Advent 3317-3617-3317a PC mobo (Trigem TG Crush 18D M)

[JEWilson]

This is the mobo that was dead and caused a PSU failure per
https://www.badcaps.net/forum/showthread.php?t=4483

I recently have built an ESR meter per
https://www.badcaps.net/forum/showthread.php?t=4476

I could do with a better pair of leads than those supplied.

The electrolytic BOM for this mobo details;

3 of Blue 1500 uF 16.0 V Rubycon MBZ 105 deg C - H 25.4 mm x Dia 10 mm EC2, EC20, EC21
2 of Blue 1800 uF 6.3 V Rubycon MBZ 105 deg C - H 25.4 mm x Dia 8 mm EC11, EC32
6 of Black 2200 uF 6.3 V Panasonic FJ 105 deg C - H 20.0 mm x Dia 10 mm EC5, EC6, EC10, EC12, EC13, EC14
12 of Black 1000 uF 6.3 V Panasonic FJZ 105 deg C - H 15.0 mm x Dia 8 mm
EC3, EC17, EC18, EC19, EC24. EC25, EC27, EC28, EC29, EC30, EC33, EC35
5 of Brown 1000 uF 10.0 V United Chemicon KZG 105 deg C - H 15.0 mm x Dia 8 mm EC22, EC36, EC41, EC44, EC45,
2 of Brown 470 uF 16.0 V United Chemicon KZG 105 deg C - H 11.5 mm x Dia 8 mm EC9, EC26
Note KZG are prone to failures per good cap list

2 of Black 330 uF 6.3 V Unknown 105 deg C - H 10.0 mm x Dia 8 mm

ESR values
Rubycon MBZ 1500 uF 16.0 V 105 deg C - H 25.4 mm x Dia 10 mm ESR @ 20 deg C - 13 milli ohms (0.013 Z) Max
Rubycon MBZ 1800 uF 6.3 V 105 deg C - H 25.4 mm x Dia 10 mm ESR @ 20 deg C - 19 milli ohms (0.019 Z) Max
Note on Rubycon MBZ Datasheet - no published can size at 25.4 x 10 1800 uF 6.3V - instead consider;
Rubycon MBZ 1800 uF 6.3 V 105 deg C - H 20.0 mm x Dia 8 mm ESR @ 20 deg C - 19 milli ohms (0.019 Z) Max
Panasonic FJ 2200 uF 6.3 V 105 deg C - H 20.0 mm x Dia 10 mm ESR @ 20 deg C - 13 milli ohms (0.013 Z) Max
United Chemicon KZG 5 1000 uF 10.0 V 105 deg C - H 15.0 mm x Dia 8 mm ESR @ 20 deg C - 28 milli ohms (0.028 Z) Max
United Chemicon KZG 5 470 uF 16.0 V 105 deg C - H 11.5 mm x Dia 8 mm ESR @ 20 deg C - 36 milli ohms (0.036 Z) Max

First, I tested all electrolytic caps valued at 470uF and above.
This mobo only has wet caps of the values that concern me.

All of the caps exhibit an ESR of between 0.17 and 0.20 but this
value is of an order of ten at least, greater than the published ESR.

I suspect this may be due to the precision of the meter.
Q - can I take it as read - there are no immediate concerns as to ESR breakdown?

Would a next plausible debug step be to put the ESR meter in parallel
with my multimeter, with the meter set to a range of 200R?
Would this eliminate any caps that have gone s/c and not due to an ESR issue?
This would require the cap under test to be taken out of cct of course, correct?

Any other suggestions what I should be looking for?

Any help, insight and advice greatly appreciated

[starfury1]

Re: Advent 3317-3617-3317a PC mobo (Trigem TG Crush 18D M)

Ok I cant see all caps reading high although possible.

All are good brands

When measuring in circuit other capacitors (or capacitances) will possibly affect readings


So now to use the meter

leads can be a problem so yeah as outline before in the other thread
clean or get quality ones

Did you calibrate the meter with the 1% resistors supplied?
(I'd recheck it just to be sure)

Also make sure you zero the meter short probes hit button

It will read but be inaccurate if you haven't, as far as a true reading goes
you need to make sure its calibrated correctly as per instructions if you want true accuracy

when the ESR of a cap fails it usually fails BIG time so you will know for sure its stuffed
They can even still show around the correct capacitance but the ESR will be sky high for what it should be

although you can take readings while the cap is in the MB (not powered)
for best accuracy you need to remove it

With practice and getting familiar with the MB you will get to know if the meter is showing a stuffed cap.

When measuring ESR its best done when caps are cold

it sound like the meter is not calibrated to me by the measurements your getting, or possible lead problems

Mine on a new Panasonic FM's 2200 uf 6V3 don't even register a reading.


In fact I had to find another cap just to check it was working

since its a newly built meter

Mistakes in construction are quite possible too.
I made a boo boo or 3 in the construction of mine so just to be sure recheck your work..its easily done

If the resistors you used were blue background colour, they are right Royle pains to read
(I checked them with a DMM before I put them in to be sure)

for constructional pics (for reference) see this thread if you haven't already...I guess you have.

Here

EVB version is very similar but note quite the same page 2 has my MK2 vesrion 3 I think has Mk1

I suppose if you suspect the meter to be out and all measurements seem to be out by about the same amount

The yeah they are ok.

the best way to check if a cap is short circuit is to use an ohm meter
0 Ohms short circuit

you just have to be aware that low ESR caps or Ultra low will look like that on the ESR meter

The original design dates back to more of GP caps era of use
not ultra low esr caps
So on good caps its not going to show much if anything

I suppose checking the calibration at times is a good idea too
(I should recheck mine to be sure on the statement on the FM's)


Lastly if the MB was dead and caused the PSU to die I would suspect more then caps and possibly a short on the MB supply rails

check this thread it may help

Testing VRM

I did a PDF of it that makes it a little clearer
(at the bottom of the thread)

HTH Cheers

[JEWilson]

Re: Advent 3317-3617-3317a PC mobo (Trigem TG Crush 18D M)

Y - thanks for the advice

Y - I did the calibration bit with the 1% resistor
Y - I do zero the meter by shorting the leads and hit the button

Y - did have a problem with two solder bridges - got these fixed ok

I will be getting a new set of leads in due course

BTW - this mobo does have a short on the 3.3V rail
Once located will try the mobo again with a PSU - thx

[starfury1]

Re: Advent 3317-3617-3317a PC mobo (Trigem TG Crush 18D M)

Are I thought you might have just, stating the obvious.

Let us know how you go with it, its nice to have follow up on issues

No worries, hope the above is of some help