DAC7750: Devices becoming damaged during testing, become low resistance from power to ground

Hi Christopher,

My colleague will get back to you on Monday.

Thanks,
Paul

Hello Christopher,

I have a few questions to better understand this issue:

1. Can you provide the schematic for review?
2. What kind of testing are you doing and at what point does the device fail?
3. Can you provide a scope capture of AVDD/DVDD at power on? This may be related to power-on such as a fast AVDD turn-on.

Thanks,
Garrett

Note that the inputs come from optical isolation powered by 5V derived from the 30 power. We have a certain limit on the amount of capacitance allowed in the circuit related to certain standards we are complying with, which explains the small power filtering values. Note the direct connections of device pins to an interface connector.

The testing we have been doing is to put 1K ohm from the output to ground, and check voltage to confirm the current level when commands are sent over the SPI. Performance has been without problem until the part fails, which is seen at some subsequent power-up and the power supply limits output because of over-current.

I will capture power supply rise scope image shortly.

Hi Christopher,

Is this measured at VDD_CUR_LOOP? Since the bypass capacitors are small I would expect a faster ramp but this may be due to the power supply used.

The capacitors C88 and C89 are smaller than I would suggest and may lead to a transient on AVDD/DVDD when the output switches. I am not certain this is the cause of the failure though.

Do you connect/disconnect the load resistor with the output enabled?

Yes, that was measured between VDD_CUR_LOOP and GND_CLP. The power up is fairly slow because that is the behavior of the bench power supply.

We did not connect and disconnect the load resistor with the output enabled. When ever failure has been noticed, it has been at power on.

Hi Christopher,

So this failure will happen just power cycling and not enabling the DAC7750 output? If you can reliably reproduce the failure, I would suggest changing C88 and C89 to a larger capacitance value. Maybe add a 1uF in parallel to each and repeat the test to see if the failure still occurs. I realize you are intentionally limiting the capacitance. Your schematic is pretty straight forward so I do not know what else could be the issue.

Thanks,
Garrett

Hi Christopher,

Any update on this thread?

Thanks,
Garrett

Next week we are going to add some transient suppressors (like those in the DAC7750 eval board) to an undamaged board

and see if it can operate continuously and go through power cycles without failing.

Hi Christopher,

Understood, let me know the results.

Thanks,
Garrett

We placed transient voltage suppressors (ST Microelectronics part number SMBJ30A-TR) between AVDD and ground, and between IOUT and ground.

We did this to five boards and have been testing them thoroughly. So far there have been no more damaged DAC7750s. We seem to have a solution.

If the TVS components have solved our problem, perhaps it was due to ESD events, or perhaps misbehavior of the power supply we were unaware of.

A sixth board was given the same modification, and then we unintentionally connected 30V power hot to the circuit. This destroyed the DAC7750 despite the presence of the TVS devices.

Is the DAC7750 known to be subject to damage from high dV/dt? Is it incompatible with a "hot-swap" situation without filter capacitors to buffer dV/dt of AVDD?