Junk Fu or not Junk Fu ...

[TELVM]

Hi guys! I'm recapping this PSU and I'm not sure about what to do with those three small guys that I've labelled with the number '3':

=> Click here to expand

They're 50V 1uF Junk Fus (thus crying out loud for replacement ), but the thing is I don't have any 50V 1uF caps at hand. However I have some Pannys FC 50V 2.2uF around.


What would you do?:

1) Replace the 1uF crap with the 2.2uF good ones.

2) Leave the 1uF Junk Fus and pray for the best long-term ...

.

[Drack]

Re: Junk Fu or not Junk Fu ...

I think you could use polyester or ceramic caps to get close to 1 if thats what one of my PSU uses, is better for reliability I think. is that a hi pro?

[jazzie366]

Re: Junk Fu or not Junk Fu ...

I'm kinda iffy about Jun Fu caps. I've had some rather interesting experiences with them (and a lot of other caps) that made me reconsider a lot of the "junk" caps.

I have a Bob Carver Sunfire True Subwoofer MKII and it is enirely capped with Jun Fu capaitors. None failed except for two which were too close to the 2700W amplifier fets that of course, got quite warm. The thing is though, all the other board caps are in fine shape and the amplifier sounds amazing after putting two Nichicons in place of the bad caps. Closer inspection revealed the caps had dried out from heat and shorted, causing the failure. Other than that it has no problems and sounds crystal.

I also had a TV box that was at least 10 years, probably more old that was made by Magnavox. It was given to me by my grandmother who thought I could still use it, (and as a kid, I actually did use it for a short time). After that, I found it again many years later, opened it up and same story. All Jun Fu caps, except none were failed or even showed signs of failure.

So that led me to believe Jun Fu's are like CapXons, keep them inside of their spec and except for a few bad series they're not terrible capacitors compared to things like Chong, which all of which I've come to encounter don't vent at all they either explode or blow their bungs out.

So personally, if the cap is just a turn on cap, or something that's just going under minuscule load, leave it it'll probably be fine. If it's a secondary cap on a shitty PSU that's probably running it out of spec, then replace it (if it hasn't gone bad already).

By the looks of the pic, there are more than just those Jun Fu's on the board, I see some on the output and the main filter is one.

Your choice, you can recap it all or leave the smalls.

[Drack]

Re: Junk Fu or not Junk Fu ...

Must say that almost all the gluxon and teapo that I have saw are in good shape and on working boards or that are damage but by other things, not caps, capxon/jun fu/G.I/KZG have failed to me almost everywhere I found them

[jazzie366]

Re: Junk Fu or not Junk Fu ...

Must say that almost all the gluxon and teapo that I have saw are in good shape and on working boards or that are damage but by other things, not caps, capxon/jun fu/G.I/KZG have failed to me almost everywhere I found them

I've encountered shit Teapos, more than once and I don't like Teapo. I posted once here about Teapos in an Antec VP-450. Well, I replaced the input filter caps and it was fine again. Then guess, what I got another of the same power supply. Same failure, resistance across the cap and low capacitance. Rated 820uF, read 690 when cold, when at operating temperature it would go down to around 500uF. The secondary caps are also Teapo, but they seem to not fail so I don't touch them. The HV primaries do all the time though.

It's weird, I guess it's just where the caps are placed. The Jun Fu's I've encountered were in a pretty expensive analog TV box. The thing when new was probably 50-100$, but was pretty small. Worked well though, never had a problem. It never really got that warm either, so maybe Jun Fu's just don't like heat?

Who knows, I guess it's all situational. But I don't judge caps by brand, I judge them by experience and reputation.

[TELVM]

Re: Junk Fu or not Junk Fu ...

... is that a hi pro?

It's a CWT GPA platform inside an Antec VP-400PC.

I think I'm going to replace the 1uF Junk Fus with the 2.2uF Panny FCs and see what happens.

.

[TELVM]

Re: Junk Fu or not Junk Fu ...

PSU works fine with 2.2uF Panny FCs in place of 1.0uF Junk Fus. No explosions.

BTW voltage regulation is much better after fitting real caps (Panny FC/FM/FR of 2200 & 3300uF) than with the original manure (Junk Fu WG/WL and Su'scam SD of a rachitic 1000uF).

.

Interestingly the only two decent caps that came OEM with this PSU are the bypass cap for the TNY176PN +5VSB PWM controller (Chemicon KY 50V 10uF), and the first +5VSB filtering cap (KY 16V 1000uF 8mm).

[Drack]

Re: Junk Fu or not Junk Fu ...

Is that a Sanyo cap?

[TELVM]

Re: Junk Fu or not Junk Fu ...

Yep , Sanyo WG 16V 1500uF.

[Drack]

Re: Junk Fu or not Junk Fu ...

Lol, Sanyo/Suncon still make those?

[TELVM]

Re: Junk Fu or not Junk Fu ...

I don't think so. It's a 'vintage' one, but in good health.

[Drack]

Re: Junk Fu or not Junk Fu ...

From what I saw they don't, but who cares if it work alright

[Behemot]

Re: Junk Fu or not Junk Fu ...

Yes they do, I plan to stock some by the end of the year. It is the last ultra-low ESR series still being made.

As such it is not much intended for SMPS.

What cap did you put after the PFC?

[TELVM]

Re: Junk Fu or not Junk Fu ...

Yes they do ... ... It is the last ultra-low ESR series still being made ...

Interesting, good to know.

... What cap did you put after the PFC?

If you mean the bulk cap, I left the OEM Elite GM 400V 180uF 85C. Rachitic in size but should do for the application.

.

[Behemot]

Re: Junk Fu or not Junk Fu ...

I am first getting some new low-ESR Rubycons for PSUs, have just payed the proforma invoice this week, but the Suncons are next (in about four months from now). Wanna get 1800/6.3 D8 and 1500/16 D10 first, don't have these or any similar on stock. It is not the lowest-ESR series, being cross-reference to KZG and MBZ, but the best available today, and still OK even for Vcore. And it is also interesting that they state longer than usual endurance, 1000-4000 hours, where most of the ultra-low ESR series only stated 1000-2000 hours.

Even for 350 W version 180uF cap is too low and does not even provide hold-up time in spec. There is little space but with some effort you may be able to squeeze some tall versions of 330uF in there (I did ), with that the HUT increases to about 14 ms. I did not check efficiency but the experiment with Corsair CX550M suggests these platforms may actually even get better efficiency if you increase the capacitance.

[TELVM]

Re: Junk Fu or not Junk Fu ...

Yep did some thinking on replacing the bulk cap.

The OEM Elite is 22x35mm. In GPA there is very little if any room for a wider cap, and height clearance is ~50mm before cap sticks its head in the fan.

There is a nice series from Rubycon that comes handy when parking slot for bulk caps is tight: MXK.

One could squeeze a MXK can of 22x50mm with 390uF 400V, or 330uF 420V / 450V.


But ultimately I chose to leave the Elite, for this PSU will power up a ~200W tops office comp lacking discrete GPU, so low uF ain't that critical (though higher voltage rating and a more reliable jap brand wouldn't harm longevity).

[Behemot]

Re: Junk Fu or not Junk Fu ...

I think I've even squeezed 25mm one in there, as I always have something laying around from other units I've replaced caps before (or ditched as bad), no need to buy them. These sausage-caps don't come cheap…

[TELVM]

Re: Junk Fu or not Junk Fu ...

... the experiment with Corsair CX550M ...

^ Very interesting and enlightening : http://www.hardwareinsights.com/wp/c...esign/7/#Bonus

[TELVM]

Re: Junk Fu or not Junk Fu ...

Addendum - Fine tuning VReg

The +12V voltage was still a bit low for my taste, beginning at 11.95V when PSU was idling (out of comp, no load, just jump-started) and dropping to ~11.7V @ ~250W load ...

so I investigated about what could be done on the subject.

In this PSU the PFC/PWM controller is a 9-pin CM6805BSX, of which I can't find a datasheet.


However a did find a datasheet for the CM6903A, which is also 9-pin and looks very similar. Scrolling down you can find this application circuit ...

As it turns out, CWT GPA uses a similar circuit for voltage feedback and regulation ...

So I replaced the 39.2K resistor with a 0~47K potentiometer ...

And adjusted the pot with a screwdriver to fine-tune the voltage regulation to my liking ...

Now the PSU idles at 12.25V and drops to ~12.0V @ 250W.

[Behemot]

Re: Junk Fu or not Junk Fu ...

CM has the datasheet on their website (they only list the basic version each time, e. g. CM6805, without all the other letters which most likely specify package and materials). I have a copy http://www.hwworld.cz/files/Datashee...cro/CM6805.rar