VOGONS


ECS P4S5A-DX+ joys and tribulations

Topic actions

Reply 20 of 30, by Repo Man11

User metadata
Rank l33t
Rank
l33t

That also occasionally happened with Socket A boards without a four pin connector; I can remember twenty plus years ago where a guy mentioned this happened to his Epox 8K3A+. A slightly out of spec, slightly loose connection leads to resistance, which leads to heat, and the heat increases the resistance, then the contacts begin to burn and the resistance spirals out of control. About ten years ago, my mother's PC had a low level electrical burning smell coming from somewhere but I couldn't find it. When it quit working, I figured out that the power cord for the power supply was out of spec and had burned up, both the cord and the contacts for the PSU.

A lot of times when you first start out on a project you think, This is never going to be finished. But then it is, and you think, Wow, it wasn't even worth it. - Jack Handey

Reply 21 of 30, by Nemo1985

User metadata
Rank Oldbie
Rank
Oldbie
rasz_pl wrote on 2026-05-26, 00:11:

Famous PCChips quality and design choices lived on in ECS 😀

The problem is insufficient carrying capacity of standard ATX connector, same thing happens to 5090 with that stupid new connector https://www.techpowerup.com/346572/nvidia-gef … 0-w-power-limit

I bet the PSU still works fine, but you will obviously need another ATX connector. Either new one with a set of crimp pins and a crimper, or a from sacrificed PSU and a set of heat shrink tube to splice it up nicely 😀

That's the sad condition of the atx connector

The attachment 003.jpg is no longer available

That's the back of the motherboard, ignore the general condition, I tried to unsolder the pins, but there was that burned pin which was ground I think?

The attachment 002.jpg is no longer available

That's the upper part of the motherboard

The attachment 001.jpg is no longer available

Ecs\pc cheaps quality strikes back huh?
I think the problem was just the loose connection between the psu pins and the motherboard? The motherboard is going to be trashed either way but I definitely want to save the psu, I opened and inside is all good, no bulging caps or anything sketchy.
Do you have any advice about how should I proceed? I need to buy a crimper, crimp pins and shrink tubes? What should I get precisely?
Thanks

Reply 22 of 30, by rasz_pl

User metadata
Rank l33t
Rank
l33t

ECS bought PCChips somewhere in 1998, what you have is a PCChips motherboard in a trench coat https://theretroweb.com/motherboards/s/pcchips-m930dalr

>I think the problem was just the loose connection

The problem occurred in the design phase when some former pcchips engineer ignored component specifications. ATX connector was never meant to push so many ampers over just four +5V pins. If not special snowflake 6 pin AUX https://www.playtool.com/pages/psuconnectors/ … ectors.html#aux then even additional stupid MOLEX would help here https://www.pcstats.com/articles/979/2.html .
6 pin AUX has been standardized in 2000 by ATX12V 1.0.
2003 ATX12V 2.0 added 24pin socket with one more +5V pin, but more importantly pushed mobo designers to use 12V rail.

some crimping tutorials
https://www.youtube.com/watch?v=xU3fQGPAZVg
https://www.youtube.com/watch?v=gYNiiL2RY5g
https://www.youtube.com/watch?v=dAiHSeoGQVQ

Personally I would find another/broken sacrificial supply or buy "Extension Cable ATX 20 pin", cut the damaegd connector off and splice the new one soldering cables one by one and sleeving in shrink tubing, but thats because soldering is super easy for me.

Mobo is also most likely ok and could be rescued with a new connector. Personally I would solder in additional molex (hdd power connector) receptacle for peace of mind. Soldering cables can be done even with $5 soldering iron, mobo will require beefy (like 100W) soldering station, and preheating to pull out those connector leftover pins because its internal ground/power planes will act as a heat sink.

https://github.com/raszpl/sigrok-disk FM/MFM/RLL decoder
https://github.com/raszpl/FIC-486-GAC-2-Cache-Module (AT&T Globalyst)
https://github.com/raszpl/386RC-16 ram board
https://github.com/raszpl/Zenith_ZBIOS Zenith Z-386 MFM-300 ZBIOS disassembly

Reply 23 of 30, by jmarsh

User metadata
Rank Oldbie
Rank
Oldbie

What is going on with the top edge of the AGP slot?

Reply 25 of 30, by jmarsh

User metadata
Rank Oldbie
Rank
Oldbie

Wasn't like that in the earlier photos. Looks like it was used as leverage.

Reply 26 of 30, by Nemo1985

User metadata
Rank Oldbie
Rank
Oldbie

Thank you all for the advices.
The agp slot connector is damaged but it's due to me.
The first pictures showed the motherboard with the connector melted together.
Then I tried to desolder the back of the connector (unsucessfully) so i relied on pliers to pull the connector from the motherboard.
The attempt may have mangled the psu connector a bit but it was already melted together with the one of the mb so (I know because before use this method I tried to pull out the connector normally for like 15 minutes).
The motherboard may still work but yes it's a bit mangled, it definitely need a new atx connector and a new agp slot (it was already damaged on both ends).
I don't care about it too much since I have a stash of them, while the PSU is definitely more precious.
I'm now buying the necessary tools from aliexpress but in the meanwhile I already bought another same model psu (just to have one in case the other plan doesn't work, I mean to use one and keep the other as a backup).
If anyone wants to try to save the mb, I'm willing to send it just at shipping cost (live in Europe, some countries delivery cost less than €10.)

Reply 27 of 30, by Nemo1985

User metadata
Rank Oldbie
Rank
Oldbie
rasz_pl wrote on 2026-05-26, 02:58:
ECS bought PCChips somewhere in 1998, what you have is a PCChips motherboard in a trench coat https://theretroweb.com/motherboar […]
Show full quote

ECS bought PCChips somewhere in 1998, what you have is a PCChips motherboard in a trench coat https://theretroweb.com/motherboards/s/pcchips-m930dalr

>I think the problem was just the loose connection

The problem occurred in the design phase when some former pcchips engineer ignored component specifications. ATX connector was never meant to push so many ampers over just four +5V pins. If not special snowflake 6 pin AUX https://www.playtool.com/pages/psuconnectors/ … ectors.html#aux then even additional stupid MOLEX would help here https://www.pcstats.com/articles/979/2.html .
6 pin AUX has been standardized in 2000 by ATX12V 1.0.
2003 ATX12V 2.0 added 24pin socket with one more +5V pin, but more importantly pushed mobo designers to use 12V rail.

some crimping tutorials
https://www.youtube.com/watch?v=xU3fQGPAZVg
https://www.youtube.com/watch?v=gYNiiL2RY5g
https://www.youtube.com/watch?v=dAiHSeoGQVQ

Personally I would find another/broken sacrificial supply or buy "Extension Cable ATX 20 pin", cut the damaegd connector off and splice the new one soldering cables one by one and sleeving in shrink tubing, but thats because soldering is super easy for me.

Mobo is also most likely ok and could be rescued with a new connector. Personally I would solder in additional molex (hdd power connector) receptacle for peace of mind. Soldering cables can be done even with $5 soldering iron, mobo will require beefy (like 100W) soldering station, and preheating to pull out those connector leftover pins because its internal ground/power planes will act as a heat sink.

I didn't know that ECS bought PC Chips (cheaps for friends), I always thought they were the same company just with different names (along with amptron).

According to your findings it seems likely that the same thing will happen with the other same motherboard I have in the long term.
I was also used to measure the power consumption of the whole system and it was always over 100 watt (100-120 if I remember right, I used the P4 3.06 ghz HT), so it seem that yes the power draw can be significant?
I watched the layout of the board and I noticed that just next to the ATX connector there is a unpopulated connector which I suppose could be the 6 pin auxiliary power cable? The problem is that my Enermax doesn't have this connector and I don't see any place where I could solder an auxiliary 4 pin molex connector on the board.

Reply 28 of 30, by rasz_pl

User metadata
Rank l33t
Rank
l33t
Nemo1985 wrote on 2026-05-26, 08:21:

According to your findings it seems likely that the same thing will happen with the other same motherboard I have in the long term.

definitely

Nemo1985 wrote on 2026-05-26, 08:21:

I was also used to measure the power consumption of the whole system and it was always over 100 watt (100-120 if I remember right, I used the P4 3.06 ghz HT), so it seem that yes the power draw can be significant?

oh YES, that CPU wasnt around when board was designed. 3.06 was released ~together with ATX12V 2.0 standard.

Nemo1985 wrote on 2026-05-26, 08:21:

I watched the layout of the board and I noticed that just next to the ATX connector there is a unpopulated connector which I suppose could be the 6 pin auxiliary

too many pins and its market VGA1. Most likely PCB was universal for variants with SiS 651 like this one https://theretroweb.com/motherboards/s/pcchips-m935dlr using same 16 pin header for integrated VGA
Additional +5V should be wired to that copper island between LPT slot and mosfets, one with 3 tall blue capacitors and inductor.

Nemo1985 wrote on 2026-05-26, 08:21:

The problem is that my Enermax doesn't have this connector and I don't see any place where I could solder an auxiliary 4 pin molex connector on the board.

Normal old school HDD power Molex https://www.playtool.com/pages/psuconnectors/ … html#peripheral is rated at 5Amps, same as 6 pin aux, and will fix the problem permanently. Buy $2 "4-Pin Peripheral Molex Extension Cable", cut in half, pull out yellow +12V pin completely, solder red to that copper between LPT and mosfets, black to ground copper for example bottom of LPT connector where its screw goes into PCB or directly to D5 D7 diodes anodes (pins 1 and 3).
verify with multimeter if correct:

The attachment solder points.jpg is no longer available

https://github.com/raszpl/sigrok-disk FM/MFM/RLL decoder
https://github.com/raszpl/FIC-486-GAC-2-Cache-Module (AT&T Globalyst)
https://github.com/raszpl/386RC-16 ram board
https://github.com/raszpl/Zenith_ZBIOS Zenith Z-386 MFM-300 ZBIOS disassembly

Reply 29 of 30, by Nemo1985

User metadata
Rank Oldbie
Rank
Oldbie

Actuallly the SIS chipset shoulds support the HT cpu, there is a bios option which has been hidden because guess what? It breaks everything and requires a reset of the bios (I think the sis chipset has some kind of bug which has been addressed in a later chipset).
Thank you so much for the precious explanation, it sounds pretty easy to do even for me!
So the plan for the motherboard is going to be the following one:
Complete Recap
Add the molex connector
Use a less power hungry cpu
Just to be sure I understood right here is a more high resolution picture:

The attachment dsc-0344-288-65e76b83845c9015962495_cr.jpg is no longer available

Is this correct? The 5v wire has to be soldered in the green circle here while the ground goes in any of the purple circles?
I just noticed a very funny thing, there is a cpu thermistone just next to the atx power connector, I do wonder if the ECS engineers knew the shit they did, still never has been triggered on my mb as a protection.

Reply 30 of 30, by rasz_pl

User metadata
Rank l33t
Rank
l33t

If you dont have serious soldering iron you wont be able to solder directly to the mosfet under green circle, but you should be able to scrape red solder mask just next to it and solder to copper there. I would solder two ground wires (as there are two in MOLEX connector) to two Diodes, one black cable to D4 one D7.

https://github.com/raszpl/sigrok-disk FM/MFM/RLL decoder
https://github.com/raszpl/FIC-486-GAC-2-Cache-Module (AT&T Globalyst)
https://github.com/raszpl/386RC-16 ram board
https://github.com/raszpl/Zenith_ZBIOS Zenith Z-386 MFM-300 ZBIOS disassembly