Operation of the 5VSB system?

[Per Hansson]

Hi, I have a simple question on the operation of the 5VSB system
Do the components running off it (chipset, NIC's, USB etc) continue to run from this rail when the system is powered on?

I know from inspecting the power circuits of the motherboards RAM slots that there is a transistor that switches the RAM from 5VSB to 5v or 3.3v when the system resumes from standby

But do the other circuits have this too?
Reason I ask is that lately my cardreader and USB stick have stopped working, so I am suspecting a malfunction amongst any of the computers I use (Including people I know)
It could of course be just a coincidence too...

[Per Hansson]

Re: Operation of the 5VSB system?

Bump
Nobody knows?
My Google skills seem to be of no help on this quest either...

[Krankshaft]

Re: Operation of the 5VSB system?

There would be an easy way to test this clip the 5vsb wire and hook a meter set to measure current across it.

Then power the PC on and see how far it drops.

If you have a DC current clamp you could do this less invasively.

I would think the latch circuitry that keeps the PC on with the momentary push button switch would remain connected to the 5VSB and some wake on lan NICs run while the PC isn't powered up. Some mobos send 5VSB power to the USB ports with the PC off. Can't think of much else that uses it.

If you're devices have stopped working check the 5V source at the USB port most mobos have a protection fuse that will blow if a device tries drawing more than 500ma.

[Per Hansson]

Re: Operation of the 5VSB system?

Thanks Krankshaft, I was actually considering going the meter route, didn't want to cut a wire in the nice PSU tho, but I have one here with a non-functioning 5VSB that I might use for testing, will force me to fix the PSU too

About the USB ports on my computer, they work fine, other devices works fine on said ports...
It's just a cardreader and USB stick that has stopped working for no apparent reason
But it seems I might be zeroing in on the culprit because today I was at a friends place and he too had a bad USB stick, I suspect it is a friends of his that has the culprit computer

[Per Hansson]

Re: Operation of the 5VSB system?

Ok, I fixed the PSU, it was easy
2x 2200uF 10v bulging caps on 5VSB secondary side, replaced with 2x Rubycon 3300uF 6.3v MBZ
(My rubies are getting kind of old so want to use up my stock, otherwise such a nice cap would have never gone into this thing )

When the computer is soft-off it draws 1.5A from 5VSB
When I start it this drops to 0.3A
When I enter S3 standby mode it also draws 1.5A
If I remove the NIC cable it draws 0.2A less in soft-off and standby, while running there is no difference...

I tried disconnecting 5VSB when the computer was running
I lost the VGA output and the CPU fan turned off but the system was still running, because my power meter showed 81w... The PSU fan was on aswell... (When running idle it draws 77w which seems to be correct, because 0.3A x 5v = 1.5w + the CPU fan)
So I don't know what it uses 5VSB for while running but it seems to be something crucial
Also I was pinging the computer from another computer over the network and it stopped replying as soon as 5VSB was gone, however the NIC LED's where still lit...

The 5VSB was very stable after my 8 year old Rubycon caps had a little time to warm up
It was at 4.0v first but I left it plugged in for 10 minutes and then it had risen to 4.97v
And it keeps this voltage stable in a transient test going from running in idle to standby mode (I don't have an oscilloscope though)
As for the original caps what are they, that S logo inside an eye?

As for the PSU this is how it looks, not half bad actually
Good input filtering (there is extra filtering on a small PCB behind the power receptable aswell, consisting of 2x Y caps, a Ferrite coil and an X cap)

Also if you want to see something EXTREMELY funny look at this old post by me, of a PSU with an identical model number, but higher ratings on all the outputs...

[POM_MJ]

Re: Operation of the 5VSB system?

Those are CS caps, known crap brand.
I recommended to replace all of them. They've much leakage when they failed then cause silicon blown out.

one of mine => https://www.badcaps.net/forum/showpo...3&postcount=95

[goodpsusearch]

Re: Operation of the 5VSB system?

This is a macron power mpt-301, a very good 300watt psu.

I have a unit like that that had bulging primary 680uF cs caps.
https://www.badcaps.net/forum/showpo...86&postcount=3
My psu also has a bridge rectifier of 6 A, not 4 diodes and larger 13009 primary transistors. It has mov too...

I replaced all caps with panasonic and rubycons and the psu works like a charm.
https://www.badcaps.net/forum/showpo...5&postcount=48

I think Rubycon MBZ are a bit too low esr for psu.

About the "CS" caps... http://capacitor.web.fc2.com/
website: http://www.12kj.com/

[Per Hansson]

Re: Operation of the 5VSB system?

Thanks for your replies POM_MJ and goodpsusearch
I will keep this as a PSU for testing only, not for usage, would it see usage I would for sure do a full recap

Anyway I noticed for example on POM_MJ's PSU there is the 115/230v selector switch, mine omits this switch (fixed @ 230VAC)
Also I see where the PFC coil is supposed to be placed on my PCB, CN3 in the attached photo (The selector switch cables go here to SW2 and SW1 hidden between the furthermost Y capacitor and the coil behind it, what does that coil do btw, just noise supression?)
It will be a good test PSU "for that other thread" if I ever find a PFC coil

Also it would be fun to try and design a safe 5VSB circuit for this PSU, which if I'm not completely off my base uses a two transistor design?
Lastly in the other pic you can see my über safe testing environment

[POM_MJ]

Re: Operation of the 5VSB system?

The coil labelled MP-EE25-1 and CX,CY is transient filter to separate noise between AC line and high voltage side of the PSU.

to mod 5VSB circuit, see Bestec ATX250-12E up and running! by everell, he's an expert.

[goodpsusearch]

Re: Operation of the 5VSB system?

Also it would be fun to try and design a safe 5VSB circuit for this PSU, which if I'm not completely off my base uses a two transistor design?

I think it does, as every psu at that price range.

[370forlife]

Re: Operation of the 5VSB system?

Yup, two transistor all right.

[Per Hansson]

Re: Operation of the 5VSB system?

POM_MJ; thanks for the clarification
I had already read the 5VSB mod thread by everell, that is what has peaked my interest

370forlife / goodpsusearch, thanks for the confirmation
Btw, that transistor on the right of the 5VSB transformer, is this also part of this circuit?

[momaka]

Re: Operation of the 5VSB system?

I also have a Macron Power mpt-301. Just recapped it about a month ago.
While it does seem to use a 2-transistor 5vsb design, I was not able to locate the so-called "critical cap". There are a bunch of 10uF caps on the secondary, but I don't think those are it. Perhaps it's a modified 2-transistor design? In most other 2-transistor designs I've seen, the critical cap is always on the primary side near the 5vsb transformer.
As for the transistor on the right of the 5vsb transformer - from what I remember, that was just part of the feedback loop for 5vsb.

By the way, if you're fixing this PSU, you might also want to change the loading resistors on the 3.3v and 5v rails. Mine had a 10 Ohm on 3.3v, and a 25 Ohm on 5v - that's almost 2W of heat being dissipated, and it's pretty close to the caps, too. I put the 25 Ohm on the 3.3v and a 100 Ohm (from another PSU) on the 5v.

[Per Hansson]

Re: Operation of the 5VSB system?

Thanks for the input momaka, but look in the picture 370forlife posted
Right above the two circled transistors there sits two capacitors, right by the 5VSB transformer
10uF 50v, don't you think those are it?

Earlier I provided this link to a better resolution shot aswell:
https://www.badcaps.net/forum/attach...1&d=1399817566

[PCBONEZ]

Re: Operation of the 5VSB system?

Hi, I have a simple question on the operation of the 5VSB system
Do the components running off it (chipset, NIC's, USB etc) continue to run from this rail when the system is powered on?.

Some do and some don't.
Depends on the specific board too.

Sounds like you figured out the RAM

USB usually switches between +5vsb and +5v based on system state too but not all motherboards do it. [Same with PS/2 sometimes.]

You need a little power on USB when it's in soft-off so USB keyboard hot-keys can boot it, but once it's running they don't want all the USB loads from 6 or 8 ports piled onto +5vsb so they switch it.

Just about anything that is "wake-on" can be set up that way if the board designer wants to.

That might include power feeds to PCI slots that need to be capable of supporting a "wake-on" NIC or modem. Not sure about that one but there aren't any PCI pins devouted to +3.3vsb or +5vsb. [Maybe that jumper provides vsb to the card but I -think- I've had WOR modems and WOL NICs with no jumper connector.]

.

[Per Hansson]

Re: Operation of the 5VSB system?

Thanks for your insight PCBONEZ
Actually that cable which connects the the NIC for WOL, atleast on this board it is not needed, because the NIC is indeed lit when the computer is off
(And it made a 0.2A difference when I unplugged it's RJ45 cable too)

The mobo I'm testing with is an old Socket 370 with a 1400Mhz Tualatin Celeron and a FIC FA15T mainboard
It does not support wake from USB, so I guess it does not have a switch for USB from 5VSB to 5v

[PCBONEZ]

Re: Operation of the 5VSB system?

FIC was pretty decent back in the P3 days.

I have a couple of FIC FR33E I held onto in case I needed to test an old ISA card. Similar to your board but these are uATX.

Except for four 470uF SI in not so critical places they had all MBZ as original. None of the SI were bad but I replaced them anyway.

Contrary to what FIC says they fully support 1.4GHz P3T-S [Tualatin S] if you use the right BIOS.
I think it's the next to last version. The last version didn't recognize those CPUs. [Odd.]
Their docs say it supports a 1400-MHz 133-FSB Copper-mine with 512k cache, which of course doesn't exist.

Same trick [older BIOS] might work for you too if a P3T-S falls in your lap and it doesn't work with the latest BIOS.

[Per Hansson]

Re: Operation of the 5VSB system?

Actually the latest BIOS for my mobo does recognize the Tualatin 1400Mhz Celeron properly... Go figure

[momaka]

Re: Operation of the 5VSB system?

Thanks for the input momaka, but look in the picture 370forlife posted
Right above the two circled transistors there sits two capacitors, right by the 5VSB transformer
10uF 50v, don't you think those are it?

Nope, those caps are part of the driving circuit for the transistors on the primary.
When I was recapping mine, I did actually trace the whole 5vsb circuit but was not able to find a critical cap. There are two polyester film caps in that area, though - perhaps one of them might be it (but I'm not sure)? Maybe Everell knows, he's an expert on those 5vsb circuits (and PSUs in general ).

[PCBONEZ]

Re: Operation of the 5VSB system?

Actually the latest BIOS for my mobo does recognize the Tualatin 1400Mhz Celeron properly... Go figure

The Celeron has 1/2 the cache of the P3-S and so the microcode in the BIOS is different.
If you've got the code for 1.4GHz then you probably have the code for the 512k cache version too, but that's not 100% for certain.

Not meaning to down-talk the CPU you have or anything.
The 1.4GHz P3-S is just one of my pet-favorite CPUs.

.