Thread
68882 query
Just like the listing states. The Radius 68020 board in my SE has a socket like that so do many Amiga and Atari Falcon/TT030 upgrade boards I've seen. Micromac boards also use that socket in their Mac upgrades and so did Daystar. I think the Mac II/IIfx boards have those sockets. I'm fairly certain my IIci and IIvx boards don't but I haven't looked at them in a while. I found a place in Germany that sells 68882's cheaper than ebay, but not much cheaper. Those things are like gold. :'(
HP made a few workstations with 50 mhz 68030s as well...
pop that one and a 50mhz 030 into a IIfx and OC it to 60mhz 
There was a clone maker that actually overclocked IIfx motherboards to 50mhz. I would imagine they wouldn't be very stable at that speed. The IIfx was frequently glitchy at 40mhz, at 50mhz it must have been a real handful.pop that one and a 50mhz 030 into a IIfx and OC it to 60mhz
that would have been the dash/30fx then 
but they used the 40mhz 030 and 68882 and OC´d them to 50mhz... there are reports of IIfx´ beeing stable at 50mhz and above with 50mhz CPU and FPU...
but they used the 40mhz 030 and 68882 and OC´d them to 50mhz... there are reports of IIfx´ beeing stable at 50mhz and above with 50mhz CPU and FPU...
I'd like to know where they got the RAM required for those machines. The IIfx RAM was unique and designed around the clock speed of the IIfx. If you overclock the motherboard then you need even faster RAM than the stock IIfx but that fits in the same slots. Since that RAM wasn't used in any other computer, where did they get faster RAM from? :?:that would have been the dash/30fx then
but they used the 40mhz 030 and 68882 and OC´d them to 50mhz... there are reports of IIfx´ beeing stable at 50mhz and above with 50mhz CPU and FPU...
Then who was RAM?
... Perhaps they simply got lucky and got the system stable on occasion at the higher speed.
... Perhaps they simply got lucky and got the system stable on occasion at the higher speed.
I think the IIfx required 80ns RAM. If they had IIfx RAM built with 60ns chips it would have had plenty of margin for the faster speeds.
It's also possible that with the asynchronous memory access, it just didn't matter that much. Or the number of clock cycles spent waiting on 80ns access may have been sufficient at 50 MHz as well as at 40 MHz because of rounding. That is, if a read takes five cycles at 40 MHz with 80ns RAM, it may be that the access is actually done in 105 ns rather than 125 ns. That 105 ns is close enough to five X 20ns cycles you'd get at 50 MHz for it to probably work, if there's a little margin in the memory chips...