[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: AE GS-RAM Plus compatibility?
Hey Guys.
>> - Revision-D board, manufactured mid-1989
>> - Has "GSPLD1A-Q" and "GSPLD2A-Q" GALs
>> - I fully populated it to 6 MB (using 1MBx1 DRAMs pulled from a 386)
>> Interestingly enough, despite the mention of these GALs being
>> only compatible with the ROM 01 firmware, the card appears to
>> function just _fine_ in my ROM 3. Is there some specific glitch
>> it has on the ROM 3?
Well, I wouldn't say it will work perfectly. I found after a little testing
that the card handled the IIgs 256k (ROM0/1) and 1024k (ROM3) a little
differently. Mainly I found that it would tend to not track the memory
correctly and tend to write over things from time to time. Although using a
RAM tester like BitBanger would test and pass the memory as if there isn't a
problem since it doesn't track the entire bank of RAM but rather tests a
small area at a time. If it were to track then entire bank then it would
realize that some of the memory is overlapping and not being reported
correctly.
>> I ran the AE diagnostic from an
>> old AW 2 Expander disk I had from years ago, and Bit Banger, yet
>> neither detected any errors.
The main thing in BitBanger is it won't report any starting or ending
numbers (in hex) for memory if there is an issue. When you use the correct
GAL in the GS-RAM Plus BitBanger will report the correct RAM addresses.
Looking at my notes, a ROM0/1 will report 02-61 (full 6 megs) and on a ROM3
it will report 10-6F. A ROM0/1 has 256k hence it starts testing at address
02 (hex) and will go to 61 (hex). 00-02 address would be 256k of onboard
RAM. On a ROM3 it starts at 10 (hex) since it has 1MB of onboard RAM.
00-0F is 1MB.
If you try and run BitBanger in a IIgs with the wrong GAL installed on the
GS-RAM Plus then BitBanger will not report any memory addresses of what it
sees or is testing. This is the best way how to identify if there is a
problem. If you run BitBanger and it doesn't report a testing address then
your RAM card won't work correctly, guaranteed.
>> I'm curious about DMA compatibility.
The GS-RAM Plus has passed all DMA tests that I could think of and the
testers of the ROM3 GAL that I have sent them for testing as well. To this
date I've only found one version of the ROM3 GAL. I would tend to think
there would only be one version too.
>> How can I be sure the DRAMs follow CAS before RAS addressing?
To be sure of the DRAMs you would need to look up their refresh schemes on a
data sheet. Most, if not all, 1MB x 1bit DRAMs do allow for CAS before RAS
refresh - aka - CBR refresh. Although the GS-RAM Plus, and all memory cards
I've examined so far for the Apple II's, don't seem to use a 'true' CBR
refresh, or at least not a really cleanly designed CBR scheme. I'm still
researching and studding this actually.
> Another issue that I found when testing both these cards is the delay
> timing chip, namely the DS1000 chip. This chip is on the regular gs-
> ram card, but not in the gs-ram Plus card as the plus came out later
> with better design. This delay chip in the gs-ram card would have to
> be changed in order for the card to operate properly. Usually it's a
> simple change of changing it from a ds1000-100 to a ds1000-125 or to a
> ds1000-150.
I think I wrote about this type of issue before when I was testing GS-RAM
cards. Not 100% sure though. I do know that the GS-RAM seems to have been
designed more for the ROM0/1 and not the ROM3. Some will work in a ROM3 but
there mostly seems to be issue with them. The timing delay chip (DS1000)
would most likely be the way to resolve this, besides the correct GAL
upgrade (if applicable). Of course those with a ROM3 would probably not use
a GS-RAM card and want something bigger like 4 Megs. When the GS-RAM cards
work, they work well. I'd recommend keeping them on ROM1s though.
If there's any specific issue you're having Mitchell feel free to contact me
directly.
--
Henry S. Courbis
www.GSE-Reactive.com
Legacy Hardware and Support - Come take a look at what we have to offer!