OPA2316-Q1: OPA2316 is damaged easily

Dam Hung,

Does this circuit share a ground with the AMC1200 circuitry?
Does the this get disconnected and reconnected often (plugged and unplugged)?

I suspect ESD (people touching) or ground shift. (separate grounds at different voltages).

This is my AMC1200 circuit.
After I replace new OPA2316, it work fine for mins. Then, the output of OPA2316 starts to float or reach +Vcc or -Vcc. I think the problem is relate to EOS. I've try place some zener diodes but it does not help!

I am thinking about change OPA2316 to another opamp!

Thanks! 

Dam Hung,

I don't know what might be happening. I do suggest measuring the voltage on all the TLV2316 pin to make sure they are as expected with a new sample. Does the IC get hot before or after failing?

Dear,
the opamp chip is not hot at all. Its output just unstable! If I re-solder it, it will works for a time, then the problem comes again!
T.T I am so tired of this!

Dam,

I believe the problem you see is caused by the violation of the AMC1200 minimum High-side power supply voltage of 4.5V - see below.

The AMC1200 datasheet specifies the VDD1 power supply range of 4.5V to 5.5 V but according to your schematic you power it from VCC=3V supply.

You may swap the two available supplies by powering VDD1 from PDS1 of 5V while powering VDD2 from VCC of 3V.

Dear Marek !

The high side of AMC1200 is isolated to the control circuit. So it isn't a reason. The low-side I supplied by 3.3V, the outputs of AMC1200 have Vcm = 1.25V. So I believe that the input voltage is not exceed the rail voltage!

Thanks!

Dam,

It does NOT matter that the high-side is isolated - if you violate the minimum supply voltage on the high-side, it will effect the performance on the low-side.

You may easily verify it by swapping the two available supplies by powering VDD1 from PDS1 of 5V while powering VDD2 from VCC of 3V.

Dear Marek!

Sorry for make you confusing! My schematic is wrong with the library of AMC1200. Please swap the side in schematic! I already swap it in PCB by rotating the AMC1200 chip upside down, it is my mistake when drawn the schematic.

However, the problem is not AMC1200 output. AMC1200 works well!

Thank you so much for your concern!

I suspect an open solder connection on pin 3 or pin 5.

Dear,

I've try so many ways to fix this but nothing works (Try use limit resistor at input pin, zener diode at Vss, increase or reduce the gain resistors.)

The 1st opamp is always saturated at -Vss even input voltage is 0V. is there any error in my schematic? Please help me, I stuck in this situation whole month!!!!

Thanks! 

I built your OPA2316 circuit and have it running for a 3 hours and it works fine.
I tried disconnecting pin 3 then tried pin 5 and the output drifted to the rail over a couple minutes.

Dear,

I've try to measure the input and in+, in-, output of OPA2316 with this circuit: Input is generated by waveform generator.

Here is the result of the input and in+ of U1A: RED is input, blue is in+

Input and output of U1A:

Then I tried remove R7 to separate U1 and U2. Here is the result: Input and IN+

Input and output

It shows that if connect with U2, the output and input has some level shifted and noises. 

Can anyone explain this. This happen after I turn on the circuit again.

last night it stills work fine!

Thanks!

Hello Dam,

Please send us the complete updated/corrected schematic and oscilloscope pictures of the input pins and output pins. There is also something else connected to U1B's feedback path that is not shown. Also, please try disconnected each amplifier stage to ensure that the inputs/outputs of each stage are as expected. For example, I would remove R5 and then measure InN, InP, and pin1 of U1 to ensure they are as expected. If not then you may want to disconnect the AMC device and provide known inputs (either dc inputs or try a differential signal function generator). I am unsure of exactly where InN and InP connected because they are only found on the amplifier schematic. If U1A is working properly then re-install R5 but remove R7 and perform the same analysis. Please let us know the results. Please be sure to use a known good device...you can sample them on ti.com.

Dear,

This is my schematic for this test:

I used a generator to generate the input signal.

1ST: I remove both R5 and R7:

 red color is input signal, blue is the in+, in- and output (they are the same as I tested) of opamp U1A.

The output of U1B is zero, the output of U2 is 1.24V. That means U1B and U2 work fine.

The problem here is the input of U1A is shifted below zero from -0.5V to -1.5V

Yesterday it stills work fine. Today I turn it on and it becomes like this. May be after I re-solder the schematic, it will work again!

 

Please tell me the problems.

I've built a PCB with 4 channels as below and it work fine for other channels:

Thanks!

Dear all,

I've tried to simulate the schematic in TINA-TI software to solve my problem. I figured out that if I add a resistor between Vin+ and Vss-, the result is similar to my problem:

 

So I wondering, is there any EOS problems, ESD problems that make this happen: has small resistance between Vin+ and Vss-?

If yes, please let me know how to prevent it!

Thank you all so much!

Hello Dam,

Thanks for the additional information. Let's just focus on U1A for now.

It appears as though it should now be connected in a simple buffer configuration, therefore the output should be equal to the input. The input appears to be a ~2.5Vpp sine wave with a frequency of approximately 1.7kHz. Also, since it appears as though you have a 50ohm (probably 49.9ohm) termination resistor please make sure the output impedance of the function generator is set properly.

The output of U1A appears attenuated, shifted, and has some interesting spikes and noise. It also appears to shift after the first few cycles.

This appears to me as though you have a grounding or reference issue. I do not necessarily think that this is an EOS/ESD issue. To rule out EOS/ESD simply sample a new device and solder it to the board.

If you have a blank PCB laying around I highly recommend checking the impedance (aka 'ohm out') the bare board (if not, please remove U1). In other words use an ohmmeter and check the impedance between each pad of U1 and its corresponding connection. Use the schematic as your guide. This will help you find any unintended opens or shorts on the PCB. I am particularly interested in making sure that the positive and negative supply references (or grounds) are properly connected to a known reference voltage (0V).

Due to the floating nature of this issue I suspect that this is ultimately a grounding issue. Please make sure that the side of the board that connects to the output of the AMC1200 has its own proper ground connection somewhere.

Concerning the re-creation of the issue with the simulation...you essentially created a summing amplifier that adds a -1.25V bias to the input, thereby shifting the output down by 1.25V. Since this is not in the original circuit I do not expect it to be the cause of the issue.

Dear,

This is my schematic of power parts:

Is there any problems with them? I've checked the output voltages result and it is good.

The P12 is the main power source, comes from outside source. The +-15V is used for current sensor LA-55P, 3.3V is used for DSP launchpad and final opamp state, +-2.5V is used for sensor signal conditioning. 

I've checked the blank PCB, it has no problem. 

I am wondering: OPA2316 can operate well with dual supply +-2.5V?

Another thing is: the IN+ and IN- of opamp has a dual diode connected to the rail power to protect the input pin (internally), is there any change that these diodes keep on and off, so that the IN+ pin floated under zero?

Thanks for your help!

Dam,

For LM317, LM337, and LM1117 support I suggest starting a new forum thread. Personally I support general purpose amplifiers, and those are power devices.

Yes, the OPA2316 can operate with a +/-2.5V supply.

The ESD diodes will only start conducting if the input goes more than 500mV above or below the supply voltages. If they start to conduct and the signal source supplies 10mA or more they can become damaged. I have been under the assumption that you have tried new samples/devices to eliminate the possibility of ESD/EOS issues.

Unfortunately I have run out of ideas for you to try. Given that 3 of the 4 channels work properly I still suspect a PCB layout/soldering/grounding issue with this particular channel. Sorry I can't be of more help. I would continue to look for any differences between this channel and the other 3 that work.