Bias modulated trem help

[jzucker]

I bought a Vibroclone amp off ebay. It's made using Doug Hoffman's AB763 board kit which is based off the Fender AB763 circuit but with the 6G16 brown face vibrato which uses a single 12AX7 tube and oscillator circuit to modulate the bias on the power tubes. I'm having a few problems with the amp and the trem seems to be acting a little funny so I thought I'd try soliciting some advise from some of the amp gurus here.

The amp was professionally done and the wiring is very neat and tidy using high quality sprague 715 or 716 caps, metal film resistors, yada-yada.

The problems I'm seeing are:

1) Vibrato is weak
2) Amp fades out occasionally

I've gone over the trem circuit and it's exactly like the 6G16 other than a 270k resistor in series with the two .1 caps. (This is listed as a performance mod on Doug's site in order to increase vibrato intensity). He's using the correct value pots including a 3M RA Taper speed control.

Using ohms law and measuring the DC mv across the 1 ohm cathode resistor with the tubes set to idle at 35ma, I'm seeing the voltage swing from about (approx) 31ma to 47ma with the trem on. I also see the odd behavior of more swing as the speed control is increased. This seems odd but it also seems like a fairly narrow swing for this type of circuit. However, since I've never owned a 6G16 amp, I'm not sure what swing to expect here...

Another interesting phenominon is that the guy who made it installed a switch which adds a 2.7k resistor in series with the stock 820ohm neg feedback resistor. When the 2.7k resistor is switched into the circuit, the audible vibrato is increased though when I measured the voltage swing, it's unchanged. Perhaps the neg feedback is "negating" some of the audible oscillation when using only the 820ohm resistor?

Regarding the amp fading, I've checked the bias supply which is giving me a very consistent -52v and I've replaced all the tubes, including swapping out the rectifier tube for a weber copper cap, all with the same behavior. I've looked for cold solder joints in the power and bias supply and could not find anything....

Just wondering what other suggestions you folks could give me to help debug the amp...

Doug does not offer a schematic for the amp but here's the layout diagram:

http://www.hoffmanamps.com/images/Ab763.gif

And here's the 6G16 diagram:
http://www1.korksoft.com/~schem/fenderamps/vibroverb_6g16_schem.pdf

[DreamTone7]

First thing to try is putting in a known good 12AX7 in the vibrato position. S/B the second one after the power tubes. If that doesn't work, grab your volt-ohm meter and take your readings for voltage and compare to the chart. They should be pretty close. If one is way off, post here what voltage and where. Poor sweep equates to lousy gain....could be the cathode bypass cap has gone south. (S/B a small electrolytic across the 4.7k resistor.)

[PaulC]

In order for the bias trems to do their thing you need a strong enough voltage swing to bias the tubes close to cut off. You need to get the tubes close to shutting off for a nice dip in volume. The problem is the same signal will bias the tubes hot when it swings the other way. If it's to hot you'll get some clipping/compression making it seem like it's not peaking as loud as it should. So for the strongest trem effect you should bias the tubes somewhat cold to begin with so you can get close to cut off at the low peak, and a good strong signal at the hi peak. This can make the sound weak when the trem is off, so then you need a bias offset to compensate for the two different modes of operation.

You're right about the feedback. It has a limiting effect on the output swing which can mask the trem effect.

As for the volume fade that's where I get the scope out. Is it a smooth fade, or a cut off type sound? Will it only do it when it wants to, or can you make it happen?

[jzucker]

Originally posted by PaulC
In order for the bias trems to do their thing you need a strong enough voltage swing to bias the tubes close to cut off. You need to get the tubes close to shutting off for a nice dip in volume. The problem is the same signal will bias the tubes hot when it swings the other way. If it's to hot you'll get some clipping/compression making it seem like it's not peaking as loud as it should. So for the strongest trem effect you should bias the tubes somewhat cold to begin with so you can get close to cut off at the low peak, and a good strong signal at the hi peak. This can make the sound weak when the trem is off, so then you need a bias offset to compensate for the two different modes of operation.

You're right about the feedback. It has a limiting effect on the output swing which can mask the trem effect.

As for the volume fade that's where I get the scope out. Is it a smooth fade, or a cut off type sound? Will it only do it when it wants to, or can you make it happen?

Hey Paul,

First of all your advice is greatly appreciated. Regarding the Bias swing, the cold idle point, etc...It all makes sense. That's probably why I've heard for years that the 6V6 amps sound better with bias modulated trem than the bigger tubes - Theory being that you're idling the 6V6 closer to 20ma which allows the swing to come closer to cutoff. Is the bias offset part of the trem circuit to begin with or are you speaking of something added to the circuit?

Regarding the fade, it's inconsistent first of all and so far when it's on the bench it hasn't happened. When it happens, it's quick but it doesn't cut all the way off. It fades to about half the normal volume and then it'll come back on. I almost wonder if there isn't a problem in the bias supply but I checked everything out with a meter and also for cold solder joints and I see nothing out of the ordinary. My guess is the original owner dumped it when he couldn't figure out these issues!

Jaz

[jzucker]

Originally posted by DreamTone7
First thing to try is putting in a known good 12AX7 in the vibrato position. S/B the second one after the power tubes. If that doesn't work, grab your volt-ohm meter and take your readings for voltage and compare to the chart. They should be pretty close. If one is way off, post here what voltage and where. Poor sweep equates to lousy gain....could be the cathode bypass cap has gone south. (S/B a small electrolytic across the 4.7k resistor.)

Hmmm - In the hoffman layout (from his site), there is no 25uf cap across the 4.7k resistor. I wonder if this is a bug. In other respects, his circuit is the same as the 6G16...

Interesting...

[jzucker]

I put a 25uf cap across the 4.7k cathode resistor and now the thing really "trems" out.

With the amp idling at 35ma, I'm seeing a swing of +- 12v.

It sounds really good now and it comes on in about 1.5 seconds and it works with the speed control all the way down too.

Thanks for your help, folks! :-)

Jaz