Unknown PSU - Uniq LC-500E - should I recap?

[amartolos]

Hello,

I own a power supply which apparently does not exist (according to a quick googling). Most of the capacitors bear only a logo, without name. The others are Canicon, G-Luxon, Koshin and (?) DON, none of them to be found among the "good" capacitors for PSU application.

Here is the full list:

values brand position
__________________________
820uF/200V (logo) 1,2
2.2uF/50V (logo) 9,10,13,A
22uF/50V Canicon 8
1uF/50V (logo) 16
4.7 or 47uF/50V G-Luxon 14
10uF/50V (logo) 23
1000uF/10V G-Luxon 15,17
470uF/16V (logo) 33
2200uF/10V Koshin-KJH B
2200uF/10V DON 24,C,D
1000uF/16V Koshin-KJH E
1000uF/16V (logo) 37

I intend to replace the high-voltage capacitors with Panasonic TS-HC of the same rating, just higher temperature (105°C), the 50V capacitors with Nichicon PW and the 10/16V ones with Panasonic FM (I am ordering from digikey, so unfortunately no Rubycons available).

Here are pictures of the patient:


general view


detail right side


detail middle


detail input caps


The logo in question is clearly visible in the picture with the input caps.

So, what to do now?

Thanks!

Dan

[Th3_uN1Qu3]

Re: Unknown PSU - Uniq LC-500E - should I recap?

The "logo" you are speaking of is Yang-Chun, they're not that bad (i have pulled a bunch of them from various PSUs and they test good), but they eventually find their death next to those hot minimum load resistors. You'll find some power resistors between the secondary caps. Do your best to move them away.

I see a "2005" controller in there - it looks like the standard Solytech (otherwise known as Deer) affair, but with half-decent input filtering and correctly sized transformer. Oughta do 400W, just don't ask it to do it continuously. The primary caps can stay, but replace all secondary caps.

Also replace the small caps on the standby rail, the ones on the right next to the secondary heatsink (3rd pic). And just to be sure, install a 5v1 1W zener between 5vsb and ground. If anything funky happens to the 5vsb rail, the zener will short and save the motherboard.

[mariushm]

Re: Unknown PSU - Uniq LC-500E - should I recap?

Honestly, if you're going to spend more than 10-15$ recapping it, it's not worth it.

Here's a psu I had two power supplies ago (had a Sirtec 460w and now a Seasonic x-650) ... it's labeled Premier 450w. Same capacitors, in the end they just decided to puke.

The design has about 65-70% efficiency, it's noisy, passive pfc, with 20-25$ you can get 80+ power supplies nowadays (granted, in US, in Europe they're a bit more expensive).

[Th3_uN1Qu3]

Re: Unknown PSU - Uniq LC-500E - should I recap?

Bwahahaha, Premier? I thought they didn't sell those anywhere else - wait, you're from Romania too. Anyway, the one he has is beefier, it's got a bigger transformer and better input filtering. His supply is worth fixing IMO.

[b700029]

Re: Unknown PSU - Uniq LC-500E - should I recap?

Those look like they're the same as this:
http://hardwareinsights.com/wp/2011/...rp-vc503-500w/
400W real wattage and 80+ efficiency.

[goodpsusearch]

Re: Unknown PSU - Uniq LC-500E - should I recap?

Premier psus are still being sold here....:

http://www.plaisio.gr/Computers/Hard...U-PR500EPS.htm

[amartolos]

Re: Unknown PSU - Uniq LC-500E - should I recap?

Recapping went without incidents.

Now, that I got the taste of it, I would like to install a DM311 circuit for standby, as I did with my Antec PSU. The original two-transistor circuit looks very similar to the circuit that everell gave as example in this thread. Some differences that I noticed:

- there is no resistor in parallel with the electrolytic capacitor

- the diode D4 is inverted in my schematic. Maybe I did not identify correctly the terminals?

Ignoring these, a schematic for the DM311 mod can be drawn using a somewhat more advanced "monkey see - monkey do" approach just by following the explanations in the above-mentioned thread. Before I put it in practice, I would really appreciate a short check-up, just to make sure I am in the right way. One question about the 4.7uF/50V electrolytic that goes in there: I only have a Jamicon, which AFAIU is not so reliable, so in case I don't find a better one, is it safe to use it?

Aside from this, I also plan to do something about the fans: replace with ball bearings and maybe hang a 120mm one at the bottom intake. There is also a temperature control for them.

[everell]

Re: Unknown PSU - Uniq LC-500E - should I recap?

Your drawings look great.....much better than the sloppy hand drawn ones I posted. I am not up to speed on those CAD software programs.

The Jamicon 4.7 uF capacitor should work fine. The nice thing about the DM311 circuit is that it works.....or if there is a problem, it doesn't work. It won't go overvoltage like the two transistor circuits. On some power supplies, I added a 100k resistor across the 4.7 uF capacitor to stabilize the circuit. Put a LED (with a series resistor about 220 ohms) across the output of the 5vsb. When on, the LED lights. When turned off, there should be a delay, and then it turns off. If when turned off it starts blinking, it needs the 100k bleeder resistor.

In retrospect, the DM311 modification was to PERMANENTLY fix a troublesome 5vsb circuit on a Bestec ATX-250 12E. I attempted to fix it several times, but only succeeded in blowing it out each time I fixed it. I was determined to fix it (or perhaps more stubborn!), so I gutted that two transistor circuit and installed the DM311 mod. That power supply now runs good, and is powering one of my computers to this day. In conclusion, if you find a Bestec ATX-250 12E, don't trash it. Do the DM311 mod, and recap it. It then becomes a great power supply which will do a full 250 watts.

The avatar picture is a Bestec ATX-250 12E with DM311 mod and added fan which runs when the computer is off but 5vsb is on.

[amartolos]

Re: Unknown PSU - Uniq LC-500E - should I recap?

Thanks for the post! You know, something has always to compensate for the lack of knowledge, in this case nice computer drawings

In the meantime I had a look over the datasheet of DM311 and some application notes; they recommend there a capacitor even in the range of 10-50 uF for the Vcc pin, so I thought I could also use the 22 uF capacitor at C8 (Nichicon) in the original circuit, not needed after modification.

I'll keep you updated about the progress.

[everell]

Re: Unknown PSU - Uniq LC-500E - should I recap?

As you increase the capacitance, you are increasing the loading effect on the circuit. I tried several different values, and I found that the capacitance can be increased to the point that the circuit will not start. The value of capacitance for reaching this critical point varies from power supply to power supply. I found a Fortron power supply with this circuit - it had 4.7 uF. My experience is that 4.7 uF worked for all of the power supplies I modified.

[amartolos]

Re: Unknown PSU - Uniq LC-500E - should I recap?

I have made the DM311 modification to the PSU. The output voltage is now 4.62 V, so obviously something must be done to correct it.

The original 5Vsb circuit I posted earlier in this thread had some omissions. Here is the correct one. As I guessed, the diode D4 had the other orientation, and there was also a resistor in parallel with the C8 electrolytic. I also put in this new drawing the secondary circuit, as I may need to change it to get the correct 5V out.

I installed most of the components of the DM311 circuit on the PCB of the PSU because I removed all the components and there were enough holes and traces. The 5.6 V Zener (SMD type) was installed directly at the opto-coupler on the solder side. I only had to cut the trace between the pin 4 of the opto and the feedback wiring of the transformer. The DM311 chip was installed on a smaller perforated board, together with the three 330k resistors suplying the startup voltage. I have fixed this board to the primary heatsink where the original 5Vsb transistor was bolted, in such a way that the DM311 rests with the top against the heatsink, and put a silicone pad in between. The thermal contact is probably the worst possible, but heatsinking should be provided by the metallic foil soldered to the pins 6,7,8 of the drain output.

I did try the 22 uF capacitor at the Vcc pin, but with it it could not start, just giving pulses of voltage. So I used the 4.7 uF Jamicon.

The charging voltage of the mains capacitors is 315 V, so no problems here. Also, the voltage at the Vcc pin is about 7.8 V, which should be within specifications.

So I guess I have to change something in the secondary side, which probably means to adjust the R17/R18 ratio that give the reference for the voltage regulator? I did not write their value, because I have to look again at the colors when I get home. I found brown-brown-green-yellow-brown for R17 and brown-brown-blue-green-brown for R18, but this seems to give non-standard values. BTW, the layout of the PSU board is very similar to this one posted by everell, so maybe he can give me some hints, considering also his expertise with DM311?

Thanks!

[amartolos]

Re: Unknown PSU - Uniq LC-500E - should I recap?

I ended up by removing the R17 and R18 resistors and measuring them off board. They have values of 1.45 kΩ and 1.50 kΩ, respectively. I remembered having read in an earlier post what resistor and by how much to modify, so I started decreasing the value of R17. For 1.18 kΩ the voltage comes up to 5.18 V, and the PSU starts when shorting the green wire, powering a 45 W resistive load.

However, a load of 3.8 Ω on the standby rail drops the standby voltage between 3 and 4 V, and consequently the PSU can no longer be started up, also when hooked up to a computer. I noticed the voltage drop also with the original R17/R18 voltage divider. So I guess I will have to troubleshoot the DM311 circuit and see why it cannot handle the power. I hope I did not damage the chip with heat when soldering

Any suggestion here is greatly appreciated.

[amartolos]

Re: Unknown PSU - Uniq LC-500E - should I recap?

I think it was the DM311 which was faulty (I probably overheated it when soldering the heatsink to the drain), because now, after several tests with different configurations, the feedback pin is shorted to ground. I somehow managed to kill it

What a bummer

[everell]

Re: Unknown PSU - Uniq LC-500E - should I recap?

Buying a replacement DM311 should be cheap enough. I really have enjoyed reading about your experiments and hope you will continue.

The 5vsb is a small circuit all to itself, so is easy to investigate and experiment with compared to the psu as a whole.

I have found that results vary from power supply to power supply, especially transformer to transformer. Some transformers work great, some not at all.

Looking at the size of the 5vsb transformer in this psu, it is smaller in size than most I have used, and I would predict that it will not put out more than one amp without causing other problems. As problems develop, the two transistor circuit tries to compensate by putting out more power and self destructs. The DM311 reduces output or stops when problems are developing. This protects the mobo.

Hope you will try the DM311 mod in several different psu. I am skeptical as to how well it will work when used in Deer/L&C/Solyteck and chip of the year power supplies. You can check the the circuit by using a handful of different resistors for loading. I use ten ohms for half amp, five ohms for one amp, and 2.5 ohms for two amp. If it won't hold a half amp load, there is a significant problem. If it will hold a one amp load, that's reasonable. If it will hold a two amp load, that's great.

[amartolos]

Re: Unknown PSU - Uniq LC-500E - should I recap?

Thanks for the reply. Actually, while it's not impossible to get another DM311, it won't be so easy, so I started to think about other chips I could use in its place. One question arises now: how do I find out what was the switching frequency of the original circuit? I see that this type of controllers work with either 67 kHz or 132 kHz. The Antec PSU I modded earlier obviously had 67 kHz, but is this the same for any SMPS for PC use? Searching for the standby transformer that this supply uses (EEL-19-2005, apparently meaning a EEL 19 core) found results pointing to 132 kHz, but also statements concerning a wider frequency range (1-200 kHz). So I can use whatever frequency I find?

PS. To moderators: the original question in the thread title has been answered. Should I open a new thread concerning the 5Vsb modification?

[b700029]

Re: Unknown PSU - Uniq LC-500E - should I recap?

It's variable-frequency, depending on the load.

Here is a good article all about it:
https://www.badcaps.net/forum/showthread.php?t=13909

[amartolos]

Re: Unknown PSU - Uniq LC-500E - should I recap?

Wow, exactly what I wanted! Thanks for pointing to it, and thanks to Th3_uN1Qu3 for sharing the paper.

[amartolos]

Re: Uniq LC-500E

After some time I managed to get back to my PSU. I managed to get the 5Vsb working using a TNY278P chip. I chose it because it needs a minimum amount of components (see schematic). It does not even need a feedback winding because it gets its power from the drain pin. Not even the two 560k resistors are needed for operation, they are required to prevent restart glitches when turning off AC power. And what's best is that the only required component is a 0.1uF cap, so it doesn't have to be electrolytic.

For implementing it into my supply, the datasheet had enough information (actually, I looked in the datasheet for TNY26x series, where circuits for 5V/10-15W standby supplies are given). Because it has so few connections, I tried to tinker around a bit so that it fits directly on the PCB, without the need of additional wires or boards, using the existing holes and traces left after removing the components of the two-transistor circuit. I've put it very close to the transformer and it is also quite close to the optocoupler. The 0.1uF cap was soldered directly on its pins, above the chip. Heatsinking should be provided by the large area of the copper trace where the source pins (5,6,7,8) connect, in my case it was a metal foil soldered to the pins in question, in a vertical position. Being grounded, no big precautions must be taken for insulating it. Mounting the chip like this should not affect too much the airflow inside the PSU.

There is however an issue now, discussed also in the datasheet: the C7 capacitor from the snubber circuit squeals, it has to do with the mechanical vibrations of the piezolectric between the plates. When inside the case and inside the computer it is not so loud, but still audible, so I will try to correct it by replacing C7 with a film capacitor (quite big at this rating, 2.2nF/1kV, I have to see that it fits), one with a non-piezo dielectric (they seem to be rare and expensive), or maybe using a different method of clamping the reverse voltages (a Zener across drain-source?)

BTW, it was easy to locate the source of noise using the trick found somewhere here, with a sheet of paper rolled into a long tube

In the output side, I did not have to modify anything (the original voltage divider resistors were OK). It gives 5.08 V with no external load, decreasing to 5.06 V at 2.6 A.

Finally, I replaced the two 80 mm fans (Xinruilian, sleeve bearing) They were actually quiet, but I did it for reliability. I used Sunons with maglev/vapo bearing (they were cheap and they seem to have long life) In normal conditions, they run below their maximum power, so they are also pretty quiet. The intake fan is now a 120 mm fan - I had to cut the panel but it fits nicely inside. The other fan (80 mm) is not mounted at the exhaust, like the original, but on the opposite side, pushing air inside the PSU (there were some mounting holes already there, and venting slits that I removed). Without modifying anything in the control circuit, there is a considerable amount of air coming out of the supply (it blows me away... ), so I think I have some headroom for lowering even more the RPM, making it really whisper quiet. I think the tip given by Th3_uN1Qu3 in this post should work here.

[Scenic]

Re: Unknown PSU - Uniq LC-500E - should I recap?

edit: nvm.