CURRENT PROTOTYPE STATUS (SEC 64MHz 68000)

[PhilC]

does it still work at 8mhz? You're certainly not having much luck tonight.

[exxos]

does it still work at 8mhz? You're certainly not having much luck tonight.

It's now a stock machine... so the boards died somehow.

[PhilC]

does it still work at 8mhz? You're certainly not having much luck tonight.

It's now a stock machine... so the boards died somehow.

Certainly feel your pain today! PSU maybe?

[exxos]

Certainly feel your pain today! PSU maybe?

its my new red one :)

Oddly I just found out the half of the databus is pulled up next to the MFP, So thats been a long running thing nobody has ever noticed.. I changed the address and databus ones, or so I thought, to 2.2K, but D0-4 is on a pull up no where near the CPU, madness.. so will change that one now.. maybe get lucky.

EDIT:

Nope. Guess I will have to change the blitter socket :(

[exxos]

Socket changed, still the same :( Never known a stock board be grumpy like this either :roll: I only just kitted this board out like a week ago because my previous one died for no reason.. now looking on starting on a 3rd board.. geeze...

[exxos]

Tried a totally different machine, and still getting the same error... I think there is some issue with the diagnostic cartridge test itself :roll:

So blitter is now working on this other board (tested wit GB6) , but the one I actually trying to test the booster on this basically resets not long after I turn the machine on, so that does have some odd fault which obviously is not actually related to the blitter :roll: But saying that, when the blitter is removed, the machine does work :pullhair: :pullhair: :stars: :stars:

EDIT:

and now the blitter is working in the machine it wasn't doing earlier , WTF is going on :shock:

Plugged the SEC back in and blitter acting up again :roll:

[PhilC]

:dizzy:

[exxos]

Plugged the original SEC back in, same blitter issues... turned on and off server times, now its working fine again..

There has to be something screwy with the MMU wakeup states or something, because the fault changes from just powering off and on again it seems.. I can press the reset button loads and it doesn't make it better or worse, its only a power up issue... I'm really starting to hate the Atari MMU these days :evil: :evil:

EDIT:

Several turn off-on's and 2 bombs after the Fuji now.. oddly that was what TF's board was doing as well.. Power off a few more times, working fine again.

I'm also wondering if these are the issues I had with the PAK boards as it didn't work on my machines, but worked on RJ's.. I bet if I powered off and on enough times that would boot up as well..

[exxos]

Progress :)

Oddly the original SEC worked fine with the blitter.. and yet this new SEC version is basically the same, only didn't play nice with the blitter.

So I added a CLK8 delay to BR, and then it loaded GB6 but did this...

IMG_4335.JPG

But it was actually repeatable that GB6 did load each time, but it was the graphics side which was going wonky...

So then I added another CLK8 delay to ST_BG... and...

IMG_4337.JPG

I dare not turn it off in case it does not turn back on again :lol:


Current setup..

IMG_4336.JPG

So oddly the bus grant needed some delays, when in all my testing on the previous booster did not need this.. Bizarre... I need to look into this some more...

[exxos]

I have been running the SEC at 50MHz as much as possible the past days.. it seems to get around 25 passes then 2 bombs to desktop. I've repeated this a few times now. I think the SEC is overheating after a while. Oddly the 50MHz osc is getting rather hot as well. So earlier I put in a 40MHz osc (doesn't get warm at all) and its just passed loop 33. So run time wise, its been going a bit longer than 50MHz.

My previous booster could run all day, so I think its likely the SEC will need some cooling if running at 50MHz. Of course not all CPU's are going to overclock as good as others. I will keep doing tests over the next few days, and of course add a heatsink onto the CPU and re-test at 50MHz.

At 40Mhz.. I'm taking the shifter can as ambient temp, 28c.. the CPU is at 38c at the top, and 33c at the bottom (as looking at it on the motherboard) :shrug:

EDIT:

It died again, I didn't see the results for a while, but it easily passed 40 loops. So its ran a lot longer than before, almost double I guess... heatsink time...