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

[Trash80toHP_Mini]

In another thread, trag mentioned that his SuperMac C600 had Slots A, B AND C along with CS2. Another tangent ensued because the SuperMac clone is based on the 6400's Alchemy architecture.

Long story short, got this riser in hopes that it would plug into Slot A and give me A, B, and the mythical C.

View attachment 24070

No joy there Slot mapping appears not to work that way.



Plugging it into Slot A of the 6500 Mobo I get Slot A in connector PC11 which ought to be Slot B per my reading of the tables. Guess I have to find a board to test it in Slot D?

Black magic on the card.





Genesis of the tangent:





Gotta organize the Slot info research I did in that thread to post here. It appears that the swaperoonies on this PCB align with the pinout info I found. If anyone has a suggestion about how the table on the riser works, please let me know. Dunno if my interpretation is at all accurate. :mellow:

 

[Trash80toHP_Mini]

@trag I'm already thinking about just in case workaround possibilities, your C600 wouldn't have crown jewels on a card in the ROM DIMM slot perchance?

View attachment 24056

This TattleTech report has me thinking that there is support for Slot C1 in the 6400, Apple appears not to have hobbled it in ROM, trusting the 6400 setup to do that a/o just not giving a schiznit about expansion capability one way or another. Comparing ROM dumps between 6400 and C600 would be good to do up front. If Slot C1 doesn't show up after much blood, sweat and tears are shed on the bench it would definitely be time to compare dumps and try a C600 ROM card hack in a freshly installed DIMM socket in the 64oo.

Links from other thread:

Apple Developer Connection: PCI Card Services


Designing PCI Cards and Drivers for Power Macintosh Computers 1996



Revised Edition 3/26/99 has corrections, additions and irrelevant New World complications.



Designing PCI Cards and Drivers for Power Macintosh Computers




http://www.kennedybrandt.com/supermac_insider/support/downloads/C600UserGuide.zip

 

[trag]

@trag I'm already thinking about just in case workaround possibilities, your C600 wouldn't have crown jewels on a card in the ROM DIMM slot perchance?

I'm not sure what the above means.   Can you be more specific.

 

[Trash80toHP_Mini]

Sorry, are  your C600 ROMs on a DIMM or on the motherboard? Idly wondering about last ditch effort in testing for Slot C function in Apple's implementation of Alchemy. If all else fails. Testing my (standard cheapo type) riser that works in the BG3 with its jumper card setup that's been modified to work without those connections to other mobo slots might need to be tried with the C600 ROM in the 6360 it won't work with its stock ROM.

Probably not a likely scenario however:

1 - TattleTech's PCI Slot report "sees" and lists (unimplemented) Slot C as being in the hardware of the my 6500 board.

-  -  possibly polling the bridge IC to find it supporting three slots?

2 - Apple's use license of the "crown jewels" almost surely forbade a clone maker's fiddling with ROM?

-  -  Black Box documentation of ROM with revocation of license for disassembly violation clause?

Step one of the process as I now see it unfolding will be to duplicate Apple's twin slot riser on the "simple three slot rackmount riser card." That may be done as simply as patching line values from Slot A pins to the "nubbin daughter card riser" header inputs to mimic the Apple riser. Just doing that completes reverse engineering the Slot A to B transition for the two slot TAM riser PCB design along with the low rider two slot riser for the 6360/5400/5500 drawer.

Step two will be wire wrapping connections on the headers for the nubbin daughtercard Slot C from Slot B headers to mimic the (nubbin jumpers not present) much more complicated ADVANTEC riser's Slot ID +1 function.

edit: I've probably gotta post pics of the "standard" rackmount riser/nubbin card jumper setup to make the jibberish above comprehensible. :-/

 

[Trash80toHP_Mini]

Well, that was a bust. Took a few hours to find all the bits needed to get the 9500 board up and running on the bench to no avail. For whatever reason the fourth slot which should be Slot D1 shows up as Slot D2. Never seen nor heard of such a thing? First test of the ADVANTECH board didn't work in what the Tables on the card should have been Slot C. Didn't really expect it to work. I wonder what that ID designation comes up as in the QS? Another time perhaps.

Buzzed the nubbin board header connectors on the "standard" three slot riser to the soldertail pins of its PCI connector last night. Gotta buzz those pins the 6400/6500 risers and try to determine what's up with that next.

 

[Trash80toHP_Mini]

Can't figure this table out at all from the results I got for the 9500.

View attachment 24290

Bottom slot works wherever I try it, but only at the host slot's ID and the top two slots are just not there. It even passes on the curious oddball D2, E2, F2 Slot ID. Looks like a pair of standard PCI bridge ICs from TI on the board. Haven't checked yet, but a bus 1 and bus 2 suffix makes sense. Does or did the PC world do that kind of thing?

The ADVANTECH board is now relegated to $35 curiosity status, oh well. Gotta put one of the old PC boards on the bensh some day to play with it.

Looking at the 6400/6500 riser from the jumper cards of the "standard" three slot riser ATM.

 

[trag]

Hey, jt.  I was traveling last week.  I'll try to pull some C600 info this weekend.  There's one in the garage that hasn't reached the attic yet, so it's possible I'll remember and get to it.

 

[trag]

I pulled the riser card from the C600 this weekend.  Here's a photo:



There is no ROM or any other logic on board.  It's a dumb board.   The connector on the right is to the floppy drive.  The white tape in the top middle is covering the COMMII Slot.  I don't know why.   The connector at the bottom has 194 pins, 97 on each side.    The pins aren't numbered in silk screen, so for my purposes, I numbered them from left to right on the front.   On the back, I did not flip the card over.   Pretend you're looking through the board, oriented as above, and the back pins are left to right, 98 - 194.     So pin 98 is opposite pin 1.  Pin 194 is opposite pin 97, in my system.

Pins run:  1 - 42, 43 - 87, 88 - 97;    98 - 139, 140 - 184, 185 - 194

where commas are the slot notches.

I found the following:

Pin 114         Slot C interrupt

Pin 113         Slot A interrupt

Pin 17           Slot B interrupt

Pin 76           Slot A REQ

Pin 77           Slot B REQ

Pin 79           Slot C REQ

Pin 174         Slot A GNT

Pin 175         Slot B GNT

Pin 177         Slot C GNT

Pin 153         Slot A IDSEL, ADD 13

Pin 55           Slot B IDSEL, ADD 14

Pin 151         Slot C IDSEL, ADD 15

I did not have a pinout for the comm slot handy, although I think it's posted around here somewhere.   However, the resources list above occupy sequential pins on the edge connector.   This isn't obvious when they are 113, 17, 114, but 17 is opposite 114.   And the gaps in sequences like 76, 77, 79 are because of intervening gnd or power pins.

Anyway for the first three above, I found a pin on the comm slot which connected to what ought to be the next pin in the sequence of interrupts, GNTs or REQs.  I didn't check IDSEL.   So I strongly suspect it's an independent PCI slot.    Umax did some unnatural stuff on the S900, where the sold a combo U2 LVD SCSI card with 10/100 ethernet on board.   It only worked in slot 1 of the S900 and there was an extra little connector at the end of the PCI slot to provide extra signals to support the ethernet portion.  Then there was a hacked to the gills driver that added an OF script when made the ethernet portion appear as slot E, which is some kind of special legacy slot in the firmware or some such.  

But it wouldn't make any sense for the C600 to use a similar system because anybody's card might be installed in the comm slot.  The E100 card, as it was called, was the only card that could use that extra connector in the S900.

Do the PM6400 and PM6500 use the same 194 pin connector for their riser cards?

 

[Trash80toHP_Mini]

OK, NOW we're cooking! Thanks, trag. I'm back from my sister's wedding after 12hrs of driving followed by 12 hrs of sleep. WONDERFUL bit of work you did while I was gone. If I wake up again any time soon (coffee's not doing the trick) I'll take a serious look at your data.

IIRC, I confirmed that a standard 32 bit PCI card works as if in Slot A1 on the Riser when directly inserted in the 6400/6500 logic board's riser slot. Setting the Digital STARION Alchemy/Gazelle testbed back up on the bench to verify that might be a good bit of sleepyhead RetroPlay for tonight.

Not sure I'd recommend trying that in your C600 board's slot even if you could though. Is the pincount for the initial section of the card and the following portion of the center section identical to Slot A1 on the riser? Looks like it's keyed very differently.

Deciphering your numerology tonight ain't gonna happen. If you buzzed it, verified it and tried it, I'd bet it'd work on a simple key correcting adapter though. No reason to buck the PCI  convention and every reason to stick to it for basic troubleshooting with that simple adapter and PCI card at that level. Very curious about that extended center portion and the CS-2 connections. Table time  .  .  .  not tonight.

 

[Trash80toHP_Mini]

Verified the Riser Slot ID/function as Slot A1 and that the CS-2 NIC is a PCI Card, discrete or a  "hardware enabled" PseudoSlot implemented at Slot ID E1.



CS2 - specifically the PCI Bus Communications Slot 4 pinout is in this post:





Pin 95 is labeled INT~ which has me thinking it's the Slot E interrupt. Which again has me wondering about the possibilities of a four slot riser?

Gotta look at the PCI bridge setup, we've got A1, B1, possibly C1 of the C600, CS2 at E1 and the ATI ChipSet at F1 on the logic board.

I have a sneaking suspicion that an interrupt for Slot D1 may have gone missing somewhere? [}]

 

[Trash80toHP_Mini]

6500 Block Diagram



Looks like A, B, and E run off the PSX+ Bus Bridge ASIC in the 6500.

6360/6400/5400 have the 40MHz PSX Bus Bridge ASIC, I'm wondering if torquing PSX up to 50MHz as PSX+ level is the root of Gazelle's PCI evils?



Is PSX the controller in your C600, trag? I can't imagine it's not a standard, licensed Apple part?

 

[Trash80toHP_Mini]

Noticed a couple of things:

oddity: There's no indication of the simple Video Out board for the 5400 or the far more complex board for the 6500 Video Out/Subwoofer.

Amazingly: The soldered RAM does the opposite of Apple's customary hobbling of a memory bank. In Alchemy, it's apparently configured as a third DIMM, giving it a higher Memory Ceiling than the high performance Gazelle Architecture!

 

[Trash80toHP_Mini]

@trag is there any chance you could do a high res scan of your C600 riser and maybe take a higher res photo of the component side that shows all the traces clearly? I'm a big fan of taking PCB pics in the shade on a "Kodak Sky" day. That's a bright day with a just enough even haze to act as a diffusor balloon for the sun.

 

[Trash80toHP_Mini]

I've been playing with cards in the 6500 bench setup.

Adaptec AHA-2930CU Mac is an Apple (Fast/Narrow?) SCSI-2 part that works a treat with one of my standard SCSI drives. It'd be just about perfect coupled with the SCSI2SD Model 6. Too bad that won't fit in a TAM, but fab for the other Gazelle machines. I'll need to use set up one of my SCA Savvios to mess around with as a boot disk.

Sonnet Tempo ATA133 shows up in ASP and TattleTech as does the HDD, but the drive is listed as unmounted. 6400 testing is up next.

Is there an IDE utility the likes of of SCSI probe that might mount the disk from a Gazelle board? :huh:

 

[trag]

@trag is there any chance you could do a high res scan of your C600 riser and maybe take a higher res photo of the component side that shows all the traces clearly? I'm a big fan of taking PCB pics in the shade on a "Kodak Sky" day. That's a bright day with a just enough even haze to act as a diffusor balloon for the sun.

I don't have a working scanner these days.    The glass fell out of my old Umax mumble 600S.   I tried to fix it with 3M 2-sided tape, as that is what it was originally held in with, but that failed.  I should have just epoxied it back in place, but I threw it out before that thought occurred to me.

 

[Trash80toHP_Mini]

Heh! I didn't either until recently. My HP AIO Ink Jet won't scan because the ink cartridges are out of date. : LEXMARK Pro915 from the grlf scans, but is VERY annoying dire warnings that it will damage itself if I don't buy $100 worth of cartridges and won't go to sleep until I do  .  .  .  so I turn off the power strip between scans.

Until very recently I was using the nice AIO Laser Printer/Copier/Scanner/FAX machine at work that emails nice .PDFs to me at home. Got one of those at work? []

I've got a MICROTEK e300 from the mid-nineties to test and a couple of their ScanMaker4 8X10 transparency capable scanners to recap and build one good working unit.

One day! :lol:

 

[Trash80toHP_Mini]

YAY! I found my packed to the gills 6360 board while I was putting things away and looking for the Quadra 630 to try out in the BenchMac test setup. Hooked up the original HDD and the L2 Crescendo/400/1M fired right up!  []

One of the pair of 400/512s from Sonnet's clearance still tested DOA in the tested good 6360 rig though. Whatever. Hoping to find that &(*&$%^%@^ Quadra 630 board for one last  playtime session before I get down to the business of borking something in this project. That'll definitely be one of the Gazelle boards, I've only got the one 6360 and oneof the  6400 Alchemy boards. I need those for the two TEMPOtrio IDE/USB/FW combo boards.

Next to get the QS up and running on the rolling standing desk to start noodling out the slot data in a format I can see. Can't grok tables. :blink:

 

[Trash80toHP_Mini]

The first difference between Slots A and B on the 6400 riser is plain as day on the solder side.



REQ for Slot B is supplied from the logic board from a pin that's RESERVED in the PCI Specification. Now it's time to meter more oddities.

View attachment SlotSearch-001.PDF

 

[Trash80toHP_Mini]

I found fifteen lines not connected between A $ B on the 6400 Riser. Five of them have traces visible (light green) on the solder side. Strange! One test board with headers on hose pins coming right up. [}]



View attachment LittleRedRiser.000.PDF