LM25116: Unwanted 2V pk-pk output with 3A load

Mark Graham

Part Number: LM25116

Hello

My design is a 24V to 5V regulator at 5A max. I see that the regulator goes into synchronous mode above 4.5A.

Below that I think it goes into diode emulation mode.

I attached some 'scope traces: Yellow is LO, Blue is switch node, Red is HO Green is 5V output.

At 5A the output seems quite stable. Switch T is 1.1uS ie 900kHz.

At 3A the output seems quite unstable. The output has 1.9V pk-pk and the switching is quite irregular. This amount of output swing is too much.

At 0A5 the output seems quite stable but the switching is very irregular.

Is there something wrong when I'm running at 3A? The board is demanding less than 3A most of the time, the 5A is just in case. I do not need to be very efficient and wonder whether it is best to force continuous
mode. I think I need to place a resistor between DEMB and CSG but I am not clear on the value.

I have attached the design and scope plots, I hope they are readable.

24V-5V@5A_40C_900kHz_webench_design.pdf

Thanks in advance

Mark Graham

over 8 years ago

0 Piotr F. over 8 years ago

TI__Intellectual 2850 points

Dear Mark,

Yes, from given schematics I can see that LM25116 is working in DEMB mode (DEMB pin pulled down to ground). With the use of an external DEMB resistor (RDEMB), the current sense threshold for diode emulation can be adjusted resulting in the gradual transition to synchronous operation

According to the datasheet full synchronous operation is obtained if the DEMB pin is always biased to a higher potential than the SW pin when LO is high (and the DEMB bias potential must always be kept below 2 V). RDEMB = 10 kΩ bias the DEMB pin to 0.45 V minimum, which is adequate for most applications. For more, please have a loot at section 7.4.1 "Soft Start and Diode Emulation" in LM25116 datasheet: http://www.ti.com/lit/ds/symlink/lm25116.pdf

Please let me know if system works properly in synchronous mode.

Piotr

0 Mark Graham over 8 years ago in reply to Piotr F.

Prodigy 245 points

Thank you for the reply.
Can you explain why there is such a large excursion from the expected 5V output? It would be nice to know my regulator will regulate for low current demand.
Possibly I can force synchronous 100% ot the time, but that still needs testing.
Thanks
Mark

0 David Baba over 8 years ago in reply to Mark Graham

TI__Mastermind 36120 points

Hi Mark,

Can we have a look at your layout? Thanks.

0 Mark Graham over 8 years ago in reply to David Baba

Prodigy 245 points

I'll attempt to attach a documentregulator_layers.docx

0 David Baba over 8 years ago in reply to Mark Graham

TI__Mastermind 36120 points

Looks like you are using Altium, can you send to me please? David.Baba@ti.com

Thanks.

0 Mark Graham over 8 years ago in reply to David Baba

Prodigy 245 points

For future reference:

The layout could be better, especially the size of the Switch node and its proximity to the controller IC. The Problem was fixed by changing the Cramp cap from 100pf to 68pf. After that, the controller switched to synchronous mode from 1A load upwards and the output remained regulated throughout the 0 - 5A load range.

Power management

Power management forum

LM25116: Unwanted 2V pk-pk output with 3A load