Jump to content
Sign in to follow this  

TWO SLOT IIsi Nubus Card hack . . . Take 2

Recommended Posts

For those who might be curious, this insanity started back in March of 2011  .  .  .


TWO SLOT IIsi Nubus Card hack . . . (SE/30 hacks)


.  .  .  and I'm just recently getting back into it. ::)


Why cram two NuBus slots into a IIsi, you ask? Just because I think I can and it would be ever so sweet to have a really good VidCard sitting underneath the Radius Rocket 33/SCSI II DaughterCard combo up in the top slot of my SuperIIsiHack™


Suffice it to say the first go round was the expected failure, but important data was produced and I managed to keep the magic smoke contained wherever it was that it belonged!


This is the DuoDock's Twin Slot "reversed riser" card that drops the Duo's NuBus Slots down and tucks them underneath the Dock's Logic Board. I'd been wondering about using what's apparently an almost passive NuBus backplane for some sort of hackage for quite some time.




Back to that Data:



There are three possible states when the hacked Gemini Card is inserted into the IIsi NuBus Adapter:

A - Radius Card installed in top slot - Unserviceable Slot Interrupt
B - No NuBus Card installed - boots normally
C - Radius Card installed in bottom slot - boots normally


A couple of interesting things here:


First off, the the lower Slot ID is coded as $D, which doesn't exist in the IIsi's hamstrung little world of mid-level Macdom: Hence, no problem booting up with the Radius VidCard installed in it, The IIsi's ROM doesn't give the Slot Manager a single clue that Slot $D exists in the upper echelons of the Macintosh pecking order.


The Top Slot ID is encoded as $C coming from the Gemini Card end of things and encoding signals for the IIsi's NuBus adapter are yelling at the NuBus Card that it's installed in Slot $9 at the same time!


In this situation the pins for slot ID signals for /ID0  and /ID2 are being pulled both high and low at the same time. NOT GOOD! Hence the Unserviceable Slot Interrupt error message bleating out as the CPU throws its hands up in disgust.


Here's what's left of the Gemini Card after I lopped off the top section and its connections to the bass ackwards and upside down interboard connection to the Dock's Logic Board.




Here's the other side, showing a top view of the "Machine Pin Hack" inflicted upon the poor NuBus connector's soldertail cones:




Here's an oblique view of this (quite effective) brutally inflicted mess, the remaining pins anyway:




This shows the machine pin interface and the standoff resting comfortably upon the FPU.




This shows layers exposed at the cut line during connector reduction surgery:




Top view of abbreviated Gemini Card installed in IIsi NuBus Adapter.






That's enough for re-documenting photos from Take 1 of this hack that were lost in the great pixel robbery, so now down to the current iteration of the project. [:)]]'>


An oblique pic of what I described at the end of the NuBus Slot ID Assignment Encoding thread:




I'm using Wire Wrap Connectors for the final version for reliability, adding Machine Pins is good for down and dirty feasibility testing, but not something I want for any kind of general use. The above was my first experiment with the process using some 2x32 Wire Wrap connectors I got by mistake: Gotta try again when I'm not so friggin' tired! Bent one pin and broke it off trying to bend it back. ::)


Here's a shot of the unbroken pin sticking straight out from the side of the NuBus connector, clearing the Gemini PCB  (use your imagination!) without interfering with mounting a NuBus Card. Doing this decouples all four Slot ID signals: /ID0 through /ID3 of both slots from the IIsi side as well as the Gemini Card side. This allows me to snatch high and low signals right off the IIsi Adapter with jumper wires for properly encoding the Gemini Card's slots as IIsi friendly Slots $9 and $A.




Now that you've seen the pics, maybe my diagrams will make things a little less murky  .  .  .  or not.




The down and dirty illustration of this Nastyhack shows the tapping of high and low signals fromthe empty connections (remember those bent pins!) and their use for encoding the Gemini Cards Slot IDs:




Confused yet? :lol:





Edited by Trash80toHP_Mini

Share this post

Link to post
Share on other sites

THX, talking to myself here gets boring sometimes. Distraction and procrastination rule my world,  that's where cyclic tenacity (borderline obsession at best) becomes important.


Meanwhile, back at the banana ranch:


These thin rectangular section wire wrap pins are so brittle they're driving me up the wall! Which makes a better insulator, hot glue or epoxy? If the new connectors come in without solid square section header pins I'll need to solder jumper wires to subsurface nubbins and glue them in place.


This might lead to a header based  Slot ID "breadboard" I can stick into any NuBus for /IDx playtime  .  .  .


.  .  .  resistance is futile. [:O]]'>

Edited by Trash80toHP_Mini

Share this post

Link to post
Share on other sites

May be of interest:


Wire Wrap Is alive And Well!


Connector in wait, need to search more for a better, less expensive candidate.


Prototype of IBM PC - 1981:






Note the bypass caps across power and ground legs of components.


No SEED PCBs or hobby level PCB Design Apps back in the day. I've never done anything complex in wire wrap. My process started with doing the hand routing and then the graphics work for silk screen printing two layer board resists on plated copper FRP blanks. Etching and then more than a bit of tedious drill press work on holes for ICs (or sockets for them) and connectors followed that.


Thankfully, I then handed the boards off to my partner for soldering, a skill I've studiously avoided until fairly recent procurement of inexpensive Chinese clones of a real (commercial grade) Hot Air Rework/Soldering Station and a vacuum pump de-soldering gun.


Playtime! [:D]]'>







Share this post

Link to post
Share on other sites

I had just been reading your other thread on this.  Nice to see you've got it completed.


OOPSIE! missed that part. It's not completed, just moving along very nicely now that I've (very likely) sorted out the Slot ID addressing issues.


update: Haven't had time to play with the de-soldering gun yet, but I've worked out some more of the physical/spatial issues using the machine pin modified card. Decided to start fresh with another of my spare Gemini PWB cards and keep the old one as a prototyping test bed, so I'm glad I procrastinated a bit.


Found a lead on a multiple slot NuBus expander board's extender/jumper board:




Creating a very simple version of that board will make things a lot less messy when/if this project leads to the SE/30/NuBus hack. I've got 8" jumper cables that do the same thing, but this will make the interconnect much more reliable over its much abbreviated length. It'll be tedious, but I can prototype it on cladded pad perf  board in wire wrap using standard headers. Alternately I can make the connections solid by soldering them as I tack the headers to the pads on the prototyping board. Wondering about doing twisted pair signal/ground connections to attenuate crosstalk. Looking very promising! [:)]]'>

Share this post

Link to post
Share on other sites

update: The Good, the Bad and the not quite as Ugly:


The Bad: found an error in my encoding, but it was an simple flip flop of the Slot ID signals for proposed Slot $A.

The Good: put a worry to rest, figuring a simple header/jumper block that'll permit playtime with Slots $C and $D in the IIcx.

The not quite as Ugly: realized I can tap high and low signals from the IIsi's NuBus Expansion Slot Adapter by direct connections to /ID1 and /ID3 which alleviates jumpers from IIsi NuBus Adapter to TwinSlot Adapter shown in the diagrams above. KISS in retrospect!


The encoding case that  .  .  .


/ID1 = HIGH = $9 - $C - $D

/ID3 = LOW  = $9 - $C - $D


.  .  .  .  lets me set jumpers to $C or $D for plugging my adapter into IIx Slot $9 just to see if the ROM derived from the IIx supports Slots $C and $D. These are 2 of 3 physical slots lopped off the IIx board for its abbreviated case conversion. That's just for shiggles and gits.

Edited by Trash80toHP_Mini

Share this post

Link to post
Share on other sites

Last update: Time to start the hardware hacking! Trimming the next Gemini PWB victim and de-soldering its connectors is next on the menu. Then I'll order up some EuroDin 96 Wire Wrap connectors  .  .  .  maybe sooner.




Worked out a jumper block board. The blue jumpers are set to $9 & $A for the IIsi (and maybe someday the SE/30) in the diagram. At some point I'd like to play around in the IIcx by setting the lower Slot ID bits to $C or $D just for fun. [}:)]]'>

Edited by Trash80toHP_Mini

Share this post

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this