-
Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.
ProtoCache1 - IIsi/SE/30 PowerCache Adapter Prototype Development
- Thread starter Trash80toHP_Mini
- Start date
Trash80toHP_Mini
NIGHT STALKER
Have you got the SuperMac adapter for using their full size SE/30 video card in the IIsi? I just buzzed mine for you.
On this dedicated, single purpose Video Card Adapter B1<->B1 in n.c. So your problem lies elsewhere. Bolle's Slot ID/interrupt conflict notion could explain it. Would you almost certainly get a bus error in POST, but I'd think you'd get death chimes for that machine state?
I'll get a dedicated RCPII/IIsi cable building thread started with references to posts in my hacks thread on the subject. That's one insanely convoluted thread. I finally managed to bork the cable from my SE/30 Pivot Card and now need to build one. :
Getting a cable set up with sense coding built into the harness for 16" resolution as well as the existing straight thru spec seems in order. Basing the design on readily available breadboard jumpers/soldertail HD-15 connectors or straight thru cables with the DA-15 connector would be the ways to go IMHO.
On this dedicated, single purpose Video Card Adapter B1<->B1 in n.c. So your problem lies elsewhere. Bolle's Slot ID/interrupt conflict notion could explain it. Would you almost certainly get a bus error in POST, but I'd think you'd get death chimes for that machine state?
I'll get a dedicated RCPII/IIsi cable building thread started with references to posts in my hacks thread on the subject. That's one insanely convoluted thread. I finally managed to bork the cable from my SE/30 Pivot Card and now need to build one. :
Getting a cable set up with sense coding built into the harness for 16" resolution as well as the existing straight thru spec seems in order. Basing the design on readily available breadboard jumpers/soldertail HD-15 connectors or straight thru cables with the DA-15 connector would be the ways to go IMHO.
Bolle
Well-known member
C40M should stand for Clock 40 MHz.
The signal is probably used only on the NuBus adapter in the IIsi. You were never going to stick a NuBus card into the MacCon so that's why they probably left it out.
I will measure the SE/30 board if this is even putting out a 40MHz signal. If not this should not be a problem unless we want to use NuBus (which I guess it is used for)
On the Pivot card passthrough it is also not connected. In fact it does not seem to be connected to anything on the pivot card. Same goes for the MacCon. I would say we do not need this one.
Last edited by a moderator:
Trash80toHP_Mini
NIGHT STALKER
Easy enough for an olde schoole patch wire fix! Iff'n a board ain't got no patches, it fails to meet the requirements to be listed in the time honored prototype board category, it's just a crappy lookin' pre-production test card. [] ]'>
Meanwhile, I've got what appears to be a very sick SE/30 on my hands here. It can't keep proper time, It won't boot with a IIfx ROM SIMM even under System 7.1 on a good HDD and a fresh installation of 16MB in Bank A. These upgrades should met the gamba prerequisites list for a "just works" machine state as I understand them at this point. Back to the docs. :
The SE/30 also fell prey to the dreaded brittle SIMM Socket Retainer Syndrome despite extreme caution. Thankfully, only Bank B was borked and I have plenty of metal clipped donors amongst the SIMMsaver collection amassed for IIsi MaxMem hack research. [] ]'>
] ]'>Meanwhile, I've got what appears to be a very sick SE/30 on my hands here. It can't keep proper time, It won't boot with a IIfx ROM SIMM even under System 7.1 on a good HDD and a fresh installation of 16MB in Bank A. These upgrades should met the gamba prerequisites list for a "just works" machine state as I understand them at this point. Back to the docs. :
The SE/30 also fell prey to the dreaded brittle SIMM Socket Retainer Syndrome despite extreme caution. Thankfully, only Bank B was borked and I have plenty of metal clipped donors amongst the SIMMsaver collection amassed for IIsi MaxMem hack research. [
] ]'>
Bolle
Well-known member
I wrote that somewhere else already but just checked again because you are having problems with two cards + accelerator.
Gamba did indeed build a stack back in the day consisting of:
Asante MacCon -> Daystar IIsi Adapter with PowerCache -> XCeed. So two cards + Accel. are definetly possible with the MacCon sitting at the bottom of the stack just as we plan to do it.
joethezombie
Well-known member
I had tried the Pivot as $9, and the Asante as $B. I didn't try any other combinations or orders. I've never tried using the Pivot before (I just got one for this testing). I'm traveling this week, but when I get back in I plan to do exhaustive testing. Who knows, maybe my Pivot card is bad?
Trash80toHP_Mini
NIGHT STALKER
Finally scheduled a little playtime in Illustrator today. [] ]'>
] ]'>
Trash80toHP_Mini
NIGHT STALKER
joethezombie made a good point about Bolle's PowerCache Adapter being a hacker's dream and an average user's nightmare because of the wrong angle hack requirement for the Asante MacCon NIC. While the three of us are chomping at the bit to get this particular version up and running in our SE/30s, some folks will be hesitant to desolder and replace the Passthru Connector on what's a fairly difficult to source NIC that can cost a significant amount of change.
While it would be fairly easy to turn his design into just another clone of the DiiMO form factor adapter by dropping bottom of the PCB design so the PDS connector meets the mobo, it's not quite as attractive a proposition.
Spent a bit of time (way too much) pulling bits and pieces from the end on views in my Illustrator files for the (early and thankfully superceded) wire wrap project. These were warped into the most complicated of three approaches possible for utilizing Bolle's Adapter with an unmodified NIC.
So, straight out of a notion that came up while working with BadGoldEagle on his vertical Color Pivot adapter: meet the WrongAngler©
As you can see, this adapter offsets Bolle's board, clearing the mounting ears. So it's not really necessary to add the lower pair of holes for using nylon hardware to bolt his adapter up to the SE/30 mounting ears.
However this WrongAngler© version for stock NIC adaptation does raise his PowerCache adapter something on the order of 7mm. If the PowerCache doesn't clash with the CRT at that altitude, this approach should work with anything short of a MaxSpec SE/30 PDS Card. The GS card would be in that category. It doesn't happen to clear the roof by at least 8mm when sitting in the ArtMix adapter?
Dunno, let me know?
The KISS approach to keeping everything within Bolle's DiiMO spec. would be to install a Vertical EuroDin120F on the bottom of his board. Then a pair of correctly offset 3x40 arrays of headers on a similar PCB would keep the boards and grounding/mounting points aligned.
< sigh! >
The UltraKiss WrongAngler variation would really be a significant rework of Bolle's design. Turning his adapter inside out and stretching it down into a WrongAngle form factor adapter board with the PCci Slot installed on its "Solder Side" would free up quite a bit of PCB real estate for TTL driver ICs and possibly even thet elusive "Third Passthru Slot" by replacing his "standard" EuroDin120-RA-Male PDS plug with a simple 3x40 array of headers. Thes would be inserted way down yonder into the Passthru Slot of an UNMODIFIED Asante MacCon NIC! [] ]'>
< sigh >
Much too bleary-eyed and tired to run with that line of thought tonight. :blink:
Thoughts?
While it would be fairly easy to turn his design into just another clone of the DiiMO form factor adapter by dropping bottom of the PCB design so the PDS connector meets the mobo, it's not quite as attractive a proposition.
Spent a bit of time (way too much) pulling bits and pieces from the end on views in my Illustrator files for the (early and thankfully superceded) wire wrap project. These were warped into the most complicated of three approaches possible for utilizing Bolle's Adapter with an unmodified NIC.
So, straight out of a notion that came up while working with BadGoldEagle on his vertical Color Pivot adapter: meet the WrongAngler©
As you can see, this adapter offsets Bolle's board, clearing the mounting ears. So it's not really necessary to add the lower pair of holes for using nylon hardware to bolt his adapter up to the SE/30 mounting ears.
However this WrongAngler© version for stock NIC adaptation does raise his PowerCache adapter something on the order of 7mm. If the PowerCache doesn't clash with the CRT at that altitude, this approach should work with anything short of a MaxSpec SE/30 PDS Card. The GS card would be in that category. It doesn't happen to clear the roof by at least 8mm when sitting in the ArtMix adapter?
Dunno, let me know?
The KISS approach to keeping everything within Bolle's DiiMO spec. would be to install a Vertical EuroDin120F on the bottom of his board. Then a pair of correctly offset 3x40 arrays of headers on a similar PCB would keep the boards and grounding/mounting points aligned.
< sigh! >
The UltraKiss WrongAngler variation would really be a significant rework of Bolle's design. Turning his adapter inside out and stretching it down into a WrongAngle form factor adapter board with the PCci Slot installed on its "Solder Side" would free up quite a bit of PCB real estate for TTL driver ICs and possibly even thet elusive "Third Passthru Slot" by replacing his "standard" EuroDin120-RA-Male PDS plug with a simple 3x40 array of headers. Thes would be inserted way down yonder into the Passthru Slot of an UNMODIFIED Asante MacCon NIC! [
] ]'>< sigh >
Much too bleary-eyed and tired to run with that line of thought tonight. :blink:
Thoughts?
Last edited by a moderator:
Trash80toHP_Mini
NIGHT STALKER
joethezombie
Well-known member
Hmm. I wonder if there would be room to do some crazy idea like this:
The DualPort clone running at TwinSpark length, but with a low 90deg for an asante backpack.
The 90degree would need extra long pins to provide clearance, or be on a small riser board.
The DualPort clone running at TwinSpark length, but with a low 90deg for an asante backpack.
The 90degree would need extra long pins to provide clearance, or be on a small riser board.
Trash80toHP_Mini
NIGHT STALKER
Trash80toHP_Mini
NIGHT STALKER
Crazy would be to flip the NIC upside down and plug the PassThru connector onto a 3x40 array of headers, keep the rows arranged correctly and it should work out. Dunno if there's enough clearance though. My Digipers give me eyeballed readings ranging between 5mm and 7mm between the top of a MacCon's Passthru connector to the FDD cage. That's really marginal for piggybacking anything down in there.
Bolle
Well-known member
I thought of this before as well.
Problem will be that you won't be able to fit the IIsi pivot this way as it reaches all the way to the front panel and has to pass the CRT. Things are pretty tight even when the stack is perfectly lined up with the original PDS slot on the Logicboard.
If we move even 1mm more inwards with the top connector it will not be able to pass the CRT easily. It already needs a little bending when everything lines up.
It might fit if you bend everything a bit more but I do not like that.
The line buffers (I used 74HCT365 for now) and a second PT will only fit if we do it without the 90° hack on the MacCon (as long as we want to keep things nice regarding the routing - which I want to as messing around with troubleshooting traces that are buried in the inner layers is no fun)
As I need the cards in a way for the IIsi pivot to fit and also do not need the second PT right now I will do a run of my initial design at first.
If everything works I will give the version with direct plug in support for the MacCon+Line drivers+second PT a run (or pass the design files to someone who wants to put one together)
This will hardly fit. I would rather not go for it
Will be too high for max sized cards.
EDIT
Like so:
Routing has to be done again more or less but as I know the issues from the first approach this shouldn't take as much time as the first one.
Last edited by a moderator:
Trash80toHP_Mini
NIGHT STALKER
Sweetness! [] ]'> Glad to hear you're all over every angle.
I threw everything I had together with the kitchen sink for this last pre-viz Illustrator session.
View attachment 14585
Pretty sure I've taken AI playtime past the point of being useful for practical purposes, just wanted to let the rest of the gang see how your long game might turn out. Accumulated errors of approximation based on scans, file exports and diagrams from DCaDftMF have turned file into GLOP! :lol:
The only thing I might take a stab at would be scaling in the zombie's scan of the actual GS cards the two of you've got so the rest of us can turn green with envy! GS display inboard and 8bit 16" resolution outboard will be to die for!
Speaking of which, does the GS Card stand straight up, clearing the CRT despite the inset induced by the inside out form factor adapter? Wire wrap connectors bolted up to the card with standoffs could set that aright. Clipping the longer pin protrusions after soldering is no biggie. Coming up with a correctly keyed RA connector might be a problem in either case, but as I said earlier, Gamba & Co. have been there and done that for us already.
Curious to see what bending the Pivot card means/how it complicates things?
Bolle_TrisSlot+Kitchen_Sink.PDF
] ]'> Glad to hear you're all over every angle.I threw everything I had together with the kitchen sink for this last pre-viz Illustrator session.
View attachment 14585
Pretty sure I've taken AI playtime past the point of being useful for practical purposes, just wanted to let the rest of the gang see how your long game might turn out. Accumulated errors of approximation based on scans, file exports and diagrams from DCaDftMF have turned file into GLOP! :lol:
The only thing I might take a stab at would be scaling in the zombie's scan of the actual GS cards the two of you've got so the rest of us can turn green with envy! GS display inboard and 8bit 16" resolution outboard will be to die for!
Speaking of which, does the GS Card stand straight up, clearing the CRT despite the inset induced by the inside out form factor adapter? Wire wrap connectors bolted up to the card with standoffs could set that aright. Clipping the longer pin protrusions after soldering is no biggie. Coming up with a correctly keyed RA connector might be a problem in either case, but as I said earlier, Gamba & Co. have been there and done that for us already.
Curious to see what bending the Pivot card means/how it complicates things?
Bolle_TrisSlot+Kitchen_Sink.PDF
Last edited by a moderator:
Trash80toHP_Mini
NIGHT STALKER
Nope, that's plenty already, the PowerCache and all three possible Slot IDs used to run the NIC and TWO video cards.
NuBus is for another day . . . someday . . . maybe. :-/
NuBus is for another day . . . someday . . . maybe. :-/
joethezombie
Well-known member
Sad news everyone!
So I tried just the Pivot IIsi card alone, and was still receiving a black screen on both $A and $9. Seeing as others have this card working with no issue, I started additional troubleshooting and found that apparently the Pivot IIsi has an incompatibility with the Mac ROM-inator II. At least, on my system it does. Anyone have that combination working? Anyway, After swapping back in the original ROM, the Pivot works. But that's where the fun ends. [V]
NO combination of Pivot, Asante, PowerCache, and Adapter work. :scrambled:
Three of the four work together without issue, as long as you can physically fit them together in their proper slots. But once all four are in play, no bueno.
I can Asante -> Adapter -> Pivot and all is well. But once you add the PowerCache, the machine starts shrieking and buzzing something fierce! The display actually comes on, but it is full of fuzz and flickers like mad. Funny thing, behind all that mess you can barely make out that the machine is trying to boot. I didn't leave it on very long for fear of the thing breaking out into fire-- it is NOT friendly sounding at all!
I tried all combinations of selectable addresses on the Asante and Pivot, with no difference. I tried every order I could physically fit together with the same results.
Asante -> Adapter + PowerCache -> Pivot :scrambled:
Asante -> Pivot -> Adapter + PowerCache :scrambled:
Adapter + PowerCache -> Asante -> Pivot :scrambled:
I couldn't try these combinations , because the orientation would jam cards into each other or into the logicboard:
Pivot -> Asante -> Adapter + PowerCache
Pivot -> Adapter + PowerCache -> Asante
Adapter + PowerCache -> Pivot -> Asante
I'm sad now. Tomorrow I will try with a Micron card and I think I also have a Farallon network card I could try.
So I tried just the Pivot IIsi card alone, and was still receiving a black screen on both $A and $9. Seeing as others have this card working with no issue, I started additional troubleshooting and found that apparently the Pivot IIsi has an incompatibility with the Mac ROM-inator II. At least, on my system it does. Anyone have that combination working? Anyway, After swapping back in the original ROM, the Pivot works. But that's where the fun ends. [V]
NO combination of Pivot, Asante, PowerCache, and Adapter work. :scrambled:
Three of the four work together without issue, as long as you can physically fit them together in their proper slots. But once all four are in play, no bueno.
I can Asante -> Adapter -> Pivot and all is well. But once you add the PowerCache, the machine starts shrieking and buzzing something fierce! The display actually comes on, but it is full of fuzz and flickers like mad. Funny thing, behind all that mess you can barely make out that the machine is trying to boot. I didn't leave it on very long for fear of the thing breaking out into fire-- it is NOT friendly sounding at all!
I tried all combinations of selectable addresses on the Asante and Pivot, with no difference. I tried every order I could physically fit together with the same results.
Asante -> Adapter + PowerCache -> Pivot :scrambled:
Asante -> Pivot -> Adapter + PowerCache :scrambled:
Adapter + PowerCache -> Asante -> Pivot :scrambled:
I couldn't try these combinations , because the orientation would jam cards into each other or into the logicboard:
Pivot -> Asante -> Adapter + PowerCache
Pivot -> Adapter + PowerCache -> Asante
Adapter + PowerCache -> Pivot -> Asante
I'm sad now. Tomorrow I will try with a Micron card and I think I also have a Farallon network card I could try.
Last edited by a moderator:
Bolle
Well-known member
Did you measure the voltages? Display issues and strange noises sounds like low voltages.
Also did you use the original GAL or one of your clones?
The IIsi pivot seemed to have issues together with the IIsi ROM as well. I will have to test that again to verify.
As said Gamba had succes getting three cards to work together, so it may just be an incompatibility with either the PowerCache or the Pivot.
With the possibility to have three cards we will have to test what works together and what does not anyways.
Also did you use the original GAL or one of your clones?
The IIsi pivot seemed to have issues together with the IIsi ROM as well. I will have to test that again to verify.
As said Gamba had succes getting three cards to work together, so it may just be an incompatibility with either the PowerCache or the Pivot.
With the possibility to have three cards we will have to test what works together and what does not anyways.
Trash80toHP_Mini
NIGHT STALKER
Copy that. I'm always thinking power requirements vs PDS power budget spec. the Radius Pivot cards in particular:
SE/30 Color Pivot form factor is a maximum card size spec - not intended for use with another card - uses close to the full PDS power budget?
Color Pivot II/IIsi form factor looks to be maximum card size? - design precludes using a third PDS card. It's likely never been tested with PowerCache, doesn't fit horizontally(?) in IIsi.IIRC I've looked into that possibility, will recheck.
Stuff ready to box up for mailing to you:
Macintosh II PowerCache Adapter
Macintosh IIx PowerCache Adapter
adding:
CPII/IIsi with wrong angle adapter sitting in position awaiting solder (untested since connector-ectomy) fairly easily adapted (pin clip, pull, whatever) to external power?
can add:
SuperMac Spectrum 8: MaxSpec SE/30 card?
IIsi board from my stack, they ALL need caps. : IIRC you don't have one, any IIcx form factor PSU resistor limited to IIsi specs should work as a parallel testing setup? You could also limit power with more resistance for variable borderline case testing?
will add:
Anything else we can think of that you might want to try testing before I box it up.
SE/30 Color Pivot form factor is a maximum card size spec - not intended for use with another card - uses close to the full PDS power budget?
Color Pivot II/IIsi form factor looks to be maximum card size? - design precludes using a third PDS card. It's likely never been tested with PowerCache, doesn't fit horizontally(?) in IIsi.IIRC I've looked into that possibility, will recheck.
Stuff ready to box up for mailing to you:
Macintosh II PowerCache Adapter
Macintosh IIx PowerCache Adapter
adding:
CPII/IIsi with wrong angle adapter sitting in position awaiting solder (untested since connector-ectomy) fairly easily adapted (pin clip, pull, whatever) to external power?
can add:
SuperMac Spectrum 8: MaxSpec SE/30 card?
IIsi board from my stack, they ALL need caps. :
IIRC you don't have one, any IIcx form factor PSU resistor limited to IIsi specs should work as a parallel testing setup? You could also limit power with more resistance for variable borderline case testing?will add:
Anything else we can think of that you might want to try testing before I box it up.
Last edited by a moderator:
Similar threads
