Some data:

Vin ~24VDC/12VDC Vout ~12VDC with worst case 4A peak current (short) or around 2A continuous. Parts: PMOS Schotky Inductance

Board is put into a switch cabinet with no direct air. Only passive. Ambient temperatures can be up to 40°C (wc)

For my project I am trying to design my own PMOS buck via LM25085 so I can get ~100 % duty cycle. The input voltage is either 12 or 24 VDC and it regulates itself to ~12 V (11.# something for 12 V in but that is fine). I took a inductance for a worst case current of 4-5 A but that will never happen unless something is wrong. I took all the parameters from the TI Webbench for the needed components so I think they are fine.

Now I am at the part where I want to put it onto my PCB. I followed the guidelines as much as I can. enter image description here "The first loop is that formed by CIN, Q1, L1, COUT, and back to CIN. The second loop is that formed by D1, L1, COUT, and back to D1." I also tried to make these possible with the following configuration. (In my picture C16 is Cin, C15 is Cvcc)

Here is the full schematic of the regulator enter image description here

My question now is the following: I see that in the first picture there is just 1 big rectangle Vin that directly connects to Cin and Cvcc and Rsen while using a trace to get to the other? Is that correct? Does that mean I need to make a big shape and try to get all these components into the shape while all the other are fine with just a trace?

For the trace : I would like to now what trace width should I use here as it's not said in the layout. I am usually using normal 0.250 mm trace width but with the relative high current that thing can generate should I change all my trace width where the voltage/current is higher to a bigger width? (30 mills or something)

And last cooling: do I need some cooling or would that work?

  • \$\begingroup\$ Why the C17 and C18 are needed. Why you put all those capacitors in series??!! \$\endgroup\$ Jun 7, 2021 at 9:14
  • \$\begingroup\$ These are Cff from the example in the datasheet but the TI Webbench also told me I need 7.4nF. But yes obviously that's wrong they need to be in parallel likewise Cout(C19-21) \$\endgroup\$
    – Daniel Do
    Jun 7, 2021 at 9:20
  • \$\begingroup\$ everything around the mosfet needs to be close. There's two loops - the charge and the discharge loops. These need to be small and done with thick tracks - 60thou or more. Why in the hell are you using through hole resistors on what is ostensibly a smt board? You're just making it harder for yourself. Do you need cooling? We don't know because you haven't told us much. What is the output current? What are all the part numbers? What is the expected environment? \$\endgroup\$
    – Kartman
    Jun 7, 2021 at 10:24
  • \$\begingroup\$ Added parts. I am using THT because in the first place this should have been a THT board with just a few smd components. I could argue that smd caps are much better because of the low ESR for the buck. Is there any advantage for SMD resistors other than size ? I.e. I read that THT resistors are better for power/heat conversion \$\endgroup\$
    – Daniel Do
    Jun 7, 2021 at 11:49
  • 2
    \$\begingroup\$ When you edit the question you should leave the old information, then you add EDIT and attach new schematics, PCB,.. Now you got answers and comments that have no meaning since your post has changed 100%. \$\endgroup\$ Jun 9, 2021 at 23:00

1 Answer 1


I'd be more concerned that your basic circuit is wrong: -

![enter image description here

Look at the direction of the bulk diode inside the MOSFET.

Better view: -

enter image description here

  • \$\begingroup\$ Oh dang! I messed that one up because I have a ton of NMOS on my project. Thanks a lot ! \$\endgroup\$
    – Daniel Do
    Jun 7, 2021 at 12:27
  • 1
    \$\begingroup\$ @DanielDo I think you need to start taking account of the goodwill shown in giving you free help. I'm not asking for money of course but I am asking you to take the 2 minute tour to understand the motivation behind people giving their time freely to help you solve your problems. \$\endgroup\$
    – Andy aka
    Aug 8, 2021 at 15:24
  • 1
    \$\begingroup\$ @DanielDo additionally you should not have completely evolved your question like you did because it makes my valid answer appear stupid. When a question is correctly answered and is pointing out an embarrassing mistake, you should take it on the chin and accept the answer given. I am assuming you are going to put this right fairly quickly. \$\endgroup\$
    – Andy aka
    Aug 8, 2021 at 15:27
  • 1
    \$\begingroup\$ @DanielDo if you have a new question then raise a brand new question and don't overwrite the original making already-provided answers seem completely inappropriate. \$\endgroup\$
    – Andy aka
    Aug 8, 2021 at 15:29
  • \$\begingroup\$ @DanielDo good work dude. \$\endgroup\$
    – Andy aka
    Aug 8, 2021 at 15:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.