Bestec ATX-300-12E Repair and Recap Success

[LDSisHere]

I needed a 20-pin ATX power supply to power an old P4 motherboard and IDE hard drive combo but I did not have any good ones to use. What I did have was a stack of bad ones that I got for learning and experimenting with so I settled on trying to repair a Bestec ATX-300-12E.

I chose this supply for several reasons. The most important being it met my primary criteria, but it had a few other things that interested me in this supply. All but one cap looked good from a visual standpoint and the 5vsb not only worked it was powered by a Viper22A which I am very familiar with but was my first time seeing one put in a power supply by a manufacturer. In the past, all the supplies I have repaired required a recap before troubleshooting could even be started, with this supply I could fix it first then, decide if it was worth a recap.

This supply uses a TPS3510 supervisor chip and a UC3842B PWM chip. I had the 5vsb voltage like I stated previously and I had PS_ON# but when this was grounded I would get nothing. I found the UC3842B was shorted out so I replaced it. I tried it again and no response. I then found a shorted Zener (ZD1) and after I replaced it still no response, the UC3842 was not getting its' Vdd. The transistor Q2 turned out to be the final culprit preventing the supply from working.

The recap was fairly simple but the glue they use is madding. This supply had three CrapXon capacitors, the two main ones and one on the 5vsb. The one on the 5vsb was the only popped cap in the entire supply. There were two KZE that I did not replace but everything else had to go. I luckily had some Chemicon main caps from a supply I was getting parts from to replace the CrapXon ones this one had in it. The rest of the caps I had on hand in my new stock.

The final picture I have attached shows the output voltages of the power supply. I put a 20 ohm load on the 12V, a 10 ohm load on the 5V and I did not load the 3.3V output. I was very pleased that I was able to find what was wrong with this supply all by myself, and then to scrounge all the parts I needed to get it working again. Anyway, if you have read this far I hope I have not board you to death with my post, thanks for reading.

[Pentium4]

Re: Bestec ATX-300-12E Repair and Recap Success

I hate to say it but I would have left the 560uF CrapXon input caps. Their caps are shit but their input caps are usable. I would use those NCC's on something else but that's just me nice recap though good job!

[retiredcaps]

Re: Bestec ATX-300-12E Repair and Recap Success

I found the UC3842B was shorted out so I replaced it.

Was Vdd shorted to ground?

The transistor Q2 turned out to be the final culprit preventing the supply from working.

How did you determine Q2 was bad? Was it shorted?

The recap was fairly simple but the glue they use is madding.

Good job in removing the glue as it turns conductive over time.

The final picture I have attached shows the output voltages of the power supply.

A minor suggestion. Since your Fluke 21 is a 3200 count multimeter and the 77IV and 175 are 6000 count, the following configuration will get you the most resolution.

Put the 21 on the 12V rail (it will show 12.11) and put the 77IV and 175 on 3.3 and 5V rail. With the 77IV or 175 measuring the 5V rail, you might see 5.000V DC.

PS. Thanks for documenting and sharing your repair notes.

[LDSisHere]

Re: Bestec ATX-300-12E Repair and Recap Success

Was Vdd shorted to ground?

Not exactly. There was less than 7 ohms or less resistance between pins 5, 6, 7 and 8. I pulled the chip and again verified the numbers and they were the same and much different than compared to a new chip. I then checked the Pins 6, 7, and 8's paths for shorts before placing the new chip in the circuit.

How did you determine Q2 was bad? Was it shorted?

IIRC pin 3 was connected to the rail with the Zener I replaced and I got a reading of around 150 ohms between pins 2 and 3 in both directions.

Good job in removing the glue as it turns conductive over time.

That stuff is very annoying to put it politely.

A minor suggestion. Since your Fluke 21 is a 3200 count multimeter and the 77IV and 175 are 6000 count, the following configuration will get you the most resolution.

Put the 21 on the 12V rail (it will show 12.11) and put the 77IV and 175 on 3.3 and 5V rail. With the 77IV or 175 measuring the 5V rail, you might see 5.000V DC.

PS. Thanks for documenting and sharing your repair notes.

Good idea, I will keep this in mind next time, I was just concerned with getting the readings setup from lowest to highest lol.

[LDSisHere]

Re: Bestec ATX-300-12E Repair and Recap Success

I hate to say it but I would have left the 560uF CrapXon input caps. Their caps are shit but their input caps are usable. I would use those NCC's on something else but that's just me nice recap though good job!

Thanks for the compliment. As far as the NCC caps, I do not know if or when I would find another use for them as they are rather a low capacity plus free. Before I stumbled upon them I was eying a pair of Teapos but I thought finding the UCC's when I did was a sign that I should use them so I did.

[Khron]

Re: Bestec ATX-300-12E Repair and Recap Success

I've seen bloated Fuhjyyu and CapXon primaries, so... I wouldn't even trust them there, not to mention on the secondary side...

[LDSisHere]

Re: Bestec ATX-300-12E Repair and Recap Success

My personal policy on Fuhjyyu capacitors is to replace them on sight, no questions asked. If CrapXon is in the same league then they may get the same treatment. This supply mostly had Jamicon and none of them were bloated and I did not even bother to test the ESR, I just decided to replace them all just so I would not have to worry about it.

[Pentium4]

Re: Bestec ATX-300-12E Repair and Recap Success

Wow nice job removing all the glue too! I just came across one of these again and forgot how much glue is in them...Did you by chance test the ripple with a scope after recapping it?

Also, couldn't see in the pic did you replace that little 0.1uF 63V Jamicon cap on the back by the main transformer? Only value I don't have but such a small capacitance value

[kaboom]

Re: Bestec ATX-300-12E Repair and Recap Success

Also, couldn't see in the pic did you replace that little 0.1uF 63V Jamicon cap on the back by the main transformer? Only value I don't have but such a small capacitance value

There are three of those in the 250 and 300W Bestecs that I've worked on.

One behind the transformer and another in the 3.3V mag amp. I forget right now where the third one is...

I've replaced them with 100V or greater TDK MLCCs. Good polypros/polyesters should also work.

Wow nice job removing all the glue too! I just came across one of these again and forgot how much glue is in them...Did you by chance test the ripple with a scope after recapping it?

I usually end up removing the primary heatsink, as it's much easier to work on. It takes me so much longer to de-glue these things than it does to recap 'em!

The recap was fairly simple but the glue they use is madding. This supply had three CrapXon capacitors, the two main ones and one on the 5vsb. The one on the 5vsb was the only popped cap in the entire supply. There were two KZE that I did not replace but everything else had to go. I luckily had some Chemicon main caps from a supply I was getting parts from to replace the CrapXon ones this one had in it. The rest of the caps I had on hand in my new stock.

The final picture I have attached shows the output voltages of the power supply. I put a 20 ohm load on the 12V, a 10 ohm load on the 5V and I did not load the 3.3V output. I was very pleased that I was able to find what was wrong with this supply all by myself, and then to scrounge all the parts I needed to get it working again. Anyway, if you have read this far I hope I have not board you to death with my post, thanks for reading.

Nice job! Give yourself a pat on the back from me. I can't stand those glue-monsters... Here's another trick for your Bestec: Partially cover 1/3 to 1/2 of the rearward portion of the vents in the top cover of that PS. All those slots let the air bypass too much. Try and go for more of a "flow-through" characteristic.

[LDSisHere]

Re: Bestec ATX-300-12E Repair and Recap Success

Wow nice job removing all the glue too! I just came across one of these again and forgot how much glue is in them...

Thanks, but I think the amount of glue may vary. I could not find the one I recapped, but I did find another one on my shelf. (See the attached pictures.) This one does not seem to have a lot of glue, but any glue is too much IMO.

Did you by chance test the ripple with a scope after recapping it?

I did not have a scope at that time so I did not check the ripple.

Also, couldn't see in the pic did you replace that little 0.1uF 63V Jamicon cap on the back by the main transformer? Only value I don't have but such a small capacitance value

I wish you had not asked me this question. If I go to the trouble of recapping a supply I like to make sure I replace them all, small ones included. So yes I replaced it with what I had on hand and in doing so I took a chance by using a 50V cap. The two scope shots are taken from this same cap on the supply I found today, the first one is the DC and the second one is the AC. It appears to be 40V DC under the load of my ATX tester. I am not sure if or how much it could change under a higher load so it may have not been a good idea to use a 50V cap. I think if I were to do it now, I would just find a .1uF 100V ceramic cap and be done with it.

[Wester547]

Re: Bestec ATX-300-12E Repair and Recap Success

Nice job! Give yourself a pat on the back from me. I can't stand those glue-monsters... Here's another trick for your Bestec: Partially cover 1/3 to 1/2 of the rearward portion of the vents in the top cover of that PS. All those slots let the air bypass too much. Try and go for more of a "flow-through" characteristic.

Are the extra vents that give the PSU more "breathing space" harmful to the path of airflow in the PSU? I thought the PSUs that employed such ventilation (whether to keep the PSU cooler or to exhaust heat from the computer case) in such a way that they were designed to, with those extra vents, properly exhaust heat out of the PSU without too much air escaping. Of course, if you're talking about keeping the path of airflow as straight and unobstructred as possible, I can understand that. The same could be said of those PSU cases that have side vents (the path of air might not be that straight, so they use transparent plastic to direct the flow of air in some spots, which I think greatly alleviates that problem, at least with regard to static air pressure). I still think good ventilation is crucial in a PSU, even if there's sometimes too much of it.

The one small capacitor I usually worry about in this design is the startup capacitor (that I believe drives UC384x) next to the +5VSB flyback auxiliary transformer (the primary side of it). In this case I think it would be a 47uF, 50V CapXon KM so it should definitely be replaced... or maybe it's a Jamicon or something. I don't like the glue Bestec use either, not just because it's conductive but because wantonly and sloppily covering so much of the secondary section in that glue can't be good for airflow.

Bestecs are still far better than the gutless lore of PSUs but from what I've seen in this particular model, I wasn't too impressed with the secondary side. I only saw a 16A ultrafast in the +12V for what was rated to be 15A and a 30A TO-247 rectifier for +5V and +3.3V. This is single switch forward (FQA9N90C I believe is the main switcher which can waste quite a bit of heat before failing) so that means a half-wave rectified output, IE only one of the diodes in a two diode rectifier package is fully conducting which curtails the amount of current the rectifier is capable of by 40%. I think they cut it a bit too close. Not impressed with the output filtering either, the use of general purpose capacitors and the rather low values per rail. I do not like the use of 20 gauge wires in a 300W unit either.

But enough of my complaining. I think LDSisHere did a great job cleaning up and recapping the PSU as well.

[Pentium4]

Re: Bestec ATX-300-12E Repair and Recap Success

There are three of those in the 250 and 300W Bestecs that I've worked on.

One behind the transformer and another in the 3.3V mag amp. I forget right now where the third one is...
I usually end up removing the primary heatsink, as it's much easier to work on. It takes me so much longer to de-glue these things than it does to recap 'em!

I usually leave them, because I never get around to ordering that value it bugs me, but I think Jamicon are much better than the likes of CapXon, Fuhjyyu, etc. Gahhh the glue drives me insane. Can't stand it! I agree it takes much longer to remove all that damn stuff! That glue that goes hard in these Bestec's seems to be exclusive to their older units. I've seen a decent amount of new (well, 2-3 years old) Bestec's where they use the rubbery white glue. It's still annoying but not as bad in my opinion

I wish you had not asked me this question. If I go to the trouble of recapping a supply I like to make sure I replace them all, small ones included. So yes I replaced it with what I had on hand and in doing so I took a chance by using a 50V cap. The two scope shots are taken from this same cap on the supply I found today, the first one is the DC and the second one is the AC. It appears to be 40V DC under the load of my ATX tester. I am not sure if or how much it could change under a higher load so it may have not been a good idea to use a 50V cap. I think if I were to do it now, I would just find a .1uF 100V ceramic cap and be done with it.

Sorry thank you very much for the follow up with pictures. These are good PSU's, I snag them whenever I see them. Although I'm more of a fan of the 300-12Z as it has 4 more amps on the 12V (on the label at least) and I believe it also uses dual TO-247 schottky's for the 5V, although not a huge deal unless you're using an older machine. Interesting, so do you think that 50V nichicon cap would not last long there?

Are the extra vents that give the PSU more "breathing space" harmful to the path of airflow in the PSU? I thought the PSUs that employed such ventilation (whether to keep the PSU cooler or to exhaust heat from the computer case) in such a way that they were designed to, with those extra vents, properly exhaust heat out of the PSU without too much air escaping. Of course, if you're talking about keeping the path of airflow as straight and unobstructred as possible, I can understand that. The same could be said of those PSU cases that have side vents (the path of air might not be that straight, so they use transparent plastic to direct the flow of air in some spots, which I think greatly alleviates that problem, at least with regard to static air pressure). I still think good ventilation is crucial in a PSU, even if there's sometimes too much of it.

The one small capacitor I usually worry about in this design is the startup capacitor (that I believe drives UC384x) next to the +5VSB flyback auxiliary transformer (the primary side of it). In this case I think it would be a 47uF, 50V CapXon KM so it should definitely be replaced... or maybe it's a Jamicon or something. I don't like the glue Bestec use either, not just because it's conductive but because wantonly and sloppily covering so much of the secondary section in that glue can't be good for airflow.

Bestecs are still far better than the gutless lore of PSUs but from what I've seen in this particular model, I wasn't too impressed with the secondary side. I only saw a 16A ultrafast in the +12V for what was rated to be 15A and a 30A TO-247 rectifier for +5V and +3.3V. This is single switch forward (FQA9N90C I believe is the main switcher which can waste quite a bit of heat before failing) so that means a half-wave rectified output, IE only one of the diodes in a two diode rectifier package is fully conducting which curtails the amount of current the rectifier is capable of by 40%. I think they cut it a bit too close. Not impressed with the output filtering either, the use of general purpose capacitors and the rather low values per rail. I do not like the use of 20 gauge wires in a 300W unit either.

But enough of my complaining. I think LDSisHere did a great job cleaning up and recapping the PSU as well.

These Bestec PSU's seem to run pretty cool as they are especially considering their low RPM fans and mediocre efficiency. I think it says a lot that I've seen Jamicon last for years and years powering Pentium D systems covered in dust. (I still replace them on site) Although it does make more sense to pull more air from farther back.

I usually replace the 1000uF 16V caps with 1500uF 16V caps, and match the rest. And when you're only talking 300W with those decent sized inductors on the output, I would say the ripple would be pretty decent

[momaka]

Re: Bestec ATX-300-12E Repair and Recap Success

Bestecs are still far better than the gutless lore of PSUs but from what I've seen in this particular model, I wasn't too impressed with the secondary side. I only saw a 16A ultrafast in the +12V for what was rated to be 15A and a 30A TO-247 rectifier for +5V and +3.3V. This is single switch forward (FQA9N90C I believe is the main switcher which can waste quite a bit of heat before failing) so that means a half-wave rectified output, IE only one of the diodes in a two diode rectifier package is fully conducting which curtails the amount of current the rectifier is capable of by 40%. I think they cut it a bit too close. Not impressed with the output filtering either, the use of general purpose capacitors and the rather low values per rail. I do not like the use of 20 gauge wires in a 300W unit either.

+1
I worked on only one of these and I wasn't impressed either.
But yeah, congrats on cleaning all of the glue out in these units. Indeed it does take much more effort to do that than the recap job itself. At least with the 300-12E it's somewhat easier because you don't have to remove the PCB from the case in order to solder. Also, the 300-12E I worked on had white glue instead of the tan conductive crap.

Here's another trick for your Bestec: Partially cover 1/3 to 1/2 of the rearward portion of the vents in the top cover of that PS. All those slots let the air bypass too much. Try and go for more of a "flow-through" characteristic.

+1 for this tip as well. I did it on a 250-12Z and airflow is much better IMO.

[Pentium4]

Re: Bestec ATX-300-12E Repair and Recap Success

I've replaced them with 100V or greater TDK MLCCs. Good polypros/polyesters should also work.

Should any issues arise from using a 0.1uF 1kV ceramic cap instead? They're easier to find when salvaging and it bugs me leaving Jamicon caps

[LDSisHere]

Re: Bestec ATX-300-12E Repair and Recap Success

I would try it. I would do a before and after shots with my scope to see if it makes any difference on the output. If you see no difference and the the supply works then it should be good.

[Pentium4]

Bestec ATX-300-12Z Repair

So, how do you get a 300W Allied to do 300W in spec? Oh, you just throw the internals of a 300-12Z in there (looks almost the same as the 12E) The ratings on the label are even almost identical to the 300-12Z's label

I like the look of the wire grille plus I thought it'd get better airflow in this chassis (Like kaboom suggested)

Those CapXon input caps looked healthy so I left them, 680uF is beefy for 300W so they should be okay there.

I also replaced those three 0.1uF 63V Jamicons with 0.1uF 100V ceramic caps (The one by the transformer, the one by the bridge rectifier, and the one by the +5VSB wire). I have not measured it with a scope, but the voltages were really stable with just a basic 12V based 125W load. I also removed the drying grease out of that Jamicon fan and put some fresh oil in it.

However, I did not hook up the rocker switch, I may do that later. The glue is such a pain in the ass, I tried to get the worst of it out of there! Even replaced a resistor with a fresh one because there was so much glue on it, you couldn't even see the damn thing!

[everell]

Re: Bestec ATX-300-12E Repair and Recap Success

Interesting solution! Did you change the label? Otherwise the next person who works on that power supply will be really confused!!!

[momaka]

Re: Bestec ATX-300-12E Repair and Recap Success

Haha, nice!
Yeah, I agree with everell. You should either change the label or put a note that this is NOT an Allied PSU. Otherwise, you (or whoever else may end up with that PSU) may end up just tossing it - and it would be a shame because it's a decent power supply.

By the way, is it the first picture or do I see some PCB burning near the 5VSB area? Just wondering. I do know this PSU uses a proper IC for 5VSB generation, but again, it's a Bestec - some of their design ideas aren't the brightest I've seen.

[Pentium4]

Re: Bestec ATX-300-12E Repair and Recap Success

Interesting solution! Did you change the label? Otherwise the next person who works on that power supply will be really confused!!!

Thanks No I didn't change the label, if I don't use it personally I definitely will. It sure would confuse them! "Why is this Allied so heavy??"

Haha, nice!
Yeah, I agree with everell. You should either change the label or put a note that this is NOT an Allied PSU. Otherwise, you (or whoever else may end up with that PSU) may end up just tossing it - and it would be a shame because it's a decent power supply.

By the way, is it the first picture or do I see some PCB burning near the 5VSB area? Just wondering. I do know this PSU uses a proper IC for 5VSB generation, but again, it's a Bestec - some of their design ideas aren't the brightest I've seen.

Yeah, that would definitely be a shame, because of the time spent, not the PSU itself I think it's just the lighting in that picture, are you looking at IMAG0178?

[momaka]

Re: Bestec ATX-300-12E Repair and Recap Success

Yes, IMAG0178