• Hello MLAers! We've re-enabled auto-approval for accounts. If you are still waiting on account approval, please check this thread for more information.

Three Slot Riser for 6400 - In search of the Mythical Slot C

@Daniël do the "PCI Reserved" traces on your riser make it a the way through Slot_A1 to Slot_B1 and Slot_C1?

Yes, if you look at the board, you can see the reserved pins I soldered to have traces going to them from all three slots.
Just as an example, I have quickly made a picture illustrating the traces for A9, A11 and A14, attached to this post (you have to open it specifically for some reason).
 

Attachments

Interesting, never thought to look at that as I've been target fixated on the main line TwinSlot Riser development threads. My board does not have passthru of the reserved lines from Slot_A1 so I'll be terminating the six signals there just as I've been thinking of for PCB design.

Thanks to your rewiring of a standard PCI riser approach I can dispense with any notion of a need for a custom TriSlot riser. Anyone interested in this brand of craziness will be doing a recasing of the Alchemy board and so will be likely be able to do the rewiring easily. ;)
 
I decided to play around with this on my 5400 today, to verify both that the information in this thread is correct and that I understood it. I took the following steps to convert my single-slot riser from slot A to slot C:
  1. Cut trace between finger A26 and pin A26; bridge pin A44 to pin A26 with 22Ω (resistor probably not necessary, but I had a 20Ω on hand, so why not)
  2. Cut trace between finger A17 and pin A17; bridge pin A17 to pin A19
  3. Cut trace between finger B18 and pin B18; bridge pin B10 to pin B18
  4. Cut traces between fingers A6, A7, B7, and B8 and their corresponding pins; bridge pin A11 to pins A6, A7, B7, and B8 (probably only necessary to connect A7, but Apple did them all, so why not)
When I say finger, I mean the gold-plated finger on the edge connector, and when I say pin, I mean the through-hole pin on the PCI slot. Here's what the bottom looked like when I was done cutting and soldering:
IMG_20250202_172204.jpg

And here are the cuts I made on the top:
IMG_20250202_172310.jpg

Result:
SlotC.png

The trackball I have plugged into the USB card in the now slot C riser shows up and works. Very nice. Thanks Trash80 and Daniël for doing this research and documenting it.

To make sure I understand what's been figured out thus far, not all of the signals necessary to wire up a slot D are present on the riser slot, correct?
 
Very nice, welcome to what's now a gang of three! ;)

Correct, you need to rob signals from CSII go full-on PCI for Slot D. I used a cheap single slot RA riser card, removed the connector and hotwired from there to my cable monster test rig just for a screen shot op in TattleTech.

Don't see much use in a Slot D hack unless you either don't need a NIC or want to play with a PCI GigaBit NIC? Doubtful you'd see much improvement over 10/100 on your machine's Alchemy architecture or even a faster Gazelle board? New build 10/100 CSII cards are available at long last.

I've got three slots set up in my Mid-Tower 6x00 testbed, but haven't yet bothered with soldering the few wires needed to complete that adaptation. Procrastination is definitely a strong suit after establishing the feasibility of one of my crazy notions. :rolleyes:

The only Mac case capable of housing a TriSlot riser is the P6360 or like case donation from a lesser PPC in Quadra 630 clothing. You need to heavily rework the mezzanine support shelf for exhaust fan, Video out board and the TV/FM Tuner Module that has to be deep sixed. You'd be limited to fairly low profile cards in the top two slots for fan clearance, but it will definitely fit, given a custom three slot riser of course.

My TwinSlot Riser for P6360/5x00 has yet to be laid out, nor has the TAM Version been. Maybe someone will jump on that part of the project at some point?
 
Last edited:
Back
Top