So far...
The Radeon is working on two of them (including the first one).
The others two have some fault-finding to do. On both the CPLD programs ok and the CTPCI is detected on the boot screen, along with the PCI bridge.
A few notes on the 'faults-to-find' pair:
i) Radeon detected, starts to run BIOS then reboots. I guess it reads bad data from the Radeon BIOS chip and thus executes garbage. Might try to read the BIOS chip from a working board and this one to see which bits are different.
ii) Radeon detected, runs BIOS then gets to the point 'press a key'. After I press a key it tries to set video mode and seems to hang. I get no screen on the Radeon.
On board (i) I found a few bad connections on the PCI1 header. I suspect that still, it was hard to hand solder the 'inner row' in-between the two headers.
REMINDER - Stay logged in for at least 2 hours a week to get whitelisted.
Also it helps build a picture where our "good traffic" is coming from for detection scripts.
:o)
Also it helps build a picture where our "good traffic" is coming from for detection scripts.
:o)
Let's build a CTPCI!
-
foft
- Posts: 346
- Joined: 28 Mar 2022 12:20
Re: Let's build a CTPCI!
So I wrote a little program to read the first bytes of the BIOS from tos to investigate what is wrong with the other two boards.
It looks like a stuck data line on two of the boards:
i) One board seems to always have 0x2000 or'ed with each 32-bit read.
ii) The other board claims to have the bios at 0x50020000, when it should be at 0x60020000
I've attached the test program, its super basic but might help others.
Hopefully from here I just need to trace the stuck line(s) from this test and then they'll all work. Fingers crossed!
Good board:
Board i:
Board ii:
It looks like a stuck data line on two of the boards:
i) One board seems to always have 0x2000 or'ed with each 32-bit read.
ii) The other board claims to have the bios at 0x50020000, when it should be at 0x60020000
I've attached the test program, its super basic but might help others.
Hopefully from here I just need to trace the stuck line(s) from this test and then they'll all work. Fingers crossed!
Good board:
Board i:
Board ii:
You do not have the required permissions to view the files attached to this post.
-
foft
- Posts: 346
- Joined: 28 Mar 2022 12:20
Re: Let's build a CTPCI!
So... on the board with 0x2000 stuck. I checked AD13 (pin 32 of PCI1) and it is connected fine. I re-soldered it just in case though! No dice though.
So I realized that the macro I copied BIOS_IN32 to read 32-bits actually does 4 8 bit reads. Now I guess that under the hood it must be doing a 32-bit read since otherwise, why would it be 1 bit out of 32 that seems broken, rather than 1 bit out of 8?! Perhaps PCI does the 8 bit reads in the respective 8 bits of the bus, I'd better check my book...
So I realized that the macro I copied BIOS_IN32 to read 32-bits actually does 4 8 bit reads. Now I guess that under the hood it must be doing a 32-bit read since otherwise, why would it be 1 bit out of 32 that seems broken, rather than 1 bit out of 8?! Perhaps PCI does the 8 bit reads in the respective 8 bits of the bus, I'd better check my book...
-
foft
- Posts: 346
- Joined: 28 Mar 2022 12:20
Re: Let's build a CTPCI!
Done some beeping out on data lines. Found it, seemed to be a bad connection on R10. (D21 to CPLD).
Now 3 boards are working :) . Last one to go, then onto the ripped pad one!
The bad connection:
The pin -> D21 -> endian stuff which is why its not D13!
Now 3 boards are working :) . Last one to go, then onto the ripped pad one!
The bad connection:
The pin -> D21 -> endian stuff which is why its not D13!
You do not have the required permissions to view the files attached to this post.
-
foft
- Posts: 346
- Joined: 28 Mar 2022 12:20
Re: Let's build a CTPCI!
And the fourth one is now working! I found a short on R40. Between A30 and A31 I think -> surprised it booted!
:yay:
:yay:
-
HigashiJun
- Posts: 2387
- Joined: 19 Jun 2020 07:21
- Location: Tokyo
Re: Let's build a CTPCI!
Glad you managed to sort that out !
:D
:D
-
foft
- Posts: 346
- Joined: 28 Mar 2022 12:20
Re: Let's build a CTPCI!
I'd like to check the IDE port on these. To use it do I just do this, or is there more to it:
i) Connect IRQ wire
ii) Move CF from Falcon IDE to the CTPCI IDE? I guess I need a power + 40 pin IDE -> 44 pin IDE adaptor...
iii) Boot
Or is there more to it?
Also I bought an RTL8139, but it doesn't even appear on the boot menu. I guess it's a broken card so I just bought a few more to try: 8139C and another 8139D. Though I was wondering, perhaps it didn't appear since I need to mod it? Are there any known fixes to make some RTL8139 cards appear?
i) Connect IRQ wire
ii) Move CF from Falcon IDE to the CTPCI IDE? I guess I need a power + 40 pin IDE -> 44 pin IDE adaptor...
iii) Boot
Or is there more to it?
Also I bought an RTL8139, but it doesn't even appear on the boot menu. I guess it's a broken card so I just bought a few more to try: 8139C and another 8139D. Though I was wondering, perhaps it didn't appear since I need to mod it? Are there any known fixes to make some RTL8139 cards appear?
-
foft
- Posts: 346
- Joined: 28 Mar 2022 12:20
Re: Let's build a CTPCI!
Today I received:
i) An IDE 40-44 pin adaptor
ii) The second RTL8139
Mixed results!
This RTL8139 shows on the boot menu, so I guess the previous one was indeed broken, or somehow otherwise incompatible. I need to work out how to get it up and running.
The IDE adaptor seems to recognise that I have the compact flash connected, but will not boot from it. I changed the CPX to select CTPCI IDE and also connected the IRQ wire. The problem I have now is that I can't deselect CTPCI IDE mode since with the CTPCI removed it vanishes from the CPX options and with it installed it won't boot!
i) An IDE 40-44 pin adaptor
ii) The second RTL8139
Mixed results!
This RTL8139 shows on the boot menu, so I guess the previous one was indeed broken, or somehow otherwise incompatible. I need to work out how to get it up and running.
The IDE adaptor seems to recognise that I have the compact flash connected, but will not boot from it. I changed the CPX to select CTPCI IDE and also connected the IRQ wire. The problem I have now is that I can't deselect CTPCI IDE mode since with the CTPCI removed it vanishes from the CPX options and with it installed it won't boot!
-
HigashiJun
- Posts: 2387
- Joined: 19 Jun 2020 07:21
- Location: Tokyo
Re: Let's build a CTPCI!
IIRC, Didier Mequignon uses a RTL8139D in his setting.
-
HigashiJun
- Posts: 2387
- Joined: 19 Jun 2020 07:21
- Location: Tokyo
Re: Let's build a CTPCI!
Isn't it a question of firmware ?foft wrote: 23 May 2023 21:32 The IDE adaptor seems to recognise that I have the compact flash connected, but will not boot from it. I changed the CPX to select CTPCI IDE and also connected the IRQ wire. The problem I have now is that I can't deselect CTPCI IDE mode since with the CTPCI removed it vanishes from the CPX options and with it installed it won't boot!
I think only the last firmware lets you make use of the CTPCI IDE, but doesn't allow you to use the Radeon...
Who is online
Users browsing this forum: britelite, CCBot and 4 guests