Replacing VRM components

[davmax]

I responded to another forum in terms of understanding the implications when replacing VRM caps and MOSFETs.

The detail is attached

[starfury1]

Re: Replacing VRM components

Well No guru on VRM design

This is true davmax if you change some parameter in someway it has to have some other effect on circuit behavior.

If this is detrimental to circuit operation or not is another question but if you don't know the implications its best to stick were possible to original specified parts.

Your paper does present an argument against the idea of blowing out the VRM output capacitance (stuffing all holes with some biggies) and sticking closer to the original total value capacitance or more specifically not pushing the ESR too low.

Thanks Davmax for the pdf

Cheers

[PCBONEZ]

Re: Replacing VRM components

Thank you Davmax.

Much needed info here.

Been getting a lot of flack (from some) in trying to tell people you can't just slap any ole cap in as a replacement or radically change values and this doc explains a lot better than I do why.

.

[Brian C]

Re: Replacing VRM components

Thank you Davmax.

Much needed info here.

Been getting a lot of flack (from some) in trying to tell people you can't just slap any ole cap in as a replacement or radically change values and this doc explains a lot better than I do why.

.

It is true... The MOSFETs in VRM circuit are usually lower gate charge type. MOSFETs part no. with the suffix L.

[davmax]

Re: Replacing VRM components

Thanks for that input Brian C. L is the usual addition to the end of the part number. I have noticed a trap in that one MOSFET I came across had LL and I first assumed it as a logic level device, however checking the data sheet it was not. Hence my warning, check the data sheet.

Changing topic.

My document delivers a very simple concept to deliver a message that there are hidden dangers in component replacement. In reality the problem is massively more complex and I plan the document more to give some idea how complex and the problems designers face. This should broaden understanding.

I guess the fundamental message is to not meddle with design unless the problem and solution are fully understood.

[PCBONEZ]

Re: Replacing VRM components

I guess the fundamental message is to not meddle with design unless the problem and solution are fully understood.

EXACTLY
EXACTLY
EXACTLY

And if you don't wanna do the math then don't change it.

[davmax]

Re: Replacing VRM components

As promised I attach an expanded version of the complexity within VRM filtering.

Same message. Beware of changed characteristics of components.