PI trimmer and tube mismatch

[FUCHSAUDIO]

You should see AC on the plate, and you should be able to measure from plate to ground. You should also be able to measure plate to plate AC as well.

I used a dual AC Voltmeter (Leader) and with output of amp at 1 volt, I then measured approx 750 MV at the plates of the driver. Since I could look at both halves, I adjusted the trimmer until they were equal. They should measure '0' when at equal but opposite phase.

[grtamp]

Hi ,which model? Leader LMV 181A o other?

[greiswig]

You should see AC on the plate, and you should be able to measure from plate to ground. You should also be able to measure plate to plate AC as well.

I used a dual AC Voltmeter (Leader) and with output of amp at 1 volt, I then measured approx 750 MV at the plates of the driver. Since I could look at both halves, I adjusted the trimmer until they were equal. They should measure '0' when at equal but opposite phase.

Man, I chased my tail for a half hour on this one. I could measure the AC signal as dead steady through V1 and (when engaged), and then things would get sloppy some time after that; the signal would waver at a low frequency, almost like a beat. I was sure something must be oscillating, so I started chopsticking.

Then it dawned on me that I still had the effects loop engaged. With effects on. Yup, a delay/reverb on a steady signal will do some odd things, I guess.

Okay, so I ended up being able to measure a steady AC signal per Andy's and Tony's description, but only after putting a 20k pot in was I able to find the matched spot on several tubes, including one that is supposedly balanced. So I need to measure the relative total resistance and match the resistors better to get to where I can use a 5-10k pot. 86k and 100k isn't cutting it.

But anyway, using this method, the DC voltage is close to about 2VDC difference when matched, and the amp sounds really good. I haven't tried the middle position Tele feedback, but my HB guitar already sounds like it's trying to feedback more. Thanks again for all the patience! From what I can tell, this method (without effects loop) works well to get you in the ballpark.

[angelodp]

Now how does it jive to your ears. Now that you have that spot, if you tweak by ear do you get back to that spot.

[talbany]

g-Glad you got it figured out..FWIW The AC balance you did will get you close however since output transformers are never perfectly balanced continue to sweep around the sweet spot see if it makes a difference..Try it with the Tele test if you feel like it..Keep us posted..

Good luck

Tony

[glasman]

A few additional comments on adjusting the PI trimmers.

Starting

1. Normally I find that 10K is about the best split between the two PI plate resistors. I like 110K/120K or 91K/100K depending on the PI style.

2. 5K or 10K trimmer. The sweet spots are very narrow in my testing over the years.

Always inject the single directly to the PI input. Anywhere prior to this point will introduce too much distortion and make the adjustment almost impossible.

Methods I have used.

Scope and Generator. Requires dual trace scope and low distortion signal generator.

1. Inject a 1Khz single directly to the input to the PI.
2. Measure Plate 1 on Channel 1 and Plate 2 on Channel.
3. Set the scope up to sum the two channels.
4. Adjust trimmer for minimum signal deflection. On either side, the signal will amplitude will increase.

Note : it will never be a straight line, but you can get very close.

This approach is similar to Andy's AC voltmeter metod.


My preferred method

Requires a spectrum analyzer.

1. Inject a 100hz single directly to the input to the PI. I use a 1v p/p signal.
2. Place the spectrum analyzer input to the speaker output. I normally grab the speaker side of the PA feedback resistor.
3. Set the analyzer up to show the frequency spectrum and measure the signal amplitude at 200hz (second harmonic).
4. Adjust the trimmer to the minimum amplitude at 200hz (second harmonic). You should be able to sweep on either side of this point and the 200hz harmonic should increase. In a skyliner style PI you should be able to achieve a level difference of at least -70db and about -55dB for a Marshall style PI.

All of the above require a properly size dummy load.

Gary

I have one to adjust later today, I will upload some pictures.

[glasman]

Forgot to mention, I select tubes that have a matched transconductance (the only thing I can measure on my old Hickok). between the halves. Ideal case is matched transconductance and matched gain.

Someday when I am rich and famous I will own an amplitrex.

[greiswig]

Now how does it jive to your ears. Now that you have that spot, if you tweak by ear do you get back to that spot.

Well, it sounds better than it did, but I have to say I have no confidence in my ability to adjust the 20k pot in small enough increments. Measuring from the wiper to the far tail of each resistor, Gary is right on the money: AC balance on mine is with a 10k difference. At the time I got the carbon film resistors I'm using, the closest values were 100k and 86k. So I need to get my hands on a 90k-ish resistor and put a smaller pot in so I have higher resolution on the adjustment.

Thanks again for the patience, guys. This method exposed something that hadn't been discussed before: that adjusting by ear, which is probably GREAT, won't work at all if your sweet spot is outside the adjustment range. So having an objective way to find out whether you're close is extremely helpful.

[erwin_ve]

Interesting topic!
In what way do you guys take the presence circuit in consideration?
The way you set the presence control has effect or is it negligible?

[glasman]

Interesting topic!
In what way do you guys take the presence circuit in consideration?
The way you set the presence control has effect or is it negligible?

I always make sure that the presence control is set to minimum. But that is me....

[angelodp]

For those that do not have a signal generator, can you please describe how one would apply this method. I realize that i can get a 400hz signal off of the net, but then I have no clue how to inject it.

Ange

[greiswig]

Ange,

I'm not sure how critical it is to inject it into the PI itself. I simply made sure I was on the clean channel, and that the signal itself wasn't so high that it was driving V1 into distortion.

Hopefully Gary will tell us if it's really critical to put the signal directly into the PI.

[FUCHSAUDIO]

I use my loop return input, and it seems to work fine. Using the input preamp potentially adds distortion or overload, or frequency changes that might affect the measurement.

[talbany]

Inject it right into the return jack on the loop... Andy beat me to it..

Tony

[angelodp]

Excellent. So simple once you know.

A