Atarian Computing's Raven Build

[PhilC]

@Atarian Computing what speed oscillator are you using? Try and use a 40mhz if you can. It maybe the ,68150 don't like being overclocked?

[Atarian Computing]

Here's what I got with the jumpers at 1x with regular TOS ROM:

20251020_190016060.JPG (281.91 KiB) Viewed 180 times

Just for fun I thought I'd put the jumpers back to 2x:

20251020_190449065.JPG (352.81 KiB) Viewed 180 times

Wow!

That was short lived though. After a few resets and shutdowns I was left with a blank output and this cycled endlessly through this:

Code: Select all

RAVEN060

CPU:   M68060R6
SIMM0: 01000000
SIMM1: 00000000
SIMM2: 00000000
SIMM3: 00200000

Warmboot
InitHeap
InitUart
InitIkbd
InitMfp
InitMidi
InitI2C
InitRtc
InitCfg
InitVbr
InitMonitor
IkbdConnect
Unknown IKBD
InitAtari
InitVbr
InitMmu
InitEmu
InitTos
InitVga

RAVEN060

CPU:   M68060R6
SIMM0: 01000000
SIMM1: 00000000
SIMM2: 00000000
SIMM3: 00200000

Warmboot
InitHeap
InitUart
InitIkbd
InitMfp
InitMidi
InitI2C
InitRtc
InitCfg
InitVbr
InitMonitor
IkbdConnect

I still don't have an eiffel or CKBD btw.

I honestly am at a loss... Here's a summary and the steps I took:

Without a video card and jumpers at 2x the serial output just cycled through RAVEN060

Without a video card and jumpers at 2x, WITHOUT the 68510, the serial output just cycled through RAVEN060

Without a video card and jumpers at 1x, the serial output using ROM_MON was normal. Output on my previous post.

WITH a video card and jumpers at 1x, using normal TOS ROM, EmuTOS loaded to desktop but output striped

WITH a video card and jumpers at 2x, using normal TOS ROM, EmuTOS loaded to desktop and output perfect (this didn't last long)

Am I to conclude that it's still a possible shit 68150? Even though it was from a reputable seller and a 40Mhz part?

What should my next steps be? Get some slower oscillators?

@PhilC I got your message just as I was about to press submit. Ok, I'll put an order for some oscillators then as I only have the 48Mhz.

[Atarian Computing]

@agranlund Could any of my issues be Nessi fw related? Are you able to pinpoint the issue based on my observations?

[agranlund]

Without a video card and jumpers at 2x the serial output just cycled through RAVEN060
Without a video card and jumpers at 1x, the serial output using ROM_MON was normal. Output on my previous post.

I think this is what needs to get resolved first, but I'm not yet sure what my theory is.
Under normal circumstances I believe there should be at least a few more debug prints before anything goes through the 68150 so at the moment I would probably not focus my attention on that one.

Video output striped (and possibly other issues) is normal when in 1x mode. Fixing 1x mode was put on hold a while back and I haven't gotten around to revisiting it again.

It does sound like the common denominator is that stuff goes bad at 2x but works (as well as can be expected) when jumpered for 1x.

There's not _that_ much stuff happening between the RAVEN060 printout and the next expected print.
But it might be an idea here to put some additional debug outputs for a better guess on what goes wrong.

I think the most exciting thing that happens before the next serial output is that it starts actually using RAM for the first time so my mind goes toward that but you mentioned having tested more than one of those simms. As well as they've been proven to work in 1x mode at least.

Is U105 all ok? Getting 96mhz on the top-most pin of J102 and 48mhz on the bottom-most pin?
Wondering if you've maybe ended up in a situation where the atf1508 is being told it's 2x mode but the cpu is getting a 1x clock?
Or something of that sort..

I still don't have an eiffel or CKBD btw.

That should be ok too I think. I tried booting mine without one just now to verify and that appeared to work.

[Atarian Computing]

Is U105 all ok? Getting 96mhz on the top-most pin of J102 and 48mhz on the bottom-most pin?
Wondering if you've maybe ended up in a situation where the atf1508 is being told it's 2x mode but the cpu is getting a 1x clock?
Or something of that sort..

I still don't have an eiffel or CKBD btw.

That should be ok too I think. I tried booting mine without one just now to verify and that appeared to work.

Thanks for this! I will probe the clocks and check U105 as you suggested tomorrow as I'm tied with work all day today.

The one thing bugs me about this issue is the seeming intermitency. I was able to boot that one brief time into TOS multiple times at 2x clock speed aka standard jumper configuration. That was the only time the CPU got warm. I'll see if it's a thermal issue as well. I'll get back on Wednesday.

[Atarian Computing]

So this is weird. Since my last post I hadn't touched the Raven in any way. I come back today, I put the jumpers to x2 to check the clocks and turned on the screen just in case. It booted up fine.

I measured clocks, they were fine. I'm like, I gotta try to use this. Finally flashed the PIC and connected a PS2 mouse and kbd. Everything works.

What about PARCP? Friggin' works. @agranlund you're a genious for including it!! Thank you!!!

What about the IDE and byteswap adapter and HDDRIVER? Everything works. Partition complete, HDDRIVER installed.

I'm over the moon but still a nagging feeling of the issues I had before.

I'm running TOS debug version so everything is slow but running for hours now. Both RAM sticks work.

I will now flash regular TOS and play around.

[Atarian Computing]

OMG this thing is fast with regular ROM. I will gradually move stuff over and do testing.

[Atarian Computing]

YES! Perfect scaling like I hoped!

20251022_185458417_preview.jpeg (95.12 KiB) Viewed 114 times

[PhilC]

Well done getting it working.

[Atarian Computing]

Well done getting it working.

Thank you! Hopefully no gremlins come to haunt me down the road. Rock solid so far.