screen and plate voltage

[Randall]

I have already done this test and commented, but for clarity's sake.... The reverb signal distorts at PI input the same way it does at speaker out. The normal channel does not and is about the same ratio larger then the other channel, same as at speaker out. I have been looking very hard at the third stage triode and the circuit around it, but it eludes me.

Following backwards it also distorts back through the 200K and .1 all the way to the plate of the 3rd stage triode. Further, it is distorted and very small on the grid, but then is clean and enormous on the other side of the 10pF/3.3M. 3vp-p vs. 100vp-p. This is what I initially was looking at. Question is, why?

[eschertron]

the 3.3M/220k (ish) voltage divider does attenuate the signal by 24dB or more. Enough to drop 100v to 3v easy. But why is it now distorted? Good question.And the right one to ask, it is. :Yoda:

edit: the test voltages on the schem for the 65 DR I'm looking at (not sure how different from the 63 DR?) show -27dB across the 3.3M resistor, so not too far off. If you're getting 3/100 out of it, the signal to the recovery stage is about -30dB, still ballpark but maybe indicates something is getting lost there?

[Randall]

Pulling the reverb driver makes no difference, anyway the signal on that side of the 10pF/3.3M is fine, on the V4b side it is not. Lifting the wire to the reverb control takes it and the 470K out of play, but no difference. The 470K, 220K, and pot all measure good. What's left, the 3.3M and 10pF? The 3.3M measures close in circuit, the 10pF I cannot measure.

I posted the schematic in post #10.

[eschertron]

I posted the schematic in post #10.

Sorry I missed that. Added to my collection.

If I had it narrowed it down to that circuit, I'd lift connections to eliminate possible side effects. This doubles as a method to do something while I'm thinking of what to do next, and may fix a dodgy solder joint as a bonus. The 10p cap is only there for 'brightness', so removing it can reduce the complexity of the voltage divider. Lifting the wiper of the reverb return should improve the gain figure a little bit. What else could be really wrong? Problem with V4? Problem with the socket?

edit: One good thing about the 65 DR schem is there are test voltages listed. 2.7vac on the plates of the second stages, 115mvac on the grid of stage 3, 3.8vac on the plate of stage 3 (the recovery stage). Are your test voltages at that amplitude? Those voltages are certainly not enough to drive to clipping, but enough for comparison between the normal and reverb channels.

[mozz]

Don't use your scope to measure, use the Fluke.

[J M Fahey]

We are working with this schematic, arenīt we?:


So please do these tests and answer my questions *only* .
Donīt tell me anything else, letīs stay FOCUSED, donīt jump all over the place.
open this schematic:

01) Set all tone controls to 7. Reverb and Vibrato to 0.
02) scope point 01
03) inject 600mVpp at point 02 (weīll work in Vpp ... for now ... since all you use is your scope)
04) start rising Normal volume until signal at point 01 starts clipping.
05) record Vpp present at that moment
06) now inject same generator signal into point 03
07) start rising Vibrato volume until signal at point 01 starts clipping.
08) record Vpp present at that moment
09) post both results.
Please this only this and nothing but this so help you God

[Randall]

Normal channel = 50 vp-p, no clipping
Vibrato channel = 30 vp-p tops a bit flat bottoms a bit rounded

yes that is the schematic

[pdf64]

I think it’s normal for the heavy load presented by the 50k term pot to reduce the max potential voltage swing at the output of the third stage.
Try lifting the term intensity pot.

[Randall]

Juan?

[J M Fahey]

The amp must work as designed, no need to lobotomize it to work (such as pulling the Tremolo pot).

Ok, NOW that we have "put some numbers into it" [tm] I see a gross problem.
No, not with the Preamp.
Both 50Vpp (16V RMS) and 30Vpp (10V RMS) are WAY over whatīs needed to fully saturate the power amp, balls to the wall and then some .... but itīs not happening.
Normal channel is enough, per previous tests, and Vibrato channel is not.

So now I suspect the power amp being incorrectly built and "VERY hard to drive".

Next tests:

10) Set all tone controls to 7. Reverb and Vibrato to 0.
11) scope point 04
12) inject 600mVpp at point 02
13) start rising Normal volume until signal at point 04 starts clipping.
14) record Vpp present at that moment at point 04
15) measure and record Vpp present at point 01
Post results of (14) and (15).

[Randall]

14) 38 vp-p
15) 42 vp-p no clipping

[eschertron]

So now I suspect the power amp being incorrectly built and "VERY hard to drive".

15) 42 vp-p no clipping

42vpp at the input to the PI? Hard to drive, indeed!

[J M Fahey]

14) 38 vp-p

Ok that amp is producing very healthy 22W RMS
38/2,83=13.4 V RMS = 22.4W RMS into 8 ohms

Only problem is that it requires ungodly signal level to reach that, which the Normal channel can barely reach and the poor Vibrato channel can not ... but itīs none of its fault, itīs amply capable to drive a power amp to full power .... a non broken Power amp that is.

15) 42 vp-p no clipping

FWIW a normal Power amp must require some 1 or 2 V RMS at the PI input, so some 3 to 6 Vpp ... tops.
Yet you found it needs .... 42Vpp? WOW!!!!

So we have 2 or 3 possibilities here:



16) you have the lossiest .001 uF capacitor on Earth or itīs not actually soldered so it or some connection between point 01 and actual PI input grid is open

OR (what I suspect) your power amp is wired with 100% NFB (unity gain) so it needs at input the same signal level it produces at the output.

How?

2 easy to make wiring mistakes:

17) R1 is shorted.
Or you soldered the speaker NFB wire to the left end, instead of the right one.

18) or R2 is open or the bottom end is NOT soldered to ground.

19) before touching anything, which will leave us with doubts about the real problem, repeat (12) to (15) but now add:
20) Vpp present at point 05 . I strongly suspect you will have around 38Vpp there

If so, check for (17) or (18) errors.

[Randall]

WOW! You have discovered the problem, which I now can see before even switching anything on! But, I shall first perform the tests you requested.

14) 38 vpp
15) 42 vpp
20) 38 vpp

18) You are correct! I omitted the ground point at R2. That turret also hosts the second cathode and LDR ground in the tremolo circuit (I don't understand how it was functioning, but it was, at least for the quick second I checked). I shall install that ground presently and will be interested to perform those tests again, as well as any others you may like to see. Still wondering about the Vibrato channel, but let's see how things look once I right this wrong.

Juan, you sir are AMAZING!!! I bow in your direction, thank you so very much!

[Randall]

Reporting back. Ground installed, the amp is now acting normal. With 60 vpp signal (as low as my generator will go) the Normal channel produces 36 vpp at the output with no clipping, and the vibrato channel clips at 36vpp at 8 on the dial. I'm happy with that. I'm not sure if the 38vpp I reported earlier has somehow changed, or it was my tired eyes (I have glaucoma).

Anyhow, thank you so much everyone for your help. You are truly a inspiring and gifted group. Now on to taming this over bearing reverb!