VOGONS


First post, by ALEKS

User metadata
Rank Newbie
Rank
Newbie

Hi,

Lately I built some SIMM memories for personal use. I am using them with 80386 machines.
I know there are other newly made SIMMs. But I would just like to also share my project with you.
Unfortunately I wasn't paying attention that SIMM sockets expect 1.2 mm PCB thickness. Mine are 1.6 mm and they are stressing the SIMM sockets. Even though these PCBs work, I will have to redo these PCBs in the future.
The PCB layout is entirely hand routed on 4 layers. Top and bottom layers are signals and inner top and inner bottom are power planes.

I tested these modules with Memtest86+ and everything works as expected.
I now have enough RAM on my 386 computer. 4 of them are 16 Mb and 8 of these things have a whopping 32 Mb.

4mx9-simm-pcba1[1].jpg
Filename
4mx9-simm-pcba1[1].jpg
File size
243.91 KiB
Views
337 views
File license
GPL-2.0-or-later
4mx9-simm-pcba3[1].jpg
Filename
4mx9-simm-pcba3[1].jpg
File size
220.86 KiB
Views
337 views
File license
GPL-2.0-or-later
4mx9-simm-pcba5[1].jpg
Filename
4mx9-simm-pcba5[1].jpg
File size
63.06 KiB
Views
337 views
File license
GPL-2.0-or-later
dram-capacity1[1].jpg
Filename
dram-capacity1[1].jpg
File size
81.46 KiB
Views
337 views
File license
GPL-2.0-or-later

Cheers,
A.

Intel 80386DX / 33 MHz | 32 Mb RAM | Tseng Labs ET4000/W32i with 2 Mb RAM | ISA I/O Interface | ISA Audio Interface | 3.5" & 5.25" FDD | 4 x CF 512 Mb | Intel EtherExpress 16

Reply 4 of 10, by ALEKS

User metadata
Rank Newbie
Rank
Newbie

Thank you!
I never thought about designing a 16 Mb module. It shouldn't be that hard. But it will be of a different size. A little bit taller. Although I know there are some 16Mx4 IC of reasonable size. With a 2 IC per PCB, they would look very compact.
However I don't have any piece of hardware that could use 16 Mb SIMMs.

Intel 80386DX / 33 MHz | 32 Mb RAM | Tseng Labs ET4000/W32i with 2 Mb RAM | ISA I/O Interface | ISA Audio Interface | 3.5" & 5.25" FDD | 4 x CF 512 Mb | Intel EtherExpress 16

Reply 5 of 10, by Tiido

User metadata
Rank Oldbie
Rank
Oldbie

I made 4MB ones not so long ago. 16MB ones are not gonna any more difficult than 4MB ones, except locating the chips since they aren't so easy to find, or at least not in good cost anyway...

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 6 of 10, by yuhong

User metadata
Rank Newbie
Rank
Newbie
ALEKS wrote on 2020-09-29, 15:45:
Hi, […]
Show full quote

Hi,

Lately I built some SIMM memories for personal use. I am using them with 80386 machines.
I know there are other newly made SIMMs. But I would just like to also share my project with you.
Unfortunately I wasn't paying attention that SIMM sockets expect 1.2 mm PCB thickness. Mine are 1.6 mm and they are stressing the SIMM sockets. Even though these PCBs work, I will have to redo these PCBs in the future.
The PCB layout is entirely hand routed on 4 layers. Top and bottom layers are signals and inner top and inner bottom are power planes.

I tested these modules with Memtest86+ and everything works as expected.
I now have enough RAM on my 386 computer. 4 of them are 16 Mb and 8 of these things have a whopping 32 Mb.

4mx9-simm-pcba1[1].jpg

4mx9-simm-pcba3[1].jpg

4mx9-simm-pcba5[1].jpg

dram-capacity1[1].jpg

Cheers,
A.

Also be careful of gold vs tin connectors as well.

Reply 7 of 10, by Warlord

User metadata
Rank Oldbie
Rank
Oldbie

I heard the gold tin thing as well,and its common knowledge based on science. However to this day no one that I have seen has ever been able to demonstrate connector corrosion even after all of these years. Despite it being in books like A+I've never seen a real world example of it. It's not just ram slots, plenty of motherboards have tin leads in their PCI connectors, yet never seen corrosion from having PCI cards with gold leads in there for years.

Reply 8 of 10, by snufkin

User metadata
Rank Member
Rank
Member

Huh, I'd never really thought about the possible corrosion. Just had a read through of this:
https://www.jpcfrance.eu/technical-informatio … ion-resistance/

It's mostly looking at aluminium corrosion, but has a good table of electrochemical compatibility (interesting that gold + copper isn't green, but I've never seen gold plating on PCB traces be a problem). It also gives a couple of methods of reducing galvanic corrosion, one of which is:

Avoid relatively small areas of the less noble metal (Aluminum) and large areas of the more noble metal (Stainless steel).

Substitute tin for aluminium and gold for stainless steel and I think that's what would happen with a card with tin plated contacts plugged in to a board with gold plated contact wires. The contact wire is close to a point contact on to a relatively large plane of tin.

Also, the table is with the metals in a 2% salt water solution. So maybe it becomes a problem for computers on boats. Or when there's a good electrolyte present, like when capacitors leak all over the place.

Looks like some people from a company in Singapore did a test for tin-gold corrosion in connectors and didn't find anything:
https://www.digipas.com/documents/Effects-of- … ed-Contacts.pdf
That's tin plated flat flexible cables inserted in to gold plated connectors.

Reply 10 of 10, by weedeewee

User metadata
Rank Oldbie
Rank
Oldbie
majestyk wrote on 2021-04-11, 16:02:

I´m still wondering why the smart guys at "PC-Chips" gold-plated the retention clips of the SIMM slots while using crappy electronic components and fake caches.

for the 'bling' factor ! 😁