Opamp runnig hot

[r2rein]

Hello everyone,
I made an attemp last days to replace an old NJM2068 opamp from my Denon PRA-1500 preamp, with some newer (an better sounding) opamps like the LT1028 , but the problem now is that the new opamp is running hot, too hot to touch, tough it seems to work fine. I also tried NE5534 which runs normally. I attached also the part of the schematic that was modified.
Any suggestions would be well appreciated.

[Fast Eddie D]

http://140.120.11.1/prophys/ael/File...eet/lt1028.pdf

Looking at the schematic, the first thing I noticed was the voltage supply of +/-17.2 V. So I looked at the product sheet and maximum specs are as follows:

-55C to 105C +/- 22 V
105C to 125C +/- 16V

which I guess are typical specs. If your op amp is running too hot to touch it is surely being run past its absolute maximum voltage rating. While there are many other things that could cause your problem (but since it is unity gain stable they shouldn't), if you really want to run this chip, you might try knocking the voltage down a few volts. Or you could find a chip that runs fine in the circuit.

[Joachim Gerhard]

Try to put a 0.1uF film cap or good quality stacked ceramic from PIN4 to PIN7 of the LT1028 as close as you can. This Opamp is much faster then the one you used before and the heat is a clear sign of oscillation. Maybe you have to play also with the compensation cap over the feedback resistor.

[lineup]

Originally Posted by Joachim Gerhard

Try to put a 0.1uF film cap or good quality stacked ceramic from PIN4 to PIN7 of the LT1028 as close as you can. This Opamp is much faster then the one you used before and the heat is a clear sign of oscillation. Maybe you have to play also with the compensation cap over the feedback resistor.

This is true.
LT1028 is not the easiest to get stable at lower gains.
It is a very good opamp. Especially for low noise and Phono RIAA amplifiers.
But personally I would not use LT1028 for 'normal' gains applications.

Joachim Gerhard's advice is probably the correct.
To try and tame the opamp and get it stable.
You may find hints about this, among the Applications Information
in the datasheet: http://cds.linear.com/docs/Datasheet/1028fa.pdf

[r2rein]

Thanks for the answers.
measured supply voltage is around +/- 16V so it is fits the specs (-55C to 105C +/- 22 V).
I also thoght about the fact the opamps are oscillating. I will try out the solution with the 0.1uF caps.

To Gerhard: do you mean Pin 4 and Pin 8 (across the supply rails) ?

What other opamps would be appropiate using in this case?

[AndrewT]

pin4 & 7 are supply rail
pin1 & 8 are input balance/comp

[Frank Berry]

Are the pinouts the same? The schematic shows pins 4 & 8 as the supply rails.

[AndrewT]

the datasheet rules.

[Bill_P]

The NJM2068 is a dual op amp and the LT1028 is a single. Lots of the pin functions do not match.

[deepthrob]

I doubt that a 5534 worked since it's a single and the schematic shows a dual. It's hard to tell the purpose of this op amp from the schematic but I'd bet it's running the optoisolator that is part of the bias circuit. Is the entire amp running hot? If so, put the original back, there's nothing to be gained by changing it since it's not in the signal path.