Page 26 of 51
Re: TF1230 Beta Interest (Open Source)
Posted: 10 Sep 2021 21:56
by 8 Bit Dreams
After flashing new FW both cards came to life on 1D4 board,
KS Screen shows up, but hangs on IDE boot, changed blue Chinese CF card to SanDisk, and it booted to WB, however, it is very unstable, hangs all the time..
Started RAM Test from Amiga Test Kit - well, it crashes...
16313074152912507898226170611338.jpg
So work was done in right direction,
and I see that Stephen already implemented Autodetection for the RAM! Correct amount was detected on 128MB build
16313077461858809397912135432207.jpg
Re: TF1230 Beta Interest (Open Source)
Posted: 10 Sep 2021 22:02
by terriblefire
This is an actual RAM issue.. the MMU issue happens when you're running MMU lib and it causes a lock up.
Re: TF1230 Beta Interest (Open Source)
Posted: 10 Sep 2021 22:05
by 8 Bit Dreams
Hmm,
I'm aware of the MMU Lockups,
no MMULibs are installed on this one,
It just freezes after loading to WB
Noticed that loading to WB from IDE is blazing fast, almost instantly :shock:
Re: TF1230 Beta Interest (Open Source)
Posted: 10 Sep 2021 22:36
by terriblefire
Ok there was a minor refresh issue with ram.. try this
<deleted>
Re: TF1230 Beta Interest (Open Source)
Posted: 10 Sep 2021 23:03
by terriblefire
Summary:
The A1200 1D4 and Rev 2 have different implementations of autovector. Rev2 ties it low all the time and the Rev1 asserts DSACK and AVEC (which the manual says you shouldnt do). I have solved that.
There was a timing violation on the refresh cause by me changing clock speeds. Fixed. I believe thats whats causing the above minor ram errors.
As for the IDE issues... its CF card dependant and i'm betting there is a way to poke the scsi.device into PIO-2 mode.
Re: TF1230 Beta Interest (Open Source)
Posted: 10 Sep 2021 23:41
by kulamario
Ram looks fixed
Still no boot from HD
20210910_233707.jpg
Re: TF1230 Beta Interest (Open Source)
Posted: 10 Sep 2021 23:47
by terriblefire
This card works …
0A365B05-E405-4ABA-BF55-3F7127989468.jpeg
Re: TF1230 Beta Interest (Open Source)
Posted: 11 Sep 2021 00:29
by kulamario
Unfortunately no have this one
Have only this 3
20210911_004701.jpg
Re: TF1230 Beta Interest (Open Source)
Posted: 11 Sep 2021 00:52
by GadgetUK164
Fanntastic work solving that autovec thing Stephen! Who knew that was a difference between boards! Makes me wonder if that's why 8MB wouldn't work on one of those boards I looked at - when testing with that AmigaKit RAM expansion.
Re: TF1230 Beta Interest (Open Source)
Posted: 11 Sep 2021 01:00
by terriblefire
GadgetUK164 wrote: 11 Sep 2021 00:52
Fanntastic work solving that autovec thing Stephen! Who knew that was a difference between boards! Makes me wonder if that's why 8MB wouldn't work on one of those boards I looked at - when testing with that AmigaKit RAM expansion.
Thanks. I had the IDE lockup on both the TF1260 and TF1230 with the cutting edge firmwares. Both happen when you are running the bus at its limit.
What happens is that the same word is read twice by the Amiga. Gayle does everything right and i've even sampled the IDE control likes when it happens. I havent checked if IDEWAIT is there.
I may change tack and have us use the onboard IDE as a stopgap until we can figure it out... and so we can get everything else sorted.