Page 20 of 28

Re: The TF330 is here

Posted: 16 Dec 2018 18:52
by alenppc
The ram is not getting detected at all on the pre_alpha now. It was working yesterday. So strange. Wonder if there is a cap value issue

Re: The TF330 is here

Posted: 16 Dec 2018 19:03
by terriblefire
alenppc wrote: 16 Dec 2018 18:52 The ram is not getting detected at all on the pre_alpha now. It was working yesterday. So strange. Wonder if there is a cap value issue
Which RAM chips are you using? Are you able to browse the the ram with DiagROM (0x40000000 - 0x44000000)? The ram appears without needing autoconfig'd.

I had something similar and it cleared itself. I cannot explain why.

Re: The TF330 is here

Posted: 16 Dec 2018 19:34
by alenppc
I am using the RAM chips you spec'd, diagrom can't access the ram starting at $40000000. Yet occasionally it boots and works fine... I double checked all the soldering

Re: The TF330 is here

Posted: 16 Dec 2018 20:23
by terriblefire
alenppc wrote: 16 Dec 2018 19:34 I am using the RAM chips you spec'd, diagrom can't access the ram starting at $40000000. Yet occasionally it boots and works fine... I double checked all the soldering
Yes. I saw this for a day yesterday then it started working again. The only thing i can think of is that something is causing the CPLD to reset during the SDRAM initialization cycle. Any ripples on the 3.3V power rail?

Re: The TF330 is here

Posted: 16 Dec 2018 21:07
by alenppc
I took out the 68030RC50 and put the 68EC030RP40 in and it's working again.

Re: The TF330 is here

Posted: 16 Dec 2018 21:15
by terriblefire
alenppc wrote: 16 Dec 2018 21:07 I took out the 68030RC50 and put the 68EC030RP40 in and it's working again.
I have had similar experiences. Different CPU and the RAM works vs doesnt. Its a bit strange.

Re: The TF330 is here

Posted: 16 Dec 2018 21:20
by alenppc
Yeah, definitely something is going on.

Stephen you should add pin1 silkscreen on the IDE connector for rev 2.

Re: The TF330 is here

Posted: 16 Dec 2018 22:10
by alenppc
I am not certain how to reproduce the boot ram issue reliably. It seems that in certain conditions the ram doesn't get initialized? Not sure what causes it.

However there's one bug that's easy to reproduce. Download PeterK's icon.library from Aminet and place it in libs:, then run

LoadResident Libs:icon.library REBOOT

This will cause the system to hang and won't boot until a power cycle.

Re: The TF330 is here

Posted: 16 Dec 2018 22:57
by terriblefire
And so the fun begins. :)
alenppc wrote: 16 Dec 2018 21:20 Stephen you should add pin1 silkscreen on the IDE connector for rev 2.
Done.
alenppc wrote: 16 Dec 2018 22:10 I am not certain how to reproduce the boot ram issue reliably. It seems that in certain conditions the ram doesn't get initialized? Not sure what causes it.
I think its caused by async logic and AS and DS arriving out of order occasionally. I've emailed you a potential fix for this issue. If this is indeed the root cause.

However there's one bug that's easy to reproduce. Download PeterK's icon.library from Aminet and place it in libs:, then run

LoadResident Libs:icon.library REBOOT

This will cause the system to hang and won't boot until a power cycle.
I'll look at that tomorrow. With these kinds of things I want to eliminate the software itself first. I wonder if its reproducible on WinUAE with an 030 & 64Mb of Fast Ram.

Re: The TF330 is here

Posted: 16 Dec 2018 23:11
by alenppc
Now it doesn't boot with any full 68030 (grey screen, tried two of them) but boots fine with EC030. Could it be something related to the MMU?