aca1233n clock dividers:
50mhz / x2 / 100mhz xtal
40mhz / x2 / 80mhz xtal
26mhz / x3 / 80mhz xtal
So, according to my assumptions, replacing a 26mhz boards Xtal with 100mhz should give a 33mhz cpu clock. But may be reported as 26mhz in software still due to cpld programming. The existing chip might actually run at 33, or may need to be replaced if not stable.
@GadgetUK164 Do you think this looks right?
Misinformation surrounding aca1233n
Moderators: terriblefire, Terriblefire Moderator
- GadgetUK164
- Posts: 445
- Joined: Fri Jan 04, 2019 2:26 pm
Re: Misinformation surrounding aca1233n
Yes, that looks correct to me ie. /3 with 100Mhz would give that board 33Mhz CPU clock. With regards to the 26Mhz reported in software - I assume that would be using the tools that come with the card. In something like SysInfo it should report the correct speed as its calculated properly there.Steve wrote: Thu Sep 17, 2020 1:17 pm aca1233n clock dividers:
50mhz / x2 / 100mhz xtal
40mhz / x2 / 80mhz xtal
26mhz / x3 / 80mhz xtal
So, according to my assumptions, replacing a 26mhz boards Xtal with 100mhz should give a 33mhz cpu clock. But may be reported as 26mhz in software still due to cpld programming. The existing chip might actually run at 33, or may need to be replaced if not stable.
@GadgetUK164 Do you think this looks right?
My YouTube Channel - www.youtube.com/GadgetUK164
Re: Misinformation surrounding aca1233n
@GadgetUK164 Yeah in their ACATool. I assume all they change in the CPLD would be the selectable 'under-clocks' (they let you clock it lower at four different stages) and they would just change the divider between 2/3.
I'm gonna try this on my board. I'll report my findings.
I'm gonna try this on my board. I'll report my findings.
Re: Misinformation surrounding aca1233n
Here is a photo of the Xtal and ram chips, I am not too familiar with Xtals and their specifications so I'm hoping someone can point me to the right item on mouser.
text on crystal:
80.000
YBCTC
.CN 1606
dimensions 5mm x 6.5mm
Looks like the SDRAM is very good and rated to 166mhz:
https://www.avnet.com/shop/apac/product ... 634227916/
text on crystal:
80.000
YBCTC
.CN 1606
dimensions 5mm x 6.5mm
Looks like the SDRAM is very good and rated to 166mhz:
https://www.avnet.com/shop/apac/product ... 634227916/
Re: Misinformation surrounding aca1233n
Interesting thread. I don't have enough knowledge to comment but according to the icomp wiki 26 & 40MHz boards made after July '19 are different to ones produced before that date.
http://wiki.icomp.de/wiki/ACA1233n_registers
http://wiki.icomp.de/wiki/ACA1233n_registers
Re: Misinformation surrounding aca1233n
I have found more info on the icomp forum where Jens goes in to quite a lot of detail surrounding the swapping of oscillators on this card:
https://forum.icomp.de/index.php?thread ... aca-1233n/
According to this it seems the card is clocking the ram close to maximum as it is, changing the oscillator could really mess with timings as well. I won't be trying this now and I hope this thread doesn't convince anyone else to try, It doesn't look like a good idea.
https://forum.icomp.de/index.php?thread ... aca-1233n/
According to this it seems the card is clocking the ram close to maximum as it is, changing the oscillator could really mess with timings as well. I won't be trying this now and I hope this thread doesn't convince anyone else to try, It doesn't look like a good idea.
Re: Misinformation surrounding aca1233n
I suspect the main reason that he discourages this is that he offers 55mhz version for an extra hundred euros. Given it doesn't cost him any more to make than the 40mhz version, that's a lot of profit he'd miss out on if you do it yourself.
He certainly doesn't have a problem with running parts out of spec when he does it. Many of his 40Mhz cards have 25Mhz 030s.
He certainly doesn't have a problem with running parts out of spec when he does it. Many of his 40Mhz cards have 25Mhz 030s.
