• Please review our updated Terms and Rules here

JX30 VL-Bus Motherboard issues

Lutiana

Lutiana

Administrator
Staff member
Joined
Mar 26, 2009
Messages
3,302
Location
Dublin, CA USA
I recently aquired a JX30 VL-Bus Motherboard with a 5x86 chip on it (133Mhz). But I am having 3 issues with it:

1. I cannot for the life of me find a battery to replace, nor does it appear to have a one of those clock chips with the built in battery as near as I can tell. Anyone have some ideas on how to find the battery?

2. The CPU comes up as 100Mhz, and that's only with the turbo on. I cannot find any information on how to configure it so that works as the full 133Mhz. The board is jumpered and dipped for the DX2-66, which I think is correct. Any ideas?

3. There is 256Kb cache on the board. The jumpers are set right for this, but as soon as I enable "External Cache" in the BIOS the board freezes on post. It says that only 64Kb was tested ok and then only a hard boot will get it to reset. I can get it to boot just fine if I set "External Cache" to disabled.

I cannot work out what the BIOS settings should be and I can't find much documentation on the board.

Thanks for the help.
 
Both the links above are the correct motherboard. I have been through every jumper and dip that is shown there and verified that they are set correctly.

I managed to find the battery header, so I will just use an external battery. Easy done, its the Motorola RTC chip (HD146818AP), which is why I could not find it. Does that chip have a batter inside it like the Dallas?

The cache jumpers are set correctly. There are 9 KM68257BP-20 chips installed. As near as I can tell these are 32kX8 chips and total 256kb Cache, and the jumpers match this estimation (I did not change them). But as soon as I set the BIOS to "External Cache" enabled, it freezes on post at 64kb.

As far as the CPU goes, running this utility tells me that it is running at 133Mhz, I am not sure if the CPU is actually running that fast and the BIOS POST is wrong, of if the utility is merely grabbing the ID from the chip and telling me what it is capable of doing. Any suggestions on a utility that will tell me the actual running core speed?
 
The BIOS will display what it is programmed to display, some models that expected DX4 processors will show the correct speed and earlier ones will show something else (default top speed).

Reseat the cache chips and see if that makes any difference, also check to see if any of them feel warmer then the others. Usualy the 9th chip is TAG SRAM and should be faster then the other 8 (15ns maybe?).
 
Unknown_K said:
The BIOS will display what it is programmed to display, some models that expected DX4 processors will show the correct speed and earlier ones will show something else (default top speed).
Click to expand...

I am guessing this is the "problem". I'll need to find some software to test this theory.

Unknown_K said:
Reseat the cache chips and see if that makes any difference, also check to see if any of them feel warmer then the others. Usualy the 9th chip is TAG SRAM and should be faster then the other 8 (15ns maybe?).
Click to expand...

Nope all 9 chips are identical. If I had to be pushed for an answer I would say that the odd one out is the one that is hotter than the others. There is a 10th chip that is the TAG chip its a TC5588P-20. See here for a motherboard layout.

They are arranged like this:
Code: 
Bank 0           Bank 1
9                  Jumpers
8                  4
7                  3
6                  2
5                  1
 
Sounds like a bad Cache RAM chip.

Do you have an EPROM programmer? Most of the current ones have the ability to test Static RAM chips.

All of the KM68257BP-20 are simply very very fast versions of a 62256 chip, a standard 32k x 8 static RAM. The pinouts are the same too.

RJ
 
channelmaniac said:
Sounds like a bad Cache RAM chip.

Do you have an EPROM programmer? Most of the current ones have the ability to test Static RAM chips.

All of the KM68257BP-20 are simply very very fast versions of a 62256 chip, a standard 32k x 8 static RAM. The pinouts are the same too.

RJ
Click to expand...

I do not have an EPROM programmer, so I will need to either leave the cache disabled or find another way to test them.

Any idea on where one can buy new chips?
 
lutiana said:
I managed to find the battery header, so I will just use an external battery. Easy done, its the Motorola RTC chip (HD146818AP), which is why I could not find it. Does that chip have a batter inside it like the Dallas?
Click to expand...

Nope--it's just the bare RTC/SRAM chip, the original animal as used in the 5170. It'll keep ticking as long as you care to replace batteries. A set of AA-sized alkaline batteries should last at least 3-4 years easily.
 
lutiana said:
I do not have an EPROM programmer, so I will need to either leave the cache disabled or find another way to test them.

Any idea on where one can buy new chips?
Click to expand...

I have some cache chips I've yanked off stuff in the past, I'll check the numbers for you when I'm home again later.

Edit:

9 chips
28-pin DIPPs

Code: 
UM61M256K-15
95330 N70218

Not sure if these could help you but if so let me know.
 
Last edited:
Chuck(G) said:
Nope--it's just the bare RTC/SRAM chip, the original animal as used in the 5170. It'll keep ticking as long as you care to replace batteries. A set of AA-sized alkaline batteries should last at least 3-4 years easily.
Click to expand...

Interesting. Must have been switched onto the board at some point, all the documentation I can find on it says:

The JX30 Motherboard does not require an external battery to be installed.

The Dallas / Benchmarq Real Time Clock Chip contains an internal battery for an approximate lifetime of 8-10 years.
Click to expand...

But no matter, the external battery I jerry-rigged is working fine. Its amazing what one can do with some electric tape, aluminum foil and a few bits of wire.
 
Chuck(G) said:
You can get a 4 AA battery holder at Radio Shack for under $2.00. Might hold together a bit better than tape and foil. :)
Click to expand...

Yes, it will. I intend to, but when I rigged it Radio shack was very much closed :D

But my tape job will most likely hold for a very long time actually, it would just be a pain to change batteries.
 
lutiana said:
Yes, it will. I intend to, but when I rigged it Radio shack was very much closed :D

But my tape job will most likely hold for a very long time actually, it would just be a pain to change batteries.
Click to expand...

Ah yes, I've done the "battery holder made of tape" thing before too. Just don't use $0.50 electrical tape, it tends to melt even at room temperature.
 
Does anyone know if these work as cache replacements?

I simply cannot find any of the chips that are actually on the board.
 
So according to this the TAG chip is by the ISA slots, and it is a different chip (a tc5588p-20). It appears to be an 8kx8 chip.

One thing I don't understand is why there are 10 slots for SRAM chips on this board. The one marked U200 seems to be the odd one out.

Do i need all 10? Could I pull the chip from U200 and put in the tag slot (and leave U200 empty?)
 
I suspect the cache is parity-checked.

You can use a 32Kx8 SRAM in the tag socket, regardless. The pinouts are compatible--it simply won't use the extra memory.
 
If you look at this picture:

MICRONICS-COMPUTERS-INC-486-JX30-VL-BUS-1.png


You will notice that there are 9 sockets by where the cache chips are and the TAG chip is up my the ISA slots.

I took the ninth chip (the odd one out) and switched it into the tag chip slot removing the 8kx8 chip. It worked! I know have 256kb of write through cache.
 
You must log in or register to reply here.
Back
Top