5f1 slight audible oscillation

[Chuck H]

One OT primary lead goes to the HV rail. The other OT primary lead goes to the power tube. Swap these two connections. If your oscillation is caused by phase in the power amp this may fix it.

To disable the NFB locate the 22k resistor that is connected to the output jack (but may not be right on the jack). Disconnect either end of this resistor to disable the feedback circuit. I'm not sure what this is supposed to reveal.

[loudthud]

Disabling the feedback should stop the feedback if it is a phase problem with the output stage. This is an easier test than switching the output transformer wires. One of the wires may be cut too short and you would have to splice a longer wire to the transformer. Just unsolder one side of the 22K resistor and bend it up in the air to see if the oscillation stops. The gain of the amp should increase, you won't have to turn the Volume control up as far for a given volume.

[martinman]

Try something around 650 ohm..Your plate (anode) voltage will change so you will have to remeasure it. (also your cathode voltage). Something else to think about(not the existing problem) is that your rectified voltage (voltage at the first filter cap) is a lot higher than what a tweed likes to run at..That voltage should be between about 340 vdc and 380 vdc..This is the fun part of building amps..as Chuck suggests, swap the output transformer primary leads, get your bias in the neighborhood and see where you are..

Ok, mac - I just swapped the 1.5k cathode resistor for an 820 Ohm that I had around (closest to 650 that I had). It didn't do much to the measurements compared to the ones I posted before, but I do notice that the amp *sounds* different. Heh, I'm sure your response is "of course it does dummy" . It went from breakup at around 4-5 (1.5K) to clean almost all the way up to 10. With the voltages being about the same, what does the drop in resistance accomplish?

V1
Pin1 163
Pin2 0
Pin 3 1.4
Pin 4/5 3.3ac
Pin6 185
Pin7 0
Pin8 1.5
Pin9 3.3ac

V2
Pin1 --
Pin2 20vdc
Pin3 415
Pin4 340
Pin5 0
Pin6 0
Pin7 20.7
Pin8 20dcv

V3 (5Y3GT)
Pin1 0
Pin2 425
Pin3 0
Pin4 370ac
Pin5
Pin6 370
Pin7 425

Also, seems like the voltages are still quite high. I wondering if I should swap out the 470/5w with a 520 or 620 in it's place. BTW, I want to make sure I'm getting this right -- the big 5w resistor in series 25uF is the "bias resistor" and the 1.5k ohm in series with the 25uF. Is this correct?

[Chuck H]

Disabling the feedback should stop the feedback if it is a phase problem with the output stage. This is an easier test than switching the output transformer wires.

I hate 'not getting it', but... How would disabling the feedback loop alter the phase relationship between the power amp and whatever it may be interacting with?

[mac dillard]

By changing the Cathode resistor from 1.5K to 820 ohm you changed the bias of the power tube. The tube is now dissipating about 8.84 watts. Meaning you are biased hotter than you were with the 1.5k cathode resistor. Per the 6V6 data sheet it has a maximum dissipation of 14 watts. You can safely run a 6V6 in a Single Ended Amp (SE) at 90% of the maximum. ie 12.6 watts. To find this you use Mr. Ohm's law..That would be the Plate Voltage x the plate current (P=VI). The actual plate voltage is the voltage you measured minus the cathode voltage. In you case 415-20=395 vdc. The plate current is the current through the cathode resistor minus the screen current. In your case I=V/R-screen current(estimate at .002 amp) = 20/820-.002=0.022 amp. Then P=395 x .022= 8.84 watts. I didn't do a very good job of explaining that but it is all about biasing the power tube. There are no dummies when it comes to build tube amps, that is unless we are all dummies.

[Chuck H]

Hi... Another dummy here... It should be mentioned that if your amp broke up sooner with the larger cathode resistor then you were likely hearing predominantly cutoff clipping. And this can sound kind of wierd compared to symmetrical clipping between cutoff and saturation. If you want the amp to break up sooner now, the thing to do is increase preamp gain rather than put the old cathode resistor back in (in case you thought about it).

Did the oscillation go away?

[mac dillard]

Need to make sure we understand what is what. The bias resistor for the 6V6 on the Weber layout is the 470 ohm resistor that is in parallel with the 25 uf cap. It connects to pin 8 of the 6V6 tube. It probably is a 5 watt resistor. That is the one I am referencing. Not the 1.5k resistor in parallel with the 25 uf cap that connects to pin 3 of the 12ax7 tube. (That is the bias resistor for the pre amp tube.)Leave it at 1.5k. Sorry for the confusion

[martinman]

Need to make sure we understand what is what. The bias resistor for the 6V6 on the Weber layout is the 470 ohm resistor that is in parallel with the 25 uf cap. It connects to pin 8 of the 6V6 tube. It probably is a 5 watt resistor. That is the one I am referencing. Not the 1.5k resistor in parallel with the 25 uf cap that connects to pin 3 of the 12ax7 tube. (That is the bias resistor for the pre amp tube.)Leave it at 1.5k. Sorry for the confusion

Mac, thanks for the clarification. In my post #18 I mistakenly swapped the 1.5k pre-amp bias for an 820ohm. So the big honkin 5W resistor is the one that controls the power tube bias -- makes total sense now (well, atleast visually...).

So, exactly what effect does changing the pre-amp bias do? I noted a quieter amp, but I'm having a hard time correlating a smaller resistance value to what I'm hearing.

I don't have another 5w resistor, but I'll order it. In which direction do I go? (I assume a larger resistance value is what I'm after, correct?)

[mac dillard]

Correct to go larger..If you have to order then order about four different values. You are going to have to play with the value to get it right.. Say one each of 520, 620, 820 and 1k. All 5 watt. If all things stayed the same we could calculate the value but they will not. Your plate voltage and the cathode voltage will change when you change the resistor. When you changed the pre amp tube cathode resistor you changed the gain of the pre amp tube.

[loudthud]

I hate 'not getting it', but... How would disabling the feedback loop alter the phase relationship between the power amp and whatever it may be interacting with?

Disabling the feedback doesn't change the phase relationship, but it will stop the oscillation if it is caused by phase reversal of the output transformer. Other possible causes of oscillation are excessive phase shift (usually a bad coupling cap), too much open loop gain, poor power supply filtering (bad filter cap), bad ground or wiring mistake.

It's hard to predict what symptoms a transformer phase reversal with create. Many times it's an audible or ultrasonic oscillation. In this case, the amp seems to have a low frequency oscillation which makes me think the amp has some other problem, perhaps in addition to a transformer phase reversal. Disconnecting the feedback is a diagnostic step that may reveal some other cause if the low frequency wavering does not go away.

[Chuck H]

Great. Thanks.

Well since it's a 5f1 I don't think the OT can be wired backwards, per se, but it's always possible that lead dress is causing oscillation. In which case reversing the OT leads may change the phase relationships in the layout and fix the problem. And the NFB loop would still be in place. Yes it can be a PITA if one lead is short, etc. But it doesn't need to be a perfect job unless it fixes the problem. Then the time can be taken to make it tidy, and for good cause.

The description of the problem is hard to pin as anything I've ever dealt with WRT oscillation. If it is motor boating then there may be a problem with the filter wiring or grounding. This seems like it's usually the cause for motorboating.

The OP doesn't say if this amp ever worked correctly or if it's a new build.

[martinman]

The OP doesn't say if this amp ever worked correctly or if it's a new build.

Chuck - it is a new build (completed last year). I've only had time recently to track it down.

I will say that I noticed the power tube wasn't seated well when I opened it up initially. The tube was just barely plugged in. I seated it (not really thinking anything of it) and concentrated on the bias issue. Having gone back and listened for the issue... it has disappeared.

[martinman]

Hi Mac, back again...

Ok, so if I understand correctly, here is the way I arrive at 14.1watts with my new 680 ohm resistor:

V2
p3 - 433
p8 - 25

433-25 = 408

(25/680ohm)-.002 = .0347

.0347 * 408 = 14.1watts

Now, going back to my original measurements in comment #8 using the 'stock' 470 ohm resistor:
V2
Pin3 415
Pin8 20dcv

415-20 = 395

20/470 = .0405

395 * .0405 = 16.01watts


So, it certainly seems I'm going in the right direction with the 680 ohm resistor. I should continue to test higher resistor values.

I understand that plate voltage changes with the bias resistance changes, thus combulating the equation. However, is there a way I can ballpark the next value? I think the highest I have is a 1.5k (but I was thinking of going with 1k.

Just wanted to make sure I'm on the right track here.

Thanks again for your help!

[mac dillard]

You are right..Sometimes with the SE type amps where you are running a really high plate voltage (which you are) you can not get down to the 12 watts by changing the cathode resistor. Go up one more step (820 ohm I think) and see what you get. I would not go any higher than that. If that does not get you down to the 12 watts then you may have to just run it at that. The result would be that the 6V6 power tube will not last as long. The other option is to try to lower the rectified voltage from the rectifier tube..That is a different process. Things like changing to a NOS 5Y3 instead of a Sovtek (if that is what you are using) and or adding Zener diodes. By the way, I checked Weber's lay out that you are using and they show 340 vdc out of the rectifier tube. They are using a Weber WO22772 transformer. That transformer has dual outputs available. One is 660 volt center taped and the other is 600 volt center taped . to get the 340 rectified voltage you would use the 600 volt connections..Can you do the same with the Heyboer? They do not show the wiring diagram for the Heyboer.

[martinman]

Added an 820 ohm resistor and it gets me *barely* into 12watt territory. I'm getting about 12.9 with a NOS Brimar 6v6GT -- 12.5 with a NOS RCA 6v6GT. This is with a NOS 5y3 too. I could try a regular 6v6 -- am I understanding correctly that they draw a little less?

Alternatively, I could swap for another PT -- Mercury Magnetics are running about $100. I could always use the heyboer for another project.

Here are the specs on the transformer:
HY022772
primary black- white 120 v
sec HV Red-Red/Yel-Red 375-0-375
fil1 yel-yel 0-5v
fil 2 green-green 6.3 v
orange- static shield between pri & sec HV, goes to ground

I suppose I should have done some research up front... I've had a heck of a time with the heyboer partially because it didn't come with a wiring diagram. I'm sure they are great iron, but prolly not so good in the paint by numbers kit. That... and I've learned a heck of a lot! Thanks Mac and everyone!