Ampeg VT-40 Redplating Solution?

[Chuck H]

You're probably not going to replace any resistors UNLESS any have drifted out of spec and are no longer doing what they should. Those resistor individual to the tubes aren't "bias resistors", they are grid loads that the bias voltage goes through. Their value won't change the bias voltage significantly so unless they are out of spec you'll be leaving those alone. Since the new cap fixed the problem it's very likely you'll check the bias and find that it's fine. But it should always be done anyway. If you find anything odd just post it here.

It might not hurt to use a touch of flux core solder to touch up the solder joints from the bias winding, through the circuit to the power tube grids in case your problem is a bad joint that might be temporarily working after the work you did. You should do this if your solder skills are decent and new solder/flux allows you to get nice, shiny joints that don't show surface tension at any of the metal contact surfaces.

[52 Bill]

I'm going to measure bias today and am planning to use the cathode resistor steps on this site: How to bias your tube guitar amplifier. I've only measured bias on two amps, and they both had adjustment pots. This Ampeg doesn't, and it uses individual bias resistors for each tube. Do the steps at the link above look right to you guys? I'll admit to being a little nervous about all the voltage in the very confined space around the power tubes! I also own a V4B but I've never had to work on it thank goodness.

So, to adjust the bias as-is I'd need to try different value resistors, yes? If I want to have adjustable bias, would I have to install an adjustment pot for each tube? Would .5w trimmer resistors be appropriate?

Don't confuse setting the bias with reading the bias voltage here.

The tubes redplate because they are dramatically under biased. What we all have been trying to get to is to first make sure that there is a stable negative voltage power supply (the bias supply), because until that is fixed the problem will continue to reoccur and you will continue to have problems.

So test the bias supply circuit parts and test the coupling caps for leakage, etc. Once you have a stable output circuit, then modify the amp for an adjustable bias control if you want to.

[neddyboy]

OK, I just finished the procedure outlined on the site linked above. Pin 5 = 64v; Pin 3 = 569v; voltage drop is 7.2mv. IF I'm doing the math correctly I get a static dissipation of about 4 watts. Do my measurements make sense? According to the site above, the voltage drop is far too low. Given these measurements, what should my next step be?

[neddyboy]

52 Bill, I'm confused about the coupling caps. C10 & C12 look like they're the coupling caps, but they're 0.33 mfd. I'm used to lower values, .042, .022, 0.1 tops.

[g1]

Verify your pin4 (screen) voltage is somewhere close to what the pin3 voltages are.
The .33's are the coupling caps, just a higher value than what you are used to.

[neddyboy]

Pin 4 = 558v. FYI it looks like one of the coupling caps has been replaced at some point if that makes a difference. I don't think I have the gear to test leakage, so I'll just replace both when I can go buy some.

EDIT: I'm an idiot. I measured voltage drop on the 47k resistor. The voltage drop on the correct resistor is 50mv on one tube and 59mv on the other.

[g1]

Which resistors are you referring to?
There are no resistors to measure idle current across unless you or someone else has installed them.
Can you tell us exactly what you are measuring to determine bias?

[neddyboy]

Hey all, I did the calculation again and got 33w. From what I've been able to find, the max for a KT88 is 42w. The amp sounds great to my ears already. Is being 9 watts under the max a good place to be? Is 33 watts with a 42 watt maximum a "goldilocks" setting?

[Chuck H]

If you installed the 1 ohm resistor correctly AND it actually measures 1 ohm, you're biased way too hot (kt88 absolute max 42W, design max 35W, you're biased near class A). It's not a bias supply I'm familiar with (anyone else?) but I'd venture increasing the value of the load resistor AFTER the reverse diode and electrically connected to the first bias cap. In your diagram it's 68k. But hold off a sec and lets see if anyone has a better idea.

[neddyboy]

Here's the outline from the steps I used:

A) Replace the ground wire on each power tube socket with a 1-ohm resistor.

B) Read the voltage drop across this resistor (in millivolts) with your DMM.

C) Read the plate voltage.

D) Use the above readings to calculate the static dissipation wattage.

I got these readings:

Pin 3: 569v
Pin 4: 558v
Pin 5: 64v
Voltage drop over 1 ohm resistors: 50mv/59mv

Tube 1: 569 x .059 = 33.57
Tube 2: 569 x .05 = 28.45

Did I get this wrong? If so, d'oh. Could the fact that each tube measured differently be due to the mismatched coupling caps? They're both .33mfd but one has been replaced at some point.

[Chuck H]

Does the hotter tube happen to be coupled to the old .33 cap? It may be leaking a little, skewing the bias. That's almost certainly why the other .33 was replaced. Normally I would replace both caps with the same type.

Your calculations are correct, except that you didn't fact check the value of the one ohm resistors. Set your meter to ohms and touch the probes together. This is your "null" figure. Now measure your one ohm resistors and subtract the null. You have to do this when measuring tiny resistances.

And you're biased too hot. You need to increase bias voltage a little.

[neddyboy]

Thanks Chuck. I had no idea about the null figure and measuring low resistance. I could only get 5% 1 ohm resistors locally, and the closest I could come was one @ 1 ohm and one @ 0.95 ohm. So, yes, what you're suggesting is likely. Next questions: will increasing the value of that 68k resistor R38 raise the bias voltage? What should I raise it to?

[Enzo]

One little nit pick when reporting voltages. I SURE hope those pin 5 bias voltages are negative.

So -64v rather than 64v.


5% is close enough.

[g1]

I'm surprised the power tubes are running that hot with -64V at the grids but I don't know what KT88's are like in there.
For sure the hot bias is why the B+ is low for this model. 586V shown on the schematic which would be even higher with modern line voltage. When you cool the bias the plate voltage will come up.
Not sure which resistor you mean by R38 as I don't think we have a correct schematic? Are you going by a label on the board that says R38? If that's the one on the anode end of the bias diode, I think you would want to change the one on the cathode end (to larger value).

[Chuck H]

Indeed! Let's identify "R38" because I don't know what that means. It's the resistor circled in red below. How high to raise it? I dunno.?. I'm not familiar with the circuit or the conditions in play. I expect it has something to do with the resistance of the winding and the impedance of the coupling cap. But it's not the typical voltage divider type circuit I'm familiar with. Maybe put a 25k or 50k pot wired as a variable resistor in series with that resistor and adjust it until the bias calculates to around 22 watts per tube (for KT88's, the diagram you posted shows 7027's). Our local guy, Merlin, has a site where he addresses the cap coupled bias supply and has a circuit illustrated to male it adjustable. The theory and math may go over some heads (he does that ) but the concept looks simple enough.