exxos's DFB1 trials

[Badwolf]

Is this one?

That's the one, thanks! @exxos -- MAPROM2.2 is here.

LED5 I saw pulsing. But on the FPU socket I did not.. So I set up a singleshot capture then I saw about 3 pulses. I checked DSACK (cannot remember if both pulsed or not) but there was activity there.

The most obvious thing for me would be DSACK0 and 1. They should both pulse but not necessarily at the same time.

I suppose it's possible the DSACK[0] connection to the CPU is broken -- running something that uses the DSP would show that up.

The other obvious candidate would be the JP5 connection to A0 (should be tied to VCC). And the two pull-ups R2 and R3 on the FPU_DSACK[x] lines.

I have tried all these, but I'll freely admit I was going crazy with it by this stage, so I may have missed something obvious.

Screenshot 2022-11-10 at 19.12.09.png

I've not done something stupid like put a 10 Ohm instead of a 10k on the lines?

From what I remember, an FPU exception throws a ton of bombs - 11 or something (LINE-F exception). It can be looked up in the Atari docs.

Ah, yes 11 bombs -- I think I said 13. That's the one.


BW

[Badwolf]

Socket out.... Place your bets....

Christ I made a mess of taking out that 030 socket and you've had that off looking like new!

I hate desoldering!

BW

[exxos]

The most obvious thing for me would be DSACK0 and 1. They should both pulse but not necessarily at the same time.I suppose it's possible the DSACK[0] connection to the CPU is broken -- running something that uses the DSP would show that up.

I will double check, but think both were pulsing.. Though I suspect that is not the problem as without the FPU it doesn't show a number.. So assume no DTACK from the FPU is seen causing no number. When the FPU is there, CPU sees DTACK and it gets "something" back , even if just garbage.

The other obvious candidate would be the JP5 connection to A0 (should be tied to VCC).

Yep.

And the two pull-ups R2 and R3 on the FPU_DSACK[x] lines.

2K, seem fine also.

I've not done something stupid like put a 10 Ohm instead of a 10k on the lines?

Checked already, all 10K and no shorts.

I have tried all these, but I'll freely admit I was going crazy with it by this stage, so I may have missed something obvious.

Maybe it is obvious, maybe these FPUs fail miserably at anything but 16Mhz. I have tried 2 in the Falcon and they are OK. But I have also swapped the 50MHz OSC for a 40MHz one, so assume 20MHz then.. close enough to 16MHz anyway I guess..

Christ I made a mess of taking out that 030 socket and you've had that off looking like new!

8-)

The new socket made no difference, still exactly the same :(

[Badwolf]

I have tried all these, but I'll freely admit I was going crazy with it by this stage, so I may have missed something obvious.

Maybe it is obvious, maybe these FPUs fail miserably at anything but 16Mhz. I have tried 2 in the Falcon and they are OK. But I have also swapped the 50MHz OSC for a 40MHz one, so assume 20MHz then.. close enough to 16MHz anyway I guess..

I put the FPU that works fine in this board at 25MHz into that one at 18MHz and nada.

The new socket made no difference, still exactly the same :(

There's something wrong with that board, or my assembly of it.

Oh well, sorry about that. It was the reason that board was set aside in the first place, but then I lost another so I thought I'd try to resuscitate it for you to have. Which has mostly worked. :)

BW

[exxos]

FPUCS

IMG_0189.JPG

DSACK0

IMG_0190.JPG

DSACK1

IMG_0191.JPG

[exxos]

There's something wrong with that board, or my assembly of it.

Oh well, sorry about that. It was the reason that board was set aside in the first place, but then I lost another so I thought I'd try to resuscitate it for you to have. Which has mostly worked. :)

I will double check the continuity from the FPU back to the CPU, aside from that im out of ideas :) Not like I'm going to be in a hurry to start using thousands of programs which use the FPU anyway :lol:

[Badwolf]

I will double check the continuity from the FPU back to the CPU, aside from that im out of ideas :) Not like I'm going to be in a hurry to start using thousands of programs which use the FPU anyway :lol:

We have Anders' new soft FPU emulator tool now, anyway. :)

http://www.happydaze.se/wp-content/uplo ... 220717.zip

BW

[dml]

I will double check the continuity from the FPU back to the CPU, aside from that im out of ideas :) Not like I'm going to be in a hurry to start using thousands of programs which use the FPU anyway :lol:

Does the FPU work in the same machine, when installed on the board without the DFB1?

Just wondering if the traces to the old mainboard FPU socket could interfere across the expansion bus with the DFB1 if something is wrong in that area.

(If the FPU works on the mainboard then the mainboard socket/traces are probably not involved)

[exxos]

Does the FPU work in the same machine, when installed on the board without the DFB1?

Yep. Tried 2 different FPUs as well.


@Badwolf Everything continuity checked fine.

The only thing I could really find is there seems to be a lot of undershoot on some of the data lines. As to that could possibly cause the problem I don't know. I'm really out of ideas now.

[Badwolf]

The only thing I could really find is there seems to be a lot of undershoot on some of the data lines. As to that could possibly cause the problem I don't know. I'm really out of ideas now.

Yeah. I couldn’t find anything wrong with that board either, but it doesnae work.

Colour me stumped.

Put it down as a demon board.

Hmm, just a thought, is the CPU seeing dsack when the FPU is ssserting, I wonder? Perhaps the CPLD has a failure and can’t read from one of the FPU dsack pins. It would therefore not assert to the CPU correctly.

Perhaps solder a wire to the bottom of the board for the dsack0 and 1 pins at the cpu and have cpu dsack0 on one channel with FPu dsack0 on the other? Same for dsack1?

BW