exxos's Atari, Amiga & retro forum

I've already received my board quite a few weeks ago, and the last missing parts (I hope) came in last weekend.

Today I could finally start this nice project, here are some photos of the process and the current state.
I think it might be quite interesting, as this seems to be the first build shown here with the SMT GLUE and MMU ;)
Also this was the first time I soldered this kind of SMT components (which shows on the MMU, but I'm still pretty sure I have no short).

Most chips seated:

Before installing MMU:

MMU installed:
Before installing GLUE:

GLUE installed:

Current state:
I'm quite happy how it went so far :D

I also have a few questions:

I'm not sure if this is right, this is located on the rear side of the 1.44MB GAL - (it may quite well be that I caused this myself when I slipped with the iron):


Also, I'm not sure if I have set the DAC jumpers here correctly?
And what is the unlabeled connection between the DAC block and the floppy-connector?


Are the connectors below the PSU connection additional power-connections (looks like it, but didn't measure them yet)?

I plan to also install a X2GS and an internal UltraSATAN, because of this I added the US- and S2-connectors, too.
Is it possible to use an internal power connection for running the UltraSATAN?
I currently don't have the ACSI- and Parallel-Ports soldered yet, I plan to have the UltraSATAN accessible through one of these openings - does anyone have experience which such a setup?

Also, what are the connections 1-4 near the S2 MIDI-Synth connector?
And finally: the 10 connections on the right side of the 5V 68k Port #4 - are they useful for something yet?

Right now I'm in the process of cleaning the board from flux, and then I need to search my RAM-box if I have some RAM which fits.
Is it possible to install a SIMM module larger than 4MB size? I know the maximum usable size is 4MB, but would it work at all - I'm not sure if I have such a small module in my collection :lol:

I'll keep you updated with the progress!

Looking good so far, eagerly awaiting your initial power up & test results!

:bravo:

CeRiAl wrote: 05 Sep 2021 19:48 I'm not sure if this is right, this is located on the rear side of the 1.44MB GAL - (it may quite well be that I caused this myself when I slipped with the iron):

Well spotted, that short should not be there!

CeRiAl wrote: 05 Sep 2021 19:48 Also, I'm not sure if I have set the DAC jumpers here correctly?

The default positions are marked with a line between the pins. Jumpers are also listed here https://www.exxosforum.co.uk/forum/viewt ... =89&t=3563

In general if they are not marked on the board they are reserved for future use.

CeRiAl wrote: 05 Sep 2021 19:48 And what is the unlabeled connection between the DAC block and the floppy-connector?

It is for a DC power jack, but I have just not had chance to see if the parts I ordered actually fit it or not yet. You can can used to power things like ultrasatan without a separate power supply etc

CeRiAl wrote: 05 Sep 2021 19:48 Are the connectors below the PSU connection additional power-connections (looks like it, but didn't measure them yet)?

Yes, they are all power.

CeRiAl wrote: 05 Sep 2021 19:48 Is it possible to use an internal power connection for running the UltraSATAN?

As said above, you can use that point, or any of the four pin power points across the board.

CeRiAl wrote: 05 Sep 2021 19:48 Also, what are the connections 1-4 near the S2 MIDI-Synth connector?

As said above, if they are not labelled, they are reserved for future use.

CeRiAl wrote: 05 Sep 2021 19:48 And finally: the 10 connections on the right side of the 5V 68k Port #4 - are they useful for something yet?

Again reserved for future use. Things which are not labelled you should just ignore them otherwise you could do damage to the board if you try and use them for something.

CeRiAl wrote: 05 Sep 2021 19:48 Right now I'm in the process of cleaning the board from flux, and then I need to search my RAM-box if I have some RAM which fits.
Is it possible to install a SIMM module larger than 4MB size? I know the maximum usable size is 4MB, but would it work at all - I'm not sure if I have such a small module in my collection :lol:

You can use 8MB, 16MB etc, but of course you are only going to get 4MB RAM.

Thank you @exxos for your fast and thorough reply! :)

I'll keep you updated about my progress and will post the usual Gembench screenshots when everything is running! :D

So, it's a new day and I have made some progress, but I also have a few new questions (I'm sorry for all these questions, but I really want to make sure everything is fine beforehand, instead of stupidly breaking something, especially with this nice hardware).

First step today was to remove the short on the rear side of the 1.44MB GAL.
I have now also fitted the "I don't have the Flashy Clock yet" jumper, I think this was the last missing modification on the board. :D

Next was the missing RAM: I have successfully searched through my RAM-Collection, smallest matching module I could find was this beautiful double-sided 32MB 60ns module. :lol:
Took it and installed it on the board:


Please take note of the Siemens 1MB modules in the background, with full parity. ;)

Moving on with my plan, the following thing was to "beep" through the connections of the MMU- and GLUE-SMTs for possible shorts between the legs.
Especially because the connections of the MMU really look funky on one side - I wanted to make sure that I don't fry anything on first power-on.

I found two shorts - but to my astonishment not on the wonky looking side of the MMU (these are all fine), but on one side which looks fine to the eye - and between two legs of the GLUE which looks flawless to the eye.
I have marked the shorts in the following pictures:


Is it possible that these are not shorts, but connections through the board, or are these possibly intentional internal connections in the package of the chips?
If these shouldn't have a connection normally, I need to rework these. :geek:

And finally:
I have this butchered v1.4 16Mhz Booster - when I removed the socket on top I accidentally ripped one pad off (red circle in the picture, bottom left). I guess before doing anything with it, I need to rewire this by hand.


Is there any way to use the Booster with the onboard PLCC CPU (without having a DIL-CPU on top of the Booster)?
And if not: can I use this butchered Booster at least for ROM decoding (after having the ripped pad rewired, if needed)?

I'm already excited to see the first boot soon! 8-)

The pins you highlighted in red you can see there are two separate tracks coming from those pins..

As for the booster, you have to use the CPU on top of the board because there is some synchronisation logic going on which cannot be done on the PLCC CPU.

exxos wrote: 06 Sep 2021 21:26 The pins you highlighted in red you can see there are two separate tracks coming from those pins..

Yes, I saw that there were separate tracks, still I thought there might be a connection somewhere through the board - I didn't have the datasheet for these chips at hand, so I really just speculated. :)

I have reworked these shorted pins on the MMU and GLUE today and could not find any shorts any more afterwards. :dualthumbup:

So finally it was time to power up this fine piece of hardware for the first time.
I took the PSU from my working 1040STF (Mitsumi, fully recapped and refreshed thanks to exxos' PSU-kit), took a floppy cable and connected the Sony MPF920 (also from exxos' shop), taking care that I twist the cable before connecting.

Power on - no magic smoke, no explosions, breakers not triggered - great! :yay2:

But it doesn't successfully boot to the desktop. :cry:
Tried resetting it, power cycled it - most of the time the floppy isn't even accessed, but sometimes it does check the floppy - the light on the keyboard and the floppy lights up like it should when booting.

I can hear the keyboard clicks when hitting keys after turning it on (but also not every time), and I also get a video and audio signal (though the audio has a slight hum, but I think this is caused by my cable).

Video wise I either get a white/gray screen, a black screen with a white screen-frame (this happens on most times), but I also saw a completely blue/turquoise or completely green screen. One or two times I got the "Bombs" displayed.

Some information about the current settings and configuration of the H5:

• DualTOS 1.04/2.06 German fitted (DualTOS selection jumper currently CLOSED)

• 1.44MB Floppy GAL fitted (JP22 set to 8/16MHz, floppies 0 & 1 set to forced 720kb mode, no drive swap)

• No blitter, so both jumpers fitted there

• Jumper fitted at the place for the RTC-chip and the middle two pins for "no flashy clock" is also fitted

• 32MB RAM fitted (60ns - couldn't find smaller modules in my collection), jumpers set to 4MB

Things I already tried:

• DualTOS jumper - I'm not sure which version is high/low?
  When the jumper is open nothing happens at all when booting (black screen only), so I guess closed means 1.04

• RAM jumpers: tried setting 1MB

• GAL jumpers: tried setting floppy 0 to 1.44MB, tried setting JP22 to 8MHz only

All these changes did not make any change in the above described behaviour ... :cry:

I have a multimeter and a scope for debugging.

I also read in another thread that the voltage on the 3.3 rail might be too high and a resistor might be needed, but I wanted to report my current findings here first before jumping to conclusions...

exxos wrote: 06 Sep 2021 21:26 As for the booster, you have to use the CPU on top of the board because there is some synchronisation logic going on which cannot be done on the PLCC CPU.

I have now fixed the broken connection to the pad - the connection is fine again now. :)
Is it possible to use this "Booster without CPU" now for TOS2.06 decoding only?

Thank you in advance for any hints and tips resolving my issue!

Diagnostic cartridge time ;). Hooked to a PC over serial is probably best. I would do the reset fix first though.

exxos wrote: 08 Sep 2021 00:55 Diagnostic cartridge time ;). Hooked to a PC over serial is probably best. I would do the reset fix first though.

I measured the voltage at the regulator, it was at 3.48V (and possibly dropped to 3.47V when doing a reset, not 100% sure though about this, I just used the multimeter).
So I did the reset fix (1k resistor on the regulator as shown in Mike's build), now the voltage is still at 3.44V and I still didn't have a successful boot.
I seem to get the bombs-screen now more often than before, mixed with "black screen with white border" and sometimes single-colored screens.

It looks that I need to do the Diagnostic cartridge dance, but I need to get one first :D