TPS3808-Q1: Device self-oscillating?

Hi Mikael,

Is this the first time for this application?  Has there been other applications with this issue?  Is this a new application?  

I see a couple of components on the schematic that I can suggest the customer to changed.  For example, is there a reason for the 10Kohm resistor to VCC33 on /MR pin? The /MR pin has an internal pullup resistor of 90Kohms.  The pullup resistor R324 can be increased to 10Kohms.  Is there a 0.1uF capacitor close to the TPS3808 VCC pin?  

One thing to note, if the monitored voltage rail VCC12 is noisy, there is a big potential that the oscillation could be coming from VCC12 being noisy.  In the datasheet, it is suggested that the SENSE pin have 10nF of bypass capacitance on it.  If the monitored rail is very noisy, my personal preference is to use 0.1uF on the SENSE pin as the bypass capacitor.

I hope my suggestions will help with the customer's issue.

Ben 

Hi Mikael,

Just curious if you have any updates.  If there are none and you are satisfied with the answers that I provided you, please click on "resolved" to close the thread.  Thanks!

Ben

Hi Ben,

Sorry for the late reponse!

Yes, it's the first time, and so far only, we see this behaviour. It's not a new application, about 7-8years, and no previous known similar issues related to this component. I don't have the full history regarding these pull-ups, but you are right. There are 0,1uF caps already close to both VCC and SENSE pins.

The VCC12 voltage is not noisy so it's still a bit of a mystery.

Normal behaviour: While SENSE is below threshold, CT pin remains low(0V). When SENSE gets above threshold CT-pin makes this cap charging curve for 174ms and then directly to 0V.

Therefore these recurrent 174ms cap charging curves looks more that the TPS3808 was in an infinite loop, than a result of a noisy voltage. Question is what could possibly get the TP3808 into such infinite state?

Best regards

Mikael

Hi Mikael,

This is the first time I have seen the TPS3808 do this.  You mentioned that you were not able to repeat the issue, correct?  It would have been great to get some scope shots of the SENSE pin when this issue occurred.  My only guess what caused the TPS3808 go into an "infinite loop" was the VCC12 rail was slightly going below the threshold voltage (1.12V) and going above 1.2V but as you mentioned the VCC12 rail is not noisy and has the bypass capacitors.  Besides that, I dont know what caused the "infinite loop" response.

Ben