Find the value of the source resistance atx BM-650W ?

dongtrien

n00b
Joined
Mar 23, 2020
Messages
3
The power supply as the picture shows, the resistors lose color code, do you have the color code of resistor values R08, R09, R10, R11 to share with me, you see the attached image
BM650A.jpg
BM650B.jpg
 

GiGaBiTe

[H]ard|Gawd
Joined
Apr 26, 2013
Messages
1,199
That power supply is garbage, it needs to go in the trash, it's a fire hazard. There's a reason those resistors failed, it's a trash design.

- It uses diodes instead of a bridge rectifier, making it likely a half wave design.
- It has no X2 or Y capacitors (has positions for Y but no X2 caps on the PCB.)
- There's no input filter choke (despite having a position for it on the PCB.)
- The transistor and mosfet heatsinks are a joke, they're paper thin and won't properly cool anything.
- The output filtering is also a joke with a tiny combined toroidal choke that makes independent voltage rail regulation impossible in cross loading.
- It appears to use Teapo or some Chineseium junk capacitors, and far too few of them, making output regulation even worse.

Buy a quality power supply from a reputable manufacturer like EVGA, Corsair, Cooler Master, Antec or Thermaltake.
 
Last edited:

ryan_975

[H]F Junkie
Joined
Feb 6, 2006
Messages
14,735
That power supply is garbage, it needs to go in the trash, it's a fire hazard. There's a reason those resistors failed, it's a trash design.

- It uses diodes instead of a bridge rectifier, making it likely a half wave design.
- It has no X2 or Y capacitors (has positions for Y but no X2 caps on the PCB.)
- There's no input filter choke (despite having a position for it on the PCB.)
- The transistor and mosfet heatsinks are a joke, they're paper thin and won't properly cool anything.
- The output filtering is also a joke with a tiny combined toroidal choke that makes independent voltage rail regulation impossible in cross loading.
- It appears to use Teapo or some Chineseium junk capacitors, and far too few of them, making output regulation even worse.

Buy a quality power supply from a reputable manufacturer like EVGA, Corsair, Cooler Master, Antec or Thermaltake.

It looks like a lot of components have been removed, so there's no telling what was missing... but it's definitely NOT a safe power supply.

Your point about the bridge rectifier is wrong. Four diode makes for a FWBR. A single component FWBR is just four diodes in a package for convenience and cost.
 

GiGaBiTe

[H]ard|Gawd
Joined
Apr 26, 2013
Messages
1,199
Your point about the bridge rectifier is wrong. Four diode makes for a FWBR.
Four diodes does not automatically mean full wave rectification in SMPS.

SWER-115-caps.gif

In old SMPS designs without PFC, only two of the diodes are used as a voltage doubler circuit in 115v mode, meaning it's a half wave design. The PSU the OP has apparently runs at 230v, which would technically run in full wave mode, but it doesn't make it any less of a garbage design.

A single component FWBR is just four diodes in a package for convenience and cost.
The only reason four diodes are used is because of cost cutting, it is in no way convenient to mount four individual components and have the extra traces and pads for them. It also means that you will get nowhere near the rated current of the supply. In 115v mode, common garbage designs like this only use 2 or 3 amp rated diodes, meaning a maximum current of 230W / 460W (115/230 for 2A) or 345W / 690W (115/230 for 3A) diodes. This stage alone already means the label of it being a "650W" PSU is a lie, because these diodes will fry long before you get even close to the rated wattage of the supply. PSUs like these are almost always in the 50-70% efficiency range, meaning at 650W, you'd be pulling 1.1kW+ from the primary.

But we already know this supply would explode long before that because of the paper thin heatsinks.
 

ryan_975

[H]F Junkie
Joined
Feb 6, 2006
Messages
14,735
Four diodes does not automatically mean full wave rectification in SMPS.

View attachment 236210

In old SMPS designs without PFC, only two of the diodes are used as a voltage doubler circuit in 115v mode, meaning it's a half wave design. The PSU the OP has apparently runs at 230v, which would technically run in full wave mode, but it doesn't make it any less of a garbage design.
Right, but I wasn't saying four diodes is always an FWBR, just that the presence of them doesn't automatically make for a half-wave. The same thing can be said for a single component BR. In an uncorrected PSU running in 115v mode (where 230v operation is an option), a single component BR will run in half wave mode too.

The only reason four diodes are used is because of cost cutting, it is in no way convenient to mount four individual components and have the extra traces and pads for them. It also means that you will get nowhere near the rated current of the supply. In 115v mode, common garbage designs like this only use 2 or 3 amp rated diodes, meaning a maximum current of 230W / 460W (115/230 for 2A) or 345W / 690W (115/230 for 3A) diodes. This stage alone already means the label of it being a "650W" PSU is a lie, because these diodes will fry long before you get even close to the rated wattage of the supply. PSUs like these are almost always in the 50-70% efficiency range, meaning at 650W, you'd be pulling 1.1kW+ from the primary.

But we already know this supply would explode long before that because of the paper thin heatsinks.
I said the single component BR was for convenience and cost. Not the four diodes.

A 2w rated diode operating 50% of the time can safely handle 4a. So they could handle 460W/920W safely. A 3w diode would be safe up to 690W/1380W.
 

GiGaBiTe

[H]ard|Gawd
Joined
Apr 26, 2013
Messages
1,199
A 2w rated diode operating 50% of the time can safely handle 4a. So they could handle 460W/920W safely. A 3w diode would be safe up to 690W/1380W.
The only situation where you could safely overcurrent a diode with a lower duty cycle is if it was properly heatsinked and cooled, neither which apply on these crap units. You may be fine for awhile until the diodes heat up, but they're going to burn eventually. I've disassembled dozens of units like this where the diode bridge has burned themselves, burned the PCB, desoldered themselves from the PCB, and a combination of all three.
 

ryan_975

[H]F Junkie
Joined
Feb 6, 2006
Messages
14,735
They're burning because they're getting into their derating curves. A diode with a 2A average or continuous forward current rating will work just fine running at 4A* with a 50% duty cycle as long as the ambient air around the diode stays below 75C (for good diodes) to (105C for great diodes). REALLY cheap diodes start derating at 25C. So that crap PSU's "2A" diodes are really only capable of maybe 0.8A (150W/300W) at typical crap PSU operating conditions. It's really not surprising that you're finding burnt boards.

*I guess I should say that proper designs will derate by 20%. So by saying 4A, I really mean 3.2A.
 
Top