No Wayne, he's saying exactly the opposite, that 500 VA separately for each channel is a better solution. Less to much less crosstalk, and enough power reserve to work impulse power peaks even into 1 Ohm loads (short term only).
If it were just one big one for both channels, you's need their composite rating plus say 20% spare so one chanel doesn't drown the other. So, the single FUNCTIONAL equivalent to 2 x 500 VA would be say 1,200 VA single.
What Nige is advocating is a THIRD smaller toroid, to feed the low power stages and sort of "balance" the two big ones, if I have it right.
No doubt it is, no argument from me on that point. Which is why I went for a dual mono construction straight off. I believe in it.
100 VA is too small? Are you kidding? My entire input and VAS stages consume princely 34 mA of current, really wild. At +/- 63V regulated, that's all of (0.034 x 63) 2.14 VA. So, at maximum loading, that third transformer would be using less than 5% of its capacity.
Sorry, you lost me at any more than 100 VA capacity. I'm not in the business of making inaudible yet more inaudible, and at some price. This one's all yours.
Actually, the whole idea of a third toroid is, in my view, really meaningless overdoing it. I plan to use separate windings on the main transformers as way more than sufficient because:
1. My toroids are custom made, and have a low regulation of just 3%, while their quiescent current of around 100 mA testimony to their above average core and winding quality (low quiescent current is an indicator of quality, the less the better, opposite of power amp output stages - btw, typical UK toroids, of some fame, waste 130 mA for their 300 VA usnits);
2. They will be able to deliver 30-40% more power in short term peaks if required without going into bad saturation, so my reserves are high, and
3. If you look closely, you might notice the unusuall large spaces provided for the capacitors smoothing out the VAS voltage and acting as ebergy resercoirs, so, if the feeling takes you, install 4,700 uF times 6 for each board, to run your VAS stage at a higher capacitance level tnam most commercial integrated amps, all the while in dual mono.
Even that is really overdoing it, but hey, if it's a ball, make it grand masked ball.
Are you a fan of regulated IPS and VAS supplies? I have wondered, separate supplies, regulated or not, and if the same diligence in the CCS is desirable. There are those who suggest simple resistors measure worse but sound better. So, stability of the supply could have a big impact. Maybe.
On that, we agree 142.7%!
Haven't made a power amp without regulated power supplies since the mid 80ies. Although it's not like I made hundreds of them, mind you, less ten of them, two of which were glorious failures.
In my case, once again, it was that German made LAS amp which convinced me this was a good idea. That was my very own revelation. The case was definitely closed by once again German made Grundig V5000, in my view, easily the best integrated amp Grundig ever made. I never owned it, but a close friend did, and he had the schematics. So, I first made it as a clean copy, and it sounded very, very good. Then, I made another version, the only difference being that the IPS and VAS stages were fed from fully regulated power supplies - and the very, very good became truly excellent. And yes, I was stupid enough to sell it a few years later. I regret that to this day.
In the meanwhile, I have tried several regulators and learnt the hard way that simple somehow always comes out the best. I learnt that Nige's idea that simple transistors have the best noise spectrum to deal with. I discovered that for me, the idea of the "virtual battery" is in fact a great idea, so I settled on it. Simple, clean and works like a charm. Essentially, a medium power transistor, with zener diodes setting its base voltage value (I usually use two in series, as this allows me to get a lot of voltages just right), preceeded and followed by some serious capacitors, in my case preferably from Nichicon (subjective choice).
I'll jot a simple schematic later on today. Technics used that in their top of the line models, but also people like Goldmund, with their usual overkills.
Last edited:
Exactly and no extra money really . 500 VA , why not ? It will sound better , Naim Audio proved that with Super Cap .
This is becoming something of a witchunt. Smearing, bloating, ringing, etc can have at least several resons for each, and claiming it's the power transformer begs the question: why didn't you get a GOOD one in the first place?
The reason why I run only my predriver off regulated supplies is that in a true class AB, ONLY the predriver is running in true, pure class A all of the time, at least in the vast majority of amps. The driver is already running in class AB, albeit somewhat enriched. If one wants the drivers under full regulation, then one must very carefully consider its impulse current requirements, which will all on their own be larger than everything else preceeding them. I expect peaks od 0.5 A could well happen and if it's pumping out the drive for evil 2 Ohm loads, possibly even more.
This starts to make one think of both sheer capacity of the reulator, but also its speed, as its response should ideally be first thing after zero. It can be done, of course, but I think one would soon discover much to one's aggrivation, that the regulator starts to become complex and in fact starts to cost as much as the amp itself.
A quick inside-the-head count leads me to believe the regulator would require 10 extra transistors per side, of which two should be medium power drivers and another two full size power transistors. This also requires real estate property on the PCBs, but more importantly, very decent heat sink area as well.
So, gentlemen, thank you, but no thank you. All class A circuitry stays stabilized, all class AB circuitry makes its living off the electrolytics, as pretty fast ones can be obtained these days. A slow voltage regulator will kill the amp's dynamics just as fast as poor circuitry.
The reason why I run only my predriver off regulated supplies is that in a true class AB, ONLY the predriver is running in true, pure class A all of the time, at least in the vast majority of amps. The driver is already running in class AB, albeit somewhat enriched. If one wants the drivers under full regulation, then one must very carefully consider its impulse current requirements, which will all on their own be larger than everything else preceeding them. I expect peaks od 0.5 A could well happen and if it's pumping out the drive for evil 2 Ohm loads, possibly even more.
This starts to make one think of both sheer capacity of the reulator, but also its speed, as its response should ideally be first thing after zero. It can be done, of course, but I think one would soon discover much to one's aggrivation, that the regulator starts to become complex and in fact starts to cost as much as the amp itself.
A quick inside-the-head count leads me to believe the regulator would require 10 extra transistors per side, of which two should be medium power drivers and another two full size power transistors. This also requires real estate property on the PCBs, but more importantly, very decent heat sink area as well.
So, gentlemen, thank you, but no thank you. All class A circuitry stays stabilized, all class AB circuitry makes its living off the electrolytics, as pretty fast ones can be obtained these days. A slow voltage regulator will kill the amp's dynamics just as fast as poor circuitry.
Last edited:
Actually, I'm overbiasing it a little, not too much. Emmiter resistors are 0.33 Ohms and bias per trannie is 105 mA, so the output stage has a total bias of 315 mA.
Into 8 Ohms, this completely drowns the amp in class A at 1W (THD 0.002%), while into 4 Ohms, the distortion at 1 W has no mean comeback, as many older Japanese amps have, but it's not quite pure class A (THD 0.03%). No high order harmonics to speak of.
At rated power into 8/4 Ohms, THD at 50 kHz is less than 0.2/0.6%, again, with a nice decay rate, most probably the result of the fact that it's full power open loop hits 55 kHz at -3 dB, and into 4 Ohms.
The usual 20 Hz - 20 kHz THD and IM measurements are less than 0.03/0.07% THD into 8/4 Ohms THD, but IM stays down fairly low at 0.02%. At a nominal impedance of 3 Ohms, which is its limit for stady state op at full power, THD rises to 0.1%, and IM goes to 0.03%.
Global NFB is just under 25 dB. Full power bandwidth extends to just over 350 kHZ, but there is a low pass filter at the input limiting this to 200 kHz. The expected Slew Rate should be around 55 V/uS. The maths says around 80 V/uS, but experience has taught me that in real life this is almost never quite so, always a bit lower when stray capacitance comes into play, try as hard as I can to reduce it, that's one resillient monster.
So, it LOOKS promising in the simulations, but the proof of the pudding will be the live test sample, and hopefully, good repeatability. It usually turns out about the same, but there have been cases when "promising" turned out to be inadeqaute. On the other hand, I don't give up easily ...
Into 8 Ohms, this completely drowns the amp in class A at 1W (THD 0.002%), while into 4 Ohms, the distortion at 1 W has no mean comeback, as many older Japanese amps have, but it's not quite pure class A (THD 0.03%). No high order harmonics to speak of.
At rated power into 8/4 Ohms, THD at 50 kHz is less than 0.2/0.6%, again, with a nice decay rate, most probably the result of the fact that it's full power open loop hits 55 kHz at -3 dB, and into 4 Ohms.
The usual 20 Hz - 20 kHz THD and IM measurements are less than 0.03/0.07% THD into 8/4 Ohms THD, but IM stays down fairly low at 0.02%. At a nominal impedance of 3 Ohms, which is its limit for stady state op at full power, THD rises to 0.1%, and IM goes to 0.03%.
Global NFB is just under 25 dB. Full power bandwidth extends to just over 350 kHZ, but there is a low pass filter at the input limiting this to 200 kHz. The expected Slew Rate should be around 55 V/uS. The maths says around 80 V/uS, but experience has taught me that in real life this is almost never quite so, always a bit lower when stray capacitance comes into play, try as hard as I can to reduce it, that's one resillient monster.
So, it LOOKS promising in the simulations, but the proof of the pudding will be the live test sample, and hopefully, good repeatability. It usually turns out about the same, but there have been cases when "promising" turned out to be inadeqaute. On the other hand, I don't give up easily ...
Our heads give about 20 dB or less . Even so above 30 db helps keep the 20 db intact . Good mono has front to back separation as does real music . Stereo in real life is rare . Where it exists it is as a byproduct of hi fi where musicians do it in stereo , if so 100 % correct . This has been demonstrated . If a blindfolded person listens to an orchestra where players swap positions . They seem to hear the musicians where they expect them to be . Layers and time difference are not to be confused with stereo . Both mono and stereo should have this . As Gilbert Briggs said we have two ears so two speakers makes sense . Often when I play mono people can not believe it is . Sometimes because it has superior front to back separation , it is more stereo than most stereo . The one below has me guessing . Says mono on my copy and sounds just like this , is it in fact stereo ( 1957 , unlikely ) ? Piano right of centre with drum to the left ? Bad mono is like having a head cold . Plenty of bad mono was recorded . Like organic food the producer has to up their game . Thus often stereo was better . The ping pong stereo for all of it's faults often was well produced . Dark Side of the Moon is ping pong in places . Brilliant in my opinion because of it .
http://www.youtube.com/watch?v=eYSbUOoq4Vg
This will be about 20 dB . The producer is still giving a good mono balance although obviously ping pong . This is the actual record that coined the term ( LP version with Edmundo Ross and Firebird rehearsal , buy the light blue label as it is cheaper and better when using a transistor lathe ) .
http://www.youtube.com/watch?v=Em1Wj4Tvpc4
Last edited:
I mentioned Edmundo Ros . This is a Decca Phase 4 in ping pong .
Edmundo Ros - The Laughing Samba - YouTube
http://www.youtube.com/watch?v=OHtGbtwT39A
Edmundo Ros - The Laughing Samba - YouTube
http://www.youtube.com/watch?v=OHtGbtwT39A
Even with cumulative 20 dB separation this was the most stereo that sound ever was . I suspect it is more than separation . Phase shift , intensity . Decca , The Journey Into Stereophonic Sound is well worth owning . The rehearsal cut is wonderful and better than the finished recording . Edmundo Ros . I have a small collection ( 4 ? ) , they go very cheaply . Most are rubbish , some are wonderful . He was a James Bond character almost . Wish he had been in one of the films as he was a sharp dresser and poise . The awfulness of them was to please his customers . All was about dance and people who probably couldn't like myself . Occasionally he just made music ( as on Journey ) .
Mono cartridge would be a mono wired M44 7 for me ( close to 1 thou ) . Had a Lyra Helikon mono to play with . It was remarkable ( Tosca , Callas , Di Stefano) .
Mono cartridge would be a mono wired M44 7 for me ( close to 1 thou ) . Had a Lyra Helikon mono to play with . It was remarkable ( Tosca , Callas , Di Stefano) .
Decca Phase 4 Stereo is my yardstick by which I mease all others. So for, nobody managed to catch up with it, including some of Deutsche Gramofon's most revered recordings.
Also, the recordings I use to explain to those not in the know why they should own a stereo system.
And it sounds wonderful from the El Cheapo untis all the way up to the top. My Dual CS604 is hardly a star model, even if it was one of the best in its time and in its price class. The only upgrade I did to it was to exchange its Dual supplied DN cartridge with an Ortofon LM30.
Also, the recordings I use to explain to those not in the know why they should own a stereo system.
And it sounds wonderful from the El Cheapo untis all the way up to the top. My Dual CS604 is hardly a star model, even if it was one of the best in its time and in its price class. The only upgrade I did to it was to exchange its Dual supplied DN cartridge with an Ortofon LM30.
The Louis Armstrong of this era is remarkable from Audio-Fidelty .
Audio Fidelity Stereodisc - Stereo Demonstration Record (LP) - YouTube
Audio Fidelity Stereodisc - Stereo Demonstration Record (LP) - YouTube
- Status
- Not open for further replies.