Tweet
I have one of our Peavey 6505 120W Amp Heads in the shop with the complaint of being noisy and having a buzz. Turned it on, centered the tone controls, then gave a listen to the two channels (Normal and Lead), first with the Pre-Gain pots down, Post-Gain pots full CW. Substantial difference. Then with the Post-Gain pots still at full CW, turned the Pre-Gain pots full CW, much greater noise difference.
I didn't have a 6505 120W schematic, but found the Peavey 5150 Schematic matches the PCB's and I presume the same circuit. I only got marginal improvement by swapping out V1, V2 & V5 tubes. V4 tube (driver) had some issues that was causing constant hum regardless of pot settings, and that was cured with another tube.
Looking at the 5150 schematic and trying to follow the signal flow thru all of the relay switching between the 1st gain stage V1B & 2nd gain stage V1A, then the switching across the voltage divider to the input of the 3rd gain stage V2B, as well as the cathode network of V1A, it looks like there are 5 cascaded 12AX7 gain stages before you get to the Post-Gain stage pots, which then go thru a cathode follower stage V3B, and then thru more relay switching & networks of the Post Gain stage pots which then finally feed V3A, the final gain stage ahead of the LTPI driver.
I gave up trying to follow that, and took my best shots swapping V1, V2, V5 & V3 tubes for lowest noise overall on the Lead Ch.
Does anyone have or have seen a block diagram of this circuit?
I then figured I'd measure the gain of each channel at several spot frequencies...100Hz, 200Hz, 400Hz, 1kHz & 4kHz. I began with the Lead Channel, and began with my Oscillator of the Amber 3501a turned down to minimum (300uV). The output with that input level was into hard clip. I changed to a different oscillator (B & K 1023), which will do 100uV full scale, to which I attenuated that another 10dB with my HP353A Xfmr coupled step attenuator (0-100dB). So, using 1mV input level for the Normal Ch, and 30uV for the Lead Ch, I got:
Channel 100Hz 200Hz 400Hz 1kHz 4kHz
Normal 7.1V/77dB 8.6V/79dB 7.5V/78dB 7.5V/78dB 2.4V/68dB
Lead 6.8V/107dB 10.8V/111dB 10.8V/111dB 11.5V/112dB 4V/103dB
That's 30 to 35dB difference in gain. That sure explains the vast amount of noise, being 5 gain stages cascaded. I'm just not seeing where there's that much attenuation in the few networks in all that relay switching, but it sure must be in there. Or, is this showing there's further problems in the lead channel, resulting in there being TOO MUCH gain?
The power tube circuit is fixed bias, and while the Ruby 6L6GC power tubes were nicely matched, they were running 40mA/20W dissipation. I ended up trimming R66 in the bias network to drop the plate current down to 25mA/12W dissipation
Peavey_5150_(EVH_120,_2-92)_Schematics.pdf
I didn't have a 6505 120W schematic, but found the Peavey 5150 Schematic matches the PCB's and I presume the same circuit. I only got marginal improvement by swapping out V1, V2 & V5 tubes. V4 tube (driver) had some issues that was causing constant hum regardless of pot settings, and that was cured with another tube.
Looking at the 5150 schematic and trying to follow the signal flow thru all of the relay switching between the 1st gain stage V1B & 2nd gain stage V1A, then the switching across the voltage divider to the input of the 3rd gain stage V2B, as well as the cathode network of V1A, it looks like there are 5 cascaded 12AX7 gain stages before you get to the Post-Gain stage pots, which then go thru a cathode follower stage V3B, and then thru more relay switching & networks of the Post Gain stage pots which then finally feed V3A, the final gain stage ahead of the LTPI driver.
I gave up trying to follow that, and took my best shots swapping V1, V2, V5 & V3 tubes for lowest noise overall on the Lead Ch.
Does anyone have or have seen a block diagram of this circuit?
I then figured I'd measure the gain of each channel at several spot frequencies...100Hz, 200Hz, 400Hz, 1kHz & 4kHz. I began with the Lead Channel, and began with my Oscillator of the Amber 3501a turned down to minimum (300uV). The output with that input level was into hard clip. I changed to a different oscillator (B & K 1023), which will do 100uV full scale, to which I attenuated that another 10dB with my HP353A Xfmr coupled step attenuator (0-100dB). So, using 1mV input level for the Normal Ch, and 30uV for the Lead Ch, I got:
Channel 100Hz 200Hz 400Hz 1kHz 4kHz
Normal 7.1V/77dB 8.6V/79dB 7.5V/78dB 7.5V/78dB 2.4V/68dB
Lead 6.8V/107dB 10.8V/111dB 10.8V/111dB 11.5V/112dB 4V/103dB
That's 30 to 35dB difference in gain. That sure explains the vast amount of noise, being 5 gain stages cascaded. I'm just not seeing where there's that much attenuation in the few networks in all that relay switching, but it sure must be in there. Or, is this showing there's further problems in the lead channel, resulting in there being TOO MUCH gain?
The power tube circuit is fixed bias, and while the Ruby 6L6GC power tubes were nicely matched, they were running 40mA/20W dissipation. I ended up trimming R66 in the bias network to drop the plate current down to 25mA/12W dissipation
Peavey_5150_(EVH_120,_2-92)_Schematics.pdf