• Please review our updated Terms and Rules here

IBM 5160 Memory

Gary C

Veteran Member
Joined
May 26, 2018
Messages
2,256
Location
Lancashire, UK
Hi

I have a quadram quadboard clone thats fully populated and as far as I can see, set correctly but the machine only steps up to the 256K memory installed on the motherboard (the original quadboard is toast due to battery acid and a lack of enthusiasm to repair evaporated tracks)

I am aware that the XT uses a quick memory pattern test to determine the size so if there is an error in the first row its knon to think the memory has ended and stop testing I believe.

So my question is, once the POST has determined the memory size, is the memory on the quadram card still accessible for testing and does anyone have a program to do it. I can't imagine a way of locking out that portion of the memory map from being tested, but you never know. What I dont want to do is mess about swapping endless numbers of chips trying to locate duff ones.

Would Checkit do the job ? I see that is on the -0° site.
 
In BASIC you can use the DEF SEG= command to PEEK and POKE anywhere in the memory map. For instance:

DEF SEG=&h4000

Would set the segment register so:

POKE 0,65

Would poke "65" into the first location after the 256k of motherboard memory and:

PRINT PEEK(0)

Will show what's stored there. You could do some ad-hoc testing with this to see if at least looks like memory is there.
 
Cheers, some useful stuff.

Always found the Intel segmented memory annoying after learning on the 6502, but as I want to try and do something with my Multibus 286, I better get used to it.
 
Well, the DEBUG method was perfect. The version of CHECKIT I could run with only 256K memory would not allow me to test any memory not detected by the POST routine.

Using debug with the parity turned off, I found Segments at 0x4000 and 0x5000 could not be written too, but memory above that was fine. I checked SW1 and they all worked so I actually suspected the settings. I had a badly damaged original Quadram Quadboard, but had managed to locate an unpopulated clone for a very good price and swapped all the RAM over.

The board looks almost identical but has about three chips absent. Setting the switches as per the Quadram manual didn't work, so I kept dumping blocks of memory in debug and changing the switches and hey presto it now gives me the 'full' 640K !

The board itself has no identifiers on it at all, and I suspect the seller had tried it themselves using the standard settings and could not get the RAM to work as it was sold with only the RTC working but 'RAM function not checked' :)
 
Interesting, the chip layout is the same, but there are no silkscreen names or identifying numbers on it at all. Looks like it was a clone of that board.

Found the manual for that version, need to check if the settings are the same.

Need to put a new battery in next.
 
Last edited:
Doh, feeling daft

This link http://www.vcfed.org/forum/attachment.php?attachmentid=7213&d=1322452261 seems to indicate that you set SW1 to match the amount of memory on the card, whereas it sets the start address and thus needs to match the memory size of the motherboard. Silly me.

Looking at the manual for the 1984 quadram quadboard, makes it obvious its the start address which actually makes much more sense, ah well found it the hard way.

Surprisingly, the original Varta battery is still working and hasn't leaked but I'm not trusting it. Got a new one on the way
 
Last edited:
The battery on my Quadboard started leaking about 4 years after purchase. I ran some leads off the board to connect the battery externally, where it could do no harm. Still have the board, but haven't used it in a system for many years...
 
Back
Top