MusicMan 2275-130 Bias question

[Prattacaster]

I am afraid that 1458 is not the same as a 4558.

If this single IC is the issue I'd by delighted. I do not know the service history of this thing, it was just in to get a tune up. I wasnt expecting to find a biasing issue but thats what I've got. This has had a 4558 in there for a years. (I do know its been since the late 80's since things been looked at).

[Prattacaster]

I'm looking over the datasheets of the 1458 and 4558. Could someone school me on the most important differences. I do not see a bandwidth spec for the 1458. According to the 4558 charts it seems the voltage gain would be close. I need some schooling.

[Enzo]

4558 will OFTEN work in place of a 1458, maybe even usually. What matters is the times it won;t. I don;t know where all those won;t instances are. I am looking at this case though. I see the problem as unbalanced bias currents. I also see the op amp as driving the two sides through capacitors. And so there is no DC connection between tube drive and op amp, so other than some RF instability or something, I don't see how the choice of op amp could upset the balance in idle DC currents between push and pull. Hell, you could probably remove the 1458 from its socket and the amp should sit there biased OK.

I don;t recall the differences, seems to me the open loop gain of the 1458 was lower than the 4558, and in SOME circuits it was enough difference to lose stability. I long ago started just stocking both types, and when I have a bad 1458 in a circuit, I replace it with a 1458. Why waste time trying to sub other parts when I don;t have to?

How does a driver measure base to ground 463 ohms but R65 measures 241 ohms? The difference in the two readings ought to be only 3.9 ohms. And a transistor ought not read a few hundred ohms resistance. Remember, when we read those 470 ohm parts, it is ACROSS each part, NOT to ground.

[Enzo]

4558 will OFTEN work in place of a 1458, maybe even usually. What matters is the times it won;t. I don;t know where all those won;t instances are. I am looking at this case though. I see the problem as unbalanced bias currents. I also see the op amp as driving the two sides through capacitors. And so there is no DC connection between tube drive and op amp, so other than some RF instability or something, I don't see how the choice of op amp could upset the balance in idle DC currents between push and pull. Hell, you could probably remove the 1458 from its socket and the amp should sit there biased OK.

I don;t recall the differences, seems to me the open loop gain of the 1458 was lower than the 4558, and in SOME circuits it was enough difference to lose stability. I long ago started just stocking both types, and when I have a bad 1458 in a circuit, I replace it with a 1458. Why waste time trying to sub other parts when I don;t have to?

How does a driver measure base to ground 463 ohms but R65 measures 241 ohms? The difference in the two readings ought to be only 3.9 ohms. And a transistor ought not read a few hundred ohms resistance. Remember, when we read those 470 ohm parts, it is ACROSS each part, NOT to ground.

[Prattacaster]

How does a driver measure base to ground 463 ohms but R65 measures 241 ohms?

Impossible right? Haha, Like I mentioned before, the components are not labeled on the board so I probably have R68 and 65 switched with R58 and R59.

Remember, when we read those 470 ohm parts, it is ACROSS each part, NOT to ground.

Yes, I know how to read a components resistance; obviously the lack of PCB lettering is making me look foolish. There are 6 - 470ohm resistors crammed into a very tight space making it difficult to keep everything straight. I should have drawn it out. Anyway nothing stuck me as odd. The resistors either read full value or half value. I will double check, maybe even uninstall the components and test.

[Enzo]

I mentioned that because you mentioned the ground connection. A ring of resistors could be grounded at any one of the nodes, and it would not effect what each measured in circuit... if they measure different, it would be because of other parallel paths.

[Prattacaster]

I've got a couple LM1458's on the way, even if I find a drifted 470ohm resistor. I like to bring this back to original specs if the parts are still available.

Hell, you could probably remove the 1458 from its socket and the amp should sit there biased OK.

I'll take you up on that one. As the amp sits it is working fine, The bias, although mismatched, is steady. I could let it slide but I can't; I wont let mysteries go unsolved. I will report back once I find the culprit.

[Prattacaster]

Hell, you could probably remove the 1458 from its socket and the amp should sit there biased OK.

I took you up on that because it made good sense.

The amp biases perfectly without IC-8 installed. I'd say that a 4558 chip, at least the ones I have, makes the bias go a little wacky. I have a couple 1458's on the way to confirm.

The amp biases much more hot and lop-sided with a JC4558 installed at IC-8.

Good Call Enzo, Thanks for all the responses so far.

[Enzo]

Just for your education, go look up the peavey Heritage VTX or Classic VTX or any of their VTX series amps. Very similar cathode drive. Look at the notes. lower right corner. It talks about voltage RANGES in expected operation.

[g1]

Might be worth changing those caps (C35 & C36). Or at least swapping them to see if the problem follows.

[woodyc]

I am afraid that 1458 is not the same as a 4558.
https://www.national.com/ds/LM/LM1458.pdf

Could you explain the difference? They look interchangable for audio -- general purpose with the same pinout. Isn't the 4558 just a 1458 with a larger bandwidth?

[Prattacaster]

Update:

The LM1458 fixed the problem! There was a 4558 chip in the amp- making the amp bias very unbalanced; more than a 30mA bias offset, between the power section sides. I ordered too of them and both of them produced an 8mA offset. This offset is acceptable and designed to reproduce the natural PI gain offset of tube amp.

EDIT: Woodyc pointed out an important typo my this last posting had. The bias offset before the chip change was over 30mV not mA.

So let me correct myself here; The mV readings are now 26mV and 32mV for the emitters. This is within 3mA dc offset. This was after I swapped in a LM1458.

The e-caps are all being changed. All bias and voltage setting components are also being replaced.

Thanks for the help.

[woodyc]

Update:

The LM1458 fixed the problem! ...

I'm glad you solved your problem, but...

- In your first post, you said the bias difference was 6.7mA. [(62mV-36mV)/3.9 Ohms] Now you say its 8mA. How is that fixed exactly?

- As Enzo pointed out, the opamp has no effect whatsoever on the bias. If it does, you must have a problem with C35 and/or C36.

[Prattacaster]

[QUOTEAs Enzo pointed out, the opamp has no effect whatsoever on the bias. If it does, you must have a problem with C35 and/or C36.][/QUOTE]

I understand what your saying however I just changed the e-caps and had the same voltages throughout as before, with the original e-caps. I am getting about .58vdc on the positive side of C36 and C35.

As far as the bias goes it is great. Less than 2mA from side to side. Haven't got a chance yet to hear it with the new tubes and filter caps. Tomorrow.

[Enzo]

Which side is the positive side? The schematic shows no polarity. Look at the circuit, those caps are connected to the driver bases. The drivers are conducting, to their bases need to be turned on. You have a +16v supply at D17, which feeds through R53 and TR1 and is held at least to one junction drop by D16, and THAT sets the bases, TR1 being the bias control. I expect a half a volt of so there for that reason. The left side - the output pins of the op amp - I expect to sit at about zero DC.


That is what those two caps are for: isolating two points at different DC voltages, while passing signal between them.