X-Power 600-TD again

[JEWilson]

Albeit a separate thread is open in respect of an X-Power 700-TD
see further http://www.badcaps.net/forum/showthread.php?t=5575

This is a 600W model
Rec'd this PSU dead
Fuse gone
Replaced fuse and blew again
Found 1 of a pair of Toshiba 2SK2611 in primary s/c
Replaced and replaced again, the T5.0A 250V fuse

PSU now starts up and the go/no-go tester tells
me all the rails are within regulation.
PSU exhibits non-start behaviour when I try to apply a minimal load
such as HDDs. The fans will spin and then stop.

Suspect as with the observations in the 700-TD post this
may be down to the WellTrend IC, its a WT751002 as
this may be a protection issue.

Any comments?

[capamatic]

Hi,

Don`t know if the 600TD has the same riser-board as my 700TW, but the protection chip is similar.
Anyway, have just ESR`d the two Teapo ESK caps on it - 100uf 25V.
One of them is 0.5 Ohm, the other 1.1 Ohm.
I suppose these are supply decoupling caps for the chips but are general purpose and may be causing some supply instability, enough to cause no-start ?
I`ve not yet changed them so can`t comment on a fix, but am hopeful.
Seems strange to use a GP Cap on a major chip, albeit 105C.
Worth testing just to discount,

BTW, thanks for your help thus far with my prob.

[i4004]

http://pdf1.alldatasheet.com/datashe...TC/WT7510.html

"The Over Voltage Detector (OVD) monitors 3.3V, 5V, 12V input voltage level. The Under Voltage
Detector (UVD) monitors 3.3V, 5V input voltage level. When OVD or UVD detect the fault voltage level,
the FPL_N is latched HIGH and PGO go low. The latch can be reset by PDON_N goo HIGH. There is 2.4
ms delay time for PDON_N turn off FPL_N.
When OVD and UVD detect the right voltage level, the power good output (PGO) will be issue."

so check fpl_n and pgo to see if it's latched to protection.
if it is, it could be output caps, or the chip itself is malfunctioning.
namely perhaps it thinks something is wrong, but it's not.
checking the voltages it monitors should establish that.
these would be vcc(supply and 12v monitor), v33 and v5.
(according to the sheet i linked...you have 02 on the end, so perhaps it's different layout....)

it'll be interesting to see if these are external components flaws, or chips themselves.
my guess is there's something fishy with these chips, but for your sake let's hope it's not.
here's another datasheet that can probably be helpfull...
http://www.eserviceinfo.com/download...d_WT7514L.html

[JEWilson]

The 700-TD has a daughter pcb with the WellTrend on it.
The 600-TD does not. The IC is on the main PSU.
From the datasheet for the WT7510, it does not have
current protection unlike the model in the 700-TD.

Prior to replacing the MOSFETs, I tested the ESR of the
O/P lytics, these are all Teapo and appear good with my
ESR meter.

As a matter of course, I replaced most of the small caps
with better series such as Nichicon YXF, Pana FC and NHG.
This is something I do with all of these 600-TDs as a
pre-emptive measure.
see prior post and suscon decouples Tiny Switch
http://www.badcaps.net/forum/showthr...hlight=X-power

I will require to test the WellTrend voltage comparator WT7510
with multimeter to determine voltage conditions.

Separately, if the PSU can only handle minimal load, does
this point to a failure in the secondary with the
rectifier diode(s) or is it more likely to be a control issue?

[JEWilson]

Removed secondary and tested the schottky power
rectifier diodes. All appear in good condition.

As I had the HS removed, found one end of a diode
leg discoloured. This was under the arc of the largest
inductor on what appears to be, in the first instance,
the -12V rail. Checked solder side (again) and
found at this device, the beginnings of a cold joint.

Took in cct readings and the diode read both ways.

Removed the device with some difficulty and
found it in two pieces once removed.
I suspect the device had obviously failed.

As a resullt getting the part number will be tricky.
From what I can discern, the part ends with R206
So this may be a 2.0A fast recovery diode (likely).

Any help appreciated. Thx

[i4004]

i think that's non-critical.
see what other psus are using there.

[capamatic]

Any update on the 600 TD ?

Was it the diode ? Did you manage to test the voltages `round the Welltrend ?

I`ve pulled the 700 TW apart, everything tests OK. I`m convinced these Welltrend devices are faulty. Only course of action is to hard-power it on. Provided the output voltages are within regulation, then this would seal it.

[JEWilson]

Well,

I replaced the diode but the problem persists with this
PSU as soon as you attach a load across it.
I tried one HDD as a load and the supply immediately
shuts down

It does come back up after a while but if you apply a
load again it will shut itself down.

Will have to start getting some readings as intended
in respect of the protection cct.

[i4004]

adapting it to work with some other ic would be too problematic, i think, so all you can do is try with known good ic of same type.

[JEWilson]

Just got round to looking at the WelTrend WT7510

Vcc, pin 7 reads 11.48V with no load on the PSU,
just the go/no-go tester across the 24 pin ATX
connector and both of the +12V rails.
Checked +12V and +5V at the PSU O/P, reads ok
The voltages are within ATX regualtion.

When I place one HDD across the PSU and have
the no-go tester as above, pin 8 of the WT7510
gets pulled low consistent with a fault condition.

I have no information on the possible supply
variations for Vcc on the WT7510, it is a bit low.

I have checked the caps of the O/P rails and these
all have good ESR. If a cap had s/c, it would
have been obvious from the ESR tester.
This may be a faulty WT7510, I will require to
obtain one of these or an equivalent and try again

Thanks

[i4004]

>If a cap had s/c, it would
have been obvious from the ESR tester.

it would?
how do you differentiate low ohms of esr vs. low ohms of short circuit?
you're presuming it has to be precisely 0ohm s.c.?
why?

[capamatic]

If the WT7510 is anything like most others in the series, then VCC can be as low as 4V.
The 7510 is a hard datasheet to find, beyond the 1-pager that is.
The WT7514 has a max VCC of 5.5V

Keep us posted.

[capamatic]

Perhaps you could monitor pin 5 and 6 (V33 and V5 respectively) before a load is applied , then after.

I was thinking of removing the IC from the 700-TW, and breadboarding it. If I fed it with the right voltages, it should indicate wether it`s faulty or not.

[JEWilson]

I measure the ESR and the resistance at the same time.
I use an ESR meter and MMeter to double up when I
take the readings.

Each meter will read what it finds.
This may be an incorrect approach but appears to work
for me.

[i4004]

actually that seems like ok approach. i though you just relied on esr meter.
with your method low ohms of same value on both meters should mean s.c.
which means even in-circuit testing should be ok most of the time.

if cap is in parallel to coil or primary of trafo then readings on both will be wrong, offcourse.
you can have such case on psus where feedback is via trafo.

[ifrythings]

If the switching MOSFETS were blown did you chaeck the current sensing network to see if anything is damaged? if so it mite be low enough current not to trip OverCurrent unit a small load is applied.

If this is a half bridge design check components on the driver transformer area, if it's a signle transistor forward converter, check for a open source resistor on the switching transistors or sensing resistor compator parts.