OPA350: Measured output is double simulated (expected) output level

Part Number: OPA350
Other Parts Discussed in Thread: OPA1632, , TINA-TI

Tool/software:

Hi there,

I am using the OPA350 essentially to buffer and distribute a Common Mode voltage of 1.5 V from a codec to the AmpRef inputs of a bunch of OPA1632 op amps.

Firstly, I made an error and the output of the OPA350 was designed to be 3V (I can't recall why), however, when I measure the output voltage no my actual circuit I am reading 6.11V.  I simulated my circuit using the OPA350 in Tina-TI and it confirms that my output should be ~ 3.14V. How on earth is it measuring 6.11V on the actual implemented circuit? I'll include my Tina-TI file and the schematic of the implementation for review.

I need to understand why i'm getting an unexpected output level. I need to modify my original circuit to achieve a simple voltage follower arrangement, but can't do so if this current problem I'm experiencing will also need to be incorporated into my modifications.

Thank you for your support!OPA350PA circuit.TSC

  • Your circuit is clearly in the gain of 2 so the expected output should be 3V. Assuming that you actually power the circuit from 0 to 5V rails, the only way for the output to be above 6V is by pulling it up with an external source. If it is possible, please remove OPA350 from the socket and probe the output with scope or DMM in order to determine the source of the fault condition. 

  • Sure. What about the only way for the output to be above 3V? Any reason for that firstly? 

  • Yes, it was answered above: the only way for the output to be above 6V is by pulling it up with an external source.  This means that you most likely have some circuitry on your board that inadvertently shorts the output to a voltage source above 6V or your measuring instrument is the cause of the problem - it is impossible for the op amp to drive its output above its rail all on its own.  

  • Great, thank you. So you were suggesting that for any output greater than what was intended could be caused by this external source. I was just searching for clarification that it applied to any voltage output greater.

    I have found the old circuit diagram for the project, see attached, please could you review this and provide me some advice regarding possible locations that the output is being pulled up? I don't see any, and am obviously not experienced enough to identify the possible fault. Thanks again!

    AD1939 Codec testbed.pdf

  • Not quite correct: DC output voltage greater than its supply voltage can only be caused by an external source; dynamically, this may also occur when an inductive load is suddenly disconnected resulting in so called inductive kickback.

    the only voltage source above V supply used to power OPA350 is 12V regulator output, OPA_PWR - see below.  Thus, it is the only voltage source that could pull the output of OPA350 above 6V.

    Below I have circled OPA_PWR (12V) and AmpRef nodes coming in close proximity to one another and thus I would begin troubleshooting the problem by measuring resistance between these two nodes - ideally you should be seeing an open circuit and if it's not, you must find the source of the low resistance path.

  • Hi Marek,

    Thank you for your reply. I just need to clarify something. The OPA350 is receiving power from the 5V regulated output, not the 12V regulator output as you have noted. The 12V regulator output powers the OPA1632 differential op amps. Does this change your approach to solving this problem?

  • No, absolutely not.  To the contrary, it’s the main reason for the debugging approach. The fact that the OPA350 is powered from 5V supply while its output is above 6V makes me think that another higher supply voltage on the board must be responsible for pulling OPA350 output above its supply of 5V. 

  • Thanks again, Marek. I have also just measured the DC Voltage at pin 2 of my OPA350, it is reading 3.05 Vdc. Just wanted to add that confirmation into the mix. I'm trying to work out a simple way to remove the OPA350 to test. Being just a hobbyist, things are tricky and take time...  Slight smile

  • OK. So I need to go slow and ask for assistance here. With DMM set to 200Ohm, I get open circuit. But when I retested with the setting on 2kOhm, I am reading ~0.43 on the DMM. I assume this is what you're suggesting I needed to find. Can you please give me some tips on how I "find the source of the low resistance path"? Many thanks! Slight smile

  • Hi , does providing the top and bottom trace layers of my PCB assist you in pointing me towards likely culprits and test points? (I've removed the ground fill layers for clarity). Thanks!

    PCB_AD1939-Codec-testbed_TopLayer.pdfPCB_AD1939-Codec-testbed_BottomLayer.pdf

  • Are we still talking about circuit below with Gain of 2? (see below)  If so, please measure the voltage directly on pin 3 and pin 6.

  • Are you using DMM to measure voltage or resistance? If the latter, you must choose the proper range.  In order to debug the problem, I would suggest you remove the OPA350 and probe TP20. What voltage do you read? 

  • I will attempt to remove the OPA350 from my PCB in the morning and then perform the test you suggest. Thank you!

  • Yes, still talking about the same circuit.

  • Mark,

    Marek is on travel for the next few weeks.  I can help you while he is out.  If the result of your test shows the resistors to be both 1k as expected, I would suggest replacing the OPA350 with a new OPA350 (perhaps it was damaged during the debug).  The 6V output with a 5V power supply makes this scenario likely.  

    Best regards, Art