Adding a choke to Ampeg V-4 power supply?

[jamesmafyew]

I know this is ludicrous, but I can't find any info on it anywhere.

I have a guy who wants me to convert his V-4 to use EL34s. No big deal, I've done this before -- shorting pins 1/8 on the octal sockets, adding a bias pot, and changing screen resistors to 1K -- and it seems to work fine. But he also wants a choke added to the screen node, replacing the 470R/7W dropping resistor. I can't find ANYBODY who has done this. Seems like it would be mostly a waste, since the screen node is not in series with the preamp nodes of the supply, so the choke would have no benefit for the earlier stages of the amp. I can't think of a reason it would have any more positive effects than simply increasing the cap values on the screen node of the supply.

But if he's insistent, is there any reason I should avoid doing it, for reliability reasons?

[Enzo]

So stick in his choke and make him happy. It may do nothing at all, but that is the result of a non-technical person making technical decisions. If someone presents such an idea to me, I would point out why I think it unwise, but if he still wants it, it's his money.

And for that matter, what would larger screen node caps even do?

[jamesmafyew]

Well he wants it because he's interested in "more punch at really high volumes." This sounds to me like the Mercury ad copy for putting chokes in new Marshalls. If increased screen filtering does in fact bring the punch, then it would be cheaper to increase two cap values than putting in a choke. I'll give him his choke, but don't want to do it if it'll compromise reliability, because then the amp comes back to me.

[Steve Conner]

With EL34s, any mod that would increase the screen voltage under load is asking for trouble.

[Leo_Gnardo]

There's a lot to be said - good - about having a well-filtered screen supply. OTOH Steve Connor's warning is not to be taken lightly. V4 puts about 550V on the plates & with nearly the same on SG's, just asking for a meltdown. Maybe old Mullards & Amperex's could take it, but anything made in the modern era can't. Don't believe current manufacturers' specs and curves - these are reprints from old manuals & spec sheets from the 60's & seldom if ever applied to currently made tubes. The only thing they get right is the filament voltage & current.

You will also need to find a choke rated for operation beyond 500V. Most are rated 500V max. Some for as little as 400V. That's for breakdown voltage between the winding & core/frame.

FWIW I've had good luck with JJ's 7027.

I've tried 6550 & KT88 in V4 but there's no room between the glass & PT - just asking for broken tubes - so gave up on that.

[pdf64]

'he also wants a choke added to the screen node, replacing the 470R/7W dropping resistor'
I've done that to a friend's Traynor YB1A, and there was no measurable or perceivable improvement to the output / bass response.
There was less ripple at the screen node, which is nice, but it wasn't a problem before so no benefit gained.
Pete

[Leo_Gnardo]

'he also wants a choke added to the screen node, replacing the 470R/7W dropping resistor'
I've done that to a friend's Traynor YB1A, and there was no measurable or perceivable improvement to the output / bass response.
There was less ripple at the screen node, which is nice, but it wasn't a problem before so no benefit gained.
Pete

Just did the same here, YBA1 MkII & had to put 6K in series with that choke to get a SG node @ 450V that wouldn't threaten to wreck KT88's never mind EL34. BUT - it sure is well filtered! No "gargle" when loud notes are played. OTOH there's a fair amount of compression when played loudly, sort of resembles tube rectification. 67 watts to 8 or 16 ohms when measured steady-state but obviously reaches considerably larger clean peak power when played dynamically - can see this on the scope. Not a bad compromise. FWIW B+ is 540V, Ip 55 mA, and 1K SG resistors. Note that this is a series type power supply string, no "branch" for SG supply.

[g1]

FWIW I've had good luck with JJ's 7027.

Save your money and use the JJ 6L6, it's the same tube, they just add the extra pin connections in the 7027 (pin 1 & pin 6), but you do not need them in the Ampeg.

[Joe L]

I would think that increasing the capacitance of the first node would be the correct mod.

[Leo_Gnardo]

Save your money and use the JJ 6L6, it's the same tube, they just add the extra pin connections in the 7027 (pin 1 & pin 6), but you do not need them in the Ampeg.

OK g-one, thanks, I'll try that soon. A V-2 & V-4 are in the queue for repair in the upcoming week. Leave it to JJ....

And Joe L, in my case (Traynor YBA1 MkII) I have series 220 uF at first and 2nd node. I'm sure the Ampeg V4 & V2 would also benefit from improved filters.

FWIW the choke in the Traynor is a MM 10 Hy, bolted in - but not wired in (?!?) - by "Toneman" when the owner sent it to him in California from NYC. Came back looking, working & sounding like the dog's breakfast. As in jamesmafyew's amp, it's there at the owner's insistence. Bolting in MM iron always makes an improvement, right? (To MM's profits, I suppose yes.)

[bob p]

> Bolting in MM iron always makes an improvement, right?

I know you already know the answer to that question.

But he also wants a choke added to the screen node, replacing the 470R/7W dropping resistor. ... Seems like it would be mostly a waste, since the screen node is not in series with the preamp nodes of the supply, so the choke would have no benefit for the earlier stages of the amp.

Steve's point about screen dissipation is well taken. Be careful, because you're inviting disaster if you don't adequately limit the screen dissipation. That said, I think that a properly sized swinging choke feeding the screens is a beautiful thing, regardless of whether or not the preamp section would benefit. A properly sized swinging choke is a low cost, high bang for the buck way to regulate screen supply, and a constant voltage screen really improves an amp's performance. My personal preference would be to size a swinging choke for the screen supply AND use the proper value of screen resistors to keep the screen dissipation at a safe level.

[Leo_Gnardo]

I think that a properly sized swinging choke feeding the screens is a beautiful thing, regardless of whether or not the preamp section would benefit. A properly sized swinging choke is a low cost, high bang for the buck way to regulate screen supply, and a constant voltage screen really improves an amp's performance. My personal preference would be to size a swinging choke for the screen supply AND use the proper value of screen resistors to keep the screen dissipation at a safe level.

bob p, I like your answer, but I think we need a tutorial in the choice & use of swinging chokes, if you don't mind. I know I do anyway. And how to go about buying & spec'ing, because nothing I see in the common catalogs is pointed out as a swinging choke. All's I know is it changes its rating in Henrys depending on the current passing through. Maybe that's not even right. I'll ask you to take it from there. Or point out where to read all about it. BIG THANKS in advance!

[jamesmafyew]

Save your money and use the JJ 6L6, it's the same tube, they just add the extra pin connections in the 7027 (pin 1 & pin 6), but you do not need them in the Ampeg.

Correct. And interestingly enough, JJ's 6L6GC datasheet max for screen voltage is 450V, yet these seem to do fine in V4s at 525Vg2. They're what I normally use. They do quite a bit better in longevity than the Sovtek 6550s, which seem to be the popular choice for bass use -- though they do exactly the same power as 6L6s in the V4s that I've measured. I have actually had great success using the JJ E34L in V4s (which also list 450V as max Vg2). You even get 10 extra watts out of the deal, not that anybody either needs or notices that.

I doubt that putting this choke in will really change much since the choke in question has at least a 300-ohm DCR.

[bob p]

Leo, check out any power supply design textbook from the golden age. Here's one:

High Fidelity Circuit Design : Norman W. Crowhurst and George F. Cooper : Free Download & Streaming : Internet Archive

[trobbins]

What are we thinking is going to happen with the choke in circuit?

If the effective resistance doesn't change between supply and screen, then idle and low signal conditions are likely to be pretty similar.

If the tube's screen current starts to peak within a signal cycle, or square waves, then the impedance of the choke (compared to a resistor dropper) would sag the screen level more (depending on bypass capacitance on screen end of dropper and choke), and the sag would take longer to rebound.