450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

[playit]

hello everyone , working zebronics power suply 450 watts , 5v ,3.3v givin short to ground black wire, also in circuit at output, resistor was charred but shows right values , after removing resistors 39ohms, 51 ohms, there was no short , no continuity beeps with ground , but after replacing new resistors (now i am not switching it on) ,again short still exists . when i put bigger ohm resistor like 100k ohm , no short beeping sound comes , lower ohm resistor gives continuity beeps ,other big ones donest , as its a working psu , i am confused what to remove now or how to go further now .
how this happend - i was putting a bad sata cable extension which blew the 12v/5v diode(not sure) in hdd pcb ( repaired it -working now) and made this short in psu , how to remove the short , thank you .

[playit]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

uploaded the pics , please ask for clear shot of any area if unclear, thank you.

[playit]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

please anyone could guide me on this , ??

[momaka]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

I believe you are misinterpreting your multimeter's continuity (beep) function.
Beeping does NOT mean short-circuit. Most multimeters will beep to show "continuity" when the resistance is relatively low. This can be anywhere from 20 to 100 Ohms, and varies between different model multimeters.

So you are getting beeping because those 39 and 51-Ohm resistors fall within the "Good continuity" range of your multimeter. But in reality, that' NOT a short-circuit... or good continuity for that matter.

Instead, measure the resistance of the 3.3V and 5V rails with respect to ground using your multimeter's resistance test function (200 Ohms scale if using a manual multimeter), so you can verify yourself about what I stated above. You should see around 39 Ohms on the 3.3V rail and 51-Ohms for the 5V rail. These are the minimum load resistors on these rails to keep the PSU regulation in spec at very low or no-load condition.

All of that aside, I actually DO NOT recommend you fix this PSU for computer use. It looks like every other cheap garbage PSU that's been showcased on the website here. 450 Watts? - NO WAY! You'd be lucky if this thing even provides 100 Watts of power with the ripple in spec. It has too few and too tiny electrolytic caps on the output. And that's not all that's wrong with it. Output toroid is too small, the heatsinks are very thin, and by the looks of the lead spacing distance on the underside, the primary caps are probably 2x 330 uF... or 470 uF units at best. Being such a cheap PSU, though, there is also a high likelyhood that they have fake ratings and are probably 220 uF each or lower. I have a few cheap PSUs like that, and their primary caps are not even reading 220 uF. Very low quality stuff.

[harp]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

uploaded the pics , please ask for clear shot of any area if unclear, thank you.

Whats is going on with green wire, ps-on... is solder ok, and cut the tail.

First of all, have you fuses ok, AC on board present, bridge ok, high dc ok... standby voltages 5vsb steady... I do not understand in which way you tested psu to come on conclusion not working? What reading you get?

On pcb, it is some dark area in primary side, check this also.

[playit]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

thank you for replying sir , everything is true what you have said , my multimeter does it exactly as you have said , and yes it gives those 51 and 39 ohms respectively.
no i am not using this for pc now , but sadly i have the same company 450 watts for pc running now . only this is mostly available here at locals shops . other things are expensive here.
i checked the values with paper clip and psu tester, the voltages are stable , the psu runs . when i connected the hdd sata cable to psu tester ,it showed 12 and 5v correct .nothing happend to psu tester ,can i use this psu to power a hdd ? , actually i plan to use this psu to power some case fans and some hdd occasionally.
by the way as its so subpar in its design , any way to keep a protection circuit or modify to some extent by changing some parts ? i know i lack too much of knowledge , but just curious to learn more if anything can be done at all .

[playit]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

Whats is going on with green wire, ps-on... is solder ok, and cut the tail.

First of all, have you fuses ok, AC on board present, bridge ok, high dc ok... standby voltages 5vsb steady... I do not understand in which way you tested psu to come on conclusion not working? What reading you get?

On pcb, it is some dark area in primary side, check this also.

yes i checked it all , they are all ok , when i put a bad sata cable it burn up 5v or 12v diode in hdd pcb at the same time , the other 2 hdd was spinning and computer was on ,everything fine , but i still was afraid so took it off and checked some , then i found this beep continuity which wasnt in other psu , and hence seeking if anything is wrong , but this psu works with all the volts stable. umm not sure whats that dark area tho on primary side..
3.38v , 5v , 11.87 v in paper clip , in psu tester full 12v and others are the same . i am just curious if i can power a hdd with this psu , as 5v and gound is in continuity .

[momaka]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

thank you for replying sir , everything is true what you have said , my multimeter does it exactly as you have said , and yes it gives those 51 and 39 ohms respectively.

Yup, that's how it should be.

no i am not using this for pc now , but sadly i have the same company 450 watts for pc running now . only this is mostly available here at locals shops . other things are expensive here.

Having been around a few places in the world, I totally understand. Nonetheless, I just wanted to warn you (and anyone else) what to look out for. In almost all cases, you'd be better off with a stock OEM 200-250 Watt PSU from the likes of Dell, HP, Lenovo, and etc. At least these can actually output the stated power on their label without going out of spec.

i checked the values with paper clip and psu tester, the voltages are stable , the psu runs . when i connected the hdd sata cable to psu tester ,it showed 12 and 5v correct .nothing happend to psu tester

Those PSU testers aren't true "testers". They are just a switch for the PS-ON, a few minimum load resistors, and a cheap IC to very roughly measure the voltage (if it's one of those testers with an LCD screen and not the even cheaper ones with the LEDs for the 3.3V/5V/12V only.) These "testers" won't show you the ripple and noise the PSU outputs on the 3.3V, 5V, and 12V rails... and high ripple/noise (R+N for short) is what can cause a system to crash or even kill PC components. The only way to properly check the R+N of a PSU is with an oscilloscope and a moderate load on the output.

can i use this psu to power a hdd ? , actually i plan to use this psu to power some case fans and some hdd occasionally.

Yes, it will be fine for that or any other low-load application under 50 Watts. Or 100 Watts if powering things that don't care about R+N, like electric DC motors, resistive heaters, and etc.

by the way as its so subpar in its design , any way to keep a protection circuit or modify to some extent by changing some parts ? i know i lack too much of knowledge , but just curious to learn more if anything can be done at all .

The protection circuits are usually OK even on very cheap PSUs like this - at least the basic ones, like short-circuit protection and under-voltage protection.

What you can improve on is the output filtering. Notice there is only 1 electrolytic cap per output rail. Most PSUs with this old half-bridge design and relatively low switching frequency typically have more and bigger caps on the output. The most common / standard arrangement back in the day of these designs was 2x 2200 uF for the 3.3V rail, 2x 2200 uF caps for the 5V rail, and either 2x 1000 uF or 1x 2200 uF for the 12V rail. Also "PI" coils (inductors with a ferrite rod core) between the caps. This greatly reduces R+N on the output, and also makes the output voltage more stable.
Other improvements you can make: add common-mode choke inductor on the put side, X2 and Y2 -class caps, and increase the capacity of the two 200V caps (470 uF minimum would be nice.) If you know of an e-scrap place near you... well probably best to look for a scrap ATX PSU from a Dell, HP, or other similar brand. Again, these would be better PSUs than what you have. But if not, look for old TV boards from the 90's and early 2000's, especially anything from Sony. These will usually have all of the parts to add to the input filtering.

i am just curious if i can power a hdd with this psu , as 5v and gound is in continuity .

Again, don't confuse continuity with short-circuit - these are two very different things.
Continuity on the multimeter just means "relatively conductive path", but NOT a short-circuit.
Short-circuit = extremely low resistance, i.e. a few Ohms at most... and often a lot less (under 1 Ohm.)

If there was a true short-circuit between 5V rail and ground, the PSU would not start and short-circuit protection would trigger. You can see if this protection works right by turning on the PSU with a paper clip and then intentionally shorting the 5V rail to ground with a thick piece of wire. If the PSU has decent short-circuit protection, it should switch itself Off right away.

[playit]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

so coool ! wow so much of teaching , thank you so much sir and about the mod i cant get it in scraps ,possibly i only could buy them , or may be i can ask with some tv repairing guys , i doubt these thieves here will help me at all lol .
"The most common / standard arrangement back in the day of these designs was 2x 2200 uF for the 3.3V rail, 2x 2200 uF caps for the 5V rail, and either 2x 1000 uF or 1x 2200 uF for the 12V rail. Also "PI" coils (inductors with a ferrite rod core) between the caps. "

sir is this for input side ?
"add common-mode choke inductor on the put side, X2 and Y2 -class caps, and increase the capacity of the two 200V caps (470 uF minimum would be nice.)"
will i have space to do these in the present board or i should buy a breadboard and extend it and wire it all soldering them.
yes i will also keep in mind to buy those scrap dell hp atx psu.
yes sir, i also learned i will never forget about this lower resistance beeper ! .

if i test short with 5v and ground , will anything happen that the psu will go bad ,i mean parts inside might go bad , because anything at all replacement i need to get here is so difficult , also india have banned aliexpress, which made spare parts too difficult and expensive to get .

by the way i would like to mention my rig i use with this 450 watt psu
xeon 5470 modded cpu
asus p5qlpro mobo - all 4 rams
120mm fan antec a40 cooler
nvidea gt730 ddr5 1gb(needs 300 watts they say- not sure whats true )
3 hdd - 2 x 7200 , 1 x 5400 rpm seagates
2 case fans ,intend to increase when summer comes.
i was messing around with a very bad sata cable which burned the diodes in hdd pcb and some searching led me to this shorting beeping confusion lol .but i learned a lot now

[playit]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

i tried to plug in dvd writer and see how it goes , my pc is running in different psu , and i powered this dvd writer with this psu with out sata cable first , it ran ok, it opened all good , but when i put sata cable in dvd writer and switch on my computer the psu fuse is blown and i first dint know its a fuse , i tried it again , the spark came again at fuse site , what could be the reason for this ? all the voltages were good and stable but when i tried with sata , this happen , could it be bad sata cable ?

i opened , i found a little screw stuck to the first pair of diode track at input side ,sighs, but how did it work with the screw stuck back at circuit touching the case down ! but only when i put sata cable from computer to dvd writer , it went spark and blown fuse and now those 2 diode pairs too gone .. please teach about this what happend, thank you.

[playit]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

could this screw have caused the blown fuse or the psu went bad with dvd writer load ?
when i supplied only the dvd writer , it was running ok, it opened the tray and closed many times when i pressed its button , but when i put sata data cable to computer , spark blown fuse happend .
the setup is -- i power this dvd writer with the psu in discussion and my computer is running in other psu .

[momaka]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

all the voltages were good and stable but when i tried with sata , this happen , could it be bad sata cable ?

i opened , i found a little screw stuck to the first pair of diode track at input side ,sighs, but how did it work with the screw stuck back at circuit touching the case down ! but only when i put sata cable from computer to dvd writer , it went spark and blown fuse and now those 2 diode pairs too gone .. please teach about this what happend, thank you.

Sorry, but nothing much I can teach you here.
You had a screw loose inside the PSU, and that is NEVER a good thing.
Most likely the screw wasn't shorting anything - VERY likely by complete luck - while you tested it with the DVD drive. And then when you tried to use the SATA cable, you either moved or bumped the PSU very slightly, and that caused the loose screw to go in the wrong place and short out... whatever it shorted out.

And no, a bad SATA cable or any other output cable should NOT cause the PSU to blow up. If by small chance it did, then the PSU wasn't worth using anyways. I understand certain things can be very challenging to find is some parts of the world, hence why some places are stuck with these cheap nasty PSUs (for example, I know many parts of Eastern Europe used to be filled with this cheap PSUs back in the 90's and early 2000's, and there was rarely anything better to buy.) But these cheap PSUs are a false economy - they don't and never were meant to power anything more than a low-end Celeron or Pentium dual core.

With that said, still keep your eyes peeled for scrap parts and/or local sites for used parts. Also don't worry that AliExpress is not available in your country - too many counterfeit/overrated transistors on there anyways. I bought a bunch of E13009's for repair of cheap PSUs, and all of them were counterfeit and couldn't hold even up to genuine E13003 spec. Total waste of time and money, but I expected that and bought them most just to see how bad they were.

sir is this for input side ?
"add common-mode choke inductor on the put side, X2 and Y2 -class caps, and increase the capacity of the two 200V caps (470 uF minimum would be nice.)"
will i have space to do these in the present board or i should buy a breadboard and extend it and wire it all soldering them.

Yes, it's for the input / high voltage side.
... though probably not applicable at this point with the PSU blown up (unless you repair it? )

if i test short with 5v and ground , will anything happen that the psu will go bad ,i mean parts inside might go bad , because anything at all replacement i need to get here is so difficult , also india have banned aliexpress, which made spare parts too difficult and expensive to get .

No, the PSU should not blow up when shorting any of the rails to ground. It should just shut down. If it doesn't, it probably wasn't safe to use anyways.

by the way i would like to mention my rig i use with this 450 watt psu
xeon 5470 modded cpu
asus p5qlpro mobo - all 4 rams
120mm fan antec a40 cooler
nvidea gt730 ddr5 1gb(needs 300 watts they say- not sure whats true )
3 hdd - 2 x 7200 , 1 x 5400 rpm seagates

I'm not all too worried about the GPU. GeForce GT730 is only rated for 38W TDP, so forget about what nVidia says about it needing a 300W PSU. It just needs 40 Watts itself. The Xeon 5470 is a bit worrisome, being rated for 120W TDP. At peak draw, it can reach 150W, which is probably right at the top limit of what this PSU can do safely in terms of raw power on the 12V rail.. and likely total too. DON'T run any CPU + GPU stress tests, or it's very likely that it will roast itself. With the CPU you got, you should have at least a 30 Amp rectifier on the 12V rail and 2x 1500 uF low ESR caps minimum to keep the R+N somewhat reasonable (though probably still might not be within ATX spec.)

[playit]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

no sir, i dont use the cpu or gpu too much intensively , its just sometimes movies and studies with youtube , then may be some movies with high resolution but i see how much cpu is running, if its more than 50-60% i try with other player or other browser to make cpu running 20-30 % .
i want to repair it , i want to fight with it, ) i will order the diode rl207 mic rectifier , i will put the input side of pics now , please tell me what else to chek so i can order other bad things too along .

[momaka]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

Yes, if you want to repair it, please post pictures so we can see better what happened.

RL207 - that's more than likely a diode from the bridge rectifier... and a bit undersized for the task. No need to replace it with same exact diode, just one with similar or better specs, i.e. 600V DC blocking voltage or more and 2 Amps (3 or more preferable) average current. Or just get something common like a single KBL406 or RS606/RS608 bridge rectifier and replace all of the RL207 diodes with that.

Since the fuse has blown, you should remove and check the primary-side BJT's. Should be NPN type and same or similar to E13007 or E13009. Sometimes they blow shorted, other times open-circuit. Have to test them out of circuit to verify. Also check the NTC thermistor on the input, if there was one (fuse blowing can sometimes cause the NTC to explode or go high-resistance before the fuse goes.) And if one or both BJTs are bad, check every resistor, capacitor, and diode connected to their Base-drive circuit.

[playit]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

uploading

[playit]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

please ask at any particular angle if these doesnt suffice , thank you

[playit]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

i am planning to make a series bulb thingy which i read that it can help in not blowing up psu components or the loader (fan or hdd or whatever ) when turning on testing psu . would this help ?

[CapLeaker]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

If you load the PSU probably not. Depending on how much of a load you are trying to put on it. For example a little dvd rom or a hdd shouldn’t be a problem. A full load however probably not.

Depending on the load you need to increase the wattage of the series light bulb, or put multiple light bulbs on switches.

[momaka]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

please ask at any particular angle if these doesnt suffice , thank you

These suffice for now, thank you.

Again, before I begin, let me just state that the repair of such cheap PSU is probably only good for learning purposes only. In reality, these cheap PSUs are just not worth repairing, because of how poorly they are designed and the low-quality shoddy components they use. For example, the power supply shown here doesn't even provide any form of independent regulation on the 3.3V rail - it is just taken from the main transformer's windings directly. This is very bad for cross-regulation, and likely the 3.3V rail will be very high and/or fluctuate a lot as a load is put on the 12V rail. Also, not only are there no EMI/RFI filter components on the input, but the manufacturer didn't even bother to cut the PCB tracks and put it on the PCB design / layout for there to be any. So it goes to show the manufacturer plain-straight didn't care to make a reasonable attempt at designing a proper PCB layout. On top of that, the main transformer has a fake "35" -class label. Clearly the manufacturer is trying to scam us here. The real size of that main transformer looks more like 28-size, despite being wider on the top. It is simply too short to contain any reasonable amount of windings. Such PSUs belong to this thread, really:
https://www.badcaps.net/forum/showthread.php?t=8490
How this PSU hasn't blown up quicker with the PC you have is almost a miracle. My guess is that every component in it is running right to its limits. Certainly don't expect your replacement PSU to last any reasonable amount of time.

But enough of my rants above. Here goes the troubleshooting.

For starters, there are a number of bad/cold solder joints that need to be fixed (resoldered). I highlighted them in red here:


I also highlighted a section on the input side of the PCB with blue to show where you can cut traces and install common-mode chokes and X2/Y2 caps (on the top side of the PCB, of course) should you get this PSU running and if you'd like to rebuild it a little better.

Next, let's focus on the primary side. If one of the RL207 diodes has blown, chances are there may be more that are marginal (even if still measuring good.) My suggestion is to replace all of these diodes at once or use a single bridge rectifier as mentioned in my last post.

Also, replace that 5 Amp glass fuse with a proper 5x20 mm ceramic fuse if you have one. You can see the manufacturer left space on the PCB for the much larger 5x20 fuse, but put a smaller one because of how cheap the PSU is.

If you have a way to measure the primary 200V caps, do that too, as one of the RL207 diodes blowing up could have sent AC through them and possibly damaged them.

Now take a look at this picture of yours:
https://www.badcaps.net/forum/attach...3&d=1693943602

The NPN BJTs on the primary heatsink look like a pair of (cheap) 13007 clones. Remove both and test them out of circuit with diode test on your multimeter. B-E and B-C junctions should show a normal diode voltage drop when the multimeter's red probe is connected to Base. C-E junction should show open-circuit. Chances are, if one blew up, both of these BJTs will need to be replaced. Be careful where you source your replacement BJTs from. I bought a bunch of counterfeit 13009 transistors on eBay, and they couldn't even hold up to the specs of 13003. For reference, real 13009 BJTs should be good for up to 400 Watts in a properly-designed half-bridge PSU, and 13007 are often good up to 300 Watts. Something like 13003 will likely croak at under 100W load... so beware and try to avoid counterfeits if you can.

With the 13007 transistors removed from the PSU, test the rest of the components connected that lead to their Bases. I'm talking about the small components shown in front of the 13007 transistors. There should be one or two small diodes (usually 1N4148), some resistors (two of them are usually 2.7k), and 2x small electrolytic caps in the range of 1 to 10 uF. The circuit normally looks either like this:
https://www.badcaps.net/forum/attach...1&d=1606198568
or this:
https://www.badcaps.net/forum/attach...1&d=1608943994
(For both of the above circuits, primary is on the right side and secondary for the BJT drive is on the left side.)

That said, after checking abovementioned components, check the base-drive components on the secondary side (the components on the left side of the above partial schematics.) Some of them will be located close to the IC on the secondary side, and some will be close to the middle transformer.

Lastly, here's a thread where I fixed a cheapo PSU based on the same half-bridge design that had a blown primary side (in case you want to read more about how I went about to troubleshoot the problem(s)):
https://www.badcaps.net/forum/showthread.php?t=90392

i am planning to make a series bulb thingy which i read that it can help in not blowing up psu components or the loader (fan or hdd or whatever ) when turning on testing psu . would this help ?

Yes, a series incandescent / halogen bulb is a MUST, especially when dealing with a PSU that has had a major blow-out.

Depending on the load you need to increase the wattage of the series light bulb, or put multiple light bulbs on switches.

+1

For general testing, use a 100W incandescent or halogen bulb. If such incandescent or halogen bulbs are not available in your country, use 2x or 3x 40W bulbs in parallel.
In the past, I used to use only a single 40W bulb, but I found out some PSU would not only not start, but also not produce 5VSB with such a low bulb (very few, though.) I then switched to a 60W bulb, which was OK to test most 5VSB circuits, but still often wasn't enough to test the PSU with even a tiny load (i.e. a single HDD.) Hence my recommendation to go with a 100W bulb.

[playit]

Re: 450watt power supply 5v, 3.3v short to ground wire,in output rail ,connectors

thank you momaka sir for taking such immense effort to teach me .i am taking notes .sure i have learned a lot from you , i wont be putting this in my pc , just have the plans to use only the fans and dvd drive (with ground connection which i have missed at past )
i have almost checkd all the other components with help of a local region youtube video with kinda same psu ,about the power transistors and 13007, he cheked withinthe circuit showing some values around 300-500 b-e , but nothing with b c or c e for all the 3 transistors,this was ok he mentioned, for other small components they all match up , no continuity beep or continuity reading in mulitmeter, resistors give reading according to thier ohms colour codes . they seem ok , but i want to buy and keep those primary side npn power transitors etc ,
the transistors attached to heat sink are - 208c5027r 13007-0706 1300- 1012b
but i am unable to get these exact things online here at my place , what parameters mainly i should be looking when getting a replacements for these power transistors.
also for secondary sides -stpr1620ct 7skwy u phl 430 stps2045ct 7sabz vw phl 009 stps2045ctc g417b vw chn 236
they are all ok with testing according to the video i saw . now still i would like to buy some of these and keep it stock just in case, but damm it , just 5 pieces(not giving 1 tho) of 5027 power transistor costs around the price i got this psu !

now i found out what other mistake i have done adding up to that screw i missed underneath the rectifier diode track which even left a mark , when using 2 psu on the same pc when adding up the drive or anything which connects the data cables ,atleast the ground should be connected if 2 psu connecting to different wall sockets . this also the reason the fuse blows with spark i learned.
now i gotta get those parts ,still finding a way to get them all .
by the way this power supply never have given any problem , may be its due to i never had anything intense goin on . the only occasions this went bad is because of me lol
i screwed with them oops. actually i remember this one lasting for kinda 4-5 years now until i messed up with a bad sata power cable molex thingy (they say molex sata lose data - i found this after i messed up).
almost 70-80 percent of computers in india must be using this psu !! everywhere they say its a cheap design and it will blow off pc , some did blow off as i checked in youtubes , others just go dead ending up in spares for repair shops, but at my place no one knows to fix this ! or ready to take up a challenge lol but other parts of india are doing great in fixing them for some good price , my place sucks ! . phew wrote too much !
wish i had the power to sue them and show they putting cheap designs and fake stickers !and yes i will put the pics in hall of shame psus ! everyone must know this cheap sssccrap.