Discussion I have fried two motherboards and maybe more I have no clue what it is.

[sebba241]

Components:
PSU: HDPLEX 250W GaN
FIRST CPU (used): AMD Ryzen 5 8500G
SECOND CPU (new): AMD Ryzen 7 7600
FIRST MOBO: Jginyue B650i Night Devil
SECOND MOBO: ASRock A620I Lightning WiFi
COOLER: Thermalright AXP90-X53

Detailed Timeline:
Initial Incident: Upon first use (by first use I mean simply connecting the cable to the wall), the PSU appeared to fry my brand-new motherboard. Assuming the issue lay with a second-hand CPU, I replaced both the CPU and motherboard with brand-new, tested components.
Second Incident: Before reconnecting the PSU, I tested it with a voltmeter in DC mode. All pins returned the correct voltage as per HDPLEX documentation, leading me to believe the PSU was not at fault. However, upon powering on the system with the new components, the same issue occurred. (Again, I did not even get to press the power on button since smoke started coming out seconds after plugging in the power cable to the wall)
What I thought was the issue: I realized the PSU was powering the motherboard even when its power switch was turned OFF. This prompted me to test in AC mode, revealing 26W and 18W on two different pins, even with the PSU switched off.


I've contacted HDPLEX about this and this is their reply:

My engineer team said this is induced voltage which is perfectly normal and does not carrier current.
Attached is a short video where my engineer is measuring the induced voltage from a regular ATX PSU.

So it seems that the PSU works as it should.

The only common denominator between tests has been the cooler, could this be causing the issue?

Thanks

 

[Jace]

https://www.amazon.com/Thermalright...t-reviews/B08S6XM69Q?reviewerType=all_reviews
Second customer review talks about mounting screws being too long and that they could cause a short-circuit frying the motherboard.

 

[vinnyoflegend]

What I thought was the issue: I realized the PSU was powering the motherboard even when its power switch was turned OFF. This prompted me to test in AC mode, revealing 26W and 18W on two different pins, even with the PSU switched off.


I've contacted HDPLEX about this and this is their reply:


So it seems that the PSU works as it should.

The only common denominator between tests has been the cooler, could this be causing the issue?

Thanks

Err, what pins were you (and HD PLEX) measuring that revealed that voltage? I just tried testing an ATX PSU plugged in and no pins returned any voltage that reached even 1v in AC mode. And only 5VSB/PS0-ON measured 5v in DC mode to ground.

 

[DASBOOT]

You did not list what PC case you are using or the material that may be touching the heatsink.
1. You should have tested out of the case with a real ATX PS and verified that the board has no issues. Before moving to HDGaN
2. The screw problem could be the issue (measure standoffs to know what clears, put Kapton tape on the metal area of the case that could conduct to prevent any problems, or future problems.)
3. It would be a good idea to purchase this tester: https://www.amazon.com/Thermaltake-...age-AC-069-OO1NAN-A1/dp/B0CZ2YF2FR/ref=sr_1_1
This is the only way to test a power-good circuit. This will also show if any pin locations are mislocated at molexs.

 

[sebba241]

Err, what pins were you (and HD PLEX) measuring that revealed that voltage? I just tried testing an ATX PSU plugged in and no pins returned any voltage that reached even 1v in AC mode. And only 5VSB/PS0-ON measured 5v in DC mode to ground.

In my case it would be this two pins. On HDPLEXs end I don't know but they did provide a video which I can show a picture of, is to note that they are using a different PSU.

You did not list what PC case you are using or the material that may be touching the heatsink.
1. You should have tested out of the case with a real ATX PS and verified that the board has no issues. Before moving to HDGaN
2. The screw problem could be the issue (measure standoffs to know what clears, put Kapton tape on the metal area of the case that could conduct to prevent any problems, or future problems.)
3. It would be a good idea to purchase this tester: https://www.amazon.com/Thermaltake-...age-AC-069-OO1NAN-A1/dp/B0CZ2YF2FR/ref=sr_1_1
This is the only way to test a power-good circuit. This will also show if any pin locations are mislocated at molexs.

I am using a Lzmod a24-v5, first time I tested it was in the case and second time it was outside. Will give a try to point 1 and 3. Thanks

https://www.amazon.com/Thermalright...t-reviews/B08S6XM69Q?reviewerType=all_reviews
Second customer review talks about mounting screws being too long and that they could cause a short-circuit frying the motherboard.

Good catch thanks.

 

[vinnyoflegend]

In my case it would be this two pins. On HDPLEXs end I don't know but they did provide a video which I can show a picture of, is to note that they are using a different PSU.

Very strange, I'm not sure how that's possible, do you get the same on your other non-HDPLEX PSU? I rechecked multiple PSUs I have and only see 5V DC as expected on +5VSB, and otherwise 0v if trying to check for AC.

 

[DASBOOT]

I believe he said watts, not volts... in post #1 right or wrong..

revealing 26W and 18W on two different pins, even with the PSU switched off.

 

[vinnyoflegend]

I believe he said watts, not volts... in post #1 right or wrong..

So assuming he measured current and multiplied voltage x current, that's 3-5a of draw on a 5v VSB line. I'm unable to see anything like that with any of my power supplies in AC or DC mode. I'm pretty confused by what HDPlex is showing in that picture as well then. Am I crazy here?