Idea for protecting DC-DC converters on laptop boards ?
Good day folks. So I've been struggling with laptop motherboards for a while now. Most of my repair attempts so far have failed one way or another, mostly due to a lack of tools rather than human error, but they did help me gain more and more experience and build up skill in the long run. This idea came from one of my failed attempts which you can read all about HERE.
To sum up, this Acer laptop had a dead MOSFET on one of its standby power rails (page 48 of the manual in the original post). The controller chip PU5 responsible for that rail was also shot. I replaced the chip but apparently failed to solder it properly, which, as I would learn later on, left the gate of the high-side FET PQ16 floating, causing it to conduct and send the B+ voltage directly into the 3.3v rail which destroyed the NB, rendering the board junk at that point.
I'm not sure if this would count as a design flaw or oversigth, since at the end of the day these are not meant to be serviced at component-level like I tried to do, but the way I'd design such a switching DC-DC converter, which brings us to the core of our discussion, is to add a 10k pull-down resistor to the gate pins of ALL the FETs which are otherwise floating (see the original vs. my drawing). This should prevent them from opening in case the controller chip fails or is not making contact like it happened to me, at least in the diagnosing phase until I make sure it's working - then they can be removed. In this particular case, pins 10 and 12 of PU5 would get a resistor to GND and also 19 and 21 on the opposite side of the chip for the 5v rail - same story there. Would this work or would it completely mess up the switching action and result in unwanted results ? Surely they would've thought of it (??), so perhaps it's just a bad idea on my part....cheers.
To sum up, this Acer laptop had a dead MOSFET on one of its standby power rails (page 48 of the manual in the original post). The controller chip PU5 responsible for that rail was also shot. I replaced the chip but apparently failed to solder it properly, which, as I would learn later on, left the gate of the high-side FET PQ16 floating, causing it to conduct and send the B+ voltage directly into the 3.3v rail which destroyed the NB, rendering the board junk at that point.
I'm not sure if this would count as a design flaw or oversigth, since at the end of the day these are not meant to be serviced at component-level like I tried to do, but the way I'd design such a switching DC-DC converter, which brings us to the core of our discussion, is to add a 10k pull-down resistor to the gate pins of ALL the FETs which are otherwise floating (see the original vs. my drawing). This should prevent them from opening in case the controller chip fails or is not making contact like it happened to me, at least in the diagnosing phase until I make sure it's working - then they can be removed. In this particular case, pins 10 and 12 of PU5 would get a resistor to GND and also 19 and 21 on the opposite side of the chip for the 5v rail - same story there. Would this work or would it completely mess up the switching action and result in unwanted results ? Surely they would've thought of it (??), so perhaps it's just a bad idea on my part....cheers.
Attached Files
But today, I’m making an exception here. Why? No idea. Perhaps only because the repair details are still “fresh” in my head… which is ironic, given this is a 16 year old monitor that hardly anyone will care about today. It is new to me, though. 
I picked it up last November from someone on my local Craigslist. It wasn’t very close to where I live, but was close to a family friend that I had to go visit anyways. So after watching the posting on Craigslist for a few weeks and seeing it getting...
Comment