RAM test programmes

[morelenmir]

I thought it would be a good idea to try to group together all the different RAM test programmes that are available in one thread.

 

I have used the built-in self test RAM checker, obviously. I have also previously used the 'XRAM' programme by -XI-/Satantronic. I have attached it below:

 

XRAM v0.21.xex

 

Are there any other that run on the A8 as opposed to the ST?

Edited September 21, 2018 by morelenmir

[+Nezgar]

A few ATR's of testers from previous threads:

• http://atariage.com/forums/topic/189604-small-thing-memtest-10-1996-jaku-b/?p=3815398
• http://atariage.com/forums/topic/255228-spartados-x-448/?p=3565997
• http://atariage.com/forums/topic/213602-ramrodxl-testing-in-a-1200xl/?do=findComment&comment=2777344

[manterola]

Is there are any test program that will perform a long term test? By that I mean a program that write to the extended banked ram, and then keep checking the data written is still integral and not corrupted after several minutes, hours, etc.

[morelenmir]

I would like something just like that myself!

 

I seem to get inconsistent results with XRAM. One run and every bank shows an 'E' and the next they are all 'O'... Another run and a handful show 'E', the next another different bank is 'E' while the rest are all 'O'. Weird. I would like it to keep looping and test over the course of a several hours.

 

The odd thing is all programmes run correctly and the machine as a whole is the most stable it has every been! Weird I say again!

Edited September 21, 2018 by morelenmir

[flashjazzcat]

Why do you need to loop the test for several hours if running it a couple of times already showed there's an issue? I can perhaps see the point of a burn-in test for a machine which apparently passes tests, on the other hand.

 

Is this the same machine which already caused months or years of grief?

Edited September 22, 2018 by flashjazzcat

[shoestring]

I plan To write a seperate version for systems with extra memory but to do it in this fashion seem like overkill.

 

The exhaustive tests are more than enough.

[+CharlieChaplin]

Is there are any test program that will perform a long term test? By that I mean a program that write to the extended banked ram, and then keep checking the data written is still integral and not corrupted after several minutes, hours, etc.

 

Hmmm, the PAGEFIND.BAS by Thorsten Karwoth has a long term test (named Deep RAM Test) for ext. RAM that lasts one or more hours ?!? Could be wrong, but load it with MyDOS 4.50 since it was written for this DOS (multiple files, one of them Basic, the other maybe ML or unknown)...

 

There is also Memdrive 1.0 by A.Bertram that does a "Dauertest" (permanent test), maybe for a fixed time, maybe as long as you wish...

 

Attached some system-test software (SYSTEST.ATR) with Pagefind.BAS, Memdrive 1.0 and a copy of good old MyDOS V4.50 (with all its known bugs).

 

There is one drawback though, the Pagefind program and Memdrive both use german language, but maybe Google or some other translator can help ?!?

MYDOS450.zip

SYSTEST.zip

Edited September 23, 2018 by CharlieChaplin

[Marius]

 

Hmmm, the PAGEFIND.BAS by Thorsten Karwoth has a long term test (named Deep RAM Test) for ext. RAM that lasts one or more hours ?!? Could be wrong, but load it with MyDOS 4.50 since it was written for this DOS (multiple files, one of them Basic, the other maybe ML or unknown)...

 

 

Yes! This is the one I had in mind when I read this thread. It is one of the best ramtests, but probably especially for those who have a TTL-based-DRAM memory upgrade. It checks also whether there are 'timing-issues', which did occur frequently on my PAL Buchholz Quarter Meg 800XL (which was for a certain period of time back in the mid-90's my only upgraded atari).

 

On SRAM memory upgrades it seems there could be other kind of issues, but not the issue that the refresh-timing thing PageFind tests for.

 

Anyhow, I have pretty fond memories of using Thorst K.'s 130XE+ Assembler. It came with an auto-config, including that memtest. Those were really good days. I have been using that assembler for 2 years. Then I moved (back) to Mac/65, and later I switched to a cartridge version of Synassembler, which is still my #1 assembler now.

[_The Doctor__]

I would like something just like that myself!

 

I seem to get inconsistent results with XRAM. One run and every bank shows an 'E' and the next they are all 'O'... Another run and a handful show 'E', the next another different bank is 'E' while the rest are all 'O'. Weird. I would like it to keep looping and test over the course of a several hours.

 

The odd thing is all programmes run correctly and the machine as a whole is the most stable it has every been! Weird I say again!

no you might be fine, xram has some issues not yet worked out, I've seen it with rambo (256) vs XE (320) mods.. the subject periodically get touched on... though it hasn't been dealt with yet...

http://atariage.com/forums/topic/59071-800xl-256k-ram-upgrade/?p=3923953

 

things get askew

Edited September 23, 2018 by _The Doctor__

[+Nezgar]

Oh yeah, because (some? All?) Rambo 256K XL, using 4 bits (for 16 combinations) have 12 actual banks of usable extended RAM (192K), and 4 banks that are actually mirrors of the base 64K, unlike 320K XE where it's 64Kbase plus a true 16 banks of extended ram.

 

The xram tester might not interpret those last 4 banks correctly.