Thinking about building a valve amp... / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Thinking about building a valve amp...

first previous 2 3 4 5 next

Move Thread

LAN_403

Proud Mary

Sun Sept 12 2010, 03:53AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

radiotech wrote ...

The whole point of a pentode is to eliminate the the miller effect of the amplification in a triode changing the plate -grid degeneration
due to capacitance. Thats what the screen grid does.

The screen grid substantially reduces but does not eliminate Ca-g1 in pentodes.

The Miller effect is given by Ca-g1(1 + A)

In the case of ECL83, Ca-g1 is given as <0.2pF, giving a Miller capacitance of 5.4pF when multiplied by (1 + A) whose derivation from the data sheet figures I showed above.

The Miller capacitance must be added to all the other shunt capacitances - V1k-a, V2g-k, V2g2, V2g3, plus typically 10pF as a nominal figure for strays - in order to estimate the total shunt capacitance, upon which the frequency response depends.

radiotech

Sun Sept 12 2010, 06:28AM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

I disagree with the terms of the Miller calculation in deriving the effective input shunt capacitance.
The only capacitive term to be multiplied by A +1 is the Cpg which for the ECF83 (pentode section is < 0.25 uufd on the Mullard sheet.

Agreed that for the input rolloff capacitance term Cgk and Cgsg
and Cstray all in shunt with Cpg which soley is multiplied by A+1 and then added to the others which aren't.
There are two reasons. Firstly the circuit hookup effectively bypasses the screen and the cathode to earth, thus removing the AC influence on their capacitance, had they been unbypassed. The second reason is the data which shows the Cgsg And Cgk at about 4 uufd each but the Cgp at < 0.1uufd due to the effective shielding.

With the pentode leak resistor, plate resistor and Rp of the
triode (75k) the rolloff is 230 kHz using ~~ 9 ufd.

. Discussions on other pentodes seem to indicate that changes in input capacitance are affected by density of the space charge as grid bias changes and are only a few pfds as contrasted
to Miller in triodes. The particular discussion was about AVC. In Class A the bias is fixed.

Proud Mary

Sun Sept 12 2010, 10:19AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

radiotech wrote ...

I disagree with the terms of the Miller calculation in deriving the effective input shunt capacitance.
The only capacitive term to be multiplied by A +1 is the Cpg which for the ECF83 (pentode section is < 0.25 uufd on the Mullard sheet.

Morgan Jones, in his influential Valve Amplifiers pp. 452-453, uses the same Miller capacitance calculation for pentodes as myself, and it would be easy to find other works supporting this view:

1284286423 543 FT95594 Valveamplifiersmorganjones

Exactly the same pentode Miller calculation is used in The Technique of Sound Reproduction, ed. Borwick, 1964, pp. 95-96, in which the operation of an ECL86 triode output pentode is discussed, and from which I cut and pasted the ECL86 amp circuit I posted above:

Clearly, you are entitled to your own view, but I would rather go with Morgan Jones and other mainstream authors on thermionic audio technology.

radiotech

Sun Sept 12 2010, 07:32PM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

many thanks for the Borwick book.

Borwick p.96 calculates pentode Miller exactly like my last post A times G1-p. which is 0.4 pfd times 53 = 21.2

The summation on p.95 sums g1-k, g2,g3 as 10 uufd.
This sum is not multiplied A

I believe we're very close on this, Only the position of the multiplier
for Miller-pentode is at issue.

The Miller effect in pentodes hinges on whether the screen an cathode are bypassed. An example from Lanford- Smith is shown for a 6V6 pentode with Cgp multiplied by A to to yield additional input capacitance.

Proud Mary

Sun Sept 12 2010, 08:13PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

radiotech wrote ...

I believe we're very close on this,

But didn't you begin by denying the existence of the Miller effect in pentodes altogether when you wrote....

Anyway, it would be uncivil to say more in Avalanche's thread.

Steve Conner

Mon Sept 13 2010, 11:00AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

radiotech wrote ...

Douglas Self went into this in the "master workshop on linear audio amplifier design" at the 124th AES and his talk snd slides can be watched and heard on their tutorial series on line.

Sounds very interesting (I'm a big fan of Douglas Self, personally) Do you have to be an AES member to access this online stuff? I was always looking for an excuse to join AES, and that might be it.

radiotech wrote ...

In disgustingly aged audio gear, if it works, the practice of re-capping it just to be sure, can alter the sound, because the capacitors were lousy when they were made, but that's what they used, particularly paper. Throwing in a slew of mylars with no leakage
could shift the operating point of valves.

This is very true

I would say that's the only substance to the capacitor sound myth. Here at 4hv I think we know more about capacitors than audiophiles: we push them to extreme voltages and RF currents where their limitations really do matter. We could probably have some great discussions on capacitor sound.

Finally: Avalanche, negative feedback is a complicated thing. But did you wind your own output transformers? I hate to sound bitchy, but if you did, that's probably the root of your bad sound problems. For best results, the NFB loop needs to include the OPT, and the worse the OPT is, the less NFB it can tolerate before it goes unstable. OPTs for single-ended tube amps are hard to design, and need good quality materials and lots of iron. And any hi-fi OPT needs lots of interleaving to get the leakage inductance down to a point where the high end will reach 20kHz with phase margin to spare for NFB.

One last thing I could suggest is: Try your power tubes connected as triodes. It'll hurt the power, but might help the tone.

Avalanche

Mon Sept 13 2010, 11:36AM

Registered Member #103 Joined: Thu Feb 09 2006, 08:16PM
Location: Derby, UK
Posts: 845

oops... my AF output transformers are actually standard mains transformers

so I might be struggling to get decent sound afterall. I haven't tried to gap them either.

I'll have a few more tries at the feedback, but I think a lot of my problem is probably also lack of headroom - I'm listening at almost full volume and sometimes it just has to clip, unless I turn it right down to almost nothing. I've actually learnt quite a bit from this amp so might have a stab at building something bigger.

I have Douglas Self's book, but I haven't read any of it yet

Steve Conner

Mon Sept 13 2010, 11:52AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Avalanche wrote ...

oops... my AF output transformers are actually standard mains transformers

Well, that fully explains the awful sound, no need to look any further!

Those are so bad for audio, you'll never get really good sound no matter how much you tweak the NFB. They're just about usable for guitar.

If you want to do this on a really tight budget, try looking for those old valve audio oscillators. They were used in school physics labs and the like. The ones I've taken apart all had a single EL84 output stage and a really big, high quality OPT with a 3 ohm secondary. Maybe you can find two the same and salvage the OPTs from them for your amp.

Or, if you can find one of the Advance Electronics ones with the big round dial, I have another OPT from one of those I can give you to make a pair.

radiotech

Mon Sept 13 2010, 05:03PM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

Do you have to be an AES member to access this online stuff? I was always looking for an excuse to join AES, and that might be it.

Some schools may have access.
I can recommend the AES since I've been a member since 1977.
Don Davis was active in boosting membership back then and he convinced a bunch of us to join at the 1977 Syn-Aud_Con in Seattle.

Even better if you can get your employer to pay! Mine did.
The chance to hear some of foundation people like
Theile in person if you are able to travel is worth it.

Avalanche

Tue Sept 14 2010, 06:13PM

Registered Member #103 Joined: Thu Feb 09 2006, 08:16PM
Location: Derby, UK
Posts: 845

I'll probably wind this project up soon, its been quite fun using valves and I'm already thinking of building something bigger

Having said that I have my eye on a nice little amp chassis on eBay at the moment so we'll see if that comes to anything.

I'll put some final pics up soon and maybe a video, I put way too much time into the case of this amp

first previous 2 3 4 5 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob