oh yes! I got it. Wow I didn't see the start of this thread. I started at the suntek part that was linked to from another site. I see a lot of trying things was done to find settings.
Good starting points for us. And that software will help a lot!

Thanks a lot Predator , It's working now!

I had some trouble figuring it out but eventually I could understand (I believe) the logic behind it

Here is the testext result

After a while.... I understood that this is showing the 2 megs memory expansion immediately after the base memory (512k detected by the Olivetti)
Then I applied the SW1 settings as you posted:
SW1: OFF OFF ON OFF OFF OFF OFF OFF

And success , It took 128K for shadow ram and another 128K to expand base memory to 640

Now, for testing purposes the first game that I played on my first PC (an AEG olympia with 286 and 1mb ram)

A final question, if I want to use EMS, apart from setting the DIPs 5 and 6 on SW2 , which drivers is recommended ? (I guess EMM386 is no option since this is a 286 system)

May I ask how the performance is? Do you notice any dramatic speed decrease?

Marco wrote on 2026-01-22, 13:15:

May I ask how the performance is? Do you notice any dramatic speed decrease?

I can't tell, I just got this PC and I'm just testing it for now. It's a 286 running at 8mhz so slow is a feature

Very good thanks for the feedback. Think you also understood how it works 😁

So with your current config the RAM-card fills the gap between 512kb - 640 kb and then again at 1024MB?

Therefore when you run testext with the current config the last 00000000 block will change to FFFFFFFF? Can you try it? Run testext 3000, its finished faster then 😉

Predator99 wrote on 2026-01-22, 13:29:

Very good thanks for the feedback. Think you also understood how it works 😁

So with your current config the RAM-card fills the gap between 512kb - 640 kb and then again at 1024MB?

Therefore when you run testext with the current config the last 00000000 block will change to FFFFFFFF? Can you try it? Run testext 3000, its finished faster then 😉

No problem I'll give it a try tonight.
About EMS drivers , any recomendations? , You used a RMS.SYS driver but I can't find it.

Predator99 wrote on 2026-01-22, 13:29:

Very good thanks for the feedback. Think you also understood how it works 😁

So with your current config the RAM-card fills the gap between 512kb - 640 kb and then again at 1024MB?

Therefore when you run testext with the current config the last 00000000 block will change to FFFFFFFF? Can you try it? Run testext 3000, its finished faster then 😉

It indeed changed to FFFFFFFF
here are the results

Please, could you point me to where i can find the RMS.sys driver or to another that can work with this card? I would like to leave some EMS space to try commander keen or maybe wolf3d in a tiny box.

Here are the settings that worked for me for 1 and 2 megs installed. Machines are IBM AT with 512k ram.

Next card: PE-5060-C with extension board PE-5061-C

There seems to be something wrong with this card. I tested in 2 different 286 boards.

The main board has 3 banks with 512 kb each.

In testext, bank 2 and 3 are shown with correct size. However, for bank 1 only 384 kb are shown. Additionally there is a gap of 384 kb between bank 1 and bank 2.

When selecting another starting adress, the whole pattern moves. The size of the gap stays the same.

Any idea what could be the reason for this..?

EDIT: spotted a small damage...will check later

Traces are all OK...I have no idea what can be wrong.

I didnt find anthing about the manufacturer "PE". Here is an offer for another "PE-200" RAM card with a label "80 data". Seems to be the same manufacturer:
https://www.recycledgoods.com/80-data-control … er-card-pe-200/

Also found an overview of EMS cards here
https://www.atarimagazines.com/compute/issue1 … sion_Boards.php

Interesting also this link for "MOVE'EM CONFIGURATION GUIDE". There are the names of EMS drivers for several card. I.e. for the CEI
Computer Elektronik Infosys Ramflex - he EMS driver provided with the Ramflex board is called IMM4.SYS
Qualitas Move'em: This software is included inside 386MAX v 6.0 installation disk.
https://github.com/user-attachments/files/243 … 75/EMSGUIDE.TXT

I didnt give up with PE-5060-C that early and tried some other DIP combinations...and I noticed when switching DIP1 the gap in the card-RAM disappears!

So why is there a gap of 384 kB when DIP1 is set to "ON"...yes, its a configuration for PCs with less than 640kb RAM. It fills the area up to the 640kb. Then there is a gap of 384 kB for the ROM space. Then it continues at 1024 kB.…

The card is not damaged and 100% OK 😀

So with DIP 1 set to "off" it was rather easy to get the other settings....

DIP 1:
ON: Backfill / OFF: No Backfill

DIP 2-4: starting adress kb
OFF OFF OFF: x.0 k (000 )
ON OFF OFF: x.128 k (100)
OFF ON OFF: x.256 k (010)
ON ON OFF: x.384 k (110)
ON OFF ON: x.640 k (101)
OFF ON ON: x.768 k (011)
ON ON ON: x.896 k (111)

DIP 5-8: starting adress mb - 0=OFF, 1=ON
0000 1.x m
1000 2.x m
0100 3.x m
1100 4.x m
0010 5.x m
1010 6.x m
0110 7.x m
1110 8.x m
0001 9.x m
1001 10.x m
0101 11.x m
1101 12.x m
0011 13.x m
1011 14.x m
0111 15.x m
1111 16.x m

So for a starting adress of 1.384 MB in a 286 with 1 MB onboard the setting needs to be...
OFF ON ON OFF - OFF OFF OFF OFF

There is no DIP left and I assume the jumpers relate to the I/O ports. Therefore this card can only do XMS but no EMS.

Next card is the DFI MF-3000. Some jumper setting are here, but not very complete.

https://theretroweb.com/expansioncards/s/dfi-mf-3000

Will try to add some data later.

I have some difficulties with the MF-3000. There are some deep scratches on the backside. as far as I can see I have fixed the broken traces, but the result is still not OK. I am not sure if there still a broken trace, a damaged IC or RAM or wrong DIP setting.
At least it seems the starting adress configuration on DIP1 corresponds exactly to the DFI AT-Rambank, so the missing ones at theretroweb can be added easily.