Crate GTX3500H Tidal Wave - Channel 3 Inoperative

[Neutronarmy]

Music Electronics Forum Folks,

I have a Crate Tidalwave GTX3500H on the bench with a pretty strange issue. It came in with two shorted power transistors that were blowing fuses. All power transistors were swapped and the amp came back to mostly regular functionality. However, channel three - the high gain channel - is not properly functioning. All part numbers referenced below relate to the attached preamp schematic (678SCH_1.pdf), but I also included the rest of the amp schematic.

The channel switching seems to function properly - the LED lights, the signal switches - but the total output when channel three is selected is barely above a whisper. Channel one and two function perfectly. I traced a signal through the circuit and it appears that when channel three becomes active it attenuates the signal to a great degree. I can inject a signal at the input and chase it strong to the junction of C37/Q9 when any other channel is selected. However, when the third channel is selected the amplitude drops drastically. When tracing said signal at the output of the level pot (P12) with channel three selected, all controls seem to affect the signal (tone/gain/level/mute), but the signal is too greatly attenuated to be usable. I tried bypassing everything prior to the first opamp stage and injecting at pin 2 of IC7 and it still attenuates the signal.

Q7/Q9/IC6/IC7 have been replaced as initial suspects and no change in performance. I've checked all of the diodes in channel three preamp and they seem fine (zeners check at ~.7V when forward biased and infinite reverse biased; also, for future reference: any theory on their exact roles in the opamp stage?). All TL072s are getting +/-14.5V (1.5V low on both rails admittedly). Q7/Q9 read 0.5VDC on gate when switched in an 14.1VDC when switched off.

Any ideas? I'm pretty stumped on this one. I greatly appreciate any help in the matter.

Take care!
Chris

[Jazz P Bass]

Where are you losing the signal?

You can replace components to your hearts content but that is not 'troubleshooting'.

[Neutronarmy]

Where are you losing the signal?

You can replace components to your hearts content but that is not 'troubleshooting'.

I apologize if my last post was not clear on the point of signal loss. The signal is never completely lost - just greatly attenuated. The point at which that is noticeable is the junction of Q9 source and pin 2 of IC7 (I stated wrong previously in my haste and put it on the other side of C37). Said attenuation is also only noticeable when channel three is selected from the front panel. When a 100mV 1KHZ signal is fed into the amplifier input it probes fine at the aforementioned junction if channel one or two is selected. When channel three is selected it is greatly attenuated and seemingly distorted. I also attempted injecting the signal directly at pint 2 of IC7 with the same result as using the input.

I definitely understand the point you are making about random parts replacement. I promise I wasn't just spitting into the wind. Perhaps explaining my train of thought/order of operations might help (and feel free to correct me in any wrong thinking):

1). Q9 was replaced as it is an active switching component in the signal path that happens to be the last component prior to signal attenuation. No change in behavior.
2). IC7 was replaced as it is the other half of the junction where signal disappears and all of its associated components in the stage seem to test okay in-circuit. No change in behavior.
3). IC6 was replaced as it is the next active component that could be potentially loading the AC signal. It was also on hand and easily replaced while the board was removed for the third time (though I was skeptical of its potential help and similarly skeptical of the next component). No change in behavior.
4). Q7 was replaced as it is the final switching FET before the signal is mixed with the other channels. I thought it might be leaky and thus attenuating the signal proceeding it. I feel this logic is pretty dubious, but it was frustrating and the part was available. Weirder things have definitely been the source before. No change in behavior.

I should also note that the board was looked over thoroughly and any suspect joint in the region reflowed - just to rule out real dummy mistakes.

Thanks again for any help,
Chris

[g1]

If you short drain to source of Q9, does the signal come back up? If so, there may be a problem with CV3 signal.

[Neutronarmy]

If you short drain to source of Q9, does the signal come back up? If so, there may be a problem with CV3 signal.

g1 - I'll give it a shot this afternoon and let you know. Thank you for the suggestion!

[MarkusBass]

Chris,

Your tests and conclusions regarding "no signal on the junction of Q9 source and pin 2 of IC7" are not correct. This is how opamp works. You shouldn't check the signal directly on pin #2 but rather on pin #1 (output). Also connecting signal generator directly to pin #2 is not a good idea. You can connect it through a resistor (let's say 10k). In fact, this test will tell you whether IC2 is OK, ro not. I would also (as suggested) check the switching voltage. You may get the problem simply because the switching voltage is not present and in this case the source of the problem is in completely different part of the circiut.

Mark

[Neutronarmy]

Chris,

Your tests and conclusions regarding "no signal on the junction of Q9 source and pin 2 of IC7" are not correct. This is how opamp works. You shouldn't check the signal directly on pin #2 but rather on pin #1 (output). Also connecting signal generator directly to pin #2 is not a good idea. You can connect it through a resistor (let's say 10k). In fact, this test will tell you whether IC2 is OK, ro not. I would also (as suggested) check the switching voltage. You may get the problem simply because the switching voltage is not present and in this case the source of the problem is in completely different part of the circiut.

Mark

Mark - thank you for the insight. I could've sworn that I tested pin 1 of IC7 and had the same result, thus I went back to the first point it was noticeable. I was incorrect (and apparently my testing logic was as well). Upon retesting I get a solid signal at pin 1. In the end it was a busted gain potentiometer (which I also tested, but merely from outside lug to outside lug, not to the wiper). Thank you very much for everyone's perspective. Sometimes getting mired in a repair leads to some really simple oversights.