Not sure what the two 2.2uF 400v caps do. My guess is that they are just for extra filtering on the primary side. Can't say if they are needed or not, though. Can you determine if any heat is coming from the two 30 kOhm resistors or the diodes in series with those resistors? Maybe something in there is underpowered for the job.
As far as the failure of the CapXon cap on 5vsb - what's the resistance of the big resistor next to it and what is it connected to? Could be a heat issue indeed, or it could be that CapXon caps aren't that great to begin with. Either way, replace the CapXon and bend the resistor away from that cap if they are touching or too close.
Not sure what the two 2.2uF 400v caps do. My guess is that they are just for extra filtering on the primary side. Can't say if they are needed or not, though. Can you determine if any heat is coming from the two 30 kOhm resistors or the diodes in series with those resistors? Maybe something in there is underpowered for the job.
As far as the failure of the CapXon cap on 5vsb - what's the resistance of the big resistor next to it and what is it connected to? Could be a heat issue indeed, or it could be that CapXon caps aren't that great to begin with. Either way, replace the CapXon and bend the resistor away from that cap if they are touching or too close.
Hi momaka, thanx for replying. I just realised links from my post about jou jye psu problems are not valid for some reason, so, not being able to edit it, I quote that post with corrected links of the psu.
Originally posted by goodpsusearch
Hi.
Replacement caps for Jou Jye Power Supply (post#66) have arrived, but before recapping, I am concerned about 2 issues with this unit:
The first thing that concerns me is the area among bridge rectifier and first big cap. There is board discoloration, due to heat, visible from both sides of the supply ( upside , downside ). Some big resistors and 2 diodes are found there and very close to them are placed 2 jun fu caps rated 2.2uF 400V.. I desoldered 1 of the caps and measured its esr and capacitance: capacitance: 0.7uF, esr: - (maybe too big to be measured?).
I think that this is bad design, as this area's veltilation is blocked by heatsink, big 1000uF capacitor and bridge rectifier, seen quite clearly here . What's your opinion? Is there anything I could do to reduce heat in there? Or just, replacing with quality caps could be enough to run safe?
Second problem is a bit more complicated. As stated in the post above, I have seen many jou jye power supplies, same model, all having the same capacitor leaking, a capxon 1000uF 16V, measured 0uF, "R only": 3ohm.
It's one of the 2 5vsb caps, this one is placed in the circuit before filter coil and the other is a jun fu 1000uF 10V placed after the coil, esr and uF tested ok.
Heat problem again? There is a big resistor near the cap, but no discoloration here.. Why they fail in all psus I have seen?
I hope I get your advice..
As you can see in the photos, there is heat in that area, coming from the big resistors and the diodes and that gets worse because of bad ventilation there.
I assume this could be just a bad design in placing those parts or a bad desing of this circuit, of which the schematic I have uploaded above.
I replaced every cap in that psu and tested it without load and out of its metal case. It worked nicely, voltages were in range and stable. I let it work for some time with power on sorted with ground and some time only 5vsb. At any case it didn't heat much. The only area that got noticeably hotter than others was the area of that circuit , but not extremely hot.
The resistor near 5vsb cap is 16 ohm and did not generate much heat.
You will be able to see in recap photos that I am going to post soon, how I placed new caps, so that the don't touch heatsinks or resistors. I bended some resistors too.
I would appreciate it if you told me your opinion..
Finally, I followed 370forlife's advice and played with the potentiometers on the riser card on the secondary.
The second pot, with "201" written on controls all the voltages together, (except 5vsb). Turning it at one side drops all voltages, at the other steps them up.
Unfortunately, I got no effect turning the other pots. Maybe they control OVP/OCP/UVP etc, so I restored them at their original places.
Well, if that circuit with the two 400v caps isn't generating that much heat now, then perhaps you can leave it alone. I wonder if putting a heavier load would make it heat up more, though.
Seems to me that the majority of the heat is coming from the two resistors (R1 and R2) behind the bridge rectifier. Are those the resistors that are supposed to be in the circuit you drew? If so, they are 100 Ohms, not 30 kOhms. That doesn't really matter, though, as I still can't think of what that circuit does. If not for extra filtering, perhaps it's for slight voltage-boosting? Overall the two 200v main caps are connected like in any other PSU, so it is possible that this circuit might not be of that much importance.
Seems to me that the majority of the heat is coming from the two resistors (R1 and R2) behind the bridge rectifier. Are those the resistors that are supposed to be in the circuit you drew? If so, they are 100 Ohms, not 30 kOhms.
Yeap, those are the resistors.
I know the colors look different, but maybe they changed due to heat. Brown is in fact orange, so the rating is orange-black-orange which gives 3-0-1000ohm = 30kOhms, that meets the measurement results with the dmm.
Originally posted by momaka
That doesn't really matter, though, as I still can't think of what that circuit does. If not for extra filtering, perhaps it's for slight voltage-boosting? Overall the two 200v main caps are connected like in any other PSU, so it is possible that this circuit might not be of that much importance.
Well, if that circuit with the two 400v caps isn't generating that much heat now, then perhaps you can leave it alone. I wonder if putting a heavier load would make it heat up more, though.
Seems to me that the majority of the heat is coming from the two resistors (R1 and R2) behind the bridge rectifier. Are those the resistors that are supposed to be in the circuit you drew? If so, they are 100 Ohms, not 30 kOhms. That doesn't really matter, though, as I still can't think of what that circuit does. If not for extra filtering, perhaps it's for slight voltage-boosting? Overall the two 200v main caps are connected like in any other PSU, so it is possible that this circuit might not be of that much importance.
I have some news about that circuit.
With power supply ON (fans working, all voltages on)...
I measured the voltage after bridge rectification and found it normal. Here, ac is 230-240V and the dc measured was 324-335V.
I noticed also that there was no much heat coming from that circuit.
With power supply OFF (only 5vsb working)...
There was much heat coming from the resistors and diodes after bridge rectifier and when measuring DC voltage, I got 350-360V!!!
This is way above normal DC voltage generated after bridge rectifier in a typical psu.
So, the circuit among big caps and bridge in that JOU JYE psus boosts voltage when PC is off. With no fans moving, a significant amount of heat is produced, that leads to the board discoloration in just every same JJ psu I have seen...
I decided to put out the small 2.2uF caps and disable that circuit.
Heat produced when psu is in 5vsb only mode is now negligible, and voltage is normal.
I also modded the fan controller, putting in parallel with thermal sensor a 2.7Kohm resistor. Psu is kept cool, and fans don't get very loud fortunately.
Finally, I tested JJ psus in that system: Intel(R) Pentium(R) D CPU 2.66GHz (with gigabyte dynamic overclocking goes up to 3.2ghz), 1x512mb ddr533, nvidia 7600GT, hdd sata 80gb 7200rpm, dvd rom & dvd rw , 2x9cm fans
So, the circuit among big caps and bridge in that JOU JYE psus boosts voltage when PC is off. With no fans moving, a significant amount of heat is produced, that leads to the board discoloration in just every same JJ psu I have seen...
Actually this circuit likely boosts the voltage all the time, not just when the PSU is off. I guess the reason you didn't see any voltage increase with the PSU on is because this circuit is too tiny to supply a boost when under a load.
Originally posted by goodpsusearch
I decided to put out the small 2.2uF caps and disable that circuit.
Heat produced when psu is in 5vsb only mode is now negligible, and voltage is normal.
That's what I anticipated would happen too. Glad you actually took the time to test what that circuit did. JOU JYE probably wanted to improve overall efficiency of the PSU with this but they ended up doing more harm than good. Removing it was definitely a good thing!
----
The voltages look pretty solid.
JOU JYE should hire you to fix their designs .
1)Changed primary caps:
560uF 200V HEC (real capacity measured 420uF) --> 680uF 200V Jun Fu
2)Removed burnt glue from psu that could have become conductive
3)Downloaded datasheet of pwm controller KA7500B and replaced a resistor rated 9.5Kohm that was connected to pin 1 with a 10Kohm trimmer in series with a resistor.
Group voltage regulation in that psu is still terrible, but now one could use it in a low power pc (better be 5v based) and could use the trimmer to make voltages as close to the nominal values.
It can also be used as a variable volt power supply. This is the way I think I will use it rather than trust a computed to it. That's the reason I didn't recap it, but checked all caps with esr micro meter and found them ok.
A month ago, a good friend of mine gave me his desktop pc. He told me that during last 3 months the computer would boot after many times pressing the power button and finally it would not boot up at all.
He bought a laptop and told me he was not going to use desktop pcs again, so he gave me that one. It's a pentium 4 3.06GHZ Prescott Hyper Threading, 512MB DDR2 533MHZ, Intel Desktop Board D102GGC2 mini atx, Gigabyte GV-RC30S128D ATI X300SE 128MB PCI Express, LG DVD RW drive, Samsung 17" TFT and a nice screwless black midi case.
All of the above working and in very good condition.
The power supply that was the cause of non booting of the system was a Jou Jye Electronic 300Watt and had again the same problems as its big brothers above, the 460Watt models
3 Capxon caps were bulging and leaking, 2 of them at 5vsb filtering:
1x680uF 16V and 1x1000uF 10V
and one at -12V output:
1x680uF 16V
This is the first time I see a -12V cap failing. I really can't explain why that cap failed, as it was not placed in heat area, not heavily used as 12V/3,3V/5V outputs and not used even when pc is off as 5vsb caps.
Maybe that Capxon 680uF 16V caps are just very bad...?
This was a lazy and cheap repair project for me.
1)I removed fried 2.2uF 400V caps from voltage boost circuit right after bridge rectifier and disabled it (more info about that circuit at posts above).
2)Replaced -12V capacitor with 1000uF 16V Jun Fu and 5vsb caps with 1000uF 10V Jun Fu and 1000uF 16V Rubycon that I had in stock from the 460Watt Jou Jye recapping and modding project.
3)Replaced 2x1500uF 16V Capxon with 2x2200uF 16V Panasonic FK that I had in stock from the Linkworld Project
4)Replaced 4x1500uF 10V Capxon with 4x2200uF 6.3V Teapo that I got from a dead Enlight 250W Psu .
5)Modded the fan controller with a 2.7Kohm resistor. Insufficient cooling plays a major role in these power supplies failing.
6)Finally, adjusted the potentiometer that group regulates voltages to get voltages as close to nominal as possible.
The psu is up and running without any problems at all. All caps that were not changed and the replacement ones were tested with esr micro 4 meter and found ok!
Nice computer! You have a really good friend there .
However that Pentium 4 Prescott is very power-hungry. Make sure this PSU has good 12v rail that can handle it. 20A schottky on the 12v should be the minimum, really. Otherwise this PSU looks pretty decent for 300 watts.
You can also always exchange it with one of the bigger Jou Jye PSUs you have (if the 12v rail is not adequate).
Nice computer! You have a really good friend there .
However that Pentium 4 Prescott is very power-hungry. Make sure this PSU has good 12v rail that can handle it. 20A schottky on the 12v should be the minimum, really. Otherwise this PSU looks pretty decent for 300 watts.
You can also always exchange it with one of the bigger Jou Jye PSUs you have (if the 12v rail is not adequate).
The 300watt model has 30A schottky for 5V line, 20A schottky for 3.3V and 2x12A for the 12V. Does that mean it has a maximum of 24A at 12V?
It does not have 2 12V lines though. That's a lie in the psu label.
Indeed for that computer I used a repaired Jou Jye 460watt psu, just to be sure it is powerful enough
I got it for free yesterday and I intend to fix it and give it to a good friend.
I was surprised to find out opening it that all its secondary caps, except one, are bulging
I was not planning to do a full recap, just change the failed caps, in other words the usual suspects: 5vsb 2x1000uF and -12V 1000uF Capxon and of course mod the fan controller so it will keep the psu (and the caps) cool.
What concerns me now is why all these Capxon caps failed.... I thought they were not bad caps, just mediocre..
from my experience with Capxons, the KM series is just as bad as fuhjyyuhuhyjujhuu..
in general, i'd replace all capxons on sight. seen waaay too many of them fail.. with the KM and GL series being the worst of 'em all..
I opened that psu again today to make a list of every cap in it to order replacement ones from rs.
That psu had shown excelent voltage regulation (unlike its brother, maxpower 300W NO Pfc), especially when running at that system: Pentium 4 1500mhz willamette, 3xSDRAM dimm total 640mb, TNT2M64 32mb agp, hdd ide 7200rpm 20gb, dvdrom & cd rw, 2x8cm fan.
Carefully examining its components I found them all sufficient, until I saw its schottky rectifiers on secondary side: It has 1x30A schottky rectifier for the 5v line, 1x30A for 3.3V and.... for 12V, that is most critical on pentium 4 computer it has BYQ28E-200, only 10A...
Can I change that rectifier with a bigger one? Will it work flawlessly? How up can I go?
I have no previous experience in doing such a mod. Putting out the heatsink is required?
Can I change that rectifier with a bigger one? Will it work flawlessly? How up can I go?
Maybe. Some power supplies have to have the ultra fast type diodes on the secondary for the 12v for some reason, other times it's because of cost (they are cheaper, also a ton less efficient than schottky)
You can try to replace it with a schottky type and see if it will work. I have done it many times with success, but if it does not work it will simply fail to start up, acting like it is shorted.
As for capacity. You can go up as high as a part that you have. I generally like to use 30A or higher for the 12v to give some breathing room.
I would definitely take out the heatsink. Get some good desoldering braid or a good solder sucker will work well.
Comment