Bought a couple of parts iisi units. Both have good egrets, and other components that are compatible with classic ii from searching google. Is it possible to do a cheap upgrade from the 16mhz to 20 68030 from iisi as well?? why not if its possible 
-
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.
Classic ii, iisi parts crossover
- Thread starter imactheknife
- Start date
Curious about possible wonkiness of NuBus Adapter/Card function in an overclocked IIsi board?
The crystal can is unmarked on the Apple IIsi NuBus adapters I have on hand. Methinks it's for the FPU which would be consistent with Accelerators I've seen.
The Radius NuBus Adapter for the IIsi halves the baseline 20MHz CPUCLK of the IIsi to derive the 10MHz NuBus Clock requirement. It "just works" in the SE/30, but at 8MHz based on the 16MHz SE/30 clock and NuBus cards tested appear to work at that rate.
So wondering if NuBus would be running at a multiple of 10MHz in an overclocked IIsi based on the chipped clock rate?
The crystal can is unmarked on the Apple IIsi NuBus adapters I have on hand. Methinks it's for the FPU which would be consistent with Accelerators I've seen.
The Radius NuBus Adapter for the IIsi halves the baseline 20MHz CPUCLK of the IIsi to derive the 10MHz NuBus Clock requirement. It "just works" in the SE/30, but at 8MHz based on the 16MHz SE/30 clock and NuBus cards tested appear to work at that rate.
So wondering if NuBus would be running at a multiple of 10MHz in an overclocked IIsi based on the chipped clock rate?
When you overclock beyond the 50MHz oscillator (CPU speed would be 25MHz at this point) the NuBus runs into a lot of weird issues.
Yep, that'd be the result I expected.
If using a NuBus Card, keep the IIsi system clock stock at 40MHz. I'm a bit surprised you don't get wonkiness with the 50MHz crystal can?
PDS cards may or may not work correctly. I'd not much expect an SE/30 form factor PDS Card to work correctly in a hot rodded IIsi. In DCaD Apple notes that some legacy 16MHz cards for the SE/30 may not be compatible the 20MHz Clock of the IIsi.
Is the IIsi's onboard Vampire Video kluge wonky at all given that 25MHz CPUCLK rate? Timings could be off for syncing with brain dead Apple fixed, silly resolution displays?
edit: this pretty much puts the last nail in the coffin of the "Apple had intended it as a 25 MHz computer, but chose to scale back the speed to avoid cutting into IIci sales" (urban) LEM legend. Given the tightly knit, flexible NuBus/PDS expansion spec. for its development, the IIsi was obviously designed from the very start to be a 20MHz system. It just happened to use available parts from the eleven months earlier IIci design that were rated for that faster 25MHz CPUCLK.
Clock chipping the IIsi: Just don't do it, find an accelerator.
If using a NuBus Card, keep the IIsi system clock stock at 40MHz. I'm a bit surprised you don't get wonkiness with the 50MHz crystal can?
PDS cards may or may not work correctly. I'd not much expect an SE/30 form factor PDS Card to work correctly in a hot rodded IIsi. In DCaD Apple notes that some legacy 16MHz cards for the SE/30 may not be compatible the 20MHz Clock of the IIsi.
Is the IIsi's onboard Vampire Video kluge wonky at all given that 25MHz CPUCLK rate? Timings could be off for syncing with brain dead Apple fixed, silly resolution displays?
edit: this pretty much puts the last nail in the coffin of the "Apple had intended it as a 25 MHz computer, but chose to scale back the speed to avoid cutting into IIci sales" (urban) LEM legend. Given the tightly knit, flexible NuBus/PDS expansion spec. for its development, the IIsi was obviously designed from the very start to be a 20MHz system. It just happened to use available parts from the eleven months earlier IIci design that were rated for that faster 25MHz CPUCLK.
Clock chipping the IIsi: Just don't do it, find an accelerator.
Last edited:
Yep, that'd be the result I expected.
If using a NuBus Card, keep the IIsi system clock stock at 40MHz. I'm a bit surprised you don't get wonkiness with the 50MHz crystal can?
PDS cards may or may not work correctly. I'd not much expect an SE/30 form factor PDS Card to work correctly in a hot rodded IIsi. In DCaD Apple notes that some legacy 16MHz cards for the SE/30 may not be compatible the 20MHz Clock of the IIsi.
Is the IIsi's onboard Vampire Video kluge wonky at all given that 25MHz CPUCLK rate? Timings could be off for syncing with brain dead Apple fixed, silly resolution displays?
edit: this pretty much puts the last nail in the coffin of the "Apple had intended it as a 25 MHz computer, but chose to scale back the speed to avoid cutting into IIci sales" (urban) LEM legend. Given the tightly knit, flexible NuBus/PDS expansion spec. for its development, the IIsi was obviously designed from the very start to be a 20MHz system. It just happened to use available parts from the eleven months earlier IIci design that were rated for that faster 25MHz CPUCLK.
Clock chipping the IIsi: Just don't do it, find an accelerator.
Apple may have considered making the IIsi a 25mhz machine, and certainly the PCB design would make no difference if it ran at 20 or 25mhz. As you have found, NUBUS compatibility with the SE/30 likely caused Apple to just proceed with a 20mhz model.
Is the PDS slot in the IIsi the same as the cache/PDS slot of the IIci?
I’m thinking it was likely easier for Apple to make the IIsi a “middle ground” at 20Mhz than to try to make it somehow compatible with SE/30 PDS and also run at 25mhz.
If the legend has any truth at all, it likely was a very early decision in the design process, not a late change like is alluded to by LEM editorials.
I have one of these 25MHz boards, be neat to test on my end and see.
You have a 25mhz IIsi that shipped as 25mhz, or that you made run at 25mhz with an oscillator swap?
I have one that shipped with a 25MHz and a weird serial number.
Not likely at all, I'd say, given overall history of development:
The SE/30 was an outgrowth of the SE line on the PDS expansion fork of Macintosh II/SE expandable Mac paths. It is literally a 16MHz Macintosh IIcx re-designed for the Compact line form factor.
NuBus was excised in the process and the new 030 PDS standard substituted. As such it was a natively 16MHz design and all kinds of havoc are wreaked upon its subsystems when that CPUCLK is tampered with, SE/30 Video (in essence a hardwired NuBus Card) being the first example. It's set up for the finely A/B CRT display subsystem
Per above, NuBus was entirely out of the scope of the SE/30's development. That the Radius NuBus Adapter works on its PDS at all is a minor miracle, but totally out of spec. with NuBus running at 8MHz
Not at all, it's exactly the same as the 030 PDS of the SE/30, but running at the higher 20MHz CPUCLK with 10MHz NuBus derived from that.
Both SE/30 and IIsi require exactly the same IIci Cache Slot adaptation for compatibility with the Universal PowerCache and other accelerators designed for the baseline IIci Cache slot implementation.
Note: IIci cache Slot is NOT a PDS Slot, Apple defined it as a new standard for expansion, the "Application Specific Slot" if memory serves. That spec defines itsuch as having no connection to the world outside the box
IIsi was targeted from the start to be the mid/low marketing entry to the 030 Macinosh II line. IIvi and IIvx were already on the horizon as stopgap measures during the Quadra line's development.
Macintosh IIci production was kept going almost forever as there was so much demand for it as the de facto DTP workhorse with the IIfx handling anything beyond its capabilities in Graphics Production until the Quadras were released.
I'd bet a shiny nickel that that thought never crossed the minds of Apple marketing and so never entered into design process at all.
Very interesting, are you positive it shipped with that 50MHz crystal?
Is the crystal can soldered or socketed?
What does the silk screen layer of the board say at that position?
Check to see if there's any crystal present capable of supporting 10MHz for the NuBus adapter, none present on stock IIsi other than the 40.0000MHz system crystal closest to SIMM Slots. The other three are for various purposes. Appearance of such (a fifth?) crystal on your board might prove it to be a prototype. Otherwise it would likely be special production run as indicated by the odd serial number?
If a special edition of the stock board, that would make it a PDS ONLY build, NuBus incompatiblilty would have been specified. I'd think the silk screen legend would say 40.0000MHz?
Last edited:
It’s a 25MHz CPU.
Attachments
CPU speed doesn't matter a bit, only its CPUCLK rate. Check the markings on the crystal with the 40.0000MHz silk screen legend. It would have to be 50MHz to make any difference in performance/full use of that CPU's capability. Just as you can overclock a CPU, underclocking works the same way, but is inherently stable which is definitely not the case when overclocking.
If it's a 40MHz crystal, you have a stock IIsi build running at 20MHz that just happened to be populated with a 25MHz CPU. 20MHz CPU may have been discontinued, become unavailable, production line may have run out of the stock proc or pricing for the 25MHz part in bulk may have been less expensive than buying batches of both processors separately?
So many possibilities, but only the clock will tell.
If it's a 40MHz crystal, you have a stock IIsi build running at 20MHz that just happened to be populated with a 25MHz CPU. 20MHz CPU may have been discontinued, become unavailable, production line may have run out of the stock proc or pricing for the 25MHz part in bulk may have been less expensive than buying batches of both processors separately?
So many possibilities, but only the clock will tell.
Perhaps when Apple was doing initial production, to save time they just used 25mhz CPUs they had on hand for the IIci in building the first few IIsi machines. Maybe there was a delay with Motorola for 20mhz marked chips.
Because this machine has non-prototype ROMs, and it has silkscreen for 40mhz clock oscillator, leads me to believe it was always a 20mhz board.
Because this machine has non-prototype ROMs, and it has silkscreen for 40mhz clock oscillator, leads me to believe it was always a 20mhz board.
I agree but it’s the serial number that really gets me as being very odd, hence why I cannot fully agree.
Attachments
What is the logic board serial number ? That is more important as the case sticker could have been changed by anyone. I might not even be an apple sticker it could be an inventory control sticker for asset documentation.
Supply chains back in the day were a lot more reliable, at least when mask size of components were stable, just slow by today's standards. Nineties went to JIT, "Just in Time" inventory practices which were a PITA from the start for all but the loudest of the largest, highest production outfits. Some organizations managed to torque that into what I call "Never Quite In Time" inventory, especially in retail.
Todays, sporadic "Estimated Delivery, If and When" bollixed Supply Chain issues . . .
edit: strange encoding, WAG would be very early unit using IIci Proc while they sorted out supply and production facilities. Numeric vs. Alpha-Numeric encoding would lean me in that direction.
Have you read the Can Spec. yet?
Todays, sporadic "Estimated Delivery, If and When" bollixed Supply Chain issues . . .
edit: strange encoding, WAG would be very early unit using IIci Proc while they sorted out supply and production facilities. Numeric vs. Alpha-Numeric encoding would lean me in that direction.
Have you read the Can Spec. yet?
Last edited:
