FET input stage problems

[EchoWars]

In another thread I asked about a substution for the FET in the input stage here, and a general consensus was that a 2SK389 should be fine as long as the servo was tweaked to allow for a larger correction range than the original.

However, the stage is oddly unbalanced... When first powering up, R741 and R743 (the 270-ohm resistors, part of the bias stage) began to get hot enough to smoke. To avoid burning up components whist troubleshooting, I removed Q709, Q715, Q713, Q717, and Q719. Figured at least I could see if the input was doing what it was supposed to.

Here's the deal: The collector of the right-hand side of Q703 is at about 62V as it should be...but the left-hand side is at about 52V.

WTH is going on here? I've been checking other components to see if there are resistors out of tolerance and so forth, but it all looks good. What's more, both channels are displaying the same symptoms....

[mirlo]

Check the critical matching of R713 and R715 (the 1K res in the emitters of Q705) and the two halves of Q705. Also maybe test or replace Q709/713 to make sure it doesn't have excessively low beta ... (?)

Hummm. But that wouldn't explain why both channels show the same behavior.

Are you sure the inputs are properly grounded?

Good luck!

Eric

[djmiddelkoop]

Check the voltage across R731.
I think it's not 0V cause the opamp Q721 is trying to correct the offset.
Short out R731 and make shure the other half is grounded the same way (short out R701) and check again

Dick.

[EchoWars]

Q705 is a brand-new 2SA995 dual transistor on both channels...I could see them being a little off, but the differences between the voltages are huge...

Q709/Q713 et al are removed from the circuit to ensure that they are not the problem...stated in first post...

I'll short out the resistors are see what happens...that won't smoke it, I hope...

BTW, I've pulled out the opamp at one time to see what it did...changed nothing so it was reinstalled.

[peranders]

Why have you changed the input jfet?

Have you changed to jfets with similar characteristics?

How much offset do you get?

You must be very careful with the jfets because they are connected to the current source.

[djmiddelkoop]

Q705 is a brand-new 2SA995 dual transistor on both channels...I could see them being a little off, but the differences between the voltages are huge...

Yes, any small offset or voltage differance between those two fet inputs can cause the unbalance you have.

I'll short out the resistors are see what happens...that won't smoke it, I hope...

Don't worry, as long as the components you mentioned are left out, nothing will smoke.

Dick.

[EchoWars]

Hi peranders..

Why have you changed the input jfet?

Had no choice...the originals were gone...as in, not even in the circuit. Someone before me completely removed them for some reason...and I had to find something that might work... I was suggested to use the 2SK389.

I'll short the inputs this evening and see what happens.

[EchoWars]

Was out of town for the last four days, so hadn't had a chance to check this thing in any more detail.

I shorted the gates to ground to see WTF happened. Made no difference. The collector on the left side of Q703 is still at about 50V, and the right at about 62V. An ohmmeter shows that the ground is indeed at ground potential, and I measure no difference of potential between chassis ground and the ground line on this board when power is applied.

Then I thought I'd carefully measure the voltages across the two 1K resistors on the emitters of Q705. They look pretty close, but I replaced the 1K resistors with a matched pair anyway. No change. The right hand side of Q705 is obviously saturated, wheras the left side has a good 10-11V from C-E. Also, I measure about -200mV on the emitter of Q703.

Seems like the key is to perhaps rebias Q703, or alter the current source. I can't find anything definately wrong, so I believe it's just that the 2SK389 is simply unhappy in it's new setting. What can I do to balance this thing out??

I know some of you all do this in your sleep...I'd appreciate some ideas on how to make this work....

[Tube_Dude]

Hi

Here's the deal: The collector of the right-hand side of Q703 is at about 62V as it should be...but the left-hand side is at about 52V.

But in a previus post you say

However, the stage is oddly unbalanced... When first powering up, R741 and R743 (the 270-ohm resistors, part of the bias stage) began to get hot enough to smoke. To avoid burning up components whist troubleshooting, I removed Q709, Q715, Q713, Q717, and Q719. Figured at least I could see if the input was doing what it was

So if the transistors are still removed the input stage do not have feedback for balanced DC conditions.

Without feedback the input stage can not be balanced...

[EchoWars]

But if I put the transistors in, Q713 and Q717 are turned on so hard that R741 and R743 will go up in smoke. Which is why they were removed....to see why the input stage is driving these transistors so hard and make sure that they were not influencing it.

The collectors of Q717 and Q719 go to the heatsink mounted bias transistor..which will at least pass a diode test. Is it possible that the bias pot was turned up so high that it was trying to smoke these resistors? I wouldn't think that it would be capable of that range. Still, I have turned the pot to zero ohms in preparation of reinstalling the transistors in this amp section, so the bias transistor should be cut off.

Feedback may be necessary to balance the stage perfectly, but is it really necessary in order to get it to within a volt or two? You are the first to suggest this... 10+ volts is a lot of imbalance.