6G3 Build Issues

[luthierwnc]

Hello All,

I'm working on a fairly stock Brown Deluxe build and am getting a really rank noise.  In tracing it I'm getting pretty good wave-forms all the way to the PI then they turn to, well, something less.  I've attached three handheld shots of the screen and the settings.  Can't say I've seen this pattern before.  I checked all the components, replaced the bottom cap, re-flowed the lugs, changed tube and it made no difference.  NFB uses the standard 56k resistor.  Volts are a little high on the B+ but within spec on these plates.  NOS Phillips 6V6 tubes are biased at 25ma.  The OT is a pull from another amp.  It didn't squeal like I had the OT leads reversed but that might be the next thing I try.

Any ideas will be welcome and appreciated, thanks, Skip

[PRR]

Your schematic?

[octal]

A bit unclear why you're scoping the "bottom" of the PI... it should be an LTP.  What you see at the bottom could be confounded by NFB. Pull the output tubes, scope both plates of the LTP one at a time and see what you get. If it looks odd or ugly, start making some measurements to verify your resistor values, log DC voltages, etc.  I'd pay particular attention to the voltage drop across the 820 ohm resistor, that sets the bias for the LTP. Looks like Fender is showing about a 2v drop there.


If the waveforms look good with the output tubes removed but get gnarly with them plugged in, you could have a bad output tube, miswired or open screen resistor, etc etc.

[luthierwnc]

I pulled the power tubes and jumpered the PI grids.  The waveform looked fine.  I'll rummage around for another pair of 6V6s but I just got these from KCA and haven't gone wrong yet.  I'll see if there are any problems with the screen resistors tomorrow.  I'm using 1 ohm bias resistors between the power tube cathodes and ground with banana jacks in the back of the amp that fit the multimeter probes.  The difference between the two is about a tenth of a volt. 

PRR:  schematic attached, thanks, sh
PS the tremolo pedal jack is grounded.

[octal]

I am confused why you would "jumper the PI grids." One of the PI grids is supposed to be the signal input and the other is supposed to be effectively at AC ground. (Assuming no nfb.) In fact, if you're feeding the same signal into both grids of the PI, you should be seeing very little output, as one input is inverting and the other is non-inverting.


Keep it simple- With the output tubes pulled, just inject signal in to the front panel input of the amp and scope the grid pins of the output tubes. No other changes. Depending on how you're attempting to inject signal directly to the PI, you could be disrupting its bias unless you are injecting the signal with a capacitor in series with the output of the signal generator. Unlike other stages in a tube amp, the PI's bias is not referenced to ground, but rather to the junction of the 1m grid resistors and the 820 ohm resistor....so if you're clipping in a signal generator that's grounded to the chassis it's going to jack with the PI's bias.


I think you're jumping to conclusions about the output tubes being bad until you dig further.

[luthierwnc]

Scope shots at the power tube grids.

[octal]

Those new scope shots are with the power tubes removed, feeding the sine wave generator in to the front panel input of the amp, and the scope is set to 5V division? And the signal looks normal if you scope the input to the PI at the .01 cap?


Can you give DC voltages of all the pins on the PI tube (except the filaments!) and also please the voltage at the junction of the 1m PI grid resistors, the 820ohm resistor and the 6800ohm resistor?

[sluckey]

What is the amplitude of the signal from your sig gen?

[luthierwnc]

Octal: Yes to all of the above.  At the PI cap it gets a little squarer at the bottom but still what I consider a fairly clean signal.  1k (more-or-less) sine wave.  Volts are; K = 27 VDC, #1 = 308 (246 with power tubes in), #2 = 18.74, #6 = 248, 7 = 20.22.  B+ without tubes is 478.
Sluckey:  signal at the jack is a shade less than 100 MV (5x divisions by 20 MV setting) with the probe on 10x.  I'm using a little Rolls generator set at around 60% gain. 

[sluckey]

Sluckey:  signal at the jack is a shade less than 100 MV (5x divisions by 20 MV setting) with the probe on 10x.

Hmm. 5 divisions at 20mV/DIV would be 100mV if you were using a 1X probe. But looks like 1V when using a 10X probe. Way too much input.

[octal]

You didn't provide the voltage to ground at the junction of the 1m PI grid resistors, the 820 ohm resistor and the 6800 ohm resistor. Alternately, just measure the voltage across the 820 ohm resistor.

On pin 1, you mention "308v, 246 with the power tubes plugged in." Then you mention 248V at pin 6. Is that with the power tubes in or out? (i.e. we need to compare apples to apples... if there's a 60v difference between the two plates #1@ 308 and #6@248  that's too much spread. If it's 246/248 that's fine.)

[luthierwnc]

Octal; the voltage across the 820R resistor is 2.12 VDC. 

Sluckey; I turned the signal generator from 4 out of 10 to 2.  Now the waveform at the power tube grids is clean.  Volumes and Tones are all at 5.  My next suspect is the output transformer.  It is an old Hammond that used to drive a pair of ECL-86 tubes for the reverb on an "E" organ amp.  The resistance between the outside legs and center-tap is about 10% out.  Not sure that makes a difference.  I have not tried different 6V6s yet.  I'll fiddle with it and report back.
Thanks gents, sh

[sluckey]

Sluckey; I turned the signal generator from 4 out of 10 to 2.  Now the waveform at the power tube grids is clean.

100mV_pp is a good level to feed into the input jack of most any guitar amp. I would just connect the sig gen to the input and use the scope to set the sig gen output to exactly 100mV_pp. Knowing the exact amplitude at the input is very useful for checking gain as you trace through the circuits. The later Fender schematics show the input signal level as well as other signal levels throughout the amp. Very handy when troubleshooting.

[luthierwnc]

Set the generator to 100 mv -- at 3 on the dial.  I also tried a pair of 6EY6 tubes since they were in the box.  They sounded like ass too -- thin, nasal, low-power.  This was the wave at the speaker jack (back to the sine) with either set.  This is with the wave in good shape at the PI.  With the power tubes back in; B+ is 422 VDC.  At the PI power I'm getting 259 on the 82k and 241 at the 100k.  Not sure why.
The OT was a last-minute decision.  I tried to get a trannie from David Allen but he said Heyboer was behind schedule on the units he needs for his own production.  Nice guy.  With ClassicTone going toes-up, it's a seller's market.  sh

 

[octal]

The voltage drop across the 820ohm resistor seems OK.  I did once build an amp that had an output waveform that looked kind of like that...I had switched the centertap and one of the end connections of the primary of the output transformer. DC bias of output tubes was Ok, but the waveform was a mess. Maybe checking the resistance from the CT to each of the ends... I would expect some difference, but it should be roughly symmetrical. Even if the color codes are OK, it could be miswired internally.


Were those scope photos of the output of the PI also using divided by ten mode on the probe?  How many volts can it swing peak to peak before things look gnarly?

[luthierwnc]

Hi Octal.  The last shot was 10x at 1 volt.  It looked the same with the 6EY6s and 6V6s.  That's with a 100mv input and both volume controls about 5. 

I ordered another output transformer -- had good luck with these little units: https://www.ebay.com/itm/154430048992?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2060353.m2749.l2649.  Looks like he's almost out of those too. 

Thanks again and I'll let you know how the new iron looks.  Cheers, sh

[luthierwnc]

Hello All,  It was a bad output transformer.  I slapped the new one in and it sounds great.  Voltages are right where they should be.
8-ish out of 10 times my build problems are something I did wrong.  It isn't often that I can blame it on a component -- although I did repurpose the Hammond OT without knowing it was working first.  At any rate, the new one sounds fine and I will start the cab this weekend.
As always, thanks for your input and best of luck with your projects.