Granlunds FrankenSTein
Re: Granlunds FrankenSTein
Thank you for that!terriblefire wrote: ↑Sun Jun 28, 2020 7:09 pm Try this for a starter for 10.
tf536r2_main_top_ATARI_FULL_PHASE_7_28_6_2020.zip
It boots to desktop no magic smoke has escaped (so far!)
I have no way of connecting a CF adapter at this moment so I'll have to build some temporary solution until I've made a relocator (In the end, I think that CF card is going to fit very nicely above the cartridge port)
Gotek appears to work fine though
The machine reset when loading Gembench if fastram is registered so my first assumption is going to be that I have a bad solder or a bridge somewhere. I'll check what Diagrom says tomorrow.
Anyways, I'm happy with the progress I made so far!
I believe you're already aware of EmuTOS and Maprom detecting 68MB fastram. This will probably make TOS eventually hand out a pointer to memory that doesn't exist but you'd have to have used up 64MB first so I recon it's going to be a while until that becomes an issue
-
- Moderator Team
- Posts: 5389
- Joined: Mon Aug 28, 2017 10:56 pm
- Location: Glasgow, UK
Re: Granlunds FrankenSTein
Its highly possible i have fastmem mapped wrong..
———
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
Re: Granlunds FrankenSTein
I believe that may be the case.
Had a peek at the contents from 0x05000000+ and it returns the same as when reading from 0x00000000+ so it looks like it wraps to physical address 0 after the fastram range.
This may explain why it think it sees 68MB fastram. After wrapping to 0, it'll have 4MB of address space which are legal reads until it hits the bus error to make it stop probing for memory.
-
- Moderator Team
- Posts: 5389
- Joined: Mon Aug 28, 2017 10:56 pm
- Location: Glasgow, UK
Re: Granlunds FrankenSTein
We want a bus error at 0x05000000?agranlund wrote: ↑Tue Jun 30, 2020 3:59 pmI believe that may be the case.
Had a peek at the contents from 0x05000000+ and it returns the same as when reading from 0x00000000+ so it looks like it wraps to physical address 0 after the fastram range.
This may explain why it think it sees 68MB fastram. After wrapping to 0, it'll have 4MB of address space which are legal reads until it hits the bus error to make it stop probing for memory.
———
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
Re: Granlunds FrankenSTein
Yep!terriblefire wrote: ↑Tue Jun 30, 2020 4:06 pmWe want a bus error at 0x05000000?agranlund wrote: ↑Tue Jun 30, 2020 3:59 pm
I believe that may be the case.
Had a peek at the contents from 0x05000000+ and it returns the same as when reading from 0x00000000+ so it looks like it wraps to physical address 0 after the fastram range.
This may explain why it think it sees 68MB fastram. After wrapping to 0, it'll have 4MB of address space which are legal reads until it hits the bus error to make it stop probing for memory.
Re: Granlunds FrankenSTein
Had a moment to spend with the TF536.
It's a hassle with just the Gotek that for sure
Here's some interesting stuff!
Gembench yields 249% on integer division with the TF536, compared to 942% on the TF534 in the same scenario
(Gembench in ST-RAM, EmuTos in FastRam - same CPU used on both cards)
I wonder if I ordered/received the wrong oscillator?
Or does this result signal something obvious to someone here?
Loading Gembench to fastram resets the machine sometime during loading. But for now, I think I'll wait for parts to arrive so I can hook up a CF card and continue testing with that. I'm getting less and less convinced the floppy is 100% on my machine no matter what, it was ages since I last used the Gotek, and I want to try and narrow down the potential causes for error.
Me having screwed up building the card is still the most likely scenario so I'm thinking I might ask if I can purchase a built and tested card from one of the builders so I can at least rule that out.
It could also be some oddities with the old style ST motherboard causing problems (or just my machine in general, it's not exactly like stock anymore..)
Worst case, the ST keeps the TF534 and the TF536 goes in another machine. Still pretty good for a worst case scenario
It's a hassle with just the Gotek that for sure
Here's some interesting stuff!
Gembench yields 249% on integer division with the TF536, compared to 942% on the TF534 in the same scenario
(Gembench in ST-RAM, EmuTos in FastRam - same CPU used on both cards)
I wonder if I ordered/received the wrong oscillator?
Or does this result signal something obvious to someone here?
Loading Gembench to fastram resets the machine sometime during loading. But for now, I think I'll wait for parts to arrive so I can hook up a CF card and continue testing with that. I'm getting less and less convinced the floppy is 100% on my machine no matter what, it was ages since I last used the Gotek, and I want to try and narrow down the potential causes for error.
Me having screwed up building the card is still the most likely scenario so I'm thinking I might ask if I can purchase a built and tested card from one of the builders so I can at least rule that out.
It could also be some oddities with the old style ST motherboard causing problems (or just my machine in general, it's not exactly like stock anymore..)
Worst case, the ST keeps the TF534 and the TF536 goes in another machine. Still pretty good for a worst case scenario
- Umberto1st
- Posts: 105
- Joined: Wed Sep 27, 2017 6:27 pm
Re: Granlunds FrankenSTein
I'm getting around 940% in TOS 2.06 and Emutos 0.9.12. Have not tried to run from fast ram. I having other problems, such as no IDE in TOS or Emutos. I was in the understanding that Emutos does not require an interrupt line, but TOS 2.06 does. I am also getting a bizarre glitch with my mouse pointer. It will not work unless I have my VOFA card installed. Lots of little problems to sort out
3)1040 stf/m, 2)Mega 4 ST, 1)STE, 1)Mega STE, 1)STacy, 2)Falcon 030
Re: Granlunds FrankenSTein
@agranlund as @Umberto1st says I get 940% with gembench running from ST ram. It sounds like you've got a few build errors as I've noticed almost a slowmo mode when id built something wrong one day.
I've also found that certain cheap CF cards won't work with th TF536 but I did manage to locate some cheap 4gb Transcend cards which work great.
Hope that helps.
I've also found that certain cheap CF cards won't work with th TF536 but I did manage to locate some cheap 4gb Transcend cards which work great.
Hope that helps.
If it ain't broke, test it to Destruction.
Re: Granlunds FrankenSTein
Thanks PhilC!PhilC wrote: ↑Wed Jul 01, 2020 8:46 am @agranlund as @Umberto1st says I get 940% with gembench running from ST ram. It sounds like you've got a few build errors as I've noticed almost a slowmo mode when id built something wrong one day.
I've also found that certain cheap CF cards won't work with th TF536 but I did manage to locate some cheap 4gb Transcend cards which work great.
Hope that helps.
Yeah I was almost certain it's on me and my building of the card and your info helps to further cement that
I'll probably have to desolder the CPLD and redo that one, with the same or order another one.
I did manage to clumsily bend a few of its pins and had to straighten them before soldering it on the first time, and well, I'm almost certain this is what's haunting me now. Although it appears fine with a photography loupe it doesn't mean a microscope wouldn't disagree.
All out of flux and solder wick at the moment so it'll have to wait until that stuff arrives.
I might be interested in buying one of your built-and-tested ones if you still have one available