Raven. A homemade Atari-like computer

[agranlund]

Tried it today for the the first time.
So programmed Nessi this worked.
Narnia gets a bit hot.
As I don’t have the correct serial cable I’m not able to see serial output (I thought I had one but it was the male version).
Don’t see any output from the et4k yet or is there no support in this rom version for the et4k and is always a driver needed.
Cf is plugged in with 060 program in the auto folder.

Yeah both of those 22V10's run quite hot.
EmuTOS is really quite good at driving many et4k's in regular mono resolution all by itself.
Setting up the Nova driver for fancy resolutions is a whole different chapter but EmuTOS alone should be good at giving you a display to get started. I have a few et4k's and can test some of my other ones too.


You have jumpers on J102 and J104 connecting their respective topmost two pins, the "x2" position?
The CPU was 060LC right? Not EC (because EC's will not work - pmmu is absolutely necessary)

Do you have disk activity LED action? Either on your cf-adapter or through the frontpanel header in case your adapter don't have a led? I'd expect at least one blink from EmuTOS as it starts up.

If not, let's assume it either hasn't progress from bootcode to EmuTOS or that it has died extremely early in EmuTOS's startup code. I would get rid of all non-crucial things to reduce variables.
EmuTOS will still boot to a desktop even with nothing connected to the IDE port so get rid of the CF adapter.
One RAM in slot0, the ROM in slot3 and the middle ones empty.

Do you see activity at all if you look at some low address line, A2 or something like that?
Doesn't say much other than it's actually running something, but a start anyway.
If you have tiny probe needles you can easily get at those from an empty Simm socket, or maybe even easier on an ISA slot.
Or perhaps the RB3RD signal which should be toggling as it accesses the rom, but you'd need fine probe to get at that on either pin51 on the PLCC or on your populated rom-simm socket.


Without serial cable it's hard to guess at how far it has gotten, if anywhere.
Do you have some LED hooked up to the PowerLed pins on the frontpanel header?
I can make a rom that shuts that LED off from software, and afterward attempt to play a sound on the YM and/or toggle one of its spare GPO pins to at least verify that some of the basics are working. I don't think I'll have a chance at that this evening though :/
I guess you don't need an actual LED connected, measuring on the pin would work too I suppose :)

But I'd say by far the easiest is a serial cable.
Will also come in handy in case we need to see what EmuTOS is up to.
Hmmn, I should add to the makefile to automatically generate a separate rom that has EmuTOS with serial debugging enabled.

The pinouts of the ATF1508 PLCC:
Super handy when going from Kicad -> where is the pin I actually need to put the probe on :)

Screenshot 2024-07-31 at 21.23.45.png

[agranlund]

If you have tiny probe needles

Speaking of, I have to promote these because even though they are expensive I think they are really good!
I can place them on pretty much any fine-pitch pin and they stay there, given I have magnification so I see what I'm actually doing.


I bought a bunch of these because of his video, and as it turned out the company that makes them were only about an hour drive from my house.

Highly recommend, but then again the only thing I have to compare against is the stock stuff that came with my cheap oscilloscope, and the even cheaper logic analyzer.

[Oldskool]

Thanks for the help.
Just ordered the correct (female) serial cable. Hope it will arrive tomorrow and get some more insights in what’s going on.

Cf led does not go on.
Removing simm 1 and 2 does not make a difference.

Tried LC and Full cpu. The full one runs at 100 mhz in an Amiga the LC is not stable @100 in an Amiga but works at 50.
So tried 1x and 2x. Setting does not make a difference (yeah in the temperature).
So we have a clock and voltage.

I however have the feeling that no code is running. Lets see if we have any serial output.

The eiffel part seems to work at least the lights blink on the keyboard.
Pressing reset however does not seem to reset the lights on the keyboard. Which is a bit strange.

[luciodra]

I love this card and really wish I had one. Who could I ask to build me one? Paid means...

[Oldskool]

Just got the serial cable and it poops out some stuff however it seems rubbish to me:

startup.PNG

Pressing reset many times poops out more random stuff but it is responding to reset:

press reset.PNG

Programming the monitor instead of the emutos version is acting a bit different.
It poops out a bit more characters at once. But it is still random stuff:

rom monitor.PNG

Could it be maybe be the 68150. I have two but both are 33 mhz. (I ordered 40 but they send me 33)
Not sure if these are an issue at 48mhz however you also used a 33 I think (tried both).

So in short there is serial output. It reacts to reset. But it's not yet making sense.

[agranlund]

Could it be maybe be the 68150. I have two but both are 33 mhz. (I ordered 40 but they send me 33)
Not sure if these are an issue at 48mhz however you also used a 33 I think (tried both).
So in short there is serial output. It reacts to reset. But it's not yet making sense.

That particular UART is connected directly to the CPU so it won't need to go through the 68150.
Change the baud rate to 750000 and I recon it should be a lot less gibberish-like :)

It's 750000 by default since I used to upload and execute code through it in the very beginning, and then I never changed it.
For what we're using it for at the moment it probably makes sense if it would instead default to 9600..

(The other serial port next to it is a standard Atari serial port connected to the MFP and behind the 68150 but the monitor will use the high-speed one)

[agranlund]

Finally got around to hooking up sounds properly.
At long last, all the audio sources are going out through the same Line-out jack. Much happiness indeed!
I struggled for quite a while, and failed, trying to activate AUX2 on the Ultrasound so in the end I hooked YM2149 to the MED3700 card instead.
It's a bit of a shame because cable management looked a lot nicer that way.
The glass side panel is darkly tinted though so it's not the end of the world.

It's now:
Motherboard -> MED3700 (YM2149)
MED3700 -> Ultrasound (YM2149, OPL3, MIDI)
Ultrasound -> Speakers (YM2149, OPL3, MIDI, GUS)

That MED3700 has Soundblaster compatible PCM too but I doubt I'll be using that with no DMA.
Should really make some kind of mixer registry similar to how your different soundcards would register their controls to Linux?
And then have that registry expose these to Audio XBIOS so that one could use the GSXB mixer program as-is, or something of that sort.

For now, I just used "autoexec-for-tos" to directly poke the MED3700 registers on startup to enable and set Line-in volume for the YM2149 audio. It has Windows Sound System compatible registers so that part is the same as tons of other cards.

IMG_2805.jpg

[Oldskool]

Cool to have all the sound coming from one output.......

That particular UART is connected directly to the CPU so it won't need to go through the 68150.
Change the baud rate to 750000 and I recon it should be a lot less gibberish-like :)

It's 750000 by default since I used to upload and execute code through it in the very beginning, and then I never changed it.
For what we're using it for at the moment it probably makes sense if it would instead default to 9600..

(The other serial port next to it is a standard Atari serial port connected to the MFP and behind the 68150 but the monitor will use the high-speed one)

Indeed I now see this (one mem module installed):

750000.PNG

[agranlund]

Indeed I now see this (one mem module installed):

Hurray!

'h' for a tiny list of commands.

It doesn't allow you to do much other than dump memory regions, read and write to addresses and so on but for debugging and testing it's really quite nice.

What happens if you do "test ym" (without the quotes) ?
This should test if it can reach the YM through the 68150.
Does it lock up?
Or do you get a familiar Atari Bell sound (if you have it connected to some active speakers, or line-in to your real computer or something).
Should be the lower of the two audio jacks, and assuming you've got four jumpers bridging J602 with J603.
But really, as long as that doesn't lock it up I'd be happy.


Even when running EmuTOS, or anything really, you can get to that monitor by pressing the NMI button (unless you have a proper hardware lockup)
Then type x to return to where you came from. Though it's slightly broken at the moment.. I should have had a proper hardware debounce on that switch so now software will have to deal with not re-entering the monitor multiple times :)

[Oldskool]

test ym

It says: ping!
And no crash

Next try the full emutos again?