ADP1031 Problem: 2 Ohm Resistance Between SVDD and SGND Pins

Hi bahadirdonmez,

There were 2 SVDDs for the ADP1031, the SVDD2 is the input supply pin for the GPIOs and the SVDD1 is for the SPI channels. I believe that in your design, both SVDDs were tied to the same supply so if the SVDD2 have low resistance to SGND, the GPIO and SPI will not work but the regulators will still work okay.

For the problem that the SGPO have low resistance to SGND, is this consistent between the two units that you said that had been damaged? If so, it might not be ESD but and we need to double check the SGPO connection, perhaps there is some electrical overstress (EOS) happened on this pins - might be supplied by a voltage higher that its abs max rating. Can you tell what is connected to SGPIOs? Perhaps try to look at the scope the behavior of the SGPIO pins especially during startup, maybe there is a voltage going back to SGPOs during startup or on operation.

Are the SPI channels still okay when you do the diode check?

Thanks and regards,

Jefferson

Dear Jefferson,

Thank you very much for your response. Yes, in my case both SVDD1 and SVDD2 are tied to the same supply.

The first unit that was damaged had 2 Ohm resistance between SVDD and SGND. And it had around 5-6 Ohm resistance between GPO2 and SGND. Both of these problems were fixed when I replaced ADP1031. The SPI pins did not have any problems.

The second unit that I am testing has GPO1 tied to SGND with 2 Ohm resistance. Other GPIOs and SPIs are fine. What do you mean exactly with diode check? Should I put my multimeter in diode test mode and check the voltage between GPIO pins and SGND? When I do this, GPO1 (the pin that is shorted to SGND with 2 Ohm) shows 0V both when forward biased and reverse biased. The other GPIO pins and SPI pins show 1.6V when forward biased and 0.6V when reverse biased.

I just checked a third unit now. It also has GPO1 tried to SGND with 2 Ohm resistance, interesting... I had not even powered on this unit yet.

The GPO pins in the circuit are tied to 3.3V over a 10k resistance. So I think they cannot be overloaded. I will still check their voltage during power on with an oscilloscope and let you know.

Thank you very much and best regards.

Here is a scope measurement of the GPO2 pin that is working and not shorted the SGND. It jumps from 0V to 5.15V (I am driving the input MGPI2 with high voltage at the input side). But as I said GPO2 is tied to 3.3V over 10k resistor. This is a mistake that I will solve in the next version of the PCB. It should be tied to 5.15V instead of 3.3V. Do you think this might be causing a problem?

Hi bahadirdonmez,

Thank you for these additional information.

May I know how many units experience this kind of issue (SGPO tied to SGND)? 

The diode check you did was correct - we only need the reverse bias which we should read about ~0.6V for good units. Do you still have some loose units (not mounted on the board) that perhaps you can also check the diode for the SGPOs to see if they exhibit the same issue. 

Also, the failing unit that you removed from the board, when you do the diode check on the SGPO pin to SGND, does it still shows a shorted response?

Can you check for possible excessive solder around the SGPOs that might cause solder bridge to SGND. 

Thanks.

Hi,

I have 5 PCBs in total. 3 of them have had this issue (1 of them is now fixed after replacing ADP1031 chip, 2 of them still have the problem). 2 of them have never had this issue at all and are working without any problems. Furthermore, the ones that have this problem now, have initially not had the problem. They started to have this problem only after some time.

So I think if I tested brand new ADP1031 chip that are not mounted on the PCB, they would also not have any problems. The ADP1031 that I removed is unfortunately in the trash but  I will definitely test the other damaged ADP1031 chips as you suggested, once I remove them.

All the soldering was done during the manufacturing when we ordered the PCBs. When I check it with my eye, I don't see any bridges between the pins that may be causing this issue

Hi, 

Having the separate supply for the pull up and the SVDDs might be the cause. There will be instances that the abs max on the SGPO pin will be violated. How about just lifting the other side of the 10kohm pull up on the SGPOs and tie it to 5.15V, I believe in that config we will not see this issue. Or maybe removing the 10k pull up for now so that the 3.3V will not be seen. But how about the part connected to SGPOs, what voltage are they powered from, I assume it is 3.3V. We need to make both of them in the same supply voltage level.

Thanks.

The GPOs are driving the Reset and Clear pins of AD5791 DAC. This DAC has VCC set to 5.15V as you suggested. So the ADP1031 has 5.15V connected to its SVDD pin and DAC also has 5.15V. But my pull up voltage for GPOs was 3.3V (which was a mistake). Maybe this is causing the damage.

I think it is a good idea to remove 10k pullup so that 3.3V will not be seen. This way I hope the PCBs that are working now will not also be broken soon.

Hi Jefferson,

I removed the ADP1031 which I thought was damaged from the PCB. I did the diode check on the chip it self. The SGPO1 pin is indeed showing a shorted response on the IC itself. The resistance between SGND2 and SGPO1 is 1.75 Ohm and the reverse-biased diode voltage is 0V.

I also tested a brand new ADP1031. It does not have this problem.

It is still a kind of mystery what is damaging the SGPO1 pin. SGPO2 has very similar connections on the PCB but it seems to be not damaged.

Best regards,

Bahadir

If ESD wwas damaging it, do you think more than one SGPO pin would be damaged? In my case it is just SGPO1 that is damaged...