VOGONS


PcChips sucks!

Topic actions

Reply 20 of 56, by xjas

User metadata
Rank l33t
Rank
l33t

I had an M919 for a long time and IIRC NEVER got EDO to work in it. They claimed it supported EDO but I don't think it actually did. PC Chips liked to claim a lot of things back in the day, like "this board has 256kB cache!" or "hay guys we're totally not these three other shill companies selling exactly the same product with different stickers on it... honest."

I eventually got rid of the 919 when it got too flakey for even a basic DOS PC. To be fair, PC Chips cleaned up their act a lot & some of their socket 7 & later stuff was just fine, but their 3/486-era products were pure schlock.

twitch.tv/oldskooljay - playing the obscure, forgotten & weird - most Tuesdays & Thursdays @ 6:30 PM PDT. Bonus streams elsewhen!

Reply 21 of 56, by JidaiGeki

User metadata
Rank Oldbie
Rank
Oldbie

Yep, my M919 doesn't support EDO either, only FPM, and even then it's picky. Also can't get serial ports to work on it, and it's not playing nice with a controller card. Otherwise it's perfect 😉

Reply 22 of 56, by BitWrangler

User metadata
Rank l33t++
Rank
l33t++
JidaiGeki wrote:

Also can't get serial ports to work on it, and it's not playing nice with a controller card. Otherwise it's perfect 😉

Did the serial headers come in the box with it? There's at least 3 different ways they can be wired, likewise any other headers, so maybe you have the wrong ones.

Unicorn herding operations are proceeding, but all the totes of hens teeth and barrels of rocking horse poop give them plenty of hiding spots.

Reply 23 of 56, by gdjacobs

User metadata
Rank l33t++
Rank
l33t++
fitzpatr wrote:

Wow.
Under no circumstances should that tone be adopted on this site. That was both light-hearted humour, and a correct observation. To react in that way is peculiar. Perhaps the PcChips boards aren't the only things that are thin.

Gdjacobs is quite correct that their boards are incredibly thin to save on the copper layers. I have a socket 423 board on an 840 chipset from them that has turned into an arc. As TOBOR indicated, though, some of their boards are quite capable, fake cache aside.

Yeah, that seemed a little over the top. Must be a Leafs fan.

Seriously, we're in bad shape if we can't poke a bit of fun at old hardware. Some of PC Chips' business and design practices really teed them up for ridicule, but that just makes the irony of the well performing PC Chips 486 boards so much more delicious.

Almost as delicious as people collecting Packard Bell computers and paying premium after a large chunk of the world spent the last 15 years trying to forget they even existed. I think at one point (groundless stat!) 20% of USENET was dedicated to PB rage.

All hail the Great Capacitor Brand Finder

Reply 25 of 56, by ShirBlackspots

User metadata
Rank Newbie
Rank
Newbie

PC Chips was a highly disreputable company. They were the only company known company (that I'm aware of) to put fake cache chips on their 386 and 486 boards so it looked like it had cache, but didn't have any. Also, they're built in the US, and the only other manufacturer from the US was Biostar.

See http://www.redhill.net.au/b/b-bad.html

Reply 26 of 56, by gdjacobs

User metadata
Rank l33t++
Rank
l33t++

Except PC Chips wasn't one company. The heart of the infection was a company called Hsing Tech, but the cancer spread wide from there due to extensive resale and rebadge relationships.
http://web.archive.org/web/20060622215827/htt … cchips/aka.html

All hail the Great Capacitor Brand Finder

Reply 27 of 56, by appiah4

User metadata
Rank l33t++
Rank
l33t++

I used a PC Chips A101 V1.4 and had zero problems with it, with an unofficial BIOS update it could even boot up to 32GB IDE HDDs.. It was my first retro PC board when I got into the hobby, in fact.

Retronautics: A digital gallery of my retro computers, hardware and projects.

Reply 28 of 56, by Pabloz

User metadata
Rank Member
Rank
Member
JidaiGeki wrote:

Yep, my M919 doesn't support EDO either, only FPM, and even then it's picky. Also can't get serial ports to work on it, and it's not playing nice with a controller card. Otherwise it's perfect 😉

yea there is a lot of info on this site about that:
http://leileilol.mancubus.net/486/m919.html

"Whether the VIP takes EDO RAM depends upon the version of the UM8881F fitted. There is a three letter code after the year/week code. If the FIRST letter of this code is 'E' then the board will take EDO. Other VIPs may have a 'B' code letter and it has been reported that during experiments with EDO and this chip version it doesn't work and the cache chips became extremely hot. Thanks to Ian for this tip."

it also says that if you put 32mb or more memory the cache gets dissabled.

Reply 29 of 56, by amadeus777999

User metadata
Rank Oldbie
Rank
Oldbie

It may be a faulty comparison but on my HOT433 board the 8881 ... -BCO version has been running fine with EDO.

Attachments

  • bco.jpg
    Filename
    bco.jpg
    File size
    351.67 KiB
    Views
    1825 views
    File license
    Fair use/fair dealing exception
Last edited by amadeus777999 on 2017-10-21, 16:13. Edited 1 time in total.

Reply 30 of 56, by feipoa

User metadata
Rank l33t++
Rank
l33t++
Pabloz wrote:
yea there is a lot of info on this site about that: http://leileilol.mancubus.net/486/m919.html […]
Show full quote
JidaiGeki wrote:

Yep, my M919 doesn't support EDO either, only FPM, and even then it's picky. Also can't get serial ports to work on it, and it's not playing nice with a controller card. Otherwise it's perfect 😉

yea there is a lot of info on this site about that:
http://leileilol.mancubus.net/486/m919.html

"Whether the VIP takes EDO RAM depends upon the version of the UM8881F fitted. There is a three letter code after the year/week code. If the FIRST letter of this code is 'E' then the board will take EDO. Other VIPs may have a 'B' code letter and it has been reported that during experiments with EDO and this chip version it doesn't work and the cache chips became extremely hot. Thanks to Ian for this tip."

it also says that if you put 32mb or more memory the cache gets dissabled.

I've seen that and I used to quote it, but found it to be an unreliable proclamation. I tried to more officially determine what is what on these m919 boards inasmuch as RAM is concerned, but became fed up as these boards are generally flakely with results not repeatable. As a general rule, FPM RAM works on all 486 motherboards.

Plan your life wisely, you'll be dead before you know it.

Reply 31 of 56, by feipoa

User metadata
Rank l33t++
Rank
l33t++
amadeus777999 wrote:

It may be a faulty comparison but on my HOT433 board the 8881 ... -BCO version has been running fine with EDO.

I have also run into this, but was unable to draw any general conclusions. For a particular set of sticks of RAM I tried at one point, a certain EDO module seemed to work best on my unreliable HOT-433 boards, while other EDO sticks did not. The issue is that a lot of EDO tests on these 486 boards can initially seem reliable, until you start running stability tests, installing NT4, etc. In general, FPM tests have proven a lot more reliable on 486 boards. It seems like there is a very late model of the SiS 496 chipset which apparently works well with EDO. I think it is only found on some Lucky Star LS486 boards.

Plan your life wisely, you'll be dead before you know it.

Reply 32 of 56, by betamax80

User metadata
Rank Member
Rank
Member

I didn't know this thread was a thing! I used to work for the "outfit" that sold through Micro Mart (UK) that was Protek (Europe) Limited. I am VERY familiar with PC Chips boards. I have built on a TX-Pro II and an M598 "SX Pro" myself too. I currently own an M811 Athlon XP DDR266 board and am using for Windows XP.

Reply 33 of 56, by cj_reha

User metadata
Rank Oldbie
Rank
Oldbie

PC Chips boards were a hit or miss. My 486 gaming PC uses an M912 v1.7 and it is a quite fast reliable board, though the tons of TX PRO boards i also have are crap.

They were a weird company for sure.

Join the Retro PC Discord! - https://discord.gg/UKAFchB
My YouTube Channel - https://www.youtube.com/channel/UCDJYB_ZDsIzXGZz6J0txgCA

Reply 34 of 56, by amadeus777999

User metadata
Rank Oldbie
Rank
Oldbie
feipoa wrote:
amadeus777999 wrote:

It may be a faulty comparison but on my HOT433 board the 8881 ... -BCO version has been running fine with EDO.

I have also run into this, but was unable to draw any general conclusions. For a particular set of sticks of RAM I tried at one point, a certain EDO module seemed to work best on my unreliable HOT-433 boards, while other EDO sticks did not. The issue is that a lot of EDO tests on these 486 boards can initially seem reliable, until you start running stability tests, installing NT4, etc. In general, FPM tests have proven a lot more reliable on 486 boards. It seems like there is a very late model of the SiS 496 chipset which apparently works well with EDO. I think it is only found on some Lucky Star LS486 boards.

The LuckyStar board that I have running with EDO performs nearly identical to the HOT433 - both run faster with EDO, reaching 90MB+ write speed at 50mhz bus speed. After I have modified the Socket4 systems for my Doom tests I will re-test the 486 boards and see if I can reproduce errors with critical software. I haven't installed NT4 but had, for example, SISofts BurnIn running non-stop without errors.

Reply 35 of 56, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Interesting. Could you provide some benchmark evidence of the HOT433 running faster with EDO compared to FPM when identical wait states are used?

Plan your life wisely, you'll be dead before you know it.

Reply 36 of 56, by amadeus777999

User metadata
Rank Oldbie
Rank
Oldbie

Love to - but earliest will be late November as I have to update the eeprom of the HOT433(via network card) and wait for the Alliance cache rams.
I have even tested the HOT433@60mhz fsb with a 50ns FPM module and it ran slower than with an EDO module. A fsb of more than 50 was not really that efficient. Contrast that to the LuckyStar which ran rather well on a "diet" of 60 but, in the end, also sped away at 50 - by help of virtually zero waitstates.

As far as I can remember, in "speedsys" there was a memory speed- and a synthetic bandwidth difference of about 10 to 30MB+ between the two types of memory(same bios settings). Can't say for sure but Doom & Co. showed different results too.

I'm giddy on retesting the whole crew(including a DX60 vs Pentium60 comparison) in a more "empirical" fashion but it will take some time.

Is there a good freeware program for generating bar graphics which can then be rasterized at a certain resolution?

Reply 37 of 56, by BitWrangler

User metadata
Rank l33t++
Rank
l33t++

Yah, the next speed up after you have to add wait states is usually not that much faster or equal to speed before, but if it gets you 2 or 3 steps up, it gets more worth it.

Unicorn herding operations are proceeding, but all the totes of hens teeth and barrels of rocking horse poop give them plenty of hiding spots.

Reply 38 of 56, by feipoa

User metadata
Rank l33t++
Rank
l33t++
amadeus777999 wrote:
Love to - but earliest will be late November as I have to update the eeprom of the HOT433(via network card) and wait for the All […]
Show full quote

Love to - but earliest will be late November as I have to update the eeprom of the HOT433(via network card) and wait for the Alliance cache rams.
I have even tested the HOT433@60mhz fsb with a 50ns FPM module and it ran slower than with an EDO module. A fsb of more than 50 was not really that efficient. Contrast that to the LuckyStar which ran rather well on a "diet" of 60 but, in the end, also sped away at 50 - by help of virtually zero waitstates.

As far as I can remember, in "speedsys" there was a memory speed- and a synthetic bandwidth difference of about 10 to 30MB+ between the two types of memory(same bios settings). Can't say for sure but Doom & Co. showed different results too.

I'm giddy on retesting the whole crew(including a DX60 vs Pentium60 comparison) in a more "empirical" fashion but it will take some time.

Is there a good freeware program for generating bar graphics which can then be rasterized at a certain resolution?

The bandwidth reading in Speedsys is inaccurate. I'm more interested in seeing Doom, Quake, and perhaps some Windows-based synthetic benchmarks.

Plan your life wisely, you'll be dead before you know it.

Reply 39 of 56, by Pabloz

User metadata
Rank Member
Rank
Member

On the weekend i tried the very same PCCHIPS m919 motherboard with the same PSU, same memory, and the same vga pci card.
it boots, and now it detects the memory correctly 16MB. These motherboards are a lottery

Then today i tried to boot it again, it hanged at the first boot screen that says the Videocard model.
Then i tried to boot like 20 times after that and it failed to POST., always black screen with the monitor on standby

Then i took the motherboard out of the case, inserted the videocard, and it booted fine. 😕 😕 😕 😕 😕

I have no clue of what the hell these pcchips boards have, could it suffer from a bend? It was very common to see that in the 90s the people that built these PCs, placed like a foam sheet between the motherboard and the PC case.

and what keeps the board in place are those small standoff plastic snap that are not very secured because on one side of the board it doesnt have a bracket