Antec SP500P "resurrection"

[Khron]

Re: Antec SP500P "resurrection"

The second H945 measured quite ok i suppose, but i replaced both of them. I also took out the two 1N4148's across them, and they measured ok, so i soldered them back in.

Now to put together the switcher.....

[Khron]

Re: Antec SP500P "resurrection"

Done with the switcher as well

I couldn't help myself and i put the IC switcher together on a small scored pcb, in an SMD-ish manner. Looks crude, but it should be effective enough.

I cleared off a ~5mm strip of copper from underneith the IC, hope it's enough creepage clearance (or whatever that's called). And if need be, i think i've got room to slip a cable-tie underneith the IC to tie on some heatsink of sorts.

With the risk of seeming bone-headed or "greedy", everell , i'd really appreciate a sort of "walkthrough", through your 11-step mod that you showed HERE .

Many thanks in advance (again)

[everell]

Re: Antec SP500P "resurrection"

Download those photos and put them in a folder. Then look at the file name of each .jpg file. The steps for the modification are numbered on each photo, along with the instruction as to what to do. See if that helps.

I mentioned removing two jumpers on top of pc board. You will also need to remove the main 5vsb switching transistor - the one I circled on the photo of your heat sink. Remove the resistor as shown in photo step 7. There are several parts in the old 5vsb circuit that could be removed - I just didn't.

Your mod board looks very clever - like some of the things I have made. Just be sure you have it securely attached to something. I don't want you to report back a big boom and smoke!

If you get nothing from the 5vsb after power up - it only means you will have to troubleshoot a problem. Much better than getting TOO MUCH as you do in the two transistor circuits (including boom, bang, poof, etc).

If you aren't sure about which color wire is which on the photos, the colors correspond to the circuit drawing I posted earlier in the same thread. Hope this helps.

[Khron]

Re: Antec SP500P "resurrection"

Regarding my board, i was thinking of drilling a hole through the lower left corner (which is the "ground" of the rectified primary, or "common" or however you wish to call it), and bolt it to the primary heatsink (tied to the same signal).

Thanks very much for the indications. I wouldn't have guessed anytime soon that the indications might be in the photo names I'll actually have to skip drilling a hole and installing the red wire; the chip i'm using has some internal current source to power itself up until it starts switching

Your help has been quite invaluable I hope i'll be able to report tomorrow that "It's alive!.."

[everell]

Re: Antec SP500P "resurrection"

If you are only testing the 5vsb mod, you do not need to install the two heat sinks and their components. BUT - examine the circuit board for any solder bridges between circuit traces. There have been times I neglected to do this, and heard the big bang! Also when powering up, if the pc board is outside of the case, be sure it is on a piece of cardboard or paper - something non-conductive. Also before powerup, don't forget to put on some safety glasses. This is very important.

I just added to an old post about a handy ATX power supply tester. I use mine often. When you power up, that 5vsb LED comes on, or there is a problem. Using this tester, you know your answer quickly, and can power down if there are problems.

[Khron]

Re: Antec SP500P "resurrection"

Ok, i hooked up the mod-board today, and did a quick test, with a lighbulb in series with the AC plug, and a 12volt fan on the 5vsb, as a load (rated at 120mA).

First try, nothing happened. A few mins later i realised that with the primar heatsink out, there was no PFC diode on the board. Soldered in a thick wire instead, plugged the whole thing in, lightbulb flashed, and the fan turned on

I measured the voltage on the second cap (of the CLC filter) and the DMM said 5.40v. A tad high, but somewhat workable. Maybe i should find and trim one of the feedback resistors? Or just to trim the 10K resistor on the phototransistor side of the optocoupler? (The VIPer12A uses current-control. And in all truth, the application note DOES use a 9.1k where i used a 10k which i had at hand)

For now, i'm just happy nothing blew up

[everell]

Re: Antec SP500P "resurrection"

Congratulations ... great work. Now for the tweaking part.

1. The 5.4 volt output is not that bad. You might want to just leave it alone.

2. The Application note does use a 9.1K instead of 10K resistor. I would change it and note the output voltage difference. My guess.....you probably won't see that much difference.

3. The voltage regulation set point is determined by a voltage divider on the reference pin of a TL431 voltage regulator (part number varies, but usually has 431 somewhere on the regulator - looks just like a 1815 transistor). The voltage is going to vary somewhat from 5.0 volts due to parts tolerance differences AND operating frequency due to transformer parameters. This is the reason I was a little skeptical of using that TNY chip running at 132 Khz. The 431 voltage regulator you are looking for is on the OUTPUT side of the 5vsb transformer. The pin 3 of the 431 regulator usually goes to the optocoupler pin 2. I'll look at your photos later and see if I can spot anything.

Success is such a great feeling, and hope you have many more. Now to get the rest of the power supply up and running.

[Khron]

Re: Antec SP500P "resurrection"

I can't take all the credit here, maybe only for the labor Thanks again for all the pointers and advice; couldn't have gotten far without those.

Yeah, i think i noticed a to-92-cased "something-431" right on the edge of the board, near the secondary of the transformer. If i can beat my laziness after a filling dinner, i'll try to find a 100k in my parts bin and put in parallel with that 10k and see what i get on the output. And if that won't help, i'll try to trace the secondary feedback circuit.

Next step, i think i'll solder in the replacement TL494, and i'll take the PSU to school, to scope out the "gate drive" section, just to be on the safe side.

Or should i check to see if the PFC is still "alive" first? I was just talking to my dad about this, and he reckons a cap went bad, high spikes went through the zener which in turn blew that resistor as a "fuse", so the PFC chip might actually still be ok. No darkened board or burn marks in that area, so it might still be ok.

[everell]

Re: Antec SP500P "resurrection"

Here is photo of top of your board showing location of that regulator chip. I can't seem to read the part number on it.

The voltage divider network which sets the output voltage is determined by the voltage divider R49 and R50. They are both surface mount resistors. That means lots of fun. Working on them means you learn a lot and become very professional in your work skills!

R49 looks like 5.10K, can't read R50. Either increase R49 slightly (which means replacing the resistor) or reduce R50 VERY SLIGHTLY (which can be done by bridging a small resistor across it - perhaps you could start with 100K and see how much that brings it down.)

[Khron]

Re: Antec SP500P "resurrection"

After many iterations and resistor combos, i ended up paralelling a 47k with the 5.1k already on the board, which brought the output down to 5.13v. That's good enough for me, considering it only had a 120mA load; i figure it's gonna go down a bit more when it'll be powering the mainboard, etc.

Thanks again so much for the assistance

Ok, so what should be first on my agenda for next week? PFC test or PWM test?

[everell]

Re: Antec SP500P "resurrection"

My experience with active PFC is limited - Toasty can probably give much better pointers.

For what I have worked on, the APFC had to be working properly before the main pwm would come on. So my best guess is do the APFC next. Problems I encountered was bad mains cap (bulging), bad driver transistor between APFC chip and shorted FET. The APFC chip was OK. I don't know if you have the driver transistor between your APFC chip and FET. If not, a bad FET will usually take out the chip driving it.

You mentioned earlier measuring Source to Drain on that APFC FET and that it had resistance. That doesn't sound right. If you were measuring it while in circuit, it is believable. So measure it again while out of circuit. Gate to Source should have a resistance reading, but source to drain should be open. Since you now have the 5vsb working, I am assuming that the main fuse is OK.

As for the burnt resistor in APFC circuit, I have no idea what value it should be.

You have indicated that the driver transistors for the main PWM chip have been replaced with 1815 transistors. That only leaves the KA7500 chip which you said you were replacing with a TL494. Put it in, double check for solder bridges, and let us know what happens.

[Khron]

Re: Antec SP500P "resurrection"

With red probe on source and black on drain, i'm getting 750-800Kohms, out of circuit, but gate to drain or source seem open-circuit. Could that be right?

The PFC chip is driving the FET directly (ICE1PCS01 and IRFP460A respectively), so no bad driver transistor in the middle.

The main caps (Fukh-yoo) WERE bulged, but i slipped in a pair of Nichicons (as can be seen in the photos).

Yeah, the fuse is ok. I checked it at the first 5vsb test, before i remembered about the PFC diode

Regarding that resistor, i'm thinking of putting a 22ohm in there. The PFC chip works from 10-11v to 21v, and i doubt it draws THAT much current.

[everell]

Re: Antec SP500P "resurrection"

I don't think you are checking that FET correctly. The pins, left to right, are gate, drain, source. Drain to source should have no reading both directions. Gate to source should have a reading one direction only. Try one more time.

[Khron]

Re: Antec SP500P "resurrection"

Just re-did a test as indicated HERE , and everything seems to be in order. There's no "tanned" board in the PFC area, nor any other indication of trouble.

[everell]

Re: Antec SP500P "resurrection"

Sounds like you are ready to put it all back together and perform SMOKE TEST #1. When you power up, don't let any smoke out of the transistors!

It must be nearly midnight there. Do you go to school and work tomorrow? I have off for holiday tomorrow. Right now we are getting lots of rain, thunder, and lightning. Guess you are probably getting snow?

Keep us posted on progress.

[Khron]

Re: Antec SP500P "resurrection"

Yes, it's 1:20 am over here, and yes, i DO have to get up at 7am

But hey, you only live once Besides, i'm not done talking to a friend

Hopefully, sometime next week i'll have a working PSU, which would only leave me to wait for the case, hdd's and ram for my "media server"

Hmmm... I was just thinking, the TL494 / gate drive test would be easier without the primary heatsink in place (ie. before testing the pfc)...

The snow's been here for a while, we're just "keeping" it -5 to -10 centigrade. Bearable, almost nice. But i could still go for a car

[everell]

Re: Antec SP500P "resurrection"

You would think so.....but.....the supervisory circuit looks for +12 volts, +5 volts, and +3.3 volts on the outputs. If not there within milliseconds, the supervisory circuit tells the pwm chip (TL494) to shut down (thinking that there is a problem). There are ways to cheat and tell the TL494 pwm chip to come on anyway, but not tonight. Tell your friend hello for me, and this will be my last post for tonight.

Have a good night (er.......good morning in your world!).

[Khron]

Re: Antec SP500P "resurrection"

Phase 2 complete and successful!

Read: i changed the small caps on the primary side, and soldered back the primary heatsink. Nothing blew up, the fan on the 5vsb line still working, around 162-164V on each of the two big caps, around 2-300mV dropping on the PFC diode, 10.8V supplying the PFC chip. But now that i check the datasheet again, that might be a tad low It says it only starts working once Vcc gets over 11.2V.

I have the feeling i should be a bit concerned with that, but... i can't really decide.

I think i'll leave the rest of the re-capping and mounting the TL494 for tomorrow.

[Khron]

Re: Antec SP500P "resurrection"

PS: I haven't measured it, but the line voltage would have to be almost 260Vac, in order to get ~160V over each of the primary caps, so... i'm guessing the PFC actually DOES work.

[everell]

Re: Antec SP500P "resurrection"

According to information I found on the web, Finland has 230 volt 50 Hz power. So that 230 volts goes thru the bridge rectifiers to give you 230 volts DC. This is applied across the two big capacitors. They charge to the peak value, which is 1.4 times the voltage from the bridge. So you should see about 325 volts across the two capacitors, or about 163 volts across each capacitor. The voltage generated by the APFC FET across that big torroid coil is then added to the 325 volts, which should bring it up to about 390 volts.

I have looked at your circuitry and it is much different from previous PFC circuits I have worked with. On mine, the pwm circuit would not work unless the pfc circuit worked and increased the voltage to about 375 volts or so. Yours looks like it will probably work even without the pfc working. Not absolutely sure, but probably. You can determine the amount of voltage generated from your APFC circuit by measuring on each side of the APFC diode. The difference between the two readings is what that circuit is doing. The negative side of that diode should have about 390 volts on it.

Keep us posted on progress.