If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

# OPA561 gets hot when E/S pin is connected thru a 402k resistor to +12v

Hi Marek,

I am definitely connecting +V to +12v and -V to GND (0v) so there should be no confusion on this.

I was originally connecting +12v thru a 402k resistor to the E/S pin.

In your initial email you advised me that this superceded the +5v maximum rating of the E/S pin

and you sent me a copy of that specification.

So then I used a voltage divider instead to affect a 2.39v at the E/S pin thru a voltage divider of 402k and 100k

from +V to -V with the center node connected to the E/S pin. Of course as I mentioned I read a higher voltage than expected which you attributed to the internal current drain of about 20ua. If you multiply 20ua x 100k = 2v. Hmmm.

Does not sound right. Putting that peripheral aside I don't see a problem.

However, you surprised me when you said

"If this does not work, it is most likely that the part has been damaged.  Connecting E/S directly to 12V would damage the part instantly but connecting it thru 402k resistor assures enough voltage drop across it so the E/S pin is within the voltage range of 2V to 5V on 0V to 12V power supplies."

I never connected the E/S pin directly to +12v. Only thru a 402k resistor. I have since measured the voltage at

the E/S pin with this connection (no 100k pulldown) and it read 5.3V. However this is indeed over the 5v maximum spec. So, is this bad or good? The part still gets hot.

Upon further study I noticed that although the Basic connection shows the 402k pullup resistor to V+.

In single supply operation which I am using thedatasheet range is limited to +7v to +15V. However the Output Enable/Status specification shows  a maximum VE/S High of (V-) +5V.  This spec seems to imply that

+V cannot be connected to the E/S pin in single supply operation since the minimum single supply voltage is listed as +7V  (> 5v)!

The data sheet is confusing. Can I ssume then that according to your latest post that if I use a 402k

ohm series resistor pullup that I need not worry about this since it limits the voltage to 5v (even though I am reading 5.3v  which is over the spec???? Even so the thermal problem is not solved.

Another question: The Flag pin is mentioned in the datasheet but no pin number is assigned anywhere.

I assume that it is the thermal pad which must be connected to -V which is the case. But why do they then allude to it as a "FLAG" pin without any pin number? Why would you call a thermal pad "FLAG" anyway?

Sorry for any  confusion.

Les

• Hi Les,

What is the output doing? Is it in the correct state? Does it show a large offset? Is it OSCILLATING? Is the supply current symmetrical on both supplies?

Most digital scopes will not show high-frequency oscillations unless you are specifically looking for them. Analog scopes would show "fuzz", but digital scopes can alias and a high frequency oscillation may just look like noise or a "rumble" - or nothing at all. Turn off averaging, and turn on "real time" (or "one shot") mode,  peak detection and full bandwidth and AC trigger, then start looking starting at the highest sweep speeds and going slower. You are trying to see a 10-100MHz signal at low amplitudes. An analog scope is still best for debugging oscillations.

My experience with these devices is that they will run warm just from quiescent current, but running hot with no load is a sign of either a high frequency oscillation or damage - or both. The outputs can also be damaged by inductive kick-backs or anything else that would pull the output beyond the supplies - even for just a few nanoseconds. Do you have the output clamping diodes as recommended in the datasheet?

If there is "damage" you will usually see a large change in either offset voltage (becomes 10's to 100's mV), and reduced output swing and increased supply current on one or both supplies.

Though the GBW of the device is fairly low, the BW of the output stage devices is in the 100's of MHZ. It is not uncommon to see oscillations in the 10's to 100's of MHz. A high frequency, high amplitude oscillation, plus load capacitance, equals a LOT of current.

Oscillations can also be caused by improper bypassing and ground currents. The audio amplifier guys have been haunted by this since he dawn of time...Do you have proper high-frequency and low frequency supply bypass caps? (as shown on the datasheet using a 0.1uF ceramic and 47uF tantalum - you cannot scrimp here...). Do the bypass caps and the input common share the same ground? You MUST use a star ground here. The input ground reference point, the load return, the bypass caps ground and supply ground should all meet at one point and should never be shared. This is critical for high-power devices since ground I/R drops can be considerable.

Regards,