SMPS MOSFET replacement?

[RHeTTRoNiCS]

I've done googling...Majority telling only 3 important things must consider then it'll work:

1.) Voltage
2.) Current
3.) Rds-on resistance

Any opinions on this? Thanks.

 

[dick_freebird]

The first two ratings need to be respected. The last, you
might get away with worse at only the cost of high-load
efficiency. However light-load efficiency is mostly switching
losses and you call out no parameters there (Ciss, Coss
or gate charge). Do you care? I don't know.

It's also possible that reliability (since you are replacing
a device) is marginal for the original, and maybe you need
something better - higher rated, or more tolerant of abuse
(avalanche rated, etc.).

There's "it'll work" and there's "it'll work forever no matter
what". Plenty of ground in between to keep you chasing
whatever blew it up the last time.

 

[D.A.(Tony)Stewart]

I've done googling...Majority telling only 3 important things must consider then it'll work:

1.) Voltage
2.) Current
3.) Rds-on resistance

Any opinions on this? Thanks.

I agree with previous posts but would like to add the following concerns that ought to be checked.

- If significantly lower RdsOn, I would expect significantly lower heat rise, but as this always increases C_ISS if the gate driver is not low enough ESR, it would result in slower RC transition times and higher commutation losses. , yet lower conduction loss. Thus the gate driver R's may also need to be scaled down if possible. Temperature rise effects depend on circuit and layout.

-if significantly RdsOn, then commutation loss reduces but conduction loss when saturated rises proportional to RdsOn, resulting in linear temperature rise.

When the MOSFET has failed, one has to consider;
- was it a design fault? ( low margin from component variables)
- was it a construction fault? ( poor connections, etc)
- was it a user fault? shorted load in a low cost non fault tolerant design.

Then you can decide if you want to make modifications for future reliability improvements.

 

[RHeTTRoNiCS]

This SMPS was running <2yrs...And it was blown during normal operation. It could be an aging...am not sure though...So in summary, googling result is true? The 3 things to consider in finding alternative mosfets are voltage, current, and RdsOn...thanks

 

[FvM]

It's not necessarily the MOSFET that primarly caused SMPS failure.

 

[RHeTTRoNiCS]

Doesn't have much time really looking the reasons why mosfets blown. I am watching >20 machines running 24/7. The best thing I can do to speed up troubleshooting is to see blown parts and replace. Maybe in the later part can find time to analyse... But now I need help for the criteria looking alternative parts for mosfets in SMPS. So it there's no objection I will finalize that Voltage,Current, and RdsOn are the only main things looking for alternative/replacement for mosfets. Thanks again to all.

 

[D.A.(Tony)Stewart]

Doesn't have much time really looking the reasons why mosfets blown. I am watching >20 machines running 24/7. The best thing I can do to speed up troubleshooting is to see blown parts and replace. Maybe in the later part can find time to analyse... But now I need help for the criteria looking alternative parts for mosfets in SMPS. So it there's no objection I will finalize that Voltage,Current, and RdsOn are the only main things looking for alternative/replacement for mosfets. Thanks again to all.

MOSFETS have amny more specs that can pose a mismatch situation. for an unknown application.
1. Vgs range for desired RdsOn
.2 CISSS @ Vgs [pF]
3. Rth(ch-a) or Rjc or Rjl ['C/W]
4. td(on) td(off) [ns]
5. tr, tf [ns]

and on occasions a few more....

If you know anything about the product quality , reliability or the design, that might be suspect, then reliability improvements can sometimes require some quick forensics with experience.

 

[RHeTTRoNiCS]

Oh men! Looks like lots of homework then....So for the moment is there anyone here can help me find the mosfet replacement for 2SK1518 and APT5025BN...thanks a lot.

 

[RHeTTRoNiCS]

Hello...please help me verify if the IXFH26N50 can be direct replacement with 2SK1518....thanks