Solytech with no 2 transistor 5vsb circuit?

[370forlife]

http://www.jonnyguru.com/modules.php...Story3&reid=65

My rosewill 500W is similar to this unit, some subtle differences like OCP, 2 actual 12v rails, the resistor that committed suicide is not there (maybe a step up model, or solytech was aware of the exploding), and the 5vsb circuit is wayyyy different.

On that, there is the usual 2 switching transistors on the main heatsink, and a transistor for the 5vsb circuit. Well on mine, it's not there. Instead, right where it would be, is a little 8 pin IC. A Infineon ICE3B0565 (http://www.datasheetcatalog.com/data...CE3B0565.shtml)

So my question is, is this what is generating the 5vsb? A couple of the pins are connected to the standby transformer.

[everell]

Re: Solytech with no 2 transistor 5vsb circuit?

So my question is, is this what is generating the 5vsb? A couple of the pins are connected to the standby transformer.

Yes

[i4004]

Re: Solytech with no 2 transistor 5vsb circuit?

>So my question is, is this what is generating the 5vsb? A couple of the pins are connected to the standby transformer.

surely, not the first time you see IC that steers +5vstby primary?

[370forlife]

Re: Solytech with no 2 transistor 5vsb circuit?

Ok so this deer is definitely worth keeping.

Same capacity on the secondary. 2 30A parts for 5v, 1 40A part for 12v, 1 30A part for 3.3v. Same 2SC3320 primary switchers. Massive secondary chokes, me thinks this is one that will actually put out it's rated wattage.

Just recapped it

12v
16v, 2200uf YC TK -> 16v, 3300uf Samxon RS
16v, 3300uf YC TK -> 16v, 3300uf Samxon RS

5v
10v, 3300uf YC TK X2 -> 10v, 3300uf Samxon RS X2

3.3v
" "

5vsb

2x 10v, 1000uf YC TK -> 2x 10v, 2200uf Rubycon MBZ

[everell]

Re: Solytech with no 2 transistor 5vsb circuit?

Increasing the capacitance on the output capacitors for the +12v, +5v, and +3.3v rails I can understand because you see such a wide variation in loading from mother board to mother board. BUT, doubling the output capacitors on the 5vsb rail doesn't make much sense to me. I don't think there is such a wide variation in loading for the 5vsb circuit. Perhaps I am wrong. Maybe some of our engineering types could elaborate on this?????

For a good working 5vsb I would think a PI filter with a pair of 1000 uF capacitors would be ideal. Doubling that capacitance would add a significant load on the 5vsb circuit. Perhaps I am overly cautious to make changes in component values of a properly working 5vsb circuit that uses a pwm chip.

One thing to remember - the older ATX power supplies had a 5vsb rated for only ONE amp. Then the spec went up to TWO amps. The newer power supplies are rated for 2.5 amps. I don't believe OVERBUILDING this circuit is as important as the DESIGN of the circuit, especially moving from the two transistor design to the pwm chip design.

[i4004]

Re: Solytech with no 2 transistor 5vsb circuit?

^^
yes

and he has put mbz, so he may have increased ripple on it...because originals probably didn't have such low esr....

[370forlife]

Re: Solytech with no 2 transistor 5vsb circuit?

They were the only 8mm caps I had and the old ones were bulged.

I am going to replace them with correct 10v, 1000uf UCC's that I have, but I don't have access to them currently.

[everell]

Re: Solytech with no 2 transistor 5vsb circuit?

I think making that change is a good idea. I seem to remember someone on this forum that changed capacitor values while working on a 5vsb pwm circuit. As I recall, it blew out the pwm chip. I don't remember where the thread is - or its name. Perhaps I will find it again some day.

[i4004]

Re: Solytech with no 2 transistor 5vsb circuit?

oh, you should have no trouble finding that lil gem toasty probably "induced" by suggesting "2x higher capacitance":
https://www.badcaps.net/forum/showth...KI50#post70064

surely the one to remember.

oh, another highlight from that thread, the "bonez' trafo":
https://www.badcaps.net/forum/showpo...8&postcount=16

umh...

[stj]

Re: Solytech with no 2 transistor 5vsb circuit?

you need to treat the 5vsb like any other line,

in the past you didnt,
but a lot of motherboards now have jumper options to feed it to the ps2 & usb connectors for wakeup sensing etc.

if some dick has it configured that way and owns a usb scanner with no external psu or they connect a mobile-fone charglead or something then they are gonna hammer that circuit!

[everell]

Re: Solytech with no 2 transistor 5vsb circuit?

Yes, that's the one. I think that post does merit some review because of discussion on:
1. 5vsb pwm circuit operation
2. Replacing parts with parts of significantly different value.

Due to the arguments involved, I suppose you could call that post as much a classic as the Kerri-Jane at the opera one. But lets not get the arguments going again. I think we all have plenty more to learn. So lets ditch the name-calling and arguments and think about how to EFFECTIVELY troubleshoot or modify the new generation of 5vsb circuits which have a pwm chip instead of the two transistor design.

As I reflect on it now, but did not suggest at the time - once the OKI50 chip was blown, he could have made a few circuit board cuts and added an outboard circuit using the DM311 as I have done many times, or even using another pwm chip. But all of the arguments - errors - and failures to mention - are now just history. So its time to move on.................

Happy New Year - 2010 - to all

[i4004]

Re: Solytech with no 2 transistor 5vsb circuit?

> suppose you could call that post as much a classic as the Kerri-Jane at the opera one.

well, not the post itself(which is just a report), but the complete thread.
only, opera is a matter of taste(and tastes differ) but troubleshooting techniques should be totally "tasteless", if you know what i mean.

offcourse, we don't know what really happened there(but overall i wouldn't suggest putting different values of components at all), one would need such psu to try few things....

i wanted to ask can you find such psu at your local bad psu supplier?
<wink>

[everell]

Re: Solytech with no 2 transistor 5vsb circuit?

At my local bad psu supplier, I did find an Antec true power 550 eps. It had a two transistor 5vsb circuit using 4N60B and H945P transistors. I have never seen anything with the OKI50 smt chip.

Just before finding this forum, I purchased a NEW Antec Earthwatts EA430. It has a ICE2A0565Z pwm chip. This is the Infineon F2 series. The Solytech in this thread is using a later version Infineon F3 series, the ICE3B0565. Same family of chips, just later model. So the Solytech probably has a better chip. The data sheet for these chips looks real good. Guess I will have to get one to experiment with. Mouser has the F2 series available, but I didn't see anyone with the F3 chips available. I will have to keep things within the two dollar budget range!

Overall, I am pleased to see the trend of the newer power supplies going to the pwm chip. Much better.

[i4004]

Re: Solytech with no 2 transistor 5vsb circuit?

yes indeed.

>I will have to keep things within the two dollar budget range!

i presume dm311 fits that budget too?

[stj]

Re: Solytech with no 2 transistor 5vsb circuit?

i read the datasheet for the dm311, but didnt see the maximum current it can pass,
is there a version in a decent package like a to220?

[everell]

Re: Solytech with no 2 transistor 5vsb circuit?

All psu I have seen using the DM311 had the 8 pin DIP package and NO heat sink. My Antec earthwatts with the ICE2A0565Z is the 8 pin DIP package and NO heat sink. Photo of the OKI50 pwm chip was a small 6 pin surface mount package with NO heat sink. My conclusion is that these chips don't generate that much heat for the 2.5 amp max load in the 5vsb rail. The TO220 package used in most two transistor 5vsb designs I have seen were probably overbuilt - could produce far more power than needed - and when they failed put on a spectacular show..............bang, boom, and sometimes fire. The burn areas around the components in the old two transistor circuits is evidence that a lot of power was being wasted. When problem is encountered in pwm chip, it simply shuts down the circuit.......therefore no wasted power, and no excessive heat generated. Some heat sink would be nice - but is it really necessary?

[everell]

Re: Solytech with no 2 transistor 5vsb circuit?

370forlife - how about posting some pictures of your Solytech Rosewill 500 watt psu.

[370forlife]

Re: Solytech with no 2 transistor 5vsb circuit?

https://www.badcaps.net/forum/showpo...20&postcount=4

I can take more pictures though

[everell]

Re: Solytech with no 2 transistor 5vsb circuit?

https://www.badcaps.net/forum/showpo...20&postcount=4

I can take more pictures though

Great - I would love to see what that 5vsb chip looks like in that psu. The chip number you gave (ICE3B0565) runs at frequency 100 Khz, whereas the DM311 I have been using runs at 67 Khz. The ICE2A0565Z in my Antec Earthwatts also runs at 67 Khz. I wonder why Solytech wants to run at the higher frequency???

i4004 - cost for DM311 is $.87, cost for ICE2A0565 is $1.07 (prices from Mouser in USA). So both are within my budget. I was disappointed to see that the ICE3 series were not stocked by Mouser or Digikey. Data sheets show that DM311 was released in 2005, ICE series in 2004. So both are about the same vintage.

[stj]

Re: Solytech with no 2 transistor 5vsb circuit?

higher frequency = smaller transformer.

also needs better inductors & caps though - so not a big cost saving.