DFB1 Support thread

Discussion and support for the DSTB1 & DFB1 boosters by BadWolf..
markus0321
Posts: 144
Joined: Sat Dec 19, 2020 11:42 am
Location: Zielona Gora

Re: DoG's DFB1 Build

Post by markus0321 »

Badwolf wrote: Wed Oct 19, 2022 7:19 pm
markus0321 wrote: Tue Oct 18, 2022 11:11 pm There is also an improvement with a soft reset, because it works when I use flash, but with EmuTOS, when I turn off flash and TOS4.04 is working from the motherboard, unfortunately, after a soft reset, TT-RAM still disappears and just like I upload TOS4.04 to Flashrom it also disappears: - (
A soft reset doesn't work with TOS4.04 hm ..
If you're using MAPROM, then I think this is a bug in the latest version -- it only works after a cold boot. That was an old bug that was fixed and appears to have come back.

Try v2.2 (attached).

BW
Thank you so much. Now it works great! :-)
What I had before was a completely different maprom program.
I downloaded it from here:
https://github.com/agranlund/tftools
I think you need to fix it in the official Viki to DFB1 or attach the new one?
Now works fine and speeds up the ROM even more.
Works fine with TOS4.04 and EmuTOS :-)
User avatar
Badwolf
Posts: 2231
Joined: Tue Nov 19, 2019 12:09 pm

Re: DoG's DFB1 Build

Post by Badwolf »

Badwolf wrote: Wed Oct 19, 2022 12:58 pm However the fact you're getting *half* an Atari logo sounds like bus conflict. I'll try to scope my ROM chip enable line either tonight or tomorrow.
I've scoped my ROM2 line -- no bus conflict evident. Once EmuTOS is booted from the flash, or once NVDI loads from TOS4, no further onboard ROM activity is seen.

I think I'm going to release this firmware officially.

BW
DFB1 Open source 50MHz 030 and TT-RAM accelerator for the Falcon
DSTB1 Open source 16Mhz 68k and AltRAM accelerator for the ST
Smalliermouse ST-optimised USB mouse adapter based on SmallyMouse2
FrontBench The Frontier: Elite 2 intro as a benchmark
markus0321
Posts: 144
Joined: Sat Dec 19, 2020 11:42 am
Location: Zielona Gora

Re: DoG's DFB1 Build

Post by markus0321 »

Badwolf wrote: Wed Oct 19, 2022 10:23 pm
Badwolf wrote: Wed Oct 19, 2022 12:58 pm However the fact you're getting *half* an Atari logo sounds like bus conflict. I'll try to scope my ROM chip enable line either tonight or tomorrow.
I've scoped my ROM2 line -- no bus conflict evident. Once EmuTOS is booted from the flash, or once NVDI loads from TOS4, no further onboard ROM activity is seen.

I think I'm going to release this firmware officially.

BW
I checked gembench and NVDI TOS4.04 graphics performance with agranlund maprom
"https://github.com/agranlund/tftools"
is much bigger. This maprom 2.2 from you works fine, but it's missing something. Maybe the agranlund one could be improved so that TT-RAM does not disappear after the reset?

IMG_20221020_011214.jpg
IMG_20221020_011214.jpg (394.54 KiB) Viewed 993 times
IMG_20221020_010756.jpg
IMG_20221020_010756.jpg (506.44 KiB) Viewed 993 times

Marko
markus0321
Posts: 144
Joined: Sat Dec 19, 2020 11:42 am
Location: Zielona Gora

Re: DoG's DFB1 Build

Post by markus0321 »

markus0321 wrote: Thu Oct 20, 2022 12:23 am
Badwolf wrote: Wed Oct 19, 2022 10:23 pm

I've scoped my ROM2 line -- no bus conflict evident. Once EmuTOS is booted from the flash, or once NVDI loads from TOS4, no further onboard ROM activity is seen.

I think I'm going to release this firmware officially.

BW
I checked gembench and NVDI TOS4.04 graphics performance with agranlund maprom
"https://github.com/agranlund/tftools"
is much bigger. This maprom 2.2 from you works fine, but it's missing something. Maybe the agranlund one could be improved so that TT-RAM does not disappear after the reset?
Marko
Sorry for the error, I am still learning about Falcon's operating system and forgot to set the boot order in AUTO. Maprom must be first, before NVDI and then the performance is at a similar level as before ;-)
User avatar
Badwolf
Posts: 2231
Joined: Tue Nov 19, 2019 12:09 pm

Re: DoG's DFB1 Build

Post by Badwolf »

markus0321 wrote: Thu Oct 20, 2022 12:23 am I checked gembench and NVDI TOS4.04 graphics performance with agranlund maprom
"https://github.com/agranlund/tftools"
is much bigger. This maprom 2.2 from you works fine, but it's missing something. Maybe the agranlund one could be improved so that TT-RAM does not disappear after the reset?
They are both Anders' program. I've just given you an older version that doesn't have this bug. When Anders has time I'm sure he'll have a look at what's going on for us. I've raised an issue on his GitHub page.

Of course, if you've any experience with software development, since it's open source, you're welcome to take a look yourself.

BW
DFB1 Open source 50MHz 030 and TT-RAM accelerator for the Falcon
DSTB1 Open source 16Mhz 68k and AltRAM accelerator for the ST
Smalliermouse ST-optimised USB mouse adapter based on SmallyMouse2
FrontBench The Frontier: Elite 2 intro as a benchmark
markus0321
Posts: 144
Joined: Sat Dec 19, 2020 11:42 am
Location: Zielona Gora

Re: DoG's DFB1 Build

Post by markus0321 »

Badwolf wrote: Thu Oct 20, 2022 10:14 am
markus0321 wrote: Thu Oct 20, 2022 12:23 am I checked gembench and NVDI TOS4.04 graphics performance with agranlund maprom
"https://github.com/agranlund/tftools"
is much bigger. This maprom 2.2 from you works fine, but it's missing something. Maybe the agranlund one could be improved so that TT-RAM does not disappear after the reset?
They are both Anders' program. I've just given you an older version that doesn't have this bug. When Anders has time I'm sure he'll have a look at what's going on for us. I've raised an issue on his GitHub page.

Of course, if you've any experience with software development, since it's open source, you're welcome to take a look yourself.

BW
Unfortunately, I'm not good at coding :-( I'm better at soldering and repairing the hardware.
Thanks for the useful information, I'll be checking from time to time if there is any software fix for DFB1.
This is a really great card for Falcon and it's very nice to build. :-)
markus0321
Posts: 144
Joined: Sat Dec 19, 2020 11:42 am
Location: Zielona Gora

Re: DoG's DFB1 Build

Post by markus0321 »

Badwolf wrote: Thu Oct 20, 2022 10:14 am
Hello.
I partially assembled the second DFB1. I haven't mounted 245 and RAM buffers yet. In this case, if I have Flash soldered, can it work without the SN74cb3t3245 chip?
markus0321
Posts: 144
Joined: Sat Dec 19, 2020 11:42 am
Location: Zielona Gora

Re: DFB1 Support thread

Post by markus0321 »

I compiled the last version of the nightly EmuTOS with Falcon settings and RAM test enabled.
Maybe it will be useful to someone (for me it works correctly with DFB1, as long as the firmware from this thread is programmed in DFB1).

viewtopic.php?p=92511#p92511
Attachments
Falc_emutos_test_TT-RAM.zip
(5.87 MiB) Downloaded 16 times
User avatar
Badwolf
Posts: 2231
Joined: Tue Nov 19, 2019 12:09 pm

Re: DoG's DFB1 Build

Post by Badwolf »

markus0321 wrote: Thu Oct 27, 2022 10:45 am
Badwolf wrote: Thu Oct 20, 2022 10:14 am
Hello.
I partially assembled the second DFB1. I haven't mounted 245 and RAM buffers yet. In this case, if I have Flash soldered, can it work without the SN74cb3t3245 chip?
Flash should work (it's 5V). TT-RAM won't as these are the level shifters on the SDRAM data lines.

BW
DFB1 Open source 50MHz 030 and TT-RAM accelerator for the Falcon
DSTB1 Open source 16Mhz 68k and AltRAM accelerator for the ST
Smalliermouse ST-optimised USB mouse adapter based on SmallyMouse2
FrontBench The Frontier: Elite 2 intro as a benchmark
markus0321
Posts: 144
Joined: Sat Dec 19, 2020 11:42 am
Location: Zielona Gora

Re: DFB1 Support thread

Post by markus0321 »

I tested the second DFB1 board and detected a problem with data transfer from scsi drives. This manifests itself in the fact that when trying to unpack a zip or lharc archive, errors occur during decompression. This only happens when the TT-RAM memory is active, I now have 2 pieces of DFB1 and the same on each. There are also problems when I tried to run large programs loaded on a SCSI drive or install NVDI on a SCSI drive. The same things work fine when run from the IDE drive. What can be the cause of bad data transfer from SCSI drives?
Post Reply

Return to “DSTB1 & DFB1 booster by BadWolf”