dml attempts a Raven

[agranlund]

That adapter is for the opposite direction. 060->040

If you go for LC or EC 060 parts they are relatively cheap but I don't know if Raven needs FPU or MMU?

It needs an MMU but not an FPU. As a gor instance, Agranlund had made a full 060 and LC version of Opentyrian and my first 060 that I used to get the Raven up and running on was an LC.

There is also an FPU emulator included in the release package, it'll will work in pinch unless you want to do things that only use a little bit of FPU, or not at all but is compiled to require one.


Someone could most likely make an 060->040 adapter but I'm going to guess it's less trivial than going the other way?
The 060 is configured at startup to select the 060 way of doing things where they differ, since Raven was designed with only 060-native operation in mind (as opposed to making it essentially an 040 design and with the 060 configured for such)

You might need changes to the cpld firmware, cannot remember off the top of my head if there was something there too that talks 060-native instead of 060-in-040-backwards-compatible-mode.
I don't remember the details, I just picked the 060-native way wherever the datasheet told of the differences in case you wanted to use the new fancy cpu in an older motherboard design. Perhaps it wasn't that much, but I would pick an 060LC over that hassle :)

The clock signals are different so you'd need to provide two clocks to the 040 instead of clock + enable as the 060 expects, but that could be on an adapter board too for sure or with bodge wire from the pll on the board perhaps.

[dml]

I somehow managed to scrape together all the parts for the main board and ram cards. Mouser was a last resort for a few weird parts that seemed to be not to be able to find myself and I think one or more of those are discontinued.... Annoyingly, I had a lot of 805 parts but very few 1206, except for a giant roll of 10k/1% I have here but at least I had that :)

I already had the CPU+socket, YM, pinheaders, PLCC sockets etc. etc. and some of the more common components. I might be missing a DIP socket size somewhere in there - maybe for the YM chip.

So the only excuse I have left now is the 'free time' thing which is still a problem but I'll start in small steps here and there and see how things go...

Probably I should do a bit of reading/searching here to see if there's an optimal order of assembly to test-early-as-we-go, or just assume its going to work and do it in smallest-and-most-annoying-components-first order.

[agranlund]

When I built my first one I did it small incremental stages testing each stage.
But back then the code was being written at the same time so test-cases came naturally.
You can still do some stuff in steps, probably, but the bootcode will quite early on expect most hardware to be there and work.

For the second one I built the whole thing up at once, exactly like you suggested with doing the smallest and most annoying components first.
I did buzz all adjacent pins for shorts with my teeny tiny probes after soldering each of the small smd components though, especially so with the buffers next to the 68150 socket. I also verified everything power related before installing the CPU.
This is probably the best way to go about it, and if something isn't working then it can be fun-ish to troubleshoot from there.

The bootmonitor can be of some help since it lets you read/write to any address and you can measure and/or draw some conclusions from there.
Access to logic analyzer or even just anything to measuring voltage on relevant pins can also be super useful.

[dml]

Thanks for those tips - looking forward to making a start on it :)

I don't have a logic analyzer but I do have everything else I should need, except maybe to flash the main IC. Plenty to do before that is on the horizon though.

When the time comes, I'll check the CPU socket for GND/VCC everywhere and shorts, then test first with an LC060 until I'm sure there are no real problems.

I have some ISA video cards, ET4000, Mach32 etc. for later although it will be a bit of a dice roll to find out if any of them are usable with this.

[PhilC]

@dml try with an et4000 first, they are a more forgiving card than the mach32.

[dml]

(Thanks for ET4k tip, PhilC)

Ok so I decided to stop procrastinating and made a start on it so I won't have an excuse to keep delaying progress :p

Started with problem child #1 and #2 which turned out to be more tricky than they looked. Wicking up the legs rather than sticking to the board etc. anyway they're on the board so it's officially started.

Hopefully the tiny extended white lines indicate pin #1. I can't imagine it means anything else!

r1.jpg

r2.jpg

r3.jpg

[PhilC]

@dml yes the extended line is pin 1.

You made the right choice getting those two chips out the way..do the two max232 chips early as well, as they are also a pain to do when other stuff is in the way. Oh and U303, small but annoying as the pin 1 indicator on the chip is non existent almost.

[dml]

@dml yes the extended line is pin 1.

\o/

Guess twice, cut once! :p

You made the right choice getting those two chips out the way..do the two max232 chips early as well,

Great - because I just did those two as well!

as they are also a pain to do when other stuff is in the way. Oh and U303, small but annoying as the pin 1 indicator on the chip is non existent almost.

I'll look at that next - although I will probably have to continue building at the weekend.

D.

[dml]

...got a bit more done this evening.

I have a question though - what is the correct orientation for this 24MHz oscillator?

I can probably work it out from the ground and the datasheet but I think I'll just ask because after checking 7 different things I'll still get it backwards :)

20250606_200525.jpg

[exxos]

I can probably work it out from the ground and the datasheet but I think I'll just ask because after checking 7 different things I'll still get it backwards :)

:lol: That's about my level of confidence in things! :fire: :hide: :fire: