Problem with LM25085.

Alon Goldstein

Other Parts Discussed in Thread: LM25085

I'm trying to build a buck DC-DC converter, using the LM25085, based on Webench design.

Input range is 14-24V, required output is 12V @ 3A max current.

The circuit seems to be working without load, but when I load it, the output voltage drops to much lower values - depends on load and input voltage.

I'm using the recommended components for all critical components. (I think)

What could be the problem ? (I already assembled 2 PCB's, and getting the same results, so I assume it's not bad assembly or defected chip)

Thank you.

Alon

over 10 years ago

0 JohnTucker over 10 years ago

TI__Guru**** 191445 points

Can you post your schematic and PCB layout?

0 Alon Goldstein over 10 years ago in reply to JohnTucker

Prodigy 115 points

Attached The Schematic & the PCB Layout.

The PCB is 2 layers. The Back layer(Green) is ground plane, except the 2 green lines shown.

Thank you,

0 Domonkos Baksa over 9 years ago in reply to Alon Goldstein

Prodigy 50 points

Any luck? I am having this exact problem. First I thought I thought I used incorrect current limiter resistances (Rsen, Radj), but after switching the latter to a much bigger one, the problem still exists. The circuit seems to be able to provide ~0.05A, after that voltage drops significantly.

0 Alon Goldstein over 9 years ago in reply to Domonkos Baksa

Prodigy 115 points

I'm afraid not.

I was very disappointing with the lack of support, after I uploaded the files as requested.

Finally, I had to give up on this...(first time in my life).

0 Domonkos Baksa over 9 years ago in reply to Alon Goldstein

Prodigy 50 points

I am really sorry. Are you still interested in this? I have to make this thing work today, hopefully I can figure something out.

Did you try altering the current limit? As a second try I short circuited the Rsen resistor, and got a little better result.

0 JohnTucker over 9 years ago in reply to Alon Goldstein

TI__Guru**** 191445 points

I am not surprised you are having difficulty. That layout does not look very good. Typically, you cannot connect the power traces with simple thin etch traces. You need to use copper fill areas. Yo can look at the EVM users guide to get an idea of how to lay out your PCB.

0 Domonkos Baksa over 9 years ago in reply to JohnTucker

Prodigy 50 points

I am not sure about Alon's layout, but I used copper fill areas for the power traces. I also don't think that the 0.06A current that my MC takes, should cause a ~3V voltage drop across any trace. Can you take a look at my design as well? I think the problem might be around the current limiting.

0 JohnTucker over 9 years ago in reply to Domonkos Baksa

TI__Guru**** 191445 points

I have alerted someone who supports LM25085 about this post. Let's see what he has to say.

0 Domonkos Baksa over 9 years ago in reply to JohnTucker

Prodigy 50 points

Thank you. Do you think he/she can get to it today? I am on a quite tight schedule with this uni project.

0 JohnTucker over 9 years ago in reply to Domonkos Baksa

TI__Guru**** 191445 points

He is not in my group or in my office, so I can't say for sure. I'll send him a follow on email.

0 Vijay86929 over 9 years ago in reply to JohnTucker

TI__Genius 11490 points

Hi Alon,

It seems John's comment are to the point. The current sense function of LM25085 may be triggering because of spikes and disturbances.

I suggest placing a CIN capacitor close to the power stage. Currently there is only on CIN cap and it is far away from the RSENSE-FET-DIODE path which carries switching current. When the converter switches, the whole rail probably droops and then recovers. This affects sensing which is dependent on VIN, VCC, ADJ, and ISEN rails.

For simple tests, I suggest manually placing a ceramic capcitor close to the RSENSE-FET-DIODE switching loop. If you see improvements then we probably are thinking in the right direction.

Thanks and regards,
Vijay

Power management

Power management forum

Problem with LM25085.