VOGONS


Biostar MB-8433UUD 2.0 Issues

Topic actions

First post, by Synoptic

User metadata
Rank Member
Rank
Member

Hi gang !
I bought this mobo a year ago. The floppy controller never worked when it was enabled (Floppy drive Failed). Even the COM ports are not working even if enabled, they do not show up in the Status Info before booting DOS/Windows.

I have very good soldering skill and equipment, but never worked on a motherboard before.

I was wondering if the COM and Floppy issues could be related to a common area on the board that I could check and fix maybe a missing cap or something.

Thank you !

FYI : It runs Feipoa's 2012 BIOS right now with EDO RAM.

Last edited by Synoptic on 2016-07-25, 13:26. Edited 1 time in total.

Reply 1 of 42, by Synoptic

User metadata
Rank Member
Rank
Member

Update : I have a missing Component, CB18 near UM8663BF.
It was broken when I got the board.

I cannot find what would go there. The other CB's on the board act like a Diode when I probe them on board.
I tried to put a very small wire, like a fuse, but the computer won't boot.
It's connected to one of the pin of the UMC chip. There is 4 volts going on this line.

I suspect this component to be the cause of my floppy and COM problems.

Reply 3 of 42, by Synoptic

User metadata
Rank Member
Rank
Member
Brickpad wrote:

I have the same board (revision 2). What revision do you have? I can take pictures of the where the missing component should be for you.

2.0 too.

Sure, a picture could help.
Are you ale to check for voltage on that component ?

Reply 4 of 42, by feipoa

User metadata
Rank l33t++
Rank
l33t++

There were two common pin-outs for the serial cable from this era. It is possible that you are using one with an incorrect pin-out? Try your cable on another UMC 8881-based board to ensure that it is correct. I find that UMC 8881 boards often use the same pinout for the serial ports, whereas SiS 496 boards use another. UM8667 is responsible for the serial ports. As of 2011, I was still able to source new, old stock of this item.

Have you confirmed that you are using the correct position on the floppy cable. I recall only one combination works. Try all 4 possibilities on the cable. At one point, I also thought I had a dead floppy controller, but it was the cable positioning. Also, ensure that your RTC is not dead. If you cannot save your CMOS settings, the floppy and serial might not work. The chip responsible for the floppy disk controller is UM8663. I also sourced NOS of these some years ago.

What is the datecode on your UMC8881/8886?

FYI, the information on which IC on this MB does what is provided in my custom manual. See World's fastest 486 link to download the pdf.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 5 of 42, by Synoptic

User metadata
Rank Member
Rank
Member

Thanks Feipoa to help me on this.

When I say I don't have COM ports, FLoppy or even Paralallel port is because when I enable them in the bios, they still show as "none" on the post information screen.

Dates Codes are :

8881 : 9611
8886 : 9615
8663 : 9614

I have modded my Dallas RTC with an external battery. It is safe to assume that it works because this morning the date and time was correct and I had the computer sit unplugged for the night. The battery is new btw.

Picture
http://imgur.com/a/96VMQ

Boot sequence
https://youtu.be/rIBIy3QVtSo

Reply 6 of 42, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Did you confirm that when you alter other CMOS settings, that they are saved? I recall that when the RTC battery is at a critical point, sometimes the clock will get saved, but the settings will not. When the battery weakens further, the clock will show some drift.

If you have an EEPROM, you might want to try my modified BIOS.

I think it would be very unlucky, and unlikely, for, both, the 8663 and 8667 chips to fail.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 7 of 42, by Synoptic

User metadata
Rank Member
Rank
Member

The settings do save. the CR battery shows 3.xx volts. If you have a more thourough test for the RTC, let me know, I'll do it.

I already have burned your 2014 Bios on the P28F001BXT using my EEPROM programmer.

What is the purpose of the pin of the 8663 that connects to CB18 ? is CB18 a Circuit breaker ?
What is the bios doing when it's setting the COM and LPT port to None ? is it checking some sort of registers ? I would like to understand the mechanism behind the fault right now.
Thank you !

Reply 8 of 42, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Is the 2014 BIOS in the MB right now? And the COM and FLOPPY still go to none? I have had 8433UUD boards with faulty floppy controllers still show A:\ set as the floppy drive. For this reason, I am somewhat suspicious that the IC's are at fault.

I am not familiar with that EEPROM part number. How about trying one of the usual ones? Maybe the ESCD on that chip is at a different address and is what is messing up the serial/floppy?

I am not sure what CB18 is. Have a photo?

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 9 of 42, by Synoptic

User metadata
Rank Member
Rank
Member

The 2014 BIOS is on the board right now, still showing None for PArallel and COM.

I even used UUDFLASH.EXE to flash it on the chip instead of using my EEPROM programmer.

Floppy still Fails (C0) if Floppy seek is enabled.
LPT and COM still to None.

I couldn't get a descent photo, so I am reusing Mau1wurf1977's picture instead, which is exactly what I have, Except that I DO have a RTC installed.

fwqTVSX.jpg

Reply 10 of 42, by feipoa

User metadata
Rank l33t++
Rank
l33t++

I'm not sure what CB18 is for. CX, where X is an integer, usually refers to capacitor. Not sure about the B, nor CB. All my boards have this component, that is, the original board revision, early 2.0 boards, late 2.0 boards, 3.0 board, and v3.1 board all have this component. Try using a non-intel style EEPROM. Also, ensure to set the EEPROM jumper voltage (EPROM, or 12 V EEPROM-intel, or 5 V EEPROM-SST+others).

EDIT: I have measured CB18 out of circuit. It is a capacitor and measured 105 nF.

Last edited by feipoa on 2016-08-04, 09:58. Edited 1 time in total.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 11 of 42, by Synoptic

User metadata
Rank Member
Rank
Member

I couldn't find anyother FLASH PROM compatible. the one I have is the one that came with the board.
I tested and I am able to program an older version of bios using UUDFLASH.EXE.

I read somewhere that CB is for Circuit Breaker.
I don't think the bios chip is the culprit here. I think the 8663AF has failed, but I would like it to be otherwise.

Reply 12 of 42, by feipoa

User metadata
Rank l33t++
Rank
l33t++

If CB is circuit breaker, then perhaps it is just a fuse. I have used SMD fuses of this size in designs before. But I thought fuses were labelled F. You could try to short the solder pads where CB is and see if it fixes things.

If that fails, try ordering some UM8663BF chips. Some boards had UM8663AF chips. I am not sure what the difference was. Newer boards 486 boards, like those produced from 1997 often contain UM8670F. Size and pin-count are the same as UM8663, but not sure what changes were made. I have not seen UM8670F on 8433UUD boards though. I have all these bare components in stock but haven't had time to fix my supply of partially defective motherboards. It shouldn't be too difficult to cut off the old ones and solder on the new ICs.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 13 of 42, by feipoa

User metadata
Rank l33t++
Rank
l33t++

When I ran my multi-meter across CB18 I did not get a short, so this component is not a fuse. In which case, do not bridge the contacts with a wire. I suspect the component is a capacitor, but it would need to be removed from the circuit to measure it properly.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 14 of 42, by Synoptic

User metadata
Rank Member
Rank
Member
feipoa wrote:

When I ran my multi-meter across CB18 I did not get a short, so this component is not a fuse. In which case, do not bridge the contacts with a wire. I suspect the component is a capacitor, but it would need to be removed from the circuit to measure it properly.

Could you do just that for me ? I probed on board, other CB's and they were acting like a diode.

also, can you check the voltage on CB18 ? I have 4.1 volts.

Reply 15 of 42, by feipoa

User metadata
Rank l33t++
Rank
l33t++

I didn't check it for a diode, but it very well could be. Often, diodes are used for over-voltage protection. Maybe that is what they mean by "circuit breaker"? If other CB's indicate a diode, then this CB is probably also a diode. You should remove your other CB component and confirm what the threshold voltage is.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 16 of 42, by feipoa

User metadata
Rank l33t++
Rank
l33t++

I removed a UM8663BF Super I/O IC from a revision 3.1 Biostar MB-8433UUD motherboard. The silkscreen shows UM8663AF/BF, so perhaps you can use either of these. This particular MB had a defective floppy controller, so I am hoping the replacement fixes the issue.

The solder-on is a little messy, which I expected because this is the first time I've soldered on an IC with such tight spacing with a hand soldering iron. When you cut the IC off, it is best to cut it off as close to the top of the IC as possible as this will reduce the risk of damaging the solder pads. Also, when you clean up the cut-ends, leave a little solder on the pads, but try to flatten the solder the best you can.

UM8663BF_1.jpg
Filename
UM8663BF_1.jpg
File size
167.59 KiB
Views
955 views
File license
Fair use/fair dealing exception
UM8663BF_2.jpg
Filename
UM8663BF_2.jpg
File size
184.35 KiB
Views
955 views
File license
Fair use/fair dealing exception

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 17 of 42, by feipoa

User metadata
Rank l33t++
Rank
l33t++

After replacing the UM8663BF IC, the onboard floppy controller remains non-functional. However, in my case, the BIOS still shows A:\ as 1.44MB 3 1/4" floppy drive. Anyway, I am left wondering which component, IC, or connection is causing this problem. The worst case scenerio would be something wrong with the southbridge chipset, which communicates with the Super I/O (UM8663).

I remember when the floppy controller on this MB suddenly died. I had the system on 24/7 with a 5 month uptime. There was no obvious sign of trouble until an NT4 BSOD message appeared. After that, the floppy controller didn't work. This happened more than 5 years ago.

There is a workaround for this problem, it is to use an ISA I/O card, preferably one with the same UM8663 Super I/O chips. On the ISA I/O card, I disable all features except for the floppy controller (via jumpers). In the motherboard's BIOS, it is important to leave the onboard FDC set to enabled and A:\ to 1.44 MB, 3 1/4" drive, and to ensure floppy seek at boot is disabled. Then connect your floppy drive to the ISA card's floppy header. It should work fine. It is still ugly to have to use an extra ISA card for this though. I would prefer to find the source of the problem.

I have some 4DPS motherboards which also cannot read from the floppy drive via the onboard floppy controller. I have spare Winbond Super I/O chips for these boards, but I am afraid that replacing them might be fruitless inlight of my recent findings. Anyone have any ideas where to start tackling this problem?

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 18 of 42, by Synoptic

User metadata
Rank Member
Rank
Member

Well, Maybe we should start from the begining. What about tracing each pin of the FDC header and check with a scope if we get something we expect.
Let's plug a drive and use the unused header on the ribbon to probe. We could compare with a working mobo and see if we :
1. get all right voltages
2. get same electrical activity at boot (We know we wont).

It could be the lines going into the drive but it could also be the lines coming back from the drive. So let's try to see if it tries correctly to talk to the drive, and check if the drive responds but the data does not get back to the FDC.

Reply 19 of 42, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Looking forward to what results you are able to uncover...

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486