Carvin Legacy NFB/Phase Inverter debug

[goldstache]

I have a Carvin Legacy on the bench with a PI issue. 


I have no Vdrop across the Anode load resistor on the V5A "top triode of PI". 


I get normal voltages when I pull the NFB coupling cap to that stages grid.  In the original schematic it appears to be a 10uF non polar. 
I then swapped that 10uF cap and was back to no VDrop across the anode resistor.  okaaaaaayyyyyyyy???????


And, to boot I had DC voltage after that cap headed to the NFB connection on the OT side.
Strangely, I only see the voltage (approx. 31VDC) on the side headed to the NFB, here is my test points along that part of the circuit.




NFB side of 10uF > ribbon cable pins H5A I/J > H5B I/J > stopping on the top side of the R91/C70 connection on the NFB board



Again when I lift the 10uF at the PI grid, phantom voltages toward the NFB seem to go away and the PI exhibits "normal vdrop from both triodes in V5. 


Any ideas here? I have swapped the PI tube to known working tube, no change. 


Thanks
-joe

[sluckey]

Measure the voltage on the grid of V5B. Then replace the 10µF with a .1µF.  What is the voltage on V5B grid now?

[goldstache]

Hey Sluckey! 


Yeah it blocked DC from the NFB side.  And I have Vdrop across the V5A anode.  Yay!!!!!  I have 80 volts at the grid. So can I go back to the 10uF or stick with the .1uf @250V?


So why did the 10uF non polar replacement not jive???  And if I needed to go polar would the cathode side of the cap be towards the NFB????   




I don't understand why your recommendation worked, but I'm sure glad it did!

[sluckey]

Based on what you said, the 10µF was leaking and upsetting the bias on that LTP PI. You can use a polar e-cap with the positive end connected to the grid (because of the positive voltage). I would use at least a 150V rating.

Using a .1µF cap will change the NFB. Don't know if that's good or bad. I would rather use the same value of cap that was originally in the amp. What was the voltage rating of the original?

[goldstache]

Yeah that what I thought.  It's a 10uF non polar @ 50v stock.   The weird part is I swapped it originally with a 10uF non polar at 100V, and it had the same issue.  But the .1uf "test cap" was fine. 


Fool me once, fool me twice, ask Sluckey to bail me out.  :)


Now that I know that's the issue, I'll go back to the 10uF till it's up and going so as not to fool with the response of the NFB. 


Thanks

[sluckey]

50v rating is insufficient in my mind. You measured 80V on the grid with a non leaky cap. The actual voltage is even higher than that. Your meter loads the voltage down.

Now that I know that's the issue, I'll go back to the 10uF till it's up and going so as not to fool with the response of the NFB.

I would approach that differently. That 10µF is making your LTP PI tube misbehave and that will probably mask other problems you may have. I would put the .1 cap in and fix any other issues then come back and address the cap issue when the amp is otherwise working properly.

[goldstache]

Great advice, but with the PI biased correctly everything else seems cool.  I'll leave it cook for a bit and check back later. 

Again, thank you.


-joe

[PRR]

The non-polar electrolytic leaks too much for this circuit. It "may" be OK now, if you wait to charge, but is sure to get leakier over time.

Non-polar also makes no sense: there's steady known DC, polar could be used, if that made sense.

10uFd here makes NO sense. That against the 360K is flat to 0.05Hz-- pointless for guitar.

Neither does 10pFd which doesn't do anything until 50KHz.

I think the drafting program defaults to a "10" and nobody put the right value in. Maybe it was quitting time, and didn't get picked-up the next morning.

Simple audio math suggests 360K wants a 0.03uFd cap to work over the audio band. 0.05u and 0.1u are reasonable values. 100X (10uFd) is NOT reasonable.

I calculate 88V here. This echoes your observations, which were meter-loaded. Prudence suggests 200V rating. Surely you have 0.05u-0.1u 400V film caps in stock?

[goldstache]

Well the amp was fine with a .1uF in that position.  As soon as I add an electrolytic. (10uF @150V)  anode toward grid of PI, it has the same problem. 


HMMMMMMMmmmmm?????????

[goldstache]

Perfect PRR.  Those calculations are a bit over my head as far as frequency response with regard to the NFB. 


I sure do like the sound of a .03-.1uF in that position to cover useful guitar freq's.   




Thanks, I'm gonna let it cool down and swap a 400V poly cap in that position.