exxos's DFB1 trials

[exxos]

After all the FPU is intended to be run faster than the CPU. Here we're doing the reverse. I found (empirically) about half clock the most I could get away with -- going below CPUCLK/2 was unreliable. We would probably be taking out a variable by locking the two clocks, but it wouldn't explain why *this circuit board* is so senstitive.

How many others have built the thing ? We really need to know if others have been having FPU issues or not. Or if its just this board I have.

[exxos]

Muahahahaha

IMG_0200.JPG

So OPTION2 jumper makes it work then it seems.

EDIT:

I removed the /2 clock on the FPU and still working.. food for thought I guess ?

[dml]

\o/ succcess! congrats.


On the topic of FPUs, I dug up another obscure thing I made to roast your FPU at various speeds ... a mandelbox raytracer.

https://www.dropbox.com/s/63zcz6099kezr ... x.tos?dl=1

Unfortunately I lost the source for it so it either works for you or it doesn't at this point. When I run it in Hatari it seems to go into a rather high truecolour resolution in VGA mode - so maybe it was meant to be RGB only, or maybe it assumes you have a 32MHz Videl clock and is aiming for 416+ pixels on VGA. Dunno. YMMV. I just know it used to work on a normal Falcon!

[EDIT]

Ok I was wrong - its 320x240 in VGA mode so it should be ok.

[exxos]

On the topic of FPUs, I dug up another obscure thing I made to roast your FPU at various speeds ... a mandelbox raytracer.

Thanks, will give that a try after lunch :)

Still need to figure out what just happened to the FPU :lol:

[Badwolf]

How many others have built the thing ? We really need to know if others have been having FPU issues or not. Or if its just this board I have.

I know of one other user that had FPU issues (Frank). He had other issues too, caused by my CPLD soldering not surving the (suprisingly long!) transit to Guetersloh, so I don't know how representative that is. That board was from the same batch as this one and my one.

Both worked before it left my shed, though.

I know Steve had an FPU problem, but that turned out to be a missing component, so I've excluded that.

Other than that, I don't know a) how many boards are out there in operation -- I've seen probably five or six builds mentioned nor b) how many use the FPU socket.

At the end of the day I've only been made aware of two genuine problems.

BW

[Badwolf]

Muahahahaha
So OPTION2 jumper makes it work then it seems.

Aha.

So an FPU failing at a speed you'd expect it to run at?

It *could* be AS and DS lines @12-MHz mentioned he put inline terminators into.

-----> @12-MHz Rob: was it just the AS and DS or the CS line too? Thanks.

I removed the /2 clock on the FPU and still working.. food for thought I guess ?

Still got one of your VRO boards handy? What can you wind it up to?

I'll *bet* it fails earlier than it ought.

BW

[exxos]

Some more tests..

With the OPTION2 jumper in place and your original firmware, it seems to pass tests fine.

However if I then remove the division on the FPU clock ( /2 ) then it seems to fail on some tests.. Though I need to repeat these test several times really.. I've only done them twice so far.. But wondering if the FPU is tripping up at 16MHz or something else is going on...

IMG_0202.JPG

[dml]

Does the onboard FPU slot have good local bypassing on the +5v? Just thinking that lower clock rates could require more capacity near the chip, even if the chip itself is ok with that frequency it might be dropping out with longer pulses?

[Badwolf]

However if I then remove the division on the FPU clock ( /2 ) then it seems to fail on some tests.. Though I need to repeat these test several times really.. I've only done them twice so far.. But wondering if the FPU is tripping up at 16MHz or something else is going on...

That's what I tend to see when it's overclocked. BUT the same chip will work at that speed in another board (or another machine).

Reckon it must be signal intergrity. Probably control line as normal data exchanges work. But odd it changes so dramatically board to board.

BW

[Badwolf]

Does the onboard FPU slot have good local bypassing on the +5v? Just thinking that lower clock rates could require more capacity near the chip, even if the chip itself is ok with that frequency it might be dropping out with longer pulses?

Has 100n on three power pins and 10u on another, IIRC.

BW