This thread has been locked.

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.

INA1650: High failure rate in commercial product

Part Number: INA1650
Other Parts Discussed in Thread: DRV135

Hello, I have a commercial audio product that uses INA1650IPWR for input line receiving. I have noticed that approximately 5% of the products have been returned for repair within 1 year of purchase. The problem is usually one channel of the INA1650 not passing audio anymore.

The products are most often used with unbalanced cables, which shorts one side to ground through the Input/Output jack. The relays are used for true-bypass. The device is a set of audio filters, but I have simplified the schematic a bit for this request.

Can you see any design flaws that would cause such a high failure rate?

INA1650 circuit.pdf

  • All 10uF capacitors are 50V X7R

  • Hi Benjamin,

    I suspect that switching the inputs into your signal source can produce large transients at the input node. A quick simulation shows that switching a 10V source can result in ~20mA of peak current through the input terminals, which is a violation of the abs max.

    INA165x_Switched_Inputs.TSC

    Can you provide more information about the failures? What is the input and output voltage for a failed unit?

    What is the voltage range of your signal source? Is this connected directly to a microphone? Is there phantom power applied?

    Regards,

    Zach

  • Hi Zach, thanks for your reply.

    I suspect that switching the inputs into your signal source can produce large transients at the input node.

    Do you mean because of the hard switching of the relay? Do you have any suggestions to clamp these transients?

    What is the input and output voltage for a failed unit?

    Do you mean when I test it myself on my bench? Unfortunately I don't have a broken one here at the moment, but it effectively goes "silent" when it breaks. Usually it happens late at night in a nightclub, so getting good information from customers about what happened is hard.

    What is the voltage range of your signal source?

    It's a commercial Audio FX product that is connected to DJ mixers, so it's hard for me to say the conditions under which it breaks. The two most popular DJ mixer models only specify "nominal" values of –12 dBu (Pioneer V-10) and –2 dBu (A&H Xone96) for their SEND output respectively. Pioneer has their own FX Unit RMX-1000 which has a "nominal" input of +4dBu.

    Is this connected directly to a microphone? Is there phantom power applied?

    I think it's very unlikely that customers are connecting it to a microphone or anything with phantom power.

    Thanks

    - Ben

  • Hi Ben,

    I see that there are many other devices in series with the INA1650 in your audio signal chain. How do you know it is the INA that is failing? I believe the channel could go "silent" if any of these devices are damaged. Are you able to acquire a damaged unit to verify? Are you able to acquire a working unit and try to replicate the fault condition while monitoring the INA1650 input and output?

    It is generally a good idea to include protection diodes at the inputs pins of a device. This is especially true if the input pins are connected to the outside world where a user can connect anything directly to the input pins through the TRS connection. Keep in mind that your C1 10μF capacitor only blocks DC voltages. Any voltage that is switched in through the relay or plugged in "hot" at the TRS jack is a fast transient voltage that passes through the capacitor as a short to the input pins. I see you have overvoltage protection diodes included at the output of your DRV135 which is connected to a another switching relay and TRS jack similar to the INA1650 inputs.

    Do you have any suggestions to clamp these transients?

    Here is a great article detailing input overvoltage protection schemes: https://www.ednasia.com/provide-robust-input-overvoltage-protection-for-amplifier-analog-input-modules/ 

    However, I would suggest the first step is to verify the failure mode on a damaged unit, and successfully replicate it on the bench. It's possible that the channel "silence" is caused by something completely different, in which case you may be revising your board with a protection scheme that does not fix the issue.

    Regards,

    Zach

  • Hi Zach,

    I see that there are many other devices in series with the INA1650 in your audio signal chain. How do you know it is the INA that is failing?

    The problem is definitely a broken INA1650. Basically a device comes back for repair, and I trace the signal through to see that it disappears at the INA output (Pin 8 or 13). After I replace the INA, it works fine.

    I don't have any of the fully broken units here, but I did some digging, and found 4 old problematic PCBs with noisy INAs on them, and I found a consistent problem. Pin4 or Pin5 shorted to the negative rail (different values between 7Ω - 660Ω). They still work, but at half volume when driven with a differential signal, and with quite a bit of white noise on the signal.

    I still can't tell you the exact conditions that cause it to break, but It looks like it's probably the high input current right? Would you suggest series resistors and then Schottky diodes to the rails? I'm also open to suggestion for other parts that might integrate better.

    Thanks,
    - Ben

  • Hi Ben,

    Based on the information given, it does seem that an overvoltage and/or overcurrent condition is violating the absolute maximum ratings of the INA1650 and causing the device to be damaged.

    Would you suggest series resistors and then Schottky diodes to the rails?

    Yes, including current-limiting resistors and Schottky diodes will help to protect the inputs of the INA1650. It is difficult to recommend specific resistor values as we do not know the magnitude or characteristics of the fault condition. As shown in the simulation below, 1kΩ series resistors will limit the input current to ~10mA for a 15V input transient. If you expect a larger input transient is possible, you may scale your series resistance appropriately.

    INA165x_Switched_Inputs_Protection.TSC

    Regards,

    Zach