derkom's attempt

[PhilC]

It certainly does @Icky

[derkom]

My solution to the pesky DMA resistor...

H4-007.jpeg

And...

H4-008.jpeg

It is done. Well, it is assembled. Done comes when it boots. :)

...which is it not doing. After checking to make sure VCC is everywhere it should be and nowhere it shouldn't be, I hooked it up and powered it on, and here's what I get:

H4-009.jpg

At least there is video! But sadly, nothing else follows. It is not running the diagnostic cartridge (no output even on serial). I probed a few things, and the CPU and GLUE are both getting 8 MHz clock just fine (can't check MMU with Falcon RAM installed). HALT looks rather curious though (should be logic high all the time I believe):

H4-010.jpg

I'll do more poking around tomorrow. Suggestions for things to look at appreciated. I do have a scope and a logic analyser and about half a brain.

Despite not getting a glorious green desktop, I am pleased to have gotten what I have gotten. :)

[Icky]

@derkom - usually the first issue is down to some solder problem. Have a look at the smt chips / components and make sure there are no hair line solder bridges. Also missing joints on the rear of the board.

[derkom]

@derkom - usually the first issue is down to some solder problem. Have a look at the smt chips / components and make sure there are no hair line solder bridges. Also missing joints on the rear of the board.

I inspected everything with a microscope earlier today, but will give it another go.

[exxos]

Can you upload some close up shots around the ROM etc so we can see whats going on better.

[derkom]

I shall get some photos to post a little bit later, but wanted to post these findings first.

I hooked up the logic analyser and read Halt, Reset, Berr, Clock, and A23, and here's what a power on, followed by power off looks like:

logic-poweron.png

Sorry for not getting the scale in my screenshot. The whole time power on to power off is 10-15 seconds there. (And yes, if I zoom in on the data, the clock is a good 8 MHz.)

And here's what it looks like when I press the reset button:

lgoic-reset.png

I've tested GLUE, MMU, Shifter, and DMA in a working ST and verified they all work. Also swapped in a different 68000, with no change.

The only chip I cannot test is the TOS ROM, which I pulled from the Alpha board given to me by Atarian Computing. Since I never got that board running, I cannot swear that that's actually a good ROM, although I would tend to imagine it is. I do not have the necessary hardware to look at the contents of that chip anywhere else. I do, however, have @exxos's TOS switcher board in my STe. What would happen if I were to stick that TOS chip into the H4? I wouldn't expect it to run STe TOS, but maybe it's enough to see if the H4 at least tries to execute it? Maybe it would be enough to make it run the diagnostic cartridge?

[exxos]

The STE ROM's won't run in the H4.

You need to check ROM_CE on the jumper above the SMT GLUE and see if that pulses low after reset goes high. Also CPU /AS should pulse low about the same time, and /DTACK should go low shortly after as well.

[derkom]

The STE ROM's won't run in the H4.

Certainly didn't expect it to run, but thought perhaps it's still a potential diagnostic tool.

You need to check ROM_CE on the jumper above the SMT GLUE and see if that pulses low after reset goes high. Also CPU /AS should pulse low about the same time, and /DTACK should go low shortly after as well.

Okay, will check that out after I finish verifying all the data and address lines. And that will have to wait a little while, because people around here are demanding to be fed. Don't they know I'm working on something important? :mrgreen:

[derkom]

Okay. Sorry for the text on top of the image. The software does that any time it loses a sample anywhere, and there seems to be no way to make it go away.

Here's what I'm seeing ...

logic2.png

Certainly looks like things are happening. Incidentally, I realised that the captures I posted earlier were made without RAM in the machine. Oops. :oops:

Should I install the floppy connector and see if it's trying to do anything there? I didn't install it yet because if I was going to need to do more hot air work near the DAC, it would need to come off. With a keyboard attached, there has been no floppy LED activity, which leads me to believe it's not looking for a floppy drive at all.

I'm guessing that since I see a white screen, the DAC is working? But that may not imply the data selector/multiplexers are working? Just kind of guessing here.

Another thought that worries me is that it actually is booting the diagnostic cart, but something is also wrong with serial. Just throwing out ideas here, since I only halfway (or less) know what I'm doing. :mrgreen:

[exxos]

Looks like the CPU is trying to access something, but it can't and GLUE issues a bus error,

I would start by resoldering the GLUE & MFP, MMU pins.. maybe even the CPU pins.

EDIT:

Without a proper image, I can't tell if you have the correct jumpers in either. You need the jumpers in place if not using the blitter for example.