when the amp is warm, the square waves are really not bad even if they are not perfect. I am wondering what parts are affected by the heat. Ceramic caps as Skem mentioned could have a value change with heat. I'm afraid to replace them because it could perhaps also change the sound of this amp.

 

Heat will increase the gain in all your transistors (including the front-end and VAS), which will increase the feedback (and thus could result in more overshoot). But as stated earlier, I don't think this is overshoot.

Heat could cause the transformer to partly saturate, and if it's not snubbed maybe this is ripple from the bridge rectifiers leaking through? Doesn't seem likely, as you'd think there would be enough stray resistance floating around to dampen that much faster.

Bottom line: I agree with you that the shape wouldn't unduly concern me. But it's still a curious puzzle that I'd love to know the answer to....

 

I have a small hot-air gun for shrink tubing. It has a narrow nozzle, about 1cm. Seems such a tool could be used to go around probing by adding heat to one component at a time. Time consuming but potentially really educational.

I know the old electrolytic filter caps on mine had bad leakage current. I imagine that’s temperature sensitive. I also have had some distortion issues with the old transistors, but that was when cold and the problems went away as they got warmer. I wound up replacing the diff pair anyway and I liked the sound improvement it gave me.

 

They now sell clone boards on Ebay
https://www.ebay.ca/itm/30W-Class-A...345405?hash=item33d84e593d:g:P2QAAOSw9fRZ2x6P

 

Question for Jeff or others

What they are trying to do with all the RC in the negative feedback loop ? They seem to work in very high frequency.s Thanks

 

Not a @JeffYoung level guru, but since that reduces the resistance into the negative feedback at HF, could that be compensation to control oscillation in the VAS (or globally)?

 

Phase lead compensation. I have a feeling that they struggled without the benefit of simulation software as their compensation is pretty heavy-handed.

That eBay schematic is pretty close to a design I've been working on. It's based on a Hitachi application note, which may be the source of the Pioneer topology as well.

 

Pffft. Only in my dreams.

 

As I'm tempted to lower NFB by increasing R10 I was asking if I had to change the value of R12 and C21

 

A little NFB is usually the least optimal -- you either want none or as much as your topology can support.

If you need more gain then that's another story. I expect you could raise R10 to 22K or 27K without adjusting the phase lead comp. But any more than that and you'd want to run sims.

(For a discussion on small amounts of NFB see section 24.3 "Spectral Growth Distortion" here: http://milas.spb.ru/~kmg/files/literature/Designing Audio Power Amplifiers.pdf.)

 

More gain should be good but also less 'sparkle' in the treble should be great.

 

I don't know what's responsible for the sparkle. If it's just frequency response you could roll off the top-end a bit by increasing the capacitance of the input low-pass filter (C22 in the eBay schematic).

If it's a high-order harmonics at higher frequency issue, then lowering the feedback is going to be counter-productive.

If it's a sharpness of transients issue then you could round them off in several places: (a) the phase lead compensation (C21), (b) the Miller compensation (C3), or (c) the compensation in the feedback leg of the second differential (C2). I haven't a clue which would work better, but (a) is the most likely to have unintended consequences with stability, so I'd start with either (b) or (c).

 

I'll perhaps take a break with the M22. I just finished to overhaul a pair of Melos tube single ended triode with 4 X 6KG6 per amp which delivers 75W each. The sound is simply marvelous. They are monsters which heat the basement.

 

Given legarem's experience with restoring an M22, I'm curious if others have come across this in their experience. It seems that the Toshiba transistors contribute a fair amount to the sonic signature of the M22 (and that magic). I'll be re-capping (and cleaning) an M22 likely this weekend. So long as there's not other major surprises, I intend on starting with just the smaller caps and input transistors. I ordered replacements for the relays and the filter caps as well, and would likely do that shortly after. I don't intend on replacing the carbon resistors unless they're problematic (in which case I may have other issues lol).

I'm interested in seeing how this impacts the "magic" of this amp. Since my current setup is Heresy III's (with a 15" Rythmik Sub & and a 6992 tube pre-amp), I'm hoping I don't run into a scenario where I find the M22 too "dry" or "bright". Just curious what others have experienced after servicing their M22's.

 

This might be a stupid comparison, but has anyone heard the M-22 and the First Watt J2? They seem pretty similar on paper and in price on the used market.

 

The J2 is generally more expensive, even second hand they often sell for over $2,000.

The M-22 should be cheaper, but then you will have to deal with vintage equipment maintenance. I've heard some First Watt and Pass amps, and personally speaking, didn't think they were better than the M-22.

I think @sphinxvc owned an M-22 and a First Watt F3 at the same time, so may have some input?

 

Yeah, I'd generally agree with you but one just sold on eBay today for $1550. I watched it like a hawk, but last second of the auction decided that this is just not a time for me to be spending unnecessary money.

Then I realized that's similar in price to non-refurbished M-22s I've seen. Refurbed, the M-22s seem to be in the $2500+ range. Just from what I've seen personally.

Thanks much for the response, and it's great to know that you didn't find the First Watt/Pass stuff to be better than the M-22. I have an old Pass design -- Threshold Stasis s/150 ii -- that I use with the HE-6 but it's a bit too noisy with more delicate acoustic music. Hence why I'm looking to upgrade. I don't know that it's worth the money to try and upgrade the caps or whatever, but I suppose that could be the issue since it's also a vintage amp.