• 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.

NuBus PPC - PDS inplications of RocketSocket overcloking

Trash80toHP_Mini

NIGHT STALKER
An interesting eBay thread sub-thread popped up about clip-on system clock enhancements from Output Enablers based on what they call the RocketSocket



Couldn't do a direct link to the Mac Products Table, it's a button click away on the Main Page

Noticed something curious there:

PowerMac 8100/100 ____123 MHz____40,
Radius 81/110.___________125 MHz ___40, 41.5384


I've got a spare 81/110 board, so I'll probably give this a try if they have any left in stock. I'll attribute the 2MHz difference between 8100 and Radius' fabulous Big Metal Can having a better cooling budget than the awful, crumbly plastic thin from Apple.

Question: I have a G3 Accelerator in the PDS and I'm wondering if its clock is tied to or could be tied to the faster system bus/PDS to overclock it as well?

Not a minor task if even possible as they're designed to run in systems of different clock speeds at their rated frequency. But it's a tiny bump and the 81/110 Can contained Clone has plenty of room for active cooling, which isn't specified as a requirement for it or the 8100. You wouldn't want to do this on a slower system. Overclocking the Accelerator by the same amount would likely be much less of a Kluge if that's a thing, but I gotta ask. :blink:

 

Unknown_K

Well-known member
G3 accelerator clocks are on the CPU card while the bus speed interface is decided by the components on the motherboard. No idea how overclockable a 601 PPC system board is but I assume not very.

 

Trash80toHP_Mini

NIGHT STALKER
Yep the first is a given. Second, not a lot, 15MHz bump in the case of my 81/110 per the output enablers page.

G3 cards use their own clock, presumably because they interface with with machines with different bus speeds. There should be a set of multipliers on board the G3 card to sync with whatever the bus clock might be as in the case of the 6100/60 and 6100/66, a 10% differential. From 110 MHz to 125MHz on the 81/110 that's a 4.5454% differential., not a lot, but interesting nonetheless. Finding a way to sync the accelerator to the bus speed of high the 110 MHz end systems would give the accelerator that same little 4.5454% nudge.

From what I recall offhand, overclocking G3 cards by changing crystals hasn't worked out very well, if at all? Just another way of looking at the prospects. Ramping 300MHz G3 card to 313.63MHz in perfect sync with a 125MHz PDS would be in the shiggles and gits class of modding. But interesting to tease apart how the cards interface with the various system clocks.

 

Unknown_K

Well-known member
Overclocking the FSB will probably affect Nubus slot timing as well as video timing and who knows how much room for overclocking the system memory bus has. A 5% overclock will probably not even be noticable other then system stability,

 

Byrd

Well-known member
I can confirm an overclocked Quadra 840AV ~ 48Mhz does not play nicely with a Radius Rocket II @ 40Mhz; it boots with the card but freezes are soon as you try to run RocketShare.

 

Unknown_K

Well-known member
I have a Daystar 040/33 running at 48mhz or something like that in a IIci and that works OK (no cache module I think) with a cooling can attached. 

 

Trash80toHP_Mini

NIGHT STALKER
I'm not sure I'd consider even a bone stock a 840AV a stable system. [;)]

Overclocking the FSB will probably affect Nubus slot timing as well as video timing and who knows how much room for overclocking the system memory bus has.
Shouldn't affect either, IIRC that's not been a reported problem in the OC pages. Isn't it the FDD controller gets wonky first as SWIM and WIM before it have timingtrucks up their sleeves that are, very, very picky. Not so worried about Video, HPV works under many different clock rates and onboard video is all but useless.

A 5% overclock will probably not even be noticable other then system stability,
Probably not, but what's the fun in not trying? It's all about the benchmarks, real world concerns are but a trifle.***

*** Says the guy who's never run a benchmark in his life and doesn't put much stock in results posted by others, especially in the Mac press. [:p]

 
Top