• Please review our updated Terms and Rules here

PCB design simm module

R

Robin4

Veteran Member
Joined
Sep 25, 2011
Messages
521
Anybody knows where this pcb design came from ? and if it is open source?

s-l1600.jpg


Really like the design.. Only the capacity of the modules are to big.. Would prefer 4x 4MB 16MB kit.
 
Slaton actually seems to sell them on ebay. They might of actually rolled their own design on this but they don't totally answer al's question on how they are safely using 3.3v parts.

www.ebay.ca

64MB 4x16MB 30-pin Non-Parity SIMMs for Apple Macintosh SE/30, Quadra, PC & More | eBay

Find many great new & used options and get the best deals for 64MB 4x16MB 30-pin Non-Parity SIMMs for Apple Macintosh SE/30, Quadra, PC & More at the best online prices at eBay! Free shipping for many products!
www.ebay.ca
These are newly manufactured 16mb 30-pin SIMMs for vintage high-end PCs and Apple systems and some synthesizers and other vintage devices. Lovingly hand-assembled and individually tested.

Unlike some of the other "new" modules you'll find online, these are built on a high quality 4-layer PCB that was designed based on extensive study of period systems and existing modules. These don't suffer from issues with poor contact due to board thickness or bad design -- many customers have now reported back to say our modules have worked in systems where other modern replacements haven't. We're using newer memory chips with an on-board voltage regulator and a custom design to ensure reduced bus loading and life-long performance that is guaranteed to meet or exceed original vintage parts.
Click to expand...
 
Al Kossow said:
The LDO regulator on there has me wondering if they are using 3.3v rams in a 5v system w/o level translators
Not good.
Click to expand...

Non stable when using LDO regulator? Can image that the ram ics needs constant 5volt?
 
Robin4 said:
Non stable when using LDO regulator? Can image that the ram ics needs constant 5volt?
Click to expand...

The problem with these memory modules is the LDO even being present. When FPM/EDO DRAM was introduced, it used TTL 5 volt logic levels for the Vcc and the bus. Later in the FPM/EDO lifecycle, manufacturers started to introduce 3.3 volt chips. This is a problem because these can't be used on older 5 volt memory buses, unless they're made 5 volt tolerant, or level shifters are used.

This memory module uses a LDO for what we think might be just for Vcc, but no level shifters for the memory bus to drop the 5 volt signals down to 3.3 volts. So the memory chips would be required to be 5 volt tolerant, if they aren't, there's the risk of the chips being damaged. They could work fine for some period of time, and eventually degrade to the point they stop working. Or they could burn out quickly and stop working.
 
The only way to know is peeling off that label and seeing what part they're using.
Every 3.3v Hitachi FPM part I could find was not 5v tolerant.
It has also been pointed out that 16mb simm support was not common.
For all the mucking about, you would probably just be better off with some used tested SIMMs
The other problem is the thickness of the pcb material for SIMMs isn't commonly offered, which
is why there are so many problems with things like Macintosh ROM SIMM replacements.
 
Actually, I have a surprising number of PC motherboards that don't require parity memory. It's so odd, since I don't remember having any devices in the 80s/90s that didn't require parity.

However, it may just be that some boards claim to require parity but don't bother to use it.

- Alex
 
Oh, the SIMM world could get nuts in the PC arena. 9 chip-vs-3 chip; high-density, etc. Matched pairs in were often required. There were some 3-chip cheapies that actually generated the parity on reads rather than incorporate another memory chip...
 
Oh definitely. My memory checker (Innoventions, I have both SIMM and DIMM varieties) reports those as “emulated parity.”

- Alex
 
How can software detect whether true parity is available (compared to "emulated" parity)?
 
The need for parity was kind of overblown though, and if you could turn the check off in the BIOS (or by jumper) then it was no real loss. How many people have ever even seen the parity error screen? PC's of that era had no way to recover from a parity error so they just dropped into a 40-col text mode, displayed an error, and locked up.
 
The danger, of course, of a undetected memory error is the propagation of the error throughout the system. Like having an undiagnosed contagious disease--you really want to stop it in its tracks.
Gee, if you hate parity, you must really despise ECC. For that matter, why bother with error-detecting codes on your hard disk?
 
FishFinger said:
The need for parity was kind of overblown though, and if you could turn the check off in the BIOS (or by jumper) then it was no real loss. How many people have ever even seen the parity error screen? PC's of that era had no way to recover from a parity error so they just dropped into a 40-col text mode, displayed an error, and locked up.
Click to expand...
I saw a parity error on a Sparc 10 workstation. I had just installed some memory and I didn't notice a piece of hair that lodge between the DIMM and the socket. System displayed the message and continued to boot. IMO parity is a good thing.
 
You must log in or register to reply here.
Back
Top