VOGONS


First post, by whitepawn

User metadata
Rank Newbie
Rank
Newbie

Hi,

I couldn't help myself and bought another single board computer which labeled as PC420-G2 by protech company. Here is SBC page in 1997:
https://web.archive.org/web/19970327052802/ht … .tw/ipc420.html

IMG_20210313_201133.jpg
Filename
IMG_20210313_201133.jpg
File size
902.98 KiB
Views
1400 views
File license
Fair use/fair dealing exception

After cleaning the board, installed another cpu Cx486 dx2 66 (because the one came with the board is not original, socket was very loose) it posted with "Bios checksum error".I searched online for this error it seemed either bios rom is corrupted or RTC battery problem.Second one is easy to fix for me so i retrofitted CR2032 to dallas chip but checksum error persisted so i tried another bios ( PSC-486 VER:C Jumper or manual information needed) with it which has same chipset and even bios stickers are the same. Also PSC-486 bios gives same error which i believe i have a jumper setting problem.(Top left bios is similiar bios from my working PSC-486.Bottom right is original bios.)

IMG_20210314_144539.jpg
Filename
IMG_20210314_144539.jpg
File size
830.69 KiB
Views
1400 views
File license
Fair use/fair dealing exception
IMG_20210314_111707.jpg
Filename
IMG_20210314_111707.jpg
File size
307.05 KiB
Views
1400 views
File license
Fair use/fair dealing exception

I can boot from A: drive and tried uniflash but bios chip model seems unknown in uniflash.

Started looking for bios backups online but couldn't find any.Meanwhile i discovered same hardware was labeled as "486/586 SBC Ver:GC".This board is very similiar to my board but also i couldn't find any information about it.

By any chance does someone have manual information for it?Or any advice?

Thanks in advance.

Last edited by whitepawn on 2021-03-23, 12:08. Edited 1 time in total.

Reply 1 of 19, by computerguy08

User metadata
Rank Member
Rank
Member

This being a 486 platform, I'm not sure Uniflash has full support for it (if at all). Furthermore, those BIOS ICs look like UV EPROMS, which need to be erased with a UV lamp, then programmed with a programmer like TL866.
Even if the chip was not UV, I don't think the SBC is able to flash the BIOS chip (probably lacks the Vpp circuitry, I could be wrong though, so not sure).

I'll look around the web and see if I can find more info.

Reply 2 of 19, by whitepawn

User metadata
Rank Newbie
Rank
Newbie

Hi,
Thanks for your reply.
You are absolutely right my bad, i didn't give full information.Let me explain with more details:

As i mentioned i have fully working PSC-486, in this board i backed up bios file with uniflash and hotflashed a spare chip (MX28F1000PPC-12C4) with a socket 7 board (Tomato 5STX) and tested on PSC-486 which works without any issues.Here is my topic quote:PSC-486 VER:C Jumper or manual information needed

whitepawn wrote on 2021-03-09, 20:39:
Hi again, I was able to backup bios with uniflash and tested backed up bios with another bios flash chip.Basicly first i backupe […]
Show full quote

Hi again,
I was able to backup bios with uniflash and tested backed up bios with another bios flash chip.Basicly first i backuped bios and hotflashed a spare chip with another socket 7 board, then i removed original bios chip and installed hotflashed spare chip. System booted fine so you can trust this bios file.Please add your database for future reference.Here is the bios and cmos backups:
psc486bios.rar
Please add any information if available.

And here is my spare chips(MX tested, Intel one is not tested i plan to use it for PC420-G2)

rsz_img_20210314_155407.jpg
Filename
rsz_img_20210314_155407.jpg
File size
624.35 KiB
Views
1382 views
File license
Fair use/fair dealing exception

Unfortunately i dont have any kind of eeprom programmer to read original eeprom.

Thanks again, please share if you have found any information.

Reply 3 of 19, by whitepawn

User metadata
Rank Newbie
Rank
Newbie

I have done my research and found two similiar boards to mine.
One of them is
"486/5X86 SBC Ver:GC"
And the other one is
"NAGASAKI IPC Embedded Computer, 960420 GCA, 486/5x86 SBC Ver:GC" aka "NASA-PC420" according to archive.org:
https://web.archive.org/web/20000307120159/ht … m.tw/NA-420.htm

Unfortunately none of them are have documentation or bios files but i found a guy who has issue with bios also on another vintage hardware forum.He used HS-4000 bios to work his card.I asked him to share bios file and he shared with me thanks to him.I flashed MX chip via socket 7 mobo with hotflash method.

Filename
HS4000.rar
File size
114.34 KiB
Downloads
44 downloads
File comment
HS-4000 Bios
File license
Fair use/fair dealing exception

Tried on my PSC-486 it posted as HS-4000 and works, i tried same bios chip on my PC420-G2 it shows again "bios checksum error".I am pretty sure i installed correctly jumpers because in this SBC 486 DX4-100 topic he uses ST486DX-100 and i am using AMD486DX-100.If i change jumpers to when i bought condition, only cyrix DX2-66 works.

Since i can boot from floppy i checked uniflash if bios chip recognized correctly ,yes it detects mx chip perfectly.

I also tested my PSC-486 bios with PC420 but still have "bios checksum error".

All this tests are lead me think one thing which is a hardware issue or are there another software in this board?Any theories?

Reply 4 of 19, by whitepawn

User metadata
Rank Newbie
Rank
Newbie

Ok i have done my homework.

Firstly it wasn't a hardware issue.I managed to work board with Intel DX2 66 Mhz.It was a jumper issue and bios issue.Also here is the interesting thing if i boot system with wrong jumper settings it posts but it gives error "bios rom checksum error".Has anyone ever came across case like this?

Here is my DX2 66Mhz jumper settings:

IMG_20210318_144740k.jpg
Filename
IMG_20210318_144740k.jpg
File size
580.18 KiB
Views
1305 views
File license
Fair use/fair dealing exception

And here is the bios file that i have used from mt777’s board(486/5x86 SBC ver:g6)Thanks to him, It works fine:

Filename
MITAC-B6.rar
File size
96.27 KiB
Downloads
50 downloads
File license
Fair use/fair dealing exception

And here is my excel worksheet to reverse engineer jumper settings.I may update this when i try more cpus:

Filename
PC420REV1.rar
File size
2.87 MiB
Downloads
46 downloads
File license
Fair use/fair dealing exception
EXCEL.JPG
Filename
EXCEL.JPG
File size
130.69 KiB
Views
1305 views
File license
Fair use/fair dealing exception

And here is my new problem.It conflicts i/o 278:

IMG_20210318_151738.jpg
Filename
IMG_20210318_151738.jpg
File size
518.5 KiB
Views
1305 views
File license
Fair use/fair dealing exception

I have checked bios settings to disable com port but none of them worked.Maybe it is related to jumpers?Any idea?

Reply 5 of 19, by mt777

User metadata
Rank Member
Rank
Member

5X86 is not 100% correct jumpers. I found this photo in local auction but it seems that 3x multiplier is set (so as DX4-100). I will try in my board set Am5x86 properly.
Voltage jumper: 1&2 5V, 2&3 = 3.3V ?
278h is lpt secondary port?

In your free time please recheck all SBC manuals and find similar jumper layout. I really believe that something similar exists there but didn't check all of them.
ftp.emacinc.com/LegacyProducts/SBC/

Last edited by mt777 on 2021-03-19, 00:48. Edited 1 time in total.

Reply 6 of 19, by whitepawn

User metadata
Rank Newbie
Rank
Newbie

Oh i see, a test will be useful.

Voltage jumper is:
1-2 closed=3.3v
2-3 closed=3.45 v
all open=3.3v from my board.

No 5v combinations i have found yet.You can confirm via middle leg of voltage regulator and a multimeter before installing cpu.

Reply 8 of 19, by whitepawn

User metadata
Rank Newbie
Rank
Newbie

Yes dx2's are 5v cpu's but works fine with 3.3 volts in my case.

My board have LX8385-00 (which is an adjustable voltage regulator, not fixed) not LT1085 so i expect voltages other than 3.3volts 😀

Could you measure your board 1-2 closed, 2-3 closed and all open if possible?Maybe my board or regutator is faulty.

Reply 9 of 19, by mt777

User metadata
Rank Member
Rank
Member

I checked VCC directly on PGA168.
JP9. 1&2 = 5.1V, 2&3 = 3.4V, none = 0V
JP4. Tried different configuration and 1&2, 5&6 = gives 120MHz. None success setup as full 133MHz.
So I helped you, hope that you find Am5x86 config 😀

Reply 10 of 19, by whitepawn

User metadata
Rank Newbie
Rank
Newbie

I see, let me check on my board from cpu socket.

About JP4 only 1-2 closed gives 80mhz 1-2 and 3-4 closed seems 66mhz on my board.

Thanks for testing.Since i don't have an AM5x86 i am unable to test it.

Also another info about i/o error i disabled com ports and parallel port no luck.

I have searched many of ftp.emacinc.com/LegacyProducts/SBC/ no similiar jumper setups.

Reply 12 of 19, by whitepawn

User metadata
Rank Newbie
Rank
Newbie

Ok I invested a TL866 Programmer and read original chip then i compared it with my PSC-486 bios.Surprisingly it was very similiar to PSC-486 bios only some of bytes are changed so i decided to flash PSC-486 bios and board is booting fine without i/o conflicts or so.I can boot DOS.

When i was repairing this board i removed cache chips to aviod further troubleshooting.I installed cache chips but now it freezes if i enable external cache.Also diagnostic board says 61/60 which translates "Try to turn on level 2 cache" i checked from diagnostic board manual.Either i have to disable L2 cache or remove second chip then system boots fine.

After some investigation i found JP11 and JP12 backside of board which i believe are jumpers for cache size or enable.Here is detailed photo of it:(It came originally both open)

DSC00710JPG.jpg
Filename
DSC00710JPG.jpg
File size
389.76 KiB
Views
1150 views
File license
Fair use/fair dealing exception
DSC00709JPG.jpg
Filename
DSC00709JPG.jpg
File size
476.04 KiB
Views
1150 views
File license
Fair use/fair dealing exception
DSC00707JPG.jpg
Filename
DSC00707JPG.jpg
File size
358.84 KiB
Views
1150 views
File license
Fair use/fair dealing exception

I tested chips with TL866 all passed test.Also tested on another 486 board works fine.I have tested only JP11 closed, only JP12 closed and both closed but no luck.

Any ideas?

Reply 13 of 19, by whitepawn

User metadata
Rank Newbie
Rank
Newbie

Ok i figured it out it was a bad solder joint on ALI M1489.I soldered it and cache is detected.Checked with cachechk4:

rsz_img_20210323_231143.jpg
Filename
rsz_img_20210323_231143.jpg
File size
525.76 KiB
Views
1131 views
File license
Fair use/fair dealing exception

Also figured out if you use 32 pin instead of 28 pin JP11 and JP12 must be enabled i believe.So it is not relevant in my case since i only have 28 pin sram.

Here is my simple diagram:

Cache.jpg
Filename
Cache.jpg
File size
134.05 KiB
Views
1131 views
File license
Fair use/fair dealing exception

DOS is not booting with WB (Write Back) only boots with WT mode which my Cyrix 486DX2-66 does not support obviously.Also this was a noob mistake i made 😀

I have checked voltages on directly on PGA socket and yes mt777's results are correct.(JP9. 1&2 = 5.0V, 2&3 = 3.45V, none = 0V)Please ignore my previous posts about it.

I figured out all my bios and cache issues so i can move to improve cpu jumper settings.I will update excel when i finish my work.Any contributions are welcomed.

Reply 14 of 19, by ericmackrodt

User metadata
Rank Newbie
Rank
Newbie

Hi Whitepawn,

I have the same SBC. Have you figured out how to connect the HDD LED, POWER LED and RESET button?

I can't find any documentation on this thing.

whitepawn wrote on 2021-03-14, 11:54:
Hi, […]
Show full quote

Hi,

I couldn't help myself and bought another single board computer which labeled as PC420-G2 by protech company. Here is SBC page in 1997:
https://web.archive.org/web/19970327052802/ht … .tw/ipc420.html
IMG_20210313_201133.jpg
After cleaning the board, installed another cpu Cx486 dx2 66 (because the one came with the board is not original, socket was very loose) it posted with "Bios checksum error".I searched online for this error it seemed either bios rom is corrupted or RTC battery problem.Second one is easy to fix for me so i retrofitted CR2032 to dallas chip but checksum error persisted so i tried another bios ( PSC-486 VER:C Jumper or manual information needed) with it which has same chipset and even bios stickers are the same. Also PSC-486 bios gives same error which i believe i have a jumper setting problem.(Top left bios is similiar bios from my working PSC-486.Bottom right is original bios.)

IMG_20210314_144539.jpg

IMG_20210314_111707.jpg

I can boot from A: drive and tried uniflash but bios chip model seems unknown in uniflash.

Started looking for bios backups online but couldn't find any.Meanwhile i discovered same hardware was labeled as "486/586 SBC Ver:GC".This board is very similiar to my board but also i couldn't find any information about it.

By any chance does someone have manual information for it?Or any advice?

Thanks in advance.

Reply 15 of 19, by whitepawn

User metadata
Rank Newbie
Rank
Newbie
ericmackrodt wrote on 2022-01-24, 12:37:

Hi Whitepawn,

I have the same SBC. Have you figured out how to connect the HDD LED, POWER LED and RESET button?

I can't find any documentation on this thing.

Hi Eric,

Does your board same hardware revision?A picture of board will be very good if possible.

Here is the CPU jumper settings that i found on forum somewhere:

Filename
PC-420.pdf
File size
259.23 KiB
Downloads
47 downloads
File license
Public domain

JP3 must be HDD led.I don't know about Reset and Power leds yet, but i will test and post here.

Could you please dump original bios with TL866?Uniflash or getrom also works.My original bios chip is corrupted.

GETROM:
http://cd.textfiles.com/microhaus/mhblackbox3 … MORY/GETROM.ZIP

UNIFLASH:
http://www.win3x.org/win3board/viewtopic.php? … 455&language=en

Thanks in advance.

Reply 16 of 19, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie

I had a look on the wayback protech website and found: https://web.archive.org/web/20030817092320/ht … /half/420m2.pdf

Which appears to be the manual for the PC-420.

Board layout looks about right. JP1 is HDD, JP2 is Turbo LED, JP6 is reset and pin 1-3 on JP7 is power LED. So I think (in that other thread: 486/5x86 SBC power/hdd led and reset button locations. ) I got pretty close.

[edit to add: seems odd that there's a turbo LED header, but I can't see anywhere that a turbo switch should be connected. The manual even says that you can set the turbo in the BIOS, but that you can toggle between them using the turbo switch. Looked up the datasheet for the clock generator SC464 and it looks like there are more speeds supported than given in the PC-420 manual. That chip has a "Doze" input on pin 15 that looks like it works like a Turbo switch. It looks like pin 15 is routed out from the chip, so it must go somewhere. Maybe there's a key combination for changing modes (maybe ctrl-alt-+ and ctrl-alt-- )?

Or possibly the Turbo LED is actually used for the Green PC modes: Doze, Standby, Suspend)

Frequency table from lazibayer's list: Re: Clock generators on Socket 7 and older motherboards
Datasheet: https://www.datasheetarchive.com/IMISC464%20-datasheet.html , first link in the context search entries]

Reply 19 of 19, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie

I had a go at comparing your spreadsheet with the manual, and it looks like the cpu types match the entries there, but not always the actual CPU. I've added a column for the type as given in the manual:

EXCEL_Manual_CPUtypes.JPG
Filename
EXCEL_Manual_CPUtypes.JPG
File size
143.47 KiB
Views
817 views
File license
Fair use/fair dealing exception

The frequency jumper JP4 is a bit odd. I've traces a few pins and it looks like it connects to the clock generator like this:

JP4
5 3 1 -> S2 G S1
6 4 2 -> ? S0 G

on the clock generator, S0 is pin 11, S1 is pin 10 and S2 is pin 17. Frequency is also affected by ST# (pin 18) but it doesn't look like anything is connected to that, and Doze# (pin 15) which looks like it's controlled by the chipset.

It looks like all 3 lines (S2 to S0) are pulled high through 3 4.7k resistors: R15, R16 and (on the back) R99 (numbers from the photos ericmackrodt posted). So S0 and S1 can be pulled low by putting a jumper on 3-4 and 1-2, but S2 only gets connected to a trace that disappears under the M5113 chip. I guess that's so that the system has some control over S2, otherwise the sleep modes might not work properly. E.g. the datasheet shows the Doze frequency can be 33MHz is S2 is '0'.

The manual doesn't list 50MHz, but the clock gen datasheet says that is selected by pulling only S0 low, so that just means fitting a jumper to 3-4. It may be that the manual doesn't mention 50MHz because the Doze frequencies in the datasheet are a bit strange compared to the other frequencies.

It should be possible to get the other weirder frequencies of 30,60,66&80MHz by playing with 5-6, or maybe 5-3 (to pull S2 to ground).