IIsi bus speed, overclocking/NuBus/PDS sync implications?

[Trash80toHP_Mini]

Has anyone tested the IIsi in overclocked state with NuBus Adaper/card or PDS card installed?

The only mention I've seen so far is this LEM blurb:
Although the IIsi was marketed as a 20 MHz computer, users quickly discovered it used parts rated at 25 MHz. (Apple had intended it as a 25 MHz computer, but chose to scale back the speed to avoid cutting into IIci sales.) Chipping the IIsi to 25 MHz – or even 28 MHz – is not unusual.

The IIsi shared some 25MHz components used in the IIci such as MDU and SWIM. It also had new ASIC designs such as VIA1, RBV and COMBO that were likely designed to run at higher clocks than 25MHz in future machines. However, that does not equate with Apple having specifically designed the IIsi as a 25MHz computer and then intentionally laming its system bus speed to 20MHz so as not to cut into sales of the IIci cash cow.

I'm one of the first to imagine such shenanigans from Apple, but I've also been one of the first to doubt information gleaned from LEM. So I've been wondering if this is true and might be tested or it might be yet another LEM fable?

There's a crystal can on Apple's NuBus adapters, but I can't read the legends on the two I've checked. It doesn't appear to control the Math CoPro clock, That's a 20MHz part so it probably runs off the CPUCLK line. The crystal could be related to a similar adaptation of the baseline C16M line to 10MHz the for NuBus as was done in the 16MHz Macintoshes II, IIx and IIcx. Very curious that.

At any rate, back to the point: has anyone noted anything hinky about running PDS or NuBus Cards in an overclocked IIsi?

 

[Fizzbinn]

I haven’t noticed any problems but I haven’t exactly put my 25MHz upgraded IIsi through a scientific ringer. Most confident the PDS Asante Ethernet card works fine, I played around with an E-machines combo display/Ethernet card that seemed to work okay but went back to the PDS Ethernet card when I got a Daystar combo PDS slot/cache slot card & Turbo 040 card. Come to think of it I wonder if the 25MHz 030 bus means much with the Turbo 040. Might be worth a test (socketed the crystal upgrade).

 

[Trash80toHP_Mini]

@Fizzbinn forgot to thank you for that input. THX! Now I'm back to this question again, more specifically:

Wondering how far the IIsi architecture system clock might be pushed while preserving the NuBus adapter/subsystem clock at 10MHz along with other clock dependent subsystems.

I'm no longer inclined to believe statements that the IIsi was originally intended to run at 25MHz and then hamstrung by Apple to milk the IIci cash cow, much as I've been eager to assign such dirty deeds to Apple.

1) the IIsi employs many components/subsystems from the IIci. So yes that much is true.
2) some were derived from and ostensibly improved upon as in the case of the COMBO ASIC meld of Serial Port and SCSI controllers.
3) a monophonic Sound Input kluge of epic proportion expanded upon the IIci architecture's ASC based stereo Sound Output subsystem.
4) the IIci introduced the Cache Slot, the first of the application specific interface standard having no breakout to the world outside.
5) the IIsi expanded upon the 030 PDS Slot with Slot Manager support for multiple cards introduced in the SE/30
6) IIsi 030 PDS Slot expanded upon that interface to support of a single NuBus card mapped to address $9
7) IIRC such leaves $C and $D available to a multi-function PDS/NuBus riser card combo

So we're looking at the interplay of several distinct subsystems reliant upon four different clocks lined up in marching order on the IIsi board.