Carvin Legacy 100w repair

[JazzGuitarGimp]

It looks like V5.A is in cutoff. You cannot measure the grid voltage directly (unless you have a VTVM), but you can get an idea of where the grid voltage is by measuring voltages around the grid. What are the idle DC voltages at the following points?:

- Point C
- The junction of R46 and the two cathodes
- The junction of R46 and R44
- The junction of R44 and R43

[xtian]

I reading about 70vdc on the grids of this PI. Is that a leak? I went looking for voltage charts that include LTP PIs, and found Marshall, but they don't list pins 2 and 7 of the PI.

[JazzGuitarGimp]

Assuming each triode is pulling 1mA at idle, I'd expect to see closer to 90V at the grids, but unless you are using a VTVM to measure the voltage at the grid, your meter's input impedance is going to pull the grid voltage closer to ground. So, 70V is in the ballpark. If you measure the voltage at the junction of R46 and R44, you'll probably see a higher (than 70) voltage.

[JazzGuitarGimp]

One other test you might consider is pulling the PI and measuring the voltages at the grids, cathodes, and plates. Both plates should be at whatever voltage point C is, while the grids and cathodes should be at 0V.

[xtian]

- Point C 455
- The junction of R46 and the two cathodes 78.7
- The junction of R46 and R44 78.5
- The junction of R44 and R43 39.5

[xtian]

One other test you might consider is pulling the PI and measuring the voltages at the grids, cathodes, and plates. Both plates should be at whatever voltage point C is, while the grids and cathodes should be at 0V.

With no PI tube, pins 1,6: 455v; pins 2,3,8: 0.008v; pin 7: 0.025v

[JazzGuitarGimp]

- Point C 455
- The junction of R46 and the two cathodes 78.7
- The junction of R46 and R44 78.5
- The junction of R44 and R43 39.5

In your second post, you said you've got 470V at V5.A plate. Yet the plate load resistor R49, only has 455V available (from point C). This discrepancy may point to the issue. I don't know if this will yield any useful data, because even if it's leaky, I don't think there will be 470V on the other side, but you might try lifting one end of C43, then measure the idle plate voltages on the PI.

[JazzGuitarGimp]

One other test you might consider is pulling the PI and measuring the voltages at the grids, cathodes, and plates. Both plates should be at whatever voltage point C is, while the grids and cathodes should be at 0V.

With no PI tube, pins 1,6: 455v; pins 2,3,8: 0.008v; pin 7: 0.025v

One more idea: With the PI removed, measure the resistance from pin 2 to pin 3 of the PI socket. It should measure ~100K

EDIT: And also measure resistance from pin 2 to GND, and pin 7 to GND. Both should be ~144K

[xtian]

I reflowed the tube pins, and, even though I still see 400+ volts on pin 6, suddenly the amp sounds good.

There will be a pause in the action while I take my daughter to lunch and orthodontia.

Thanks, gents!

[gingertube]

AC wise things don't look too bad.
The full input appears at the top grid
The bottom grid should be at or near zero ac
The 2 cathodes should be at 1/2 of the iput signal
So that top triode grid to cathode is 1/2 the input signal
AND
bottom triode grid to cathode is 1/2 the input signal (but opposite phase)
That is afterall how the thing operates.

AFTER I HAD TRIED ANOTHER TUBE then
I'd be looking for a DC bias probelm, Check all of R43 thru' R47 for an open circuit.

The last Carvin Legacy "blatty distortion" problem I had turned out to be a microphonic tube - spent bloody hours chasing a non-existent circuit problem. That is not likely to be causing your problem with the PI but a tube with one dud triode section could be the problem. Hence the shouting captials above.

Cheers,
Ian

[JazzGuitarGimp]

I reflowed the tube pins, and, even though I still see 400+ volts on pin 6, suddenly the amp sounds good.

There will be a pause in the action while I take my daughter to lunch and orthodontia.

Thanks, gents!

I am thinking you're most of the way there. It still looks like there is a bit of a bias problem at V5.A. Some little bit of unintended resistance pulling the grid down towards GND. The grid is a very high impedance circuit, so even a large additional resistance (high impedance) to GND can make a big difference. It could even be a carbon track on the tube socket between pins 7 and 8. It may even be residual solder flux on the PCB between those two pins.

[xtian]

With the PI removed, measure the resistance from pin 2 to pin 3 of the PI socket. It should measure ~100K

Yes, I read 98.5k.

And also measure resistance from pin 2 to GND, and pin 7 to GND. Both should be ~144K

Confirmed.

@gingertube: I checked all the resistor values in and around the PI, and couldn't find anything odd.

@Lou, I cleaned up the board, looking for anything suspect. The reverb tube (V4) had a resistor burn out and was patched up by Carvin, or so says my customer.

Anyway, this bird is baked. Sounds good, plenty of power, etc. Buttoned up and calling the customer.

[martin manning]

I reflowed the tube pins, and, even though I still see 400+ volts on pin 6, suddenly the amp sounds good.

BTW, did this get solved?