Stuart's DFB1 Build

[sety]

The quick firmware fix would be to recompile with

Code: Select all

assign ROM_A19 = A[19];

changed to

Code: Select all

assign ROM_A19 = 1'b0;

Thanks BW. I went with this option as it seemed easier. But I haven't had a lot of luck yet. Am I right in thinking that I should flash the rom in normal 030 (dfb disable mode) the same way tos is flashed to the ct60?

Thanks!

[Badwolf]

The quick firmware fix would be to recompile with

Code: Select all

assign ROM_A19 = A[19];

changed to

Code: Select all

assign ROM_A19 = 1'b0;

Thanks BW. I went with this option as it seemed easier. But I haven't had a lot of luck yet. Am I right in thinking that I should flash the rom in normal 030 (dfb disable mode) the same way tos is flashed to the ct60?

Thanks!

Hi Sety,

Yep, that's the plan. So does it flash and verify, but still dump incorrectly at F80000?

BW

[sety]

omg I have messed up. I'm so sorry!

In my usual haste I never looked at the code for dfbflash.c

I saw "Verified" or something flash up on the screen and a return to desktop and had assumed all okay. I didn't notice that there was a program exit if the verification wasn't successful. So I added getchar and recompiled...

Screen Shot 2022-05-07 at 7.12.03 pm.png (27.2 KiB) Viewed 2980 times

It turns out that it did not verify and probably never did

IMG_3031.JPG (52.37 KiB) Viewed 2980 times

my bad

[Badwolf]

Aha!

OK, so yeah, the firmware probably isn’t really doing the flash thing properly.

Or the program isn’t compatible with the latest wiring for the chip.

Or I’ve goofed the layout.

Or…

Well, we have something new to work on!

I’ll see if I can salvage a flash chip and solder one on today. Been getting carried away with the STE. The Falcon needs bringing back to the bench again.

BW

[Badwolf]

Hi @sety,

I've reclaimed an old (and very battered!) 512k flash chip. It wouldn't need the A19 change for me, but I made it anyway to make sure that isn't the problem.

I've also found an ancient flashing program on my hard disc that seems to successfully write and verify the output -- maybe give this a go?

At present it's not booting, which I'm investigating, but that could be unrelated to your inability to flash.

Ta,

BW.

[Badwolf]

More updates, Sety.

There's a 'resetblock' variable commented out on the 'lowspeed' flag line. That seem crucial for getting the flash memory booting up. Basically it stops the CPU accelerating until a short time after boot. Not sure why it's needed, but I think Exxos found the same in the past too.

I obviously removed it for simplicity's sake when I wasn't thinking about flash. Putting it back in does the job.

So with flash.ttp as above, and this JED, I think you should be golden.

https://github.com/dh219/DFB/blob/main/ ... 200507.jed

BW

(PS: It looks like I shipped the firmware with OPTION2 stoping acceleration. With that set the flash would boot, which is what reminded me of this acceleration holdoff logic).

Here's EMUTOS ROM image booted:-

IMG_5353.jpeg (123.19 KiB) Viewed 2957 times

IMG_5354.jpeg (237.34 KiB) Viewed 2957 times

[Badwolf]

Typically, today I thought I'd enjoy EmuTOS from ROM, pulled the jumper and was greeted with gibberish again.

FFS.

BW

[sety]

Typically, today I thought I'd enjoy EmuTOS from ROM, pulled the jumper and was greeted with gibberish again.

FFS.

You're doing better than me, F80000 is still blank I'll pull it out and check all my solders again when I get a chance

[Badwolf]

What did FLASH.TTP say? Faliling at verification?

BW

[sety]

I cant fault dfbflash.c no matter how many times I look at it. The erase and write sequence is correct. It should just work.

I'm leaning towards a shitty solder on my behalf.