Chaintech Apogee 7VJL Unstable

[bauto601]

Hello everyone,

I'm having a problem with a rather old board, a Chaintech Apogee 7VJL.

The pc setup is:

- Athlon XP 3000+ (175x13)
- 3x 512MB Corsair XMS PRO @350MHz
- ATI Radeon X1950PRO AGP
- Chaintech Apogee 7VJL
- Levicom VP550 (with an acrylic top, is actually fairly decent and recapped
on the secondary side with Panasonic caps)
- Club3D Theatron sound card
- USB 2.0 card with ALi chipset
- 2x 40GB Seagate HDD's
- 2x CD/DVD Combo drive
- 1x Floppy drive

Now, the problem is that the board randomly does not boot (about a third of the times). Sometimes runs rock solid and a day later it is an unstable piece of crap.

It is overclocked from a 166MHz FSB to a 175Mhz FSB, but it's got the same problems at the 166MHz FSB.

The instability comes mostly in the form of memory errors, CRC errors while extracting zip files, random rebooting, windows that can't find one of it's many "important" files at startup, hanging while posting and i think that the experienced users can fill this list with all the other problems that come with that.

But, sometimes it runs flawlessly, gaming, extracting and copying is not a problem at all then.

Raising the DRAM voltage, Chipset voltage, CPU Voltage and even the AGP voltage doesn't help. Sometimes it looks like it runs stable, and other times it runs only stable when the voltages are set to default.

Changing ram dividers, timings, different ram, nothing helps. Also swapping the GPU for 9600XT didn't help.

And now i hear you thinking, "It's obviously that nasty Levicom unit that can't even power a led strip while staying in-spec". But, i also tried it on a Supermicro ATX PSU with 550W that i use for extreme overclocking and even that didn't help a bit. Voltages on the Levicom unit are 12.6V, 4.99V and 3.33V on the major rails. Ripple shouldn't be a problem since it's recapped with pannies. The PSU also doesn't produce any coil whine or other noises that indicate oscillation.

You still here? Now comes the important part. The board is filled with C-grade caps. For the CPU vcore output purple G-Luxon (LW) caps, and for the rest green TEAPO with gold lettering. Ranging from 1500uF for the vcore input caps to 470uF caps for other smaller power delivery and filtering circuits.

None of the caps are bulged or lifted, they all look brand new actually. Could they still be faulty?

[stj]

Re: Chaintech Apogee 7VJL Unstable

yes, caps fail inside without external signs.
you can try geating them with a hair-dryer to see if it becomes more stable - marginal caps work better when warm.

[cpt.charlie]

Re: Chaintech Apogee 7VJL Unstable

Check VRM caps.

[bauto601]

Re: Chaintech Apogee 7VJL Unstable

So i did some further research on the interwebz and i found out that this board does not have the 166fsb dividers.

So i will change the board to a DFI NF2 Ultra B, but that one definately needs a full recap.

It uses the following caps:
CPU VRM output: 5x 3300uF 6.3V KZG
CPU VRM input: 3x 1500uF 16V KZG

And the following caps are spreaded over the board and i don't know exactly the purpose of each one:

- 7x 470uF 16V OST RLG
- 10x 1000uF 10V OST RLG
- 5x 220uF 10V OST RLG (around the ram slots)
- 15x 100uF 16V OST RLG

Now, i want to do a full recap. For general caps, i was thinking of replacing them with the following (the same order, number 1 replaces number 1 and so on..):

- 7x Nichicon RNE1C471MDN1 470uF 0.010ohm 16V Poly caps
- 10x Kemet A755KS827M1AAAE014 820uF 0.014ohm 10V Poly caps
- 5x Panasonic 220uF 0.018ohm 6.3V Poly caps (6.3V since they must be really flat and they probably only get 5V max)
- 15x Rubycon 16ZLQ150MEFC5X11 150uF 0.21ohm 16V Poly caps

I went with poly caps because the original caps seem to have such a low ESR spec that only poly caps can match or exceed them nowadays.

For the CPU VRM i am still deciding between poly's or normal electrolytics. Which option would be better?

CPU VRM OUT:
- Poly: 5x Nichicon PLG0G272MDO1 2700uF 0.008ohm 4V
- Normal: 5x Rubycon 6.3ZLQ4700MEFC10X25 4700uF 0.015ohm 6.3V

CPU VRM IN:
- Poly: 3x Nichicon RNL1C222MDS1 2200uF 0.008ohm 16V
- Normal: 3x Rubycon 16ZLQ2200MEFC10X25 2200uF 0.015ohm 16V

For the poly caps, i sometimes have to choose for a lower capacitance model than the original cap since the poly caps seem to have a larger diameter at the same capacitance as a normal cap. And thus, do not fit on the board.

I really want to have the board perform as good as when it was new, or even better. Money is not that big of an issue, the difference between the full poly caps or semi poly caps recap is only €8 in total.

I want to know if the capacitors i picked as a replacement will work in this board and if i should also choose poly caps for the CPU VRM, altough i would lose some capacitance there.

What do you guys think?

[RJARRRPCGP]

Re: Chaintech Apogee 7VJL Unstable

So i did some further research on the interwebz and i found out that this board does not have the 166fsb dividers.

So i will change the board to a DFI NF2 Ultra B, but that one definately needs a full recap.

It uses the following caps:
CPU VRM output: 5x 3300uF 6.3V KZG
CPU VRM input: 3x 1500uF 16V KZG

Yep, the KZG-series caps on that motherboard, especially from 2004 or 2003, are most likely bulging, if not already spewing like a volcano!

Like my 2004 Asus A7N8X-X had! (nForce2)

[bauto601]

Re: Chaintech Apogee 7VJL Unstable

Yep, the KZG-series caps on that motherboard, especially from 2004 or 2003, are most likely bulging, if not already spewing like a volcano!

Like my 2004 Asus A7N8X-X had! (nForce2)

Yup, all the KZG caps are already bulging and leaking. And i don't really trust those OST caps either. They look good but i don't think they will be in-spec anymore. These were bad times for motherboards capacitor wise.