Mac IIci Success Story

Kay K.M.Mods

Active Tinkerer
Sep 23, 2021
295
647
93
Tokyo
www.kerosmm.com
I don’t know much about the 475, but @Kay K.M.Mods created the Spicy O’Clock and he lists the 475 as an overclocking candidate.
This is my LC475 running 50MHz, no issue...
IMG_9708.jpegIMG_9707.jpegIMG_0097.jpeg

Carrera040 and Turbo040 can also be overclocked. Theoretically Booster40 should be possible. This is my SE/30 *The screen is green because I stole the tube from my IIc's broken monitor and installed it.
In the case of an accelerator, the timing limit is early. The limit is around 45MHz😅
IMG_1194.jpegIMG_1195.jpeg
 
Last edited:

Kai Robinson

TinkerDifferent Board President 2023
Staff member
Founder
Sep 2, 2021
1,153
1
1,162
113
42
Worthing, UK
3rd party Cache cards were a thing - it would be more than possible to design a new one - SRAM chips are now pretty cheap - it's just a matter of designing the controller, presumably for a CPLD like an ATF1508 / XC95144XL.
 

rikerjoe

Tinkerer
Oct 31, 2021
146
220
43
Out of curiosity @Kai Robinson i recall a guide you posted a while back showing memory chip equivalents, but I couldn’t find it. Could you share a link? I’m curious if it has the 24- and 28-pin static RAM chips that my cache cards have such as the CY7C185 and P4C116.
 

Garrett

Tinkerer
Oct 31, 2021
103
108
43
MN, USA
Yes, that’s it, Kai. Thank you. The list is proving handy in identifying equivalents for the 24- and 28-pin SRAM chips. I’ve started my troubleshooting of the dead cache card and hope to have more to share soon.
Did you ever have any luck with the cache card troubleshooting? I have one that behaves in the exact manner you had previously mentioned (boots to a black screen).
 

rikerjoe

Tinkerer
Oct 31, 2021
146
220
43
Did you ever have any luck with the cache card troubleshooting? I have one that behaves in the exact manner you had previously mentioned (boots to a black screen).
Not yet, Garrett. I wondered if the MT-branded 24-pin TAG SRAM on my bad cache card might be the problem since MT-branded RAM is so often a failure point in 8-bit computing. I used the guide Kai shared to identify a replacement with the same pinout and installed them on my cache card. Unfortunately, no change. I stopped at that point for now. I may pick it up again down the road since I have a working card and can do a component swap between it and the bad cache card to determine the failure point. I may also try to reverse engineer a schematic because I can’t fine one and would like to see if this is something I can do, just for the heck of it.
 

Garrett

Tinkerer
Oct 31, 2021
103
108
43
MN, USA
Good to know! If I ever get around to working on my cache card I'll update the group.

Hard to have motivation to work on them when we're both running more "souped up" cards for the IIci. I got ahold of a 50MHz PowerCache (which your Booster 040 would absolutely blow out of the water)!
 
  • Like
Reactions: rikerjoe

speakers

Tinkerer
Nov 5, 2021
98
76
18
San Jose, CA
peak-weber.net
I too have a non-functional Iici cache card that I'm debugging. Same failure mode at boot. It's SRAM chips are:
  • CY7C128A-20VC 2kx8 24-pin TAG SRAM
  • MT5C6408DJ-15 8kx8 28-pin SRAM
I also have a good cache card which uses:
  • MT5C1608DJ-20 2kx8 24-pin TAG SRAM
  • CY7C185A-25VC 8kx8 28-pin SRAM (note the slower 25ns access time)
So far, I've removed and tested the 3 Cypress TAG SRAM chips - and only one tests good. Utsource has used replacements which I've ordered.

As a note to testing these chips: my trusty TL866-II Plus is up to the task although doing so is challenging. Firstly, you need to use the Windows GUI Xgpro program. So, without a PC, that means firing up a VirtualBox VM under (Intel) Ventura and running Windows 10 with Xgpro installed. I usually can get away with the macOS command-line minipro utility to drive the TL866 -- but SRAM tests aren't currently supported. Secondly, the SOJ package is awkward. I couldn't find a reasonably-priced SOJ test socket - so I'm (literally) pressing into service a Proto Advantage SMT to DIP adapter. I clamp the SOJ onto the adapter plugged into the TL866 -- very Heath Robinson / Rube Goldberg!
 
  • Like
Reactions: rikerjoe

rikerjoe

Tinkerer
Oct 31, 2021
146
220
43
I’ll add my own SRAM chip list for comparison.

28-pin (4) CACHE24-pin (3) TAG
Card 1 (working)P4C164 - Performance SemiconductorP4C116 - Performance Semiconductor
Card 2 (non-working)CY7C185 - Cypress SemiconductorMT5C1608 - Micron Semiconductor

i wondered if the non-working card was due to failure of the MT-branded chips. Worth a try, I figured. I replaced with Renesas 6116SA20SOG from Mouser, for which I had to bend the legs since the chips were different packages. Pinout matched MT5C1608, and overall characteristics looked close enough. Unfortunately, this didn’t resolve the bad cache card and I set it aside for another day. As Garrett mentioned, since I have the Interware Booster card the motivation is lower, other than solving the riddle which I may try to do someday.
 

speakers

Tinkerer
Nov 5, 2021
98
76
18
San Jose, CA
peak-weber.net
So far, I've removed and tested the 3 Cypress TAG SRAM chips - and only one tests good. Utsource has used replacements which I've ordered
Three replacement (used) Cypress chips arrived from China. All tested good on the TL866, so I installed them and plugged in the card. And the machine booted!

But the cache isn't functional and, in fact, slows the machine a little. :(

Usual diagnostics (like Snooper) don't seem to recognize or test the cache card - so the I dug out a copy of MacTest Pro since this includes it. That reported a cache failure but gave no diagnostic details or clues. Ho hum.

Looks like I'll need to pull the CACHE SRAM chips and test them next.

But before I do that, I note from a Iici tech note that the cache card should map its SRAM areas (both cache and tag) in the range $5200000-$52ffffff and I wonder whether software can examine this. Sadly, MacsBug doesn't seem to see anything there (even with a working cache card installed).
 
  • Like
Reactions: rikerjoe

speakers

Tinkerer
Nov 5, 2021
98
76
18
San Jose, CA
peak-weber.net
But the cache isn't functional and, in fact, slows the machine a little. :(
Hah. On re-checking my work, I found some unconnected pins on TAG SRAM chips. My SMD soldering is flaky and I'd been so fixated with checking for bridged pins that I'd not checked for missed pins. With this fixed, the card works fine and passes its diag in MacText Pro. Victory!