Page 41 of 44

Re: STE V1.5 32MHz BOOSTER - STATUS

Posted: Fri Nov 30, 2018 9:19 pm
by exxos
The second ATF GAL has hit 76 passes so far (About 5 hours run time). So looks like the ATF chips are a winner.

For the hell of it, I am going to try "turbo mode" and 40MHz again on this board.. not expecting it will work.. but I just want to be sure...

Re: STE V1.5 32MHz BOOSTER - STATUS

Posted: Fri Nov 30, 2018 9:55 pm
by PhilC
Sounds like you've got much more of a chance now. Well done.

How do the atmel gals compare price wise? And would they be pin compatible with the likes of a 20V

Re: STE V1.5 32MHz BOOSTER - STATUS

Posted: Fri Nov 30, 2018 10:09 pm
by exxos
Forgottenmyname wrote:
Fri Nov 30, 2018 9:55 pm
How do the atmel gals compare price wise?
No price difference really.

Forgottenmyname wrote:
Fri Nov 30, 2018 9:55 pm
would they be pin compatible with the likes of a 20V
They are the same as a 22V10 GAL, just with a couple extra features. I never looked into any other GAL pinouts.

Re: STE V1.5 32MHz BOOSTER - STATUS

Posted: Fri Nov 30, 2018 10:23 pm
by PhilC
Sorry, that's the one I meant, pub got in the way of my memory banks.

Will have a look at those for use with my PAKs then I stead of the gals.

Re: STE V1.5 32MHz BOOSTER - STATUS

Posted: Fri Nov 30, 2018 10:32 pm
by exxos
Forgottenmyname wrote:
Fri Nov 30, 2018 10:23 pm
Sorry, that's the one I meant, pub got in the way of my memory banks.

Will have a look at those for use with my PAKs then I stead of the gals.
You have to be careful though as pin 4 is "power down" not all programmers have the ATF22V10 (thankfully mine did) but I didn't risk it and hardwired pin 4 to 0v just to make sure... really needs to be tested more, but I got "stung" with all this back in 2015 :roll:

Re: STE V1.5 32MHz BOOSTER - STATUS

Posted: Fri Nov 30, 2018 10:52 pm
by exxos
LOOP 25 ~ hour and half with "fast mode".. The whole machine has been on all day already as well..

Really makes me wonder about the GAL's, as they couldn't even run with a 16MHz clock... bizzare...

Going to try 40MHz on this board now.. not expecting it will work, but no harm in trying....

Re: STE V1.5 32MHz BOOSTER - STATUS

Posted: Fri Nov 30, 2018 11:35 pm
by sandord
exxos wrote:
Fri Nov 30, 2018 10:52 pm
LOOP 25 ~ hour and half with "fast mode".. The whole machine has been on all day already as well..

Really makes me wonder about the GAL's, as they couldn't even run with a 16MHz clock... bizzare...

Going to try 40MHz on this board now.. not expecting it will work, but no harm in trying....
Fingers crossed!

Re: STE V1.5 32MHz BOOSTER - STATUS

Posted: Fri Nov 30, 2018 11:49 pm
by exxos
Loop 11 (does about 17 loops per hour) suprised its run for this long...

Will try and leave it on all day tomorrow...

Also looks like my pink TV is having issues as its constantly going out of sync and its not the cable or ST :( Will have to take it to bits monday and do some recapping see if I can bring it back to life...

Re: STE V1.5 32MHz BOOSTER - STATUS

Posted: Fri Nov 30, 2018 11:55 pm
by exxos
Its died.. so 40MHz is still a no-go. Saying that, I will put back "slow mode" to rule out ROM speed issues..

Re: STE V1.5 32MHz BOOSTER - STATUS

Posted: Mon Dec 03, 2018 1:14 pm
by exxos
Another quick update, 40MHz was left running all day Saturday, I'm not at home on weekends, so someone else was keeping a eye on it for me.. We don't actually know how long it was for, somewhere between 4-7 hours. This was without ROM running faster as I just wanted to test the raw CPU speed.. But basically I am not happy that is going to be reliable.

The ROM clearly is having issues at 40MHz. I am not really surprised about this as mentioned before. We are really down to a fine line of hardware tolerances on everything. So to keep a happy medium with everything, boosters will be running at 36MHz.

Users can of course remove the wait states on the ROM, and users could even try changing the oscillator for a 40MHz one. Clearly some people get lucky with this but it is to hit and miss to manufacture the booster with a 40MHz osc. Really though, other than 34% extra on integer division, there is not really any advantages to this speed.

In any case, it is pretty clear that these ATF chips are working a lot better than the original GAL's I was using.

So hopefully, finally, I can finish off a few of these boards and forget about this project for a while and work on something else :)