• Please review our updated Terms and Rules here

256K ISA 8-bit RAM card problem

Q

QuantumII

Veteran Member
Joined
Aug 12, 2008
Messages
503
Location
Oslo,Norway
Hi all

Just recently my luggable Philips Logic Analyzer PC (xt clone with extras)'s 256K ISA card stopped working.

It began with a memory error on boot, and on the next reboot it only counted up to 256k (wich is on-board)

The 8-bit ISA card does not seem to be alive at all, and I cannot find any documentation it..

However, if I move around on the DIP switches on the card the PC does not boot at all, so it does _something_

First I thought that is was just a bad connection between the motherboard and the ISA card, but moving it around in other slots makes no results. the machine still counts up to only 256k both with the card in the slot or without.

I think I may need to replace the card.

There are pictures of the card here: it has 4 large IBM branded chips, each at 64k (I believe) making up a total of 256K.

Front

Back

Have any of you seen this card before, or have any info on it? Is there any other cards of the same type you can recommend ?
 
He says the machine counts up to 256K with the expansion board in or out, so, it probably has 256K on board, T.

Get some non-lubricating component cleaner and spray the switches then slide them back and forth and set then up again.

Sounds like it's trying to map the RAM into the same place as the RAM on the MB.

Also, clean the edge connector with a pencil eraser.

If this doesn't work, you may have a bad decoding chip on the card.
 
I've got the O&A chapter on that board. Pretty straightforward in spite of the IBM 64Kx9 modules.

You might want to look at the output of VR1 (LM317T). It should measure 8.5v.

The high side of C12 should measure 2.2v

If that checks out, try swapping the modules and see if the POST RAM test goes up higher.
 
Chuck(G) said:
I've got the O&A chapter on that board. Pretty straightforward in spite of the IBM 64Kx9 modules.

You might want to look at the output of VR1 (LM317T). It should measure 8.5v.

The high side of C12 should measure 2.2v

If that checks out, try swapping the modules and see if the POST RAM test goes up higher.
Click to expand...

Thanks, I'll look at this today. Do you have the Q&A chapter in some electronic sendable format ?

And yes, the machine has 256k onboard, and this extra card. it used to count up to 512k before the card died.
 
QuantumII said:
Thanks, I'll look at this today. Do you have the Q&A chapter in some electronic sendable format ?

And yes, the machine has 256k onboard, and this extra card. it used to count up to 512k before the card died.
Click to expand...

I can scan it, if you need it. Just PM me and I'll do the deed and send you a URL to download it.
 
I've measured the output from the voltage regulator. It's 10v 5v 10v

C12 is 2,19v

I tried to measure the pin B1 (8,5v) and J1 (GND) on the ram sockets. It's 3,05v (Supposed to be 8,5v.. so this is bad)

J1 and C1 reads 1,1v (Supposed to be -2,2v)

J1 and H1 reads 6,7v (Supposed to be 5v)

Something is screwed up royal here is seems..
 
Use the machine chassis as your ground reference (not J1) or the black leads on the main motherboard power connector. Redo the measurements. The values you've given don't make sense.

Check the resistance both ways across CR1 (with the card out of the machine). Should be infinity in one direction, but not the other.
 
QuantumII said:
I've measured the output from the voltage regulator. It's 10v 5v 10v

C12 is 2,19v

I tried to measure the pin B1 (8,5v) and J1 (GND) on the ram sockets. It's 3,05v (Supposed to be 8,5v.. so this is bad)

J1 and C1 reads 1,1v (Supposed to be -2,2v)

J1 and H1 reads 6,7v (Supposed to be 5v)

Something is screwed up royal here is seems..
Click to expand...

If you can't repair it, you should considering getting a RAM card that uses 4164s. This is because they use easy-to-find RAM IC's (I got over 100 spare just in cause of failure), and that they use the same voltage as the rest of the system so all of them can't fail at the same time, unless your PSU goes crazy.
 
QuantumII said:
I've measured the output from the voltage regulator. It's 10v 5v 10v

C12 is 2,19v

I tried to measure the pin B1 (8,5v) and J1 (GND) on the ram sockets. It's 3,05v (Supposed to be 8,5v.. so this is bad)

J1 and C1 reads 1,1v (Supposed to be -2,2v)

J1 and H1 reads 6,7v (Supposed to be 5v)

Something is screwed up royal here is seems..
Click to expand...

Umm, is this some sort of special regulator?

Shouldn't one one those pins be at ground?
 
Chuck(G) said:
Use the machine chassis as your ground reference (not J1) or the black leads on the main motherboard power connector. Redo the measurements. The values you've given don't make sense.

Check the resistance both ways across CR1 (with the card out of the machine). Should be infinity in one direction, but not the other.
Click to expand...

Ok, will do it tomorrow, as the machine has been put back into it's storage space for the evening.

btw, the motherboard has many free rows for RAM IC's.. The problem is, I have no documentation whatsoever on that motherboard. The only manual I have is a Quick Reference guide (it's a program on the HDD) which says that the motherboard is a combined XT-turbo and FDD controller. What is a bit weird is that an XT has only one DIP bank. This MoBo has two... That leaves it to the imagination on how to correctly configure the board for more onboard RAM...
 
When you've got the system out tomorrow, why not grab a photo of the motherboard while you're at it? Maybe someone will recognize the configuration...
 
Chuck(G) said:
When you've got the system out tomorrow, why not grab a photo of the motherboard while you're at it? Maybe someone will recognize the configuration...
Click to expand...

Yes, I'll do that.

here's a picture of the MoBo:

mobo_s.jpg
 
Last edited:
The DRAM setup looks familiar. 2 rows of 256K (41256) DRAM, 4x 64Kx4 (4464) DRAM and 2 64Kx1 (4164) DRAMs.

Have you tried populating the second row of 256K? You should be able to get rid of the RAM card entirely.
 
Last edited:
Chuck(G) said:
The DRAM setup looks familiar. 2 rows of 256K (41256) DRAM, 4x 64Kx4 (4464) DRAM and 2 64Kx1 (4164) DRAMs.

Have you tried populating the second row of 256K? You should be able to get rid of the RAM card entirely.
Click to expand...

If only I had RAM chips, I could do that.
 
You must log in or register to reply here.
Back
Top