Need help debugging a PR trem

[martin manning]

But the frequency looks like 1/35mS...about 28-29 Hz?

[Phil_S]

What is the purpose of the diode across the intensity pot? What happens when you lift one side of it?

[Luthierwnc]

Phil S: That's a mod to get more out of the trem but I haven't added it. I want to get things running stock before tweaking (if needed).

Martin: Hi Martin, I must have missed the class on estimating the sine wave frequency on the oscillator plate. What should I be doing? Thanks sh

PS guitar built since my last amp in the pic. It's Ergonomic (a la Adrian Legg) and designed for both passive and MIDI using a Roland pickup. Haven't had much of a chance to air it out but a recently arrived Boss GP-10 is ready for some day-job down-time. sh

[Phil_S]

Have you got a known good strong 12AX7 in the oscillator hole? Also, it must be a 12AX7 (or 6EU7) and not an AT7, 5751 or something like that. The tremolo needs a high plate resistance and a good strong tube. Weak tube, low plate resistance, won't oscillate!

I am thinking you are looking hard at the build and chances are it was done correctly. I'm looking elsewhere. The only other thing that comes to mind right now is an unintended ground, even just a whisker.

I sure hope you fix this!

[Luthierwnc]

Thanks Phil, I've had three 12AX7s in there. All were known good- a Tungsram, a Ruby 5th generation and a GE gray plate.

[tubeswell]

Its got to be a wiring error. Quadruple check everything again.

[JazzGuitarGimp]

Martin: Hi Martin, I must have missed the class on estimating the sine wave frequency on the oscillator plate. What should I be doing? Thanks sh

Take the reciprocal of one cycle of the sinewave. One cycle looks to be about 8 divisions at 5mS per division. One cycle is then 8 x 5mS = 40 mS. Now take the reciprocal:

1 / 0.040 = 25Hz

I think the range you're looking for is a few Hz at the slowest setting, and maybe as much as 10Hz at the fastest setting.

[tubeswell]

...

1 / 0.040 = 25Hz

I think the range you're looking for is a few Hz at the slowest setting, and maybe as much as 10Hz at the fastest setting.

Makes me wonder whether the resistors and caps in the LFO RC network in the ballpark correct values?

[martin manning]

In the pics showing a trace of a 1kHz signal I see about 900 Hz on the screen (close), but the red knob labeled "Swp Var" is not set in the "cal" position either. I don't think that would explain the LFO frequency appearing to be ~5x too high. Put the red knob on cal and put the probe on your heater voltage and see if that shows 60 Hz.

Thanks for filling in for me Lou.

[Luthierwnc]

Thanks for the interest. I've attached two more files. One is the heater wave for calibration. The second is another plate reading.

Of note, this waveform is completely dependent on the speed knob. Turning the knob even slightly flatlines the wave. The it slowly returns. Sometimes the wave holds this relatively stable form. Sometimes it keeps increasing until it won't fit on the screen at the same volts/div setting. Then it will go back to this relatively stable form. This pic only catches a piece of it since it flashes in parts rather than showing a consistent reading.

I'm not above ripping the board and shotgunning all the components. The mystery is losing its allure.

Cheers, sh

[Phil_S]

I'm not above ripping the board and shotgunning all the components. The mystery is losing its allure.

I wouldn't be too quick to do this. Assuming you have a mistake, you could make the same mistake twice. Patience will probably get you there faster. Turtle and the hare.

[sluckey]

Of note, this waveform is completely dependent on the speed knob. Turning the knob even slightly flatlines the wave. The it slowly returns. Sometimes the wave holds this relatively stable form. Sometimes it keeps increasing until it won't fit on the screen at the same volts/div setting. Then it will go back to this relatively stable form. This pic only catches a piece of it since it flashes in parts rather than showing a consistent reading.

This erratic/intermittent behavior indicates a problem in the oscillator circuit. The waveform should change frequency as the speed pot is changed. And the amplitude will usually change some also but it should not drop out (flat line). You want a really big amplitude waveform. Sounds like it's trying. Change all 3 of the feedback caps. Check the two 1M resistors in the feedback path. Check the speed pot and 100K resistor also. Check the 3.3K cathode resistor and the 25µF bypass cap.

[martin manning]

Skip your heater trace looks like 116 Hz, so your scope seems to be off by almost a factor of two. You haven't moved that red knob at all, though.

If the speed pot is acting weird and killing the LFO signal, I think I'd replace it.

[cbass]

I just had a similar problem with my PR type build although I'm not using bias trem I had a similar problem it would try to start and stop turned out to be the caps in the oscillator circuit I was using some little cheap yellow caps even though they were new and seem to work elsewhere they wouldn't work on the oscillator.

[Luthierwnc]

I recalibrated the scope with the red knob on CAL and got 58 hz on the heater - close enough. On the oscillator plate I got 6.5 divisions at 10 ms per so a bit over 15 hz. Still too much for the ranges suggested.

It will probably be the weekend before I do any ambitious desoldering. I'm pretty sure I have enough caps and resistors to tear out the trem components. A 3 meg pot might be a different story.

Continued ideas welcomed, Skip