LM324-N: Absolute maximum power supply

Hi Anand,

1000V? 

Kai

Hello Anand,

Snapping the device in half is an easier way to make sure it won't work anymore. 

1000V would work, and so would 100V or -32V

Could you rephrase your question with your application details?

Sure. I've purchased a product with a supply voltage rating of 24vac +/-25% . When I connected a transformer with an output of 28vac, the device failed.

When traced the circuit,  I noticed that the half wave rectified voltage is connected to the supply terminals of LM324N.LM324N has an absolute max rating of 32V in single supply operation. When I pointed out this to manufacturer, he said he has been selling this for years and never had such a problem. I am curious whether the manufacturer is right or not. Can LM324N really survive more than 32V?

Hi Anand,

Anand P S said:

When I pointed out this to manufacturer, he said he has been selling this for years and never had such a problem.

Which is the standard answer of a customer which feels caught...

A transformer delivering 28VAC shows a peak output voltage of 28V x 1.414 = 39.6V. Also, transformers always come with a "no load output voltage" which can be considerably higher. The "VB 0.5/2/24" from Block, e.g., shows a "no load output voltage" of 180%. Such a transformer would generate a "no load peak output voltage" of 41.4V. And then 110% mains voltage fluctuation should be kept in mind, which gives up to 45.5V...

Looks like the manufacturer has made a mistake.

Kai

Anand,

Line powered switching supplies are very low cost, small and light. Why are you still using a heavy and bulky transformer? I save all my left over power bricks. A few years ago, I got rid of my direct line transformers as I consider them dinosaurs.  

Consider adding a regulator after filter and before op amps.  

Hi Ron,

Using an SMPS is not an option in this case and I certainly cant make changes to a product designed by someone else. I just want to figure out what percentage of what could be the failure rate If buy a lot of this product. 

Was it just a mistake or a brilliant way to cut costs? I am pretty sure its a successful product in his market. May be people are using it with good transformers which gives 24V. I used a cheap one with poor regulation. 

Hi Anand,

can you tell more about the product? Is it sold without power supply? And the customer has provide the power supply?

Kai

Hi kai,

Product comes without power supply. Its specification says 24V ac/dc +/- 25% . I don't want to disclose brand name or model. I hate troubles. I am just curious to know why the op amp works above its absolute max rating.

Hi kai,

Product comes without power supply. Its specification says 24V ac/dc +/- 25% . I don't want to disclose brand name or model. I hate troubles. I am just curious to know why the op amp works above its absolute max rating.

Hi Anand,

24VDC + 25% would give 30V, which is below the absolute maximum rating of LM324-N of 32V. So this would be no problem. But 24VAC + 25% would definitely be a problem.

I think there has been made a mistake in the specification. It should not say "24V ac/dc +/-25%" but "24V peak or 24V dc +/-25%".

The absolute maximum rating must never be exceeded, not even briefly.

Kai

Even when the device is not immediately damaged, overvoltage causes electromigration and will degrade the electrical characteristics.

I guess that the circuit does not actually require the exact characteristics of the LM324-N.

Its not a mistake on the datasheet. I've confirmed it with the manufacturer. I even tried applying  36VDC at the power supply. The voltage across the supply pins of the op amp was only a diode drop less than the power supply. The product still works.!! 

Anand,

Could you change the op amp? TLV9304 is 40V

Hi Anand,

of course will the LM324-N survive a supply voltage of >32V, for a certain period of time. Or do you think that there's a switch in the OPAmp which kills the OPAmp, when the supply voltage is 32.00001V?

No, you need lots of headroom for wafer to wafer manufacturing tolerances. Also, the absolute maximum ratings of 32V additionally covers the performance at higher temperatures and other insane operating conditions like short-circuits etc. Maybe at room temperature and without short circuit the OPAmp might withstand 35V for an hour or so. But do it at elevated temperatures and force a short lasting short-circuit at the output and the OPAmp will immediately be destroyed. A torture it would have safely survived with a supply voltage of <32V.

Also think in terms of statistics. 100% of chips will survive 32V, but only 70% might survive at 35V and 10% at 40V (just a guess). What abolute maximum rating would you give in the datasheet? 40V?

No, a limit is a limit is a limit. You might want to exceed the supply voltage over 32V. Ok. But do it on your own risk. Nobody will give you any guarantee then.

Kai

Ron, 

Thank you for the suggestion. But I don't work for the manufacturer. I am supplying this " standard product" to my customers. If it was my choice I would have used a Zener(or zeners)with a series resistor to clamp the supply to 32V. I just want to have a rough idea of number of units that may return as faulty.

Yes. In my designs I will never go below or above the recommended specifications. I prefer a boring life :)

Information about the exact behaviour above 32 V is probably not known, and even if it were, it would not be published.

Could you correct the specification? Or clarify that "VAC" means peak-to-peak?

Its rms. Normally transformer ratings are mentioned in rms values. 

Its rms. Normally transformer ratings are mentioned in rms values. 

Anand P S said:

I prefer a boring life :)