Ampeg VT-40 overhaul questions

I'm working on rehabilitating an Ampeg VT-40 for a friend and I have a couple questions.  Any advice would be appreciated. Schematic is attached below.  Also keep in mind that I'm trying to stay away from doing any serious modifications to the amp, just looking to get it healthy again.

1. Replacing PSU filter caps - I'm having trouble deciding between using new multi-cap cans (to keep it as original as possible), or just throwing up a few turret strips and using single caps.  I realize either way would work, just wondering if there are other pros and cons I'm not thinking of.  Individual caps will probably be cheaper, but probably a bit more work to install, and they'll clutter up the inside of the chassis.  Cap cans are pricey and some people seem to think the quality is inferior to single caps.  I'd most likely use the 100uf+100uf multicaps that Weber Speakers sells.  Those CE cap cans are nice and shiny but I can't imagine they're worth $40 each.  Any opinions here?

2. Output tubes.  Schematic calls for 7027A, and it seems like the modern version of these are more or less the same as the 6L6GC. The B+ in this amp (almost +600V) is above what both of these tubes are rated for, but the general consensus in internet world is that they both work just fine, so I'm leaning towards getting some 7027As as called for and calling it good.  Other people seem to like 6550, but that seems to come with it's own set of concerns.  Any recommendations one way or the other on this?

3. I'm having trouble figuring out what type of diodes are used in the power supply section (D1-D6).  The schematic says F10 and F8, but internet searching hasn't given me much usable information about these.  Anyone know the scoop?  I have tons of 1n4007s sitting around so I'm hoping that's a reasonable substitute.

Thanks!

In general: find an Ampeg group. The VT-40 is a monster, and there's considerable opinion on every side of every detail.

1N4007 is not a good bet. The peak inverse voltage can be double the DC voltage.

1,500V 2A diodes are sold by the better guitar-amp shops.

Me I would NOT use anything less than good Russian 6550. The 7027 installed at the factory were, IMHO, "starter tubes". *When* they fail, they may fail dead-short and take out the PT also. But the kids who bought these amps would lose interest in a year, so the fact that 7027 don't hold-up at VT-40 abuse didn't matter. When you find a VT-40 still in use, it is high time to use 6550 in it.

> worth $40 each.  Any opinions here?

If the VT-40 were throw-away crap, I'd patch it with modern snap-caps.

Except in that chassis, this is ugly and difficult.

And the VT-40 has turned out to be one of the most notorious large amps over the last four decades. If a VT-40 came to me again, I'd be real inclined to fix it "right", and damn the few extra bucks.

PRR said:

Except in that chassis, this is ugly and difficult.

Click to expand...

The current caps in there are all rubber banded together and slathered in hot glue, stuck to the chassis.  yuck.

Mainly I can't tell whether it's necessary to do some minor circuit modifications to make this thing a reliable amp, or if just getting it squarely back to stock would be fine.

Stock, *with 6550*, it was a reliable amp.

The last one I worked on, the can-caps were still good. That's pushing it. But I think new can-caps are a justifiable investment in the amp's future.

Music Man amps run the 6550 tubes at 700 volts at high power, so they should take it,

going to unwind a few sec turns off this V4 to get the pwr trans down below 500,

will post results on the V4 thread,

Just an idea but the VT 40 is worth the can caps but another possible fix is keep the old can and take the insides out and slide in 2 individual caps so it looks original.  It might be to small to work with but people do this with the old paper dual electrolytic caps on the cathode of old fender amps so they look original inside. 

just ordered parts for this thing, I'll post some details as the work starts.  I couldn't justify buying the CE cap cans for $40 each, but I'm using some multi-caps that Weber sells (https://taweber.powweb.com/store/dcan.jpg), mounts the same way, should work well. 

the plan is:
replace the output tubes and a few of the preamp tubes (it came to me with several incorrect tube types installed)
new power tube sockets
replace all electrolytic capacitors (big PSU caps will be a little bigger than the originals but nothing too crazy)
replace all rectifier diodes (looks like some of these burned up at some point)
rebuild the whole bias section
replace a lot of the parts in the phase inverter/output section (possibly change screen resistors from 470r to 1k)
disable the polarity switch and remove the cap connected to it
give it a good cleaning (tube sockets, jacks, switches, the PCBs themselves, etc)

you can stick the polarity cap directly across the pri, it will still filter out some hash but not zap you when you touch the mic,

almost finished up with this, but I could use some advice.  Everything is working fine but I'm trying to eliminate the bit of LF hum that's coming through.  It's not overbearing but it seems like I should be able to get this thing quieter.  Looking for some advice on tracking down the source of the hum, as well as some insight into how much hum I should expect to just have to live with. Here's the detective work I've done:

I measure about 30mVAC of what looks like 60Hz hum on the output
Hum is there even when all volume pots are turned down, so it's after the first stages of both channels.
If I remove the PI tube and just listen to the output stage, the hum is gone. 
The hum balance pot helps a lot but does not completely eliminate it
The Bass EQ pot does not affect the hum
Moving leads around within the chassis doesn't have a noticeable effect

I'm suspicious of the 6K11 tube since I did not replace it, what's the likelihood that it's the cause of this?

I'm also concerned that the hum might be caused by the changes I made to how the power transformer primary is connected (see attached drawing).  I removed the polarity switch and associated capacitor, so the PTX primary is now completely isolated from the rest of the circuit.  There's also one point where the original schematic is not clear, just below the pilot lamp, where the blue lead may or may not connect to the black.  I'm assuming there should be a connection there, although either way does not affect hum.

probably 60 Hz, not pwr supply 120 Hz,

move the heater wires around and see if it changes the hum, if it does, then it is coming from the heater circuit,

there is a hum balance pot but this seems no good when trying to cancel 60 Hz heater noise,

you probably will not get rid of it completely, but you can get it down to a low level by possibly re-routing some wires,

Definitely try another 6k11.  Can easily be one tube. 

the saga continues, and now I have an output transformer question.  I measured the DC resistance of each side of the OT primary, one side is 146r, the other is 137r.  Not a huge difference, but is it enough to think that maybe there is a shorted winding in there or something? 

there are two symptoms that make me suspect this. one is the LF humming, which I know could come from several sources, but an unbalanced output section could be one of them, correct? (I tried moving heater wires and plugging in a different 6k11, neither had any effect on hum)  The other is that when I play at loud volumes for a while, ONE of the output tubes starts some slight red-plating.  The problem stays with that tube socket when I swap tube positions.  The side that red-plates is the one with the lower DC transformer resistance.  Idle current was also a couple mA higher on this side regardless of which tube was installed.  Otherwise the DC voltages I measured on the output tube sockets were identical between sides, and all very close to the reference voltages printed on the schematic.

am I off base here?

those DCR readings are probably normal, unless they really tried hard to balance the windings, which is rare in a guitar amp,

when tubes start to red plate, they conduct more current, if one side starts to red plate first, it will run away like a transistor that gets hot, so that tube will steal all the current from the other tube because it offers the path of least resistance,

what is your plate voltage? you might need to reduce it or get some tubes that will handle the voltage you have now, or install cathode resistors for self bias which will reduce your B+ by probably just enough to keep the tubes happy, this type of bias sounds very nice compared to fixed bias, you need to disconnect the neg bias and ground the grid resistors,

Thanks for the input

B+ is right around 590V right now. 

I bought a pair of 6550 tubes to try out in place of the 7027As, but if this red-plating issue is indicative of a problem in the surrounding circuitry, I don't wanna just use these heavier duty tubes as a band-aid to cover that problem up.  But the fact that the DC voltages are so spot on has me baffled.

Cathode bias is interesting but I'm hesitant to go down that road unless absolutely necessary, since it's not something I've designed and installed before.

tried putting the 6550s in today, seems to be working pretty well, but I'm struggling to figure out the best way to adjust the bias supply to suit the new tubes. 

it looks like R38 is the spot to adjust the negative bias level, but then I'm not sure what purpose R39 serves.  any tips here?

With the bias supply at stock values, the idle current was way too high.  As I have it right now, with R38 at 100k, the idle power dissipation is right around 27W (48mA through 570V), which seems to be right on the edge of being too high, so I'd like to bring it down just a bit to make sure I'm not over-stressing the tubes, just wanna make sure that adjusting R38 is the right way to go about this. 

> looks like R38 is the spot to adjust the negative bias level, but then I'm not sure what purpose R39 serves.  any tips here?

Don't think about it too hard.

This is a HIGHLY non-linear circuit and everything affects it.

Raise both R38 and R39 by about 20% (91K, 68K) and see where it gets you.

Maybe better: make R39 about 75K and then replace R38 with a 50K pot plus a 47K fixed resistor to ground. This will probably put you near -80V max trimmable to -40V.

There's a much bigger issue with this amp. None of the big tubes are happy with 590V on the Screens. So much so that a later "V4" used a separate winding to get around +350V for the screens. 500V-600V is not only over the ratings, electrons hammering the screens HARD, it is far more than these tubes "need" to make peak current.

In one such case I added a high-voltage transistor emitter-follower to drop from 470V down to 300V just for the screens. In that case the power waste was only 3 Watts so a small heatsink was ample. With this bigger amp and hungrier screens, you could be over 10 Watts, and of course past the 500V rating of the trasistors I used then (today we have 1000V MOSFETs).

As G1 bias is proportional to G2 supply, a lower screen means smaller G1 bias, so the stock supply may be ample.

PRR said:

Stock, *with 6550*, it was a reliable amp.

Click to expand...

So lemme get this straight ... I have a V-4, sitting dead for 20 years. I can replace the 7027As with 6550s without any other mods?

I'd like to get that big boy working again!

-a

6550 will  draw 1.6 A. heater current
7027 will only need 0.9,

so maybe an aux 6.3 xfmr will be required,

tube sockets are a bit different, so check that out,

you'll need to re-bias it, which is what I'm dealing with right now.  With the bias circuit at stock values, the 6550s were drawing about 3x more idle current than they need. 

So far the heater current hasn't been an issue.  I haven't had a chance to run the amp for very long to see if the PT gets too hot, but the heater voltage is still right on at 6.3vac despite all the extra current.

The other thing I noticed is that 6550s have a max grid 1 resistance of only 50k.  In the stock 7027 circuit, g1 resistance is 147k.  I'm not too familiar with the issues this can cause but I imagine it can destabilize the bias?

The V/VT-series amps' PTs were wound to support the 1.6A tubes, and sockets wired for same. 6550 IS the design tube, and a factory option, but buyers were cheap/poor so nearly all went to work with 7027 "starter tubes".

There may be a bias issue; *especially* with the variability in current-production bottles. Should not be large.

> 6550s have a max grid 1 resistance of only 50k.  In the stock 7027 circuit, g1 resistance is 147k.

7027 Max G1 (fix bias) is 100K.

EVERYBODY violated this spec. 99% of tubes did not leak so much that cheaters got burned. The 1% that did burn-up, well, obviously "bad", eh? And in particular, large guitar amps are often fix-biased well on the cool side. If grid-drift takes them to a less-cool point, it probably is not a problem. (One reason the 70% Pdiss rule works well: current can go up 40% and the tube doesn't die fast.)

4-jug amps have added factor of safety. While one tube may be leaky, odds are against both tubes on a side having max-spec leakage. (Unless you get a crate of junk tubes.)

Many of these cheating designs originated when RCA and Tung-Sol made GOOD tubes by the case-load, on machines largely funded by military work. We are right to re-think this now that tubes are made in faraway places under brutal cost pressure. However in many cases the driver lacks enough grunt to handle spec-legal grid resistors.