AD797 overheating (ANZ flavor)

[michouthefirst]

I am testing an AD797ANZ in a filter circuit and I noticed that on quiescent state the amp is drawing around 60mA from the PSU (+/- 15V). This makes it heat because it has to dissipate 2W.

I built a simple circuit of inverting amp (2 10-K resistors, 0V input, 0V output) and it still does the same.

I have a bunch of 10 which I bought off eBay. They all do that.

Is this normal?

 

[JonSnell Electronic]

This is not normal. I would tell your supplier that they are faulty and ask for a refund. Always buy from a trusted supplier like Mouser, Farnell, RS Components and the like.

 

[Michael Bean]

That chip is very sensitive to layout and PS bypassing issues which can make it prone to HF oscillation, and it's not really the best choice for unity gain operation. I personally don't like to buy parts from unknown/untrusted sources, lots of fake components out there. Having said all that, what kind of filter were you trying to build and what do you want it to do? A schematic would be helpful.

Mike

 

[jean-paul]

Open door: how is the power supply decoupling arrangement ? As posted by other members AD797 is very sensitive and I have seen them oscillating more than once.

 

[michouthefirst]

The power supply is a lab power supply (Agilent). I think the parts are probably fake. This happened to me a couple of times. The sellers had a very positive feedback.

I will try decoupling the power supply and see how it goes.

For the filter, it was in fact a tone control circuit. I am new at DIY audio and still at the learning stage.

Thanks all for your answers, really appreciated.

 

[jean-paul]

Decoupling the supply pins straight at the AD797 would be wiser.

 

[macboy]

Decoupling the supply pins straight at the AD797 would be wiser.

Absolutely necessary, actually.

It seems that the OP didn't bother to even look at the AD797 datasheet other than to drool over the low noise and distortion figures. It spells out some very specific recommendations for stability, among them are some strict decoupling requirements.

 

[jan.didden]

Please post a schematic. This is groping around in the dark. There are many ways this can happen. For instance, is the 60mA from both rauils or only from one rail? How hot does it reaaly get, just hot to the touch or too hot to keep your finger on it?

Jan

 

[jan.didden]

Absolutely necessary, actually.

It seems that the OP didn't bother to even look at the AD797 datasheet other than to drool over the low noise and distortion figures. It spells out some very specific recommendations for stability, among them are some strict decoupling requirements.

This is too harsh - we don't even know what the cause is! You judge without having a clue.

Jan

 

[michouthefirst]

What are you suggestions for capacitors values for effective decoupling?

I tried 10uF with 100nf straight at the pins, same result. Please bear with me as I am a noob.

 

[davidsrsb]

A photo would be helpful. Lead length is very important on decoupling capacitors.
Looking at the datasheet step photos, this chip is only marginally stable at unity gain

 

[macboy]

This is too harsh - we don't even know what the cause is! You judge without having a clue.

Jan

In retrospect, a little harsh. But he did say...

I will try decoupling the power supply and see how it goes.

Which I interpretted as meaning that the chip had no local decoupling and he was relying only on clean power from the PS. It is a common enough mistake for a newbie especially since many simplified schematics don't show the decoupling (it's assumed) or even the power supply connections.

In any case, the reading the applications information in the datasheet for AD797 is a little scary. It frequently both states and implies that instabilty can be easily achieved.