VOGONS


Reply 20 of 37, by soviet conscript

User metadata
Rank Oldbie
Rank
Oldbie

don't feel about the M919 board. its still a decent workable board and runs fine without the L2 cache. when it comes to PCI 486 boards one should just be happy to of found one.

regarding the COAST module. they usually say "FOR 486 M919 USE ONLY" on the back of the stick BUT this isn't always the case. I say a user here on the forum that had a stick that did not have this printed on it and that looked different so there are variations out there. don't even try to use a COAST module in this board unless your 100% sure its for the m919. They pop up from time to time. personally I wouldn't pay more then $40 for one. It took me well over a year until one poped up for a reasonable price on a vintage PC selling forum. It did add a nice boost to my board but honestly it ran pretty much OK without it.

One bit of advice. The board can be a bit buggy. My M919 cant seem to "see" the L2 cache module if I have EDO RAM installed. If I use FPM RAM it works just fine so keep that in mind.

here, this thread may help Bought a 486 motherboard

Reply 21 of 37, by Turboman

User metadata
Rank Member
Rank
Member

Thanks for all the discussion guys appreciate it, I do feel better now. 😀

So, that Cache stick I found on ebay should not be tried in the M919?

http://www.ebay.com/itm/IN332Q4AO84XG15-Smart … =item1c0d4628b3

What else should I do in the bios for settings to test the M912 cache? and is there any other programs I should use?

Reply 22 of 37, by Turboman

User metadata
Rank Member
Rank
Member

Guys, I did a continuity test that was suggested and found that on all the cache chips the last pin to the bottom row right is the only one that beeps everything else I get a reading and on the Tag chip I get two beeps
top side row pin 7 and pin 9 everything else is gives readings. So, this means that the cache is all real and just configured wrong?

Reply 23 of 37, by soviet conscript

User metadata
Rank Oldbie
Rank
Oldbie
Turboman wrote:
Thanks for all the discussion guys appreciate it, I do feel better now. :) […]
Show full quote

Thanks for all the discussion guys appreciate it, I do feel better now. 😀

So, that Cache stick I found on ebay should not be tried in the M919?

http://www.ebay.com/itm/IN332Q4AO84XG15-Smart … =item1c0d4628b3

What else should I do in the bios for settings to test the M912 cache? and is there any other programs I should use?

No, thats not the right kind anyways.

If it doesnt look like this https://ancientelectronics.files.wordpress.co … /win31cache.jpg

or if the seller doesn't 100% state the stick goes to the m919 board and you trust thier word I wouldn't risk killing the board.

Reply 24 of 37, by buyerninety

User metadata
Rank Newbie
Rank
Newbie

My intention with the continuity test was mainly just to see if the pins on
a single fake cache chip had zero ohm continuity to the other pins on
that single fake cache chip. Perhaps it would be easier if we just enquire if
anyone else has broken open a fake 28 pin cache chip, and if so, are the pins
all standalone separates from each other (most probably yes).

Incidently, does anyone know who the chipmaker is of the L2 cache chips in
this picture;
http://ru.pc-history.com/wp-content/uploads/M … 12_v14_mama.jpg

Reply 25 of 37, by Turboman

User metadata
Rank Member
Rank
Member

I tried it on the M919 and I get zero ohm all the way across pin to pin all 4 sides. So, I'd assume they are all connected to each other inside at least on that style of chip.

Reply 26 of 37, by FGB

User metadata
Rank Oldbie
Rank
Oldbie

Why waste your time with this? Just look at the traces around the cache rectangles on the M919. They are going NOWHERE. It's a well known fact that these are fake. You won't get cache on the M919 until you put the proprietary M919 cache stick into the damn slot 😁

www.AmoRetro.de Visit my huge hardware gallery with many historic items from 16MHz 286 to 1000MHz Slot A. Includes more than 80 soundcards and a growing Wavetable Recording section with more than 300 recordings.

Reply 27 of 37, by buyerninety

User metadata
Rank Newbie
Rank
Newbie

All my comments have been in regard to the M912.
Unfortunately no-one seems to know what chipmaker made those
chips in the pic I cited above with a symbol like a 'Pisces' sign...

Reply 28 of 37, by Turboman

User metadata
Rank Member
Rank
Member

Well, I tried jumping it to 256k and the system would not go past the post screen so, I put it back the way it was and it works again so, bank 1 looks to be fake then, it's odd that it displays 256k but is jumped for 128k.

Reply 29 of 37, by HighTreason

User metadata
Rank Oldbie
Rank
Oldbie

You have to wonder what would happen if you got adventurous and removed the fake second bank, installed sockets and inserted real ICs.

Not saying you should, just saying "I wonder, what if?" for the hell of it.

My Youtube - My Let's Plays - SoundCloud - My FTP (Drivers and more)

Reply 30 of 37, by buyerninety

User metadata
Rank Newbie
Rank
Newbie

{Grin} Oh, hello, there's HighTreason down there on the low road animated-smileys-waving-093.gif...

Yes, if we did remove those four fake SRAM chips in bank 1 of that M912,
what replacement chips are reasonably likely to be pin compatible with
those bank zer0 'TM 11256a' chips (which don't seem to be too common)?
Well, I guess we could see if there is any other M912 v1.4 board that has
the same 'TAG RAM UM61256' chip with different (though probably pin
compatible) SRAM chips (instead of TM 11256A) - now what would be the
maker/model identification of those different SRAM chips be?.. er.. hel-lo ,
now we're back up my previous posts... thinking.gif 😊

Reply 31 of 37, by HighTreason

User metadata
Rank Oldbie
Rank
Oldbie

No need to be snarky.

Judging to the available information on the Googles, it is a regular 32k*8 SRAM chip and is pin compatible with others, such as W24M257AK, CY7C199 and the offerings from UMC and Toshiba that seem to be more prevalent. You can mix manufacturer, the motherboard neither knows nor cares, I have three different brands of chip in my 386 board, including some of those TM Tech ones.

If you have a multi-meter, try probing the pins and see if they match up with what you'd expect on any other chip as per their datasheets; http://pdf1.alldatasheet.com/datasheet-pdf/vi … 24257AK-15.html

Oh, wow, look, I managed to explain all of that without having to resort to rather immature use of emoticons.

My Youtube - My Let's Plays - SoundCloud - My FTP (Drivers and more)

Reply 32 of 37, by FGB

User metadata
Rank Oldbie
Rank
Oldbie

Very mature to judge maturity by the use of emoticons. Sigh.

@Turboman: The odd false display was part of the game. Its programmed into the BIOS. It is part of the old PC-CHIPS fraud. If you swap the BIOS with one of another board with the same UMC chipset you will see the correct cache size. There are other boards, I remember M919s, that were showing 256KB in the POST screen but had no cache at all, just the plastic rectangles with the WRITE BACK CACHE engraving.

www.AmoRetro.de Visit my huge hardware gallery with many historic items from 16MHz 286 to 1000MHz Slot A. Includes more than 80 soundcards and a growing Wavetable Recording section with more than 300 recordings.

Reply 33 of 37, by buyerninety

User metadata
Rank Newbie
Rank
Newbie

Snarkiness Index: zero.
Apparently, Hightreason, you've taken no roads from Hull as high north as Loch Lomond.
Anyways, despite differing paths, we have arrived at the same point - to provide
Turboman with advice that he could try using 4 of e.g. any of these type of chips;
Winbond W24M257AK or Cypress CY7C199 or Crosslink CL63C256N to replace
the fake cache chips on his M912 in bank 1 (that's replacing components on the
M912 silkscreened as U28 , U29 , U30 and U31) and then resetting the applicable
L2 cache jumpers for 256KB of L2 cache..

Reply 34 of 37, by HighTreason

User metadata
Rank Oldbie
Rank
Oldbie

Assumption is the mother of all f---ups. You are clearly unaware that my mother's side of the family is from the far north, though the majority were south of the border in Corbridge by then and my Grandma was the one which moved south. Prior to that I think they started out mostly as a bunch of Gypsies (Tinkers to you) and slowly moved away from their roots. As it happens, I've probably seen more of this country than you have as well as parts of Europe because I spent most of my early life traveling with my father and I've seen a large portion of Scotland, it's like a more impressive version of Yorkshire, but colder.

Which would explain why I do not speak with a local accent or dialect (Nobody here knows what a spelk is and they don't pronounce the letter H whereas I do for example... They have a hard time understanding me with their primitive language, bless them, they can't even pronounce my name properly; apparently my name is 'Pole' to people around here), i fry every food that can be fried and if you don't nail down the contents of your garden it will probably end up in my kitchen or as an eyesore on my lawn... Or traded in to be melted down as scrap metal so I can buy some tiles/gravel/pegs and try to sell them to you... Do you want your roof tiling?

Edit: Actually, I think that last one is more from my Dad's side of the family; Carnies, travellers and your average big council estate family who get evicted for blowing their neighbors front door open with a stolen cannon using home made gunpowder and ammunition - no joke, my Dad and my Uncle did this once when someone owed them money.

My Youtube - My Let's Plays - SoundCloud - My FTP (Drivers and more)

Reply 35 of 37, by soviet conscript

User metadata
Rank Oldbie
Rank
Oldbie
FGB wrote:

Very mature to judge maturity by the use of emoticons. Sigh.

@Turboman: The odd false display was part of the game. Its programmed into the BIOS. It is part of the old PC-CHIPS fraud. If you swap the BIOS with one of another board with the same UMC chipset you will see the correct cache size. There are other boards, I remember M919s, that were showing 256KB in the POST screen but had no cache at all, just the plastic rectangles with the WRITE BACK CACHE engraving.

on my m919 w/ fake cache chips when no COAST module was installed it said something like "write back cache enabled" or something along those lines at post but never stated an amount. Once I installed a proper COAST module it changed to actually state "256kb cache enabled" though sometimes if it wasn't seated correctly in the slot it would report 128kb. Also as I stated before my board would report the 256kb cache but would not be able to access it. took me forever to figure out if I swapped EDO for FPM RAM it then detected the L2 correctly and could actually use it.

Reply 36 of 37, by Turboman

User metadata
Rank Member
Rank
Member
HighTreason wrote:

You have to wonder what would happen if you got adventurous and removed the fake second bank, installed sockets and inserted real ICs.

Not saying you should, just saying "I wonder, what if?" for the hell of it.

Oh, I am thinking of doing it, just not anytime soon, 🤣

Reply 37 of 37, by Turboman

User metadata
Rank Member
Rank
Member
FGB wrote:

Very mature to judge maturity by the use of emoticons. Sigh.

@Turboman: The odd false display was part of the game. Its programmed into the BIOS. It is part of the old PC-CHIPS fraud. If you swap the BIOS with one of another board with the same UMC chipset you will see the correct cache size. There are other boards, I remember M919s, that were showing 256KB in the POST screen but had no cache at all, just the plastic rectangles with the WRITE BACK CACHE engraving.

Would a bios update change it by any chance or if I updated the bios from that of non fake board?