TPS2080 not Setting the OC Pin to Low when current exceeds 1A

We have seen this behavior before with parts with an external current limit set pin, but not with a fixed internal part. The issue with the externally set current limit will occur when the GND connection from the Rilim resistor is not properly routed/connected back to IC GND.

Your issue could be related to a GND connection or layout related issue. Can you try the same test using the EVM? Assuming the OC flag works properly on the EVM then you could simply swap the part from your board to the EVM and prove whether it is board or IC related.

Hi Eric,

thanks for the reply. Haven carefully gone through the design, I was not able to detect any possible ground issue.

In your reply, you stated parts with Internal/External OC-Limit functions. In our case it is the TPS2080D, whereby the External Limit function would not be applicable! Or am I getting something wrong here?

I do not have the EVM and do not intend to buy one, moreover there is no difference in our Schematics and that of the EVM.

Perharps you could find the ground issue that  I was not able to detect in our Schematic, here is a piece of the Schematic.

I will appreciate further Support, thanks

Hi Richard,

Our EVM uses a 33kohm pull-up for the /OC pin (8). On your schematic you have a 22kohm, but it is labelled as do not populate (DNP). Is this the case on your actual board? A pull-up is needed, so can you try adding that in and testing?

As a side note:

I noticed your output capacitors have a 6.3V voltage rating. Given that this is a 5V rail, I would suggest using 10V caps on the output (same as our EVM) for a more robust solution. In case you are unaware of capacitor derating due to a DC bias, please see the following threads:

http://e2e.ti.com/support/power_management/non-isolated_dcdc/f/196/p/145858/526880.aspx#526880

http://e2e.ti.com/support/power_management/non-isolated_dcdc/f/196/p/86844/300762.aspx#300762

Both of these threads are a different use case than yours, but both somewhat explain output capacitance calculations and the DC bias effect. The result is, using 100uF, 6.3V caps with 5V applied may yield ~10-20uF of effective capacitance depending on your capacitor specs. You can view the derating curve provided in your capacitor's datasheet

Also, here is an article from Sparkfun on the subject:

https://www.sparkfun.com/news/1271

Do you have an oscilloscope available to take a look at the signals? Can you post a screen shot of an overcurrent condition with the /OC signal?

If you are still having issues, I would suggest ordering and testing on our EVM.

Thanks,

Alex

Hi Alex,

First I did like to thank you for taking some to check the Schematic and giving a Feedback.

You are right, the Pull-Up is labelled DNP. However, we are using an IC that has an internal pull-Up Configuration. on the other hand, while trying to get the OC-Issue solved, I have populated the Resistor and this did not lead to any significant difference.

I have an oscilloscope and I am using it. Only the internal current limiting feature is seen on the oscilloscope. This is the reason why I asked the question earlier "If the OC-Feature was tested? "

With almost no hope of getting any good News, I will Change the 22KR to a 33KR Resistor to see if I will get any better result.

Once again thanks for showing some concern.

Richard

Hi Richard,

Richard Enajite said:

This is the reason why I asked the question earlier "If the OC-Feature was tested? "

Our device has been in production for 7 years and has been tested and used by our customers such that the OC signal functions as described in the datasheet. It should closely resemble the graphs demonstrated in Figures 10 and 11, on page 11 of the datasheet.

Can you describe this in more detail:

Richard Enajite said:

Only the internal current limiting feature is seen on the oscilloscope.

How did you see the internal current limiting was working? Are you saying it limited the current on the output, yet the OC signal did not pull low?  Did you verify the current was being limited using an ammeter?

Richard Enajite said:

I will Change the 22KR to a 33KR Resistor to see if I will get any better result.

As you know, changing the pull-up resistor should not affect the results. In order to narrow down the problem, we need to isolate the variables.

• How are you using the OC signal, is it going to an MCU?

•  Is it going to a solder down or a socketed IC device?

• If socketed, try removing the MCU/IC, use the external pull-up resistor and connect an oscilloscope to the OC pin to see if it pulls low in an overcurrent condition.
• If solder-down, is this a prototype board and are you comfortable cutting the trace between the MCU and our TPS2080 device? Then you can use the external pull-up resistor and make sure that the OC signal goes low in an overcurrent condition (using an oscilloscope).

Thanks,

Alex

As a side note:

In the future, we highly recommend using our evaluation modules (EVMs) to test our devices and verify they are performing as you expect before designing them into your system. We spend a lot of time and resources developing these evaluation platforms and strive to offer them for nearly every product in our vast portfolio and make them available at an affordable price-point.

Hi Alex,

Thanks for still spending some time on this issue. However, i must confess to you that i havent had the time to probe further on the possible bug on these boards.

A)  " How did you see the internal current limiting was working? Are you saying it limited the current on the output, yet the OC signal did not pull low?  Did you verify the current was being limited using an ammeter?"

Right Alex, by saying only the ínternal limiting feature functions, i meant the current on the Output and yet the OC signal still remains high.  And  yes i did the verification with an Amperemeter.

B)    As you know, changing the pull-up resistor should not affect the results. In order to narrow down the problem, we need to isolate the variables.

• How are you using the OC signal, is it going to an MCU?

•  Is it going to a solder down or a socketed IC device?

• If socketed, try removing the MCU/IC, use the external pull-up resistor and connect an oscilloscope to the OC pin to see if it pulls low in an overcurrent condition.

The OC Signal is going to a USB2517 chip, are there any irregularites known using this combination?

C)    If -down, is tsolderhis a prototype board and are you comfortable cutting the trace between the MCU and our TPS2080 device? Then you can use the external pull-up resistor and make sure that the OC signal goes low in an overcurrent condition (using an oscilloscope).

Lucky enough it is still a Prototype and cutting off the trace of one of the boards would not be a Problem. I will get back at you a soon as i get further in the search. fornow,  please bear with me that i now have a task with a higher priority.

Thanks, Richard