DIY Transistor Amp Using NVA Boards

For the three and more legged things
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#46 Re: DIY Transistor Amp Using NVA Boards

Post by Nick »

Les W at Avondale Audio specifies a 1N4004 and 220R in the DC power lines between output and input stages.
And I assume a cap to ground after the decoupling?
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
karatestu
User
Posts: 138
Joined: Fri Mar 03, 2017 1:34 pm
Location: Leeds/York

#47 Re: DIY Transistor Amp Using NVA Boards

Post by karatestu »

Nick wrote: Tue Apr 10, 2018 10:57 am
Les W at Avondale Audio specifies a 1N4004 and 220R in the DC power lines between output and input stages.
And I assume a cap to ground after the decoupling?
Oh yeah, forgot to mention that. 100uf electro and 0.1uf film :D
:bigsmurf:
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#48 Re: DIY Transistor Amp Using NVA Boards

Post by Nick »

karatestu wrote: Tue Apr 10, 2018 11:04 am
Nick wrote: Tue Apr 10, 2018 10:57 am
Les W at Avondale Audio specifies a 1N4004 and 220R in the DC power lines between output and input stages.
And I assume a cap to ground after the decoupling?
Oh yeah, forgot to mention that. 100uf electro and 0.1uf film :D
Yep, thats the important part, the diode works to isolate the driver stage from the voltage variations (drops) as the output stage dumps current into the load. Without the cap the voltage on the driver stage would move with the output stage rail.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
karatestu
User
Posts: 138
Joined: Fri Mar 03, 2017 1:34 pm
Location: Leeds/York

#49 Re: DIY Transistor Amp Using NVA Boards

Post by karatestu »

Nick wrote: Tue Apr 10, 2018 12:06 pm
karatestu wrote: Tue Apr 10, 2018 11:04 am
Nick wrote: Tue Apr 10, 2018 10:57 am

And I assume a cap to ground after the decoupling?
Oh yeah, forgot to mention that. 100uf electro and 0.1uf film :D
Yep, thats the important part, the diode works to isolate the driver stage from the voltage variations (drops) as the output stage dumps current into the load. Without the cap the voltage on the driver stage would move with the output stage rail.
Thanks for the explanation Nick. This is not my thread so excuse me Steve. When you say driver stage I assume you mean Long tailed pair and voltage amplification stages ? The driver transistors get their DC feed before the diode and resistor and so are surely still affected by the voltage drops caused by the output transistors. Sorry if I have this all wrong.
:bigsmurf:
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#50 Re: DIY Transistor Amp Using NVA Boards

Post by Nick »

Yes, bad terminology in my post, I meant the front end of the amp, as opposed to the bit that drives the speakers and possibly the bit that drives the bit that drives the speakers.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#51 Re: DIY Transistor Amp Using NVA Boards

Post by Cressy Snr »

This is an interesting question. In my KISS version the stages are talking to each other. I've never asked this question before as in my valve amps, short of using seperate power supplies, the stages have been decoupled as a matter of course. I've done this unquestioningly as that's how everyone does it.

So as someone who knows enough to build something (ye Gods I've built enough!) but is a bit thick when it comes to these sorts of details, what is the rationale behind preventing stages communicating and how does the music benefit?
Sgt. Baker started talkin’ with a Bullhorn in his hand.
karatestu
User
Posts: 138
Joined: Fri Mar 03, 2017 1:34 pm
Location: Leeds/York

#52 Re: DIY Transistor Amp Using NVA Boards

Post by karatestu »

Nick wrote: Tue Apr 10, 2018 12:42 pm Yes, bad terminology in my post, I meant the front end of the amp, as opposed to the bit that drives the speakers and possibly the bit that drives the bit that drives the speakers.
:lol: :lol: :lol:
Cressy Snr wrote: Tue Apr 10, 2018 1:05 pm This is an interesting question. In my KISS version the stages are talking to each other. I've never asked this question before as in my valve amps, short of using separate power supplies, the stages have been decoupled as a matter of course. I've done this unquestioningly as that's how everyone does it.

So as someone who knows enough to build something (ye Gods I've built enough!) but is a bit thick when it comes to these sorts of details, what is the rationale behind preventing stages communicating and how does the music benefit?
I am not sure how it affects the front end "musically". The output stages dumps a lot of current in to the speakers (has a large current draw) so makes the voltage rails sag for the front end. That will mean the front end clips sooner. Also because the output stage draws a lot of current this will produce ripple on the power supply which is passed to the input stage. I think the input stage is usually more sensitive to noise on the power rail as it is preamp like. I suppose it depends on the power supply rejection ratio of the design.

I will now wait for Nick to correct what I have written :D Also don't forget we are in NVA land here and nothing is quite as it seems :bigsmurf:
:bigsmurf:
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#53 Re: DIY Transistor Amp Using NVA Boards

Post by Cressy Snr »

Thanks Stu I get the bit about isolating one stage from the other. I've never understood PSRR.

I wonder if isolating the stages is like having members of the band in different studios on different coasts.
Sgt. Baker started talkin’ with a Bullhorn in his hand.
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#54 Re: DIY Transistor Amp Using NVA Boards

Post by Cressy Snr »

The reason I say the above is that with this particular piece of equipment, I've had more lump-in-throat and grin moments than with anything else, and the thing, in terms of the power supply, is so simple; one transformer one bridge package, two caps and two rails. It does me nicely how it is. Can't fault it.
Sgt. Baker started talkin’ with a Bullhorn in his hand.
karatestu
User
Posts: 138
Joined: Fri Mar 03, 2017 1:34 pm
Location: Leeds/York

#55 Re: DIY Transistor Amp Using NVA Boards

Post by karatestu »

Cressy Snr wrote: Tue Apr 10, 2018 2:00 pm The reason I say the above is that with this particular piece of equipment, I've had more lump-in-throat and grin moments than with anything else, and the thing, in terms of the power supply, is so simple; one transformer one bridge package, two caps and two rails. It does me nicely how it is. Can't fault it.
Then you have found the holy grail Steve. Be happy, play music and get yourself a nice shrubbery :lol:

The thing is once you start assessing changes / modifications in a musical sense rather than a hifi one then most (not all) past experiences are null and void. I am learning just like you. The NVA amp boards respond like no others I have encountered especially since I allowed myself to stop thinking in hifi terms. They respond to taking unnecessary components away rather than adding them. I suspect that if I did that to some Avondale amps they would not respond like the NVA at all.

Things have literally been turned on their head. In my system that has been virtually from start to finish. Started with speakers, cables and then the amps. Only the sources left to go :D

I for one relish the simple life and simple things. If my sound system (i am trying to stop using the term hifi) ends up that way too then i could not be happier.

Stu
:bigsmurf:
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#56 Re: DIY Transistor Amp Using NVA Boards

Post by Cressy Snr »

You're dead right Stu about 99% of past "hi-fi" experiences being null and void once you start assessing in terms of music. Once you start assessing like this, then you can't "re" hear your system in hi-fi terms; it's as if that way of listening has been wiped from memory, as if it it had never existed.......Strange phenomenon.

I've started a little experiment with my 2A3 amplifier to see what removing everything except what's absolutely necessary for it to work will do to its musical rendition. I'll put that in my 2A3 thread. It could be fascinating.
Sgt. Baker started talkin’ with a Bullhorn in his hand.
karatestu
User
Posts: 138
Joined: Fri Mar 03, 2017 1:34 pm
Location: Leeds/York

#57 Re: DIY Transistor Amp Using NVA Boards

Post by karatestu »

I'm not 100% with psrr either but it is basically the circuits level of immunity to power supply noise. Output stages tend to be less fussy about noise (ripple) on the psu. That is my understanding any way.

A thought that occurred to me was that when the output stages have a separate psu to the input stages then maybe the diode, resistor and cap to ground could be put in the dc line between the output transistor and the driver transistor. That is adding complication that may not be necessary however. RD did say not to change anything on the board. :D
:bigsmurf:
simon
Thermionic Monk Status
Posts: 5600
Joined: Thu May 24, 2007 11:22 am
Location: People's Republic of South Yorkshire

#58 Re: DIY Transistor Amp Using NVA Boards

Post by simon »

I get the concept of PSRR (I think) and, for instance, it's a major selling point of Broskie's Aikido. But whilst clever cicuit design is to be applauded I can't help thinking a good PS solves the problem anyway? I s'pose good PSRR helps the PS design but a choke or two should slove the problem?

Maybe someone who knows what they're doing can fill in lack of understanding?
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#59 Re: DIY Transistor Amp Using NVA Boards

Post by Nick »

simon wrote: Wed Apr 11, 2018 10:39 pm I get the concept of PSRR (I think) and, for instance, it's a major selling point of Broskie's Aikido. But whilst clever cicuit design is to be applauded I can't help thinking a good PS solves the problem anyway? I s'pose good PSRR helps the PS design but a choke or two should slove the problem?

Maybe someone who knows what they're doing can fill in lack of understanding?
IMHO, a good power supply never hurts. PSRR is not a simple subject, and its rarely constant with frequency. Allen Wright makes a very important point in his book, along the lines of "points in the circuit that should stay fixed, should never move, and points that should move, should have nothing stopping them moving."

If you let me know what you don’t know I will try and help :-)
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#60 Re: DIY Transistor Amp Using NVA Boards

Post by IslandPink »

simon wrote: Wed Apr 11, 2018 10:39 pm I s'pose good PSRR helps the PS design but a choke or two should slove the problem?
Well it seems to me from theory and practice that you need three main things from a power supply . I'm thinking most of a PS for an SE amp here, because that tends to be most demanding. -
1. Low hum & noise
2. Low output impedance down to as low a frequency as possible
3. Clean return path for the signal as the current is modulated
If you have a passive supply with caps & choke it should do No.1 if done with care, no.3 to some extent but you need very good caps like big Poly/oils or Black Gates , at least on the final cap. It won't do No.2 .

Output impedance is important , it basically means the supply is 'stiff' . Let's take a passive 300B supply at 420V . Due to the rising impedance of the caps at low frequencies, and the choke DCR's, it will not be as low as you want. If output impedance ( Z-out ) is low - say 5 ohms, then if you swing +/-50ma in the output valve, it'll give you a B+ of 420 +/- 0.25V, which is good. Any instability of the power supply voltage appears as an extra signal on the output valve and the driver valve. The problem is that in a passive P.S., Z-out is not constant - it'll be in the single ohms from the midrange and upwards, but it tends to rise quite a lot in the bass , particularly by 20Hz or so - might be 100 ohms down there. This means the 'power supply feeding back into the signal' effect is not just there but it's frequency-dependant - and the big problems in the bass go with a phase shift effect as well - all this amounting to poor timing, muddle and indistinct bass tone/texture.
Often the best options in this respect might be a very simple passive supply with a big low-DCR choke and a big cap at the end. I think Steve has gone this route but Dave Dove ( Elson ) also did this with a big 'swinging choke' and 470uF cap on his 300B amp. Aim No.1 might suffer a bit ..

If you want to go better , you can go series-reg - like the Verdier and Thorsten 'Legacy' amp circuits : this will more or less address No. 2 but add some problems re. No.3, because the upshot of controlling the voltage by the error amplifier there causes the voltage swings to be increased on the passive bit before the regulator.
The best way to deal with it altogether is a shunt-regulated supply like the T-Rex one designed by Bench/Broskie or the one Paul did years ago with VR tubes and a 572B (?) that was also a Steve Bench design. Here the output shunt valve goes up and down in current as the amp's output valve draw goes down/up in current. The output impedance can be single ohms down to 10Hz or lower. The passive bit of the supply before the shunt has hardly any demand variance - due to the shunt valve taking up small to moderate current swings - hence it sits there smoothing the ripple and keeping nice & quiet . The signal current return path is through the shunt valve - only the cathode by-pass cap here is modulated - and this can be a very high quality part.

Now this circuit doesn't look simple on paper - but what happens from the point of view of the signal circuit IS simpler - better controlled and predictable with less noise. Hence it is 'as simple as possible but not too simple' .

Please ask if you need clarification or more detail. Nick can answer your queries ! :D
"Once you find out ... the Circumstances ; then you can go out"
Post Reply