Bestec ATX100-5 no +5VSB

[Wester547]

Re: Bestec ATX100-5 no +5VSB

LOL so Bestec put in a 10A capable part but slapped a 2A max on the sticker! This part can supply more power than the 100 watts this old thing is rated for! I have another one of these in another computer in storage. It needs some capacitors replaced. I'll make a video of the inside of it whenever I get around to it. One thing that shocks me is the size of those little heatsinks... maybe that's why they put such small ratings on it. But, just hook up a leaf blower Delta fan and you might be able to play Crysis with it

One reason why PSU makes may seem underrating with the specs on the label is because the amount of amps any one schottky (or MOSFET, if you would) can deliver depends heavily on the rest of the components, like the width of the traces, the coils, the switching transistors, the rectifying bridge, the transformer, the topology, the heatsinks, the fan, the capacitors, the other schottky diodes on the other rails, which rail is loaded, etc. If you consider that this PSU is probably limited to 100W max given that Bestecs (not having OPP) explode when significantly more than their rating is pulled overall, even though the part by itself can do 10A (though that too depends on what temperature it's rated at) and maybe even more with good cooling, +12V * 10A = 120W. The +12V rail isn't the only rail in use, naturally, and that Bestec may not even be able to do 120W if the rest of the hardware inside isn't spec'd to deliver such.

Without ranting too much, if 100W is the limit of that PSU, you won't be pulling 10A from the +12V rail of that Bestec. Maybe more than 2A, but probably not 10A. With all the rails combined you'll only be able to pull so much. Not to totally discredit what you're saying, though - good PSU makes do overbuild their PSUs and rate their PSU at a continuous and seemingly low rating rather than the honest maximum.

[cubecompmtdx]

Re: Bestec ATX100-5 no +5VSB

LOL so Bestec put in a 10A capable part but slapped a 2A max on the sticker! This part can supply more power than the 100 watts this old thing is rated for! I have another one of these in another computer in storage. It needs some capacitors replaced. I'll make a video of the inside of it whenever I get around to it. One thing that shocks me is the size of those little heatsinks... maybe that's why they put such small ratings on it. But, just hook up a leaf blower Delta fan and you might be able to play Crysis with it

[GnatGoSplat]

Re: Bestec ATX100-5 no +5VSB

Are there really no markings at all, or are they just too small to see?

I've found the pure light from a blue LED and viewing obliquely can reveal numbers on parts that otherwise look blank. The markings are often laser-etched and very shallow, so it's more like a faint engraving.

The infinite resistance does suggest a MOSFET. Try testing it as if it were one.

No markings at all, nothing visible, not even under a magnifying glass. I even tried the UV blacklight LED from my A/C leak detecting kit and nothing shows up. I tried scratching at the top in case there's a varnish coating filling in the numbers and making them disappear, but still nothing.

I tried this test:
http://www.4qdtec.com/mostest.html
It does kind of look electrically like a MOSFET turned on, with the 1st 2 pins being source and drain, but even shorting all the pins together doesn't seem to turn the MOSFET off. Also, from what I've found on Google, it looks like SOT-23 MOSFETs should have the pinouts GDS so that makes the voltage drop/low ohms reading between the 1st 2 pins look rather odd.
Just to see if that site's testing method is sound, I tried testing a brand new TO-220 MOSFET and was able to reproduce his "good" test.

LOL the ATX100-5, also known as "The Little Fighter."

Glad to see you guys experimenting with this power supply. I'm curious to know what the 12v secondary silicon is rated for on this thing. The sticker states 2A for +12v yet my buddy pulled at least 20 outta it... (he did let some magic smoke out and had two extra fans stacked on top of it ) and the unit in my video has been running my 60w (5A minimum) peltier fridge for 9 months non-stop without a problem. The capacitors (I believe JPCE-TUR) in it all appeared fine with no bulging tops. Hopefully you can get the 5VSB fixed on this one because this power supply is apparently underrated. Good Luck!

LOL! Yeah, your video made me feel more like I should fix it!
The 12V rectifier is a CTX12SL. I can't find a datasheet on that exact part, but I was able to find several websites that say it crosses to a STPR1020CT which is rated for 10A.

[cubecompmtdx]

Re: Bestec ATX100-5 no +5VSB

I saw a video of this about 1 year ago on the CubeComputerChannel

http://www.youtube.com/watch?v=Prmmd1tHV0Y&feature=plcp

LOL the ATX100-5, also known as "The Little Fighter."

Glad to see you guys experimenting with this power supply. I'm curious to know what the 12v secondary silicon is rated for on this thing. The sticker states 2A for +12v yet my buddy pulled at least 20 outta it... (he did let some magic smoke out and had two extra fans stacked on top of it ) and the unit in my video has been running my 60w (5A minimum) peltier fridge for 9 months non-stop without a problem. The capacitors (I believe JPCE-TUR) in it all appeared fine with no bulging tops. Hopefully you can get the 5VSB fixed on this one because this power supply is apparently underrated. Good Luck!

[b700029]

Re: Bestec ATX100-5 no +5VSB

Are there really no markings at all, or are they just too small to see?

I've found the pure light from a blue LED and viewing obliquely can reveal numbers on parts that otherwise look blank. The markings are often laser-etched and very shallow, so it's more like a faint engraving.

The infinite resistance does suggest a MOSFET. Try testing it as if it were one.

[ben7]

Re: Bestec ATX100-5 no +5VSB

Looks like it doesn't exactly have a 2-transistor 5VSB. It looks more like 5-transistors. I've been too tired from work to draw the schematic, but I might tackle it on the weekend. Anyway, just at a glance, I see there are 4 SOT-23 style SMD transistors. I'm assuming they really are transistors and not diodes because they are labeled on the PCB as Qxx (plus, there are already SMD diodes on this board and they used the round glass type package). The base of the switcher is connected to the collector of 2 of the SOT-23 transistors, and their both of their emitters are also connected together. However, I don't believe they are 2 in parallel doing the job of one because their bases are not in parallel.

I pulled one of the SOT-23s because it looked shorted (what a pain). Got it out and it's not actually shorted, but it doesn't measure like a regular NPN or PNP transistor. Assuming this site (http://talkingelectronics.com/projec...eMount-P1.html) is correct about all SOT-23s having the same BCE pinout, I'm getting about 224-ohms between B-to-C and infinity between B-E and C-E. In Diode Check mode, there's 0.3V drop between B-C both directions and infinity between any other pins. That doesn't look to me like any diode or transistor I'm aware of, so I'm guessing it's a bad part? Still looks like there's a short in the circuit somewhere, so something else is either bad or other components are interfering with my multimeter readings.

There are NO numbers or any kind of markings whatsoever on these SOT-23 parts, so I think the only way I will have a clue as to what they are supposed to be is by drawing a schematic.

Hm, maybe its a small mosfet?

[GnatGoSplat]

Re: Bestec ATX100-5 no +5VSB

Looks like it doesn't exactly have a 2-transistor 5VSB. It looks more like 5-transistors. I've been too tired from work to draw the schematic, but I might tackle it on the weekend. Anyway, just at a glance, I see there are 4 SOT-23 style SMD transistors. I'm assuming they really are transistors and not diodes because they are labeled on the PCB as Qxx (plus, there are already SMD diodes on this board and they used the round glass type package). The base of the switcher is connected to the collector of 2 of the SOT-23 transistors, and their both of their emitters are also connected together. However, I don't believe they are 2 in parallel doing the job of one because their bases are not in parallel.

I pulled one of the SOT-23s because it looked shorted (what a pain). Got it out and it's not actually shorted, but it doesn't measure like a regular NPN or PNP transistor. Assuming this site (http://talkingelectronics.com/projec...eMount-P1.html) is correct about all SOT-23s having the same BCE pinout, I'm getting about 224-ohms between B-to-C and infinity between B-E and C-E. In Diode Check mode, there's 0.3V drop between B-C both directions and infinity between any other pins. That doesn't look to me like any diode or transistor I'm aware of, so I'm guessing it's a bad part? Still looks like there's a short in the circuit somewhere, so something else is either bad or other components are interfering with my multimeter readings.

There are NO numbers or any kind of markings whatsoever on these SOT-23 parts, so I think the only way I will have a clue as to what they are supposed to be is by drawing a schematic.

[b700029]

Re: Bestec ATX100-5 no +5VSB

Your schematic appears to be missing the startup resistor - it's several hundred k and usually connected from HV+ to the base of the switching transistor.

[everell]

Re: Bestec ATX100-5 no +5VSB

Here is a schematic of the 5vsb in a Bestec ATX-250-12E. It should be similar although not exact. Maybe this will help...........

[ben7]

Re: Bestec ATX100-5 no +5VSB

Thanks, I gave that a try and it does the trick. I happened to have an old isolation transformer that I hadn't touched for years.




I did put my scope on it to see there is nothing going into the Base pin, V+ on Collector (through transformer primary), and V- on Emitter. No pulses at all. Guess I'll have to draw up a schematic diagram to determine whether those SMD parts are diodes or transistors.

Yes you should. There is a small capacitor -usually 10uF, sometimes up to 47uF- on the primary side, measure the voltage across that.

Either something in the feedback has gone bad, or the driver transistor is blown.

[GnatGoSplat]

Re: Bestec ATX100-5 no +5VSB

Isolate the DUT using the isolation transformer, not the scope.

More information here:
http://www.eevblog.com/forum/beginne...ge-to-a-scope/

Thanks, I gave that a try and it does the trick. I happened to have an old isolation transformer that I hadn't touched for years.

Buy a differential probe for your scope. Cost is about $200....but a Bestec psu is worth it!

I did put my scope on it to see there is nothing going into the Base pin, V+ on Collector (through transformer primary), and V- on Emitter. No pulses at all. Guess I'll have to draw up a schematic diagram to determine whether those SMD parts are diodes or transistors.

[everell]

Re: Bestec ATX100-5 no +5VSB

How can I isolate the ground on an AC-powered earth-grounded oscilloscope so I can see what kind of signal is going into the base pin of the transistor?

Buy a differential probe for your oscilloscope. They cost about $200.....but a Bestec psu is worth it!

[everell]

Re: Bestec ATX100-5 no +5VSB

Buy a differential probe for your scope. Cost is about $200....but a Bestec psu is worth it!

[b700029]

Re: Bestec ATX100-5 no +5VSB

Isolate the DUT using the isolation transformer, not the scope.

More information here:
http://www.eevblog.com/forum/beginne...ge-to-a-scope/

[GnatGoSplat]

Re: Bestec ATX100-5 no +5VSB

Isolation transformer.

Is that safe? Tektronix has a PDF that says isolation transformers are a big no-no.

Are you sure you put the transformer back together the right way?

Check the power transistor and the transistor that drives it, one or both are likely bad. If they measure good, one or both could be open circuit.

I think I did, I took lots of pics during the unraveling, although I did break the ferrite core trying to get that apart. I've read they will still work even when broken though.

I'll pull the transistors to test out of circuit later on when I feel like messing with it again.

[ben7]

Re: Bestec ATX100-5 no +5VSB

I soldered the little wires back together and re-wrapped the transformer. I put it back in, expecting the PSU to either fry or work. It did neither. No magic smoke, no voltage from 5VSB, nothing, just DOA. Checked primary coil resistance again and it looks fine. Looks like the switcher isn't switching, although I'm not sure if you can tell that with just a multimeter.

How can I isolate the ground on an AC-powered earth-grounded oscilloscope so I can see what kind of signal is going into the base pin of the transistor?

Are you sure you put the transformer back together the right way?

Check the power transistor and the transistor that drives it, one or both are likely bad. If they measure good, one or both could be open circuit.

[b700029]

Re: Bestec ATX100-5 no +5VSB

Isolation transformer.

[GnatGoSplat]

Re: Bestec ATX100-5 no +5VSB

I soldered the little wires back together and re-wrapped the transformer. I put it back in, expecting the PSU to either fry or work. It did neither. No magic smoke, no voltage from 5VSB, nothing, just DOA. Checked primary coil resistance again and it looks fine. Looks like the switcher isn't switching, although I'm not sure if you can tell that with just a multimeter.

How can I isolate the ground on an AC-powered earth-grounded oscilloscope so I can see what kind of signal is going into the base pin of the transistor?

[ben7]

Re: Bestec ATX100-5 no +5VSB

It's open-circuit between the two primary input pins. Actually, it's open-circuit between one pin and the center-tap nub, so I think it's that specific winding.

I was curious, so I decided to perform a transformer autopsy. I had to unwind the main secondary coil to get to the primary winding, but it happened to be the winding directly below that. There was a darkened burned looking spot and both ends of the coil were severed at that spot. It appears to be where the wires were cris-crossed and touching. Only thing I can think of is the enamel wore through at that spot causing that coil to short and it generated enough heat to burn through both wires, but that's just a guess. I've attached a photo, but I took the pic at night so camera flash washed it out pretty badly.

Smoking transformers are no fun, I've burnt out a few myself xD

[GnatGoSplat]

Re: Bestec ATX100-5 no +5VSB

Why do you think the transformer is burnt out then? Its extremely rare for them to burn out.

It's open-circuit between the two primary input pins. Actually, it's open-circuit between one pin and the center-tap nub, so I think it's that specific winding.

I was curious, so I decided to perform a transformer autopsy. I had to unwind the main secondary coil to get to the primary winding, but it happened to be the winding directly below that. There was a darkened burned looking spot and both ends of the coil were severed at that spot. It appears to be where the wires were cris-crossed and touching. Only thing I can think of is the enamel wore through at that spot causing that coil to short and it generated enough heat to burn through both wires, but that's just a guess. I've attached a photo, but I took the pic at night so camera flash washed it out pretty badly.