LM25085 dropping output at cold

Hey John,

Webench design creation assumes at room  temperature so the generated BOM may not be rated for cold temperature operation.  However; Webench does allow the user to specify a desired operating  temperature at the input design panel.  You can specify the  temperature at -40 degree C to allow Webench taking into account that your design needs to operate at below the room temperature.

Also, this device offers three different output ripple options from the Advance Option Panel, I would suggest selecting the Low Output Ripple solution to minimize the impact of Cout ESR value for cold temperature operation.  

Let us know if this helps.

Regards,

Howard Chen

Webench Application Team

Thanks Howard,

I appreciate the prompt response. We are going to implement that low ripple configuration and I'll let you know how it goes. We did specify the temp range when originally designing the circuit with Webench, however we did not utilize the low ripple version.

Thanks and Regards,

John

Thanks Howard for the suggestions. I am adding myself in the loop with this post, so I can keep track of the development.

Good morning Howard/Vijay,

The low ripple configuration did not work. In addition we changed all our Caps to ceramic on Vout in all distributed points on our board to reduce ESR but still had no effect. We are testing the unit after cold soaking for 30 -45 mins at -40C.  Any more suggestions would be welcome.

Thanks and Regards,

John

Hey John,

Are you using the Q grade of the IC? Only the Q grade is temperature tested that guarantees the datasheet specifications.

What about the other components on your board (PFET, Diode, Inductor); are they also rated for -40 degree C operation? 

I'll refer you to Vijay and maybe he can come up with more suggestions for you.

Regards,
Howard

Hi Howard,

Initially we were not using the Automotive part(Q grade), however we had replaced the original part with the Q grade part and no change occurred. The rest of our circuit parts are all rated for -40 C or lower.

Regards,

John

Hello John,

The part should work at -40C. Can you share the schematic or the webnech design?

Thanks and regards,

Vijay

Hello John,

Have you tried observing the SW node, VIN, and IL. I am just trying ot make sure that the input is not too low, and that the part is not simply in drop out. Sharing those waveforms will be useful in determining the root cause of this behaviour.

Thanks and regards,

Vijay

Hi Vijay,

We have observed the Vin node each time. It has never dropped below 4.97V at cold. Our nominal Vin is 5V. We have observed the output Vout switching when the part is cold. The waveform stays positive for 7mS and then drops down for 3mS. When it drops off its like an inverse sawtooth not a square wave. I will share the webench design with you in a short while. Does anyone have an answer to the Current overhead question posted earlier?

Regards,

John

Silly question probably, but how do I share a webench Design on this forum? The share option only allows E-mail or branch code of which I have neither as the e-mails come from noreply@.....

Regards,

John

Hey John,

Sharing a Webench design does require an email address.  Here is my email:  Howard.chen@ti.com and I'll forward your design to Vijay. 

Probe the Gate node, does it also exhibit an inverse sawtooth waveform?

Regards,
Howard

You can paste pictures on this post. That is if you choose to do so.

Vijay, I have forwarded the design to you.

-Akshay

Thanks Askhay. I will look at the design you shared. I was, infact, asking for some experimental waveforms from the user, since on the surface there appears to be nothing wrong in the original design.

Vijay and Askhay,

I never received any response from you guys after sharing the WebBench design. We resolved the issue by replacing all the caps at Vout to ceramic and modifying the circuit configuration to a 1A output from 3.5A. This resolved the immediate problem we were having with your part but does not answer the question as to why this was occurring with our original design. The original question in my opinion is still open and I would like a more detailed explanation for us to continue to use this part in current and future designs.

Thanks and Regards,

John

Hello John,

Thank you for sharing this information. I was looking for some experimental waveforms to understand the problem better. Can you shoot the original and the new circuit? If I understand correctly, you changed the current limit setting (1A from 3.5A). Use my email id:

vijay.choudhary@ti.com

Regards,

Vijay