Antec "Stupid" Power rebuild

[LDSisHere]

Re: Antec "Stupid" Power rebuild

Thanks for this information, it makes me fell much better. I was not sure how accurate the motherboard sensors really were or if what they read was directly off of the power supply when I posted. The readings from my DMM are great and the system has been completely stable so I will put it into service and see how it holds up.

Thanks again,
Lloyd

[mariushm]

Re: Antec "Stupid" Power rebuild

The values shown by Hardware Monitor are the same values shown by the BIOS.

The difference is that in BIOS, most hardware is uninitialized or idle, so it uses little power. In Windows, video card and other things initialize and use more power.

Voltages will change with the load.

HOWEVER, the voltage measurements in both BIOS and Hardware Monitor are NOT accurate.
Best thing is to put the probes of the meter inside one of those hard drive connectors while the system runs and measure the voltages directly. 5v and 12v are most important, 3.3v isn't.
As long as the 12v on one of the power connectors is close to 12v, then you're fine.

[goodpsusearch]

Re: Antec "Stupid" Power rebuild

Motherboard sensors are often unreliable. Not a great issue

[LDSisHere]

Re: Antec "Stupid" Power rebuild

Yesterday, I was able to get the OS loaded on the system I am powering with this supply. Since I was curious as to how the board was performing, I installed "Hardware Monitor" to see what it would display. (screen shot attached) When I saw the voltage numbers I thought my supply was not performing properly under load. Before I jumped to any conclusions I decided to do some more testing and compare the results to what was being reported.

I checked power at the ATX connector while Hardware Monitor was running on the PC so I could do a direct comparison. My DMM was showing the main voltages to be spot on, 3.31V, 5.01V, and 11.99V. Even my -12V is reading -11.5V when the supply is under a real load. I will also say that the BIOS hardware monitor reflects very closely to my tested values.

Is the bad voltage reading a problem with the software, the on-board voltage testing output, or does the board have a genuine problem with its' on-board voltage that could be caused by a problem on the motherboard?

Other than this voltage issue, the system runs great and I have not had any trouble since I put it together. I am just not sure if I have a real problem or if I am just getting false readings so I would like to get some input from others that may have ran into this type of problem.

Thanks,
Lloyd

[momaka]

Re: Antec "Stupid" Power rebuild

I fell much better about it, now to see how long it will last.

With the fan mod and the new caps, probably a lot more than it did before.

[LDSisHere]

Re: Antec "Stupid" Power rebuild

I just tore into a Antec SP350. The resistor in question is R54. The painted color code looks like 150 ohms, but the measured value is 1500 ohms. Based upon what I have seen here and in other brands with the same circuit, 1500 ohms sounds about right. Reducing this value should not stop the power supply completely, but will reduce drive signal level going to the switchers. Try changing the resistor to 1500 ohms without changing anything else and lets see what happens.

Well I knew if I did not go ahead and change out the resistor now, it probably would not happen. So I tore back into it and replaced the 330 with a 1200. I did not have any 1500 and I figured if it ran with a 330 then it should be fine with a 1200. I put it back together and tested it, it is working just fine. I fell much better about it, now to see how long it will last.

[retiredcaps]

Re: Antec "Stupid" Power rebuild

My least favorite is threads where the OP quits halfway though the thread to never report back what happened, especially if someone just typed a long post with a possible solution.

Yep, I'm sick of those as well. I spend a lot of time writing out a response/research and we never hear back.

That and people who don't even say a simple "thank you" after receiving help.

[LDSisHere]

Re: Antec "Stupid" Power rebuild

I will change the resistor out in the very near future, but I have already reassembled it, into its' case and I have a complete computer ready for OS installation. I also have a few other projects to finish up before I have to go back to work on Saturday, so I will have to address this one a bit later.

I am attaching pictures of the case / computer / power supply assembly that was saved. I am quite pleased with the results. I will post back when I change the resistor out or if something comes up before I get to it.

[everell]

Re: Antec "Stupid" Power rebuild

I just tore into a Antec SP350. The resistor in question is R54. The painted color code looks like 150 ohms, but the measured value is 1500 ohms. Based upon what I have seen here and in other brands with the same circuit, 1500 ohms sounds about right. Reducing this value should not stop the power supply completely, but will reduce drive signal level going to the switchers. Try changing the resistor to 1500 ohms without changing anything else and lets see what happens.

[momaka]

Re: Antec "Stupid" Power rebuild

I am sure that those of you that are tired of seeing this thread pop up will be glad to know that this supply is back alive.

Indeed we are. Congrats on the fix!
As for "tired of seeing this thread pop up" - not really. In fact, I do appreciate it when people report back on their findings. My least favorite is threads where the OP quits halfway though the thread to never report back what happened, especially if someone just typed a long post with a possible solution.

Sorry I couldn't be of more help. When it comes to PSU controllers, I am in the dark as well, especially with these older half-bridge controllers. The newer ones are a lot easier, IMO.

The -12V rail does seem a bit low. Not sure if that's typical for your PSU, but like goodpsusearch I'm usually not too worried about the -12V rail either. Very little hardware (if any) uses the -12V rail, so it should be okay.

[LDSisHere]

Re: Antec "Stupid" Power rebuild

I have read that thread many times the past few weeks. The resistor that looked burned on my supply is in a different location. I have circled the replacement in the picture. I may need to up it to 1500 and see if the supply turns on...., the way it is I fear could be in the future. I am not sure which way I should go really because the old one may read what it does due to being burnt.

Everell, I very much apprecite it when I can get assistance from members of this forum, but you should not feel that you are obligated to respond. I very much understand having real life things to do, I just happen to be on Staycation and getting this supply going was one of my goals so I had more time than usual to play. As long as I do not electrocute myself nor burn my house down it is all good.

[everell]

Re: Antec "Stupid" Power rebuild

Read this thread:

https://www.badcaps.net/forum/showthread.php?t=8656

Khron666 had similar problems with his Smart Power repair. The burnt resistor on his power supply was 22 ohms, and there is a picture of the location. I don't know if that is the same as what you found.

Glad you got the power supply working again. Sorry I was so slow in responding. I have been very busy this week with church activities in observance of the Feast of Tabernacles.

[goodpsusearch]

Re: Antec "Stupid" Power rebuild

Congrats! I wouldn't care about -12V. The reading you got was without any load?

[LDSisHere]

Re: Antec "Stupid" Power rebuild

I am sure that those of you that are tired of seeing this thread pop up will be glad to know that this supply is back alive. I am not sure which part was the magic part exactly. I replaced the two H945's with scavenged C1815. The diodes turned out to be 4148's and I had those on hand so they were replaced. The thing I did that I feel uncomfortable with is replacing R51 which looked burnt and had a tested value of 1500 Ohms. I did a search on SmartPowers and "R51" and it came up with Everell saying he replaced R51 with 270 Ohm resistor. Thinking my burnt one had a higher than rated resistance I put in a 1/2W 330 Ohm resistor. I do not know what this resistor is for so I am playing with fire in a way, but it works. If anyone knows the real value for this resistor (R51) I would be grateful if they would share it and the life of my power supply may depend on it.

The Fan now runs a decent speed with the resistor mod. I have to change out the main caps for the new Chemicon's that I have for it. I also still have to do the 5.6V Zener / 5VSB mod and I thought I should use two 1 watt Zener's instead of one.

I did notice that my -12V was only slightly above -10.xxV, I do not have the exact numbers with me but they were closer to -10 than -11. Is this normal for these supplies or do I still have a problem?

I want to thank everyone for their input, it was priceless as I was really lost. This has been an excellent learning experience. I still have a lot of power supply gray or just plain black areas but at least bad supplies do not have to be instant garbage fodder for me any longer.

[LDSisHere]

Re: Antec "Stupid" Power rebuild

I found a couple of C1815 that I was going to use in place of the H945, unless there is a reason I should not. Would BAT85 diodes work to replace the hole mount glass ones (D18, D19,..) tied to the two transistors I am replacing? Should I also replace that vertical resistor next to the burnt spot and transformer, D18 is pointing up at it. It seems to test good with a DMM but I am not 100% of its' value.

Thanks

[LDSisHere]

Re: Antec "Stupid" Power rebuild

I finally have pictures. I removed the two transistors attached to pins 8 & 11. They and the diodes all test good with the DMM diode test. The transistors are H945 and I do not have any on hand with that pin-out so I decided to check the voltages with them out. I put the voltage readings directly on the first image by the pin location.

Pin 16 on the KA7500 came up to 0.73V and pins 8 & 11 went to 3.3V. I still do not understand what all the pins on this chip are supposed to do so I do not know what the voltages should be on them. The datasheet says the typical voltage on 8 & 11 is supposed to be 30V. Is this supposed to come from a transformer or does this chip generate that voltage somehow?

If I need to take pictures of something else or differently let me know and I will do my best to take them as directed. Any guidance on what I should do next would be appreciated very much.

Thanks,
Lloyd

[LDSisHere]

Re: Antec "Stupid" Power rebuild

Thanks, I will see if I can find the components you are talking about and test/replace them. If I am lucky I will get it going, if not I will post back with some pictures shortly.

[everell]

Re: Antec "Stupid" Power rebuild

Pins 8 and 11 are the output from the 7500 chip. Notice that before changing the 7500 chip you had a noticeable difference in voltage. They should have been the same. After replacing the 7500 you still have a slight difference. It is common that when the 7500 goes bad one or both of the driver transistors also goes bad. I recommend replacing both driver transistors AND the small (usually glass) diodes in the driver circuit. There should be 3 or 4 of them. Also check resistor values in this circuit - sometimes a resistor has changed value due to burn damage.

From your voltage readings, it looks like your psu is trying to come back to life. If you still don't get a twitch of the fan when attempting to turn it on, I would look at that driver circuit between the 7500 chip and the small driver transformer. Two transistors, 3 or 4 diodes, and a few resistors.

[LDSisHere]

Re: Antec "Stupid" Power rebuild

I apologize for not posting any pictures. It was raining all day and my parts came in so I was hoping to have this unit working but it did not happen. I did see some improvements but it still will not start.

The KA7500 and TPS3510 have both been replaced with new identical replacements. During voltage testing I intermittently tried to start the supply with a load. After testing I noticed that the 5VSB transformer was warm to the touch. Upon farther investigation I noticed the heat sink between the transformers and the main caps was even warmer than the transformer. I think the only three devices connected to this heat sink is a transistor (Fairchild FQPF 5N606 is what I could make out.) and two diodes. I think the transistor is for the 5VSB and the diodes are for the other two transformers but I have not confirmed this. It seems strange that the heat sink and transformer would get this warm from what I was doing as I would not think it would be a lot of load to be in standby and do some intermittent attempts to start but for all I know this could be normal.

Here are the voltage readings I took after installing the new components. The first number is standby and the second is the reading while attempting to start the supply with a load.

TPS3510

1 0V / 0V
2 4.5V / 4.5V
3 0V / 0.19V
4 2.6V / 0.10V
5 0V / 0.17V
6 0V / 0.08V
7 4.5V / 4.5V
8 0V / 0V


KA7500

1 0V / 0.14V
2 2.48V / 2.48V
3 0.08V / 0.08V
4 4.12V / 0.02V
5 1.57V / 1.57V
6 3.51V / 3.51V
7 0V / 0V
8 2.16V / 1.45V
9 0V / 0V
10 0V / 0V
11 2.15V / 1.44V
12 12.35V / 12.35V
13 4.96V / 4.96V
14 4.96V / 4.96V
15 4.96V / 4.96V
16 0.13V / 0.5V

Pins 9-10 are tied together.
Pins 13,14, and 15 are tied together.

I will try to get some shots posted in the morning, hopefully the sun will be out. I do appreciate the help on trying to figure this out. It is really turning into a good learning experience, but I have no idea where to go from here. Do I have a bad transistor but one that is not shorted? I did not have a fan attached, will this keep the supply from starting?

Thanks,
Lloyd

[everell]

Re: Antec "Stupid" Power rebuild

The reference voltage (internally generated) on pin 14 of the KA7500 should be 5 volts. Since you have 0 volts on this pin, this chip has been toasted. Anyone else agree?

The TL494 chip can be used as a substitute for the KA7500 chip