Donix H5 C5B build

[donix]

This is my H5 C5B that arrived a some time ago which after som procrastination have started working on. I am really looking forward to this!

H5-donix1.jpeg (851.43 KiB) Viewed 389 times

Audio chip has been desoldered and refitted correctly as per the warning sticker. Apologies for the scratched PLCC chips. I am not a fan of that rubbish PLCC extractor. In fairness though, the ST customs were properly stuck in the donor 520ST. Pins were not damaged though.

Audio chip has been flipped as per instructions, connectors and chips installed and jumpers hopefully set to sensible positions. Alas, I have yet to provoke any other signs of life than a white screen with some pink splatter on it and nothing I do seems to have any effect at all.

I have had a look at the guide and while I have tested the basics that are outlined there. Reset and all clock checks described there check out with the scope. Pull-ups verified and there are no shorts. I cannot find the "LS11" but maybe that is an ealier rev and I would assume that shorted IC legs while not impossible, although is a bit more unlikely for the prepopulated stuff. Voltages are also fine.

Although the reset seems to work, I can observe that pin 17 (HALT) is low which I cannot imagine is any help. I have tried two different PLCC 86k that I purchased from an european eBay seller from which I previously have bought chips from but I cannot know for certain that they work but I have also tried with a DIP 68k from exxos's store.

Another observation with the scope is that when looking at address line I can see one or two pulses on the address lines but there is absolute silence on the data lines: This is A1 with the scope in single trigger mode after reset:

a0.jpeg (585.44 KiB) Viewed 389 times

Just recently I noticed the silk screen markings on the H5 that have the part numbers for the custom chips and unfortunately it seems that the chips I harvested have different part numbers which I guess is because they are from a more recent revision? Are these incompatible with the H5? I did find some discussions in forums indicating that mixing chips from different revisions was a bad idea but that they should be interchangeable otherwise. The chips that I have are:

C100110-001 (DMA)
C07013-002 (SHIFTER)
C100109-001 (MMU)
C070714-001 (GLUE)

Another issue is that also have no way of knowing that my ROM is OK. I have programmed an empty 27c406 with one 1.04 image plus some padding of 0xff up to 256k and then a 2.06 image after that. I hope that I made a correct assumption that I needn't do anything about endianness.

I do also have a diagnostic cartridge which I do know works just fine in my other STs but when trying it in the H5 there is no change in behaviour. A related question: Should he H5 boot from that diagnostic cartridge even with no ROM installed?

Any idea on what I could try next? I guess that the chips could have been damaged on removal or during handling. I am entertaining the idea to remove the chips from one of my "nice" STs and fit sockets so that I have something that I can test with. I also believe that at least one of the other machines have custom ICs with the part numbers referenced on the H5.

[olivier.jan]

As you suspected, there is something to be done with endianness: you need to byte-swap the ROM images

[exxos]

You have the ROM jumper in the wrong place and the 2 middle header pins on the flashy clock port need to be linked.

Also you can solder a VGA connector directly on the H5 so you don't need a adapter.

[donix]

Thank you @exxos and @olivier.jan!

Regarding jumpering the two middle pins of the clockport header: I did jumper the pads to the left of that socket with a solder blob and that seems to be the same as jumping these pins at the header, right?

I will try byteswapping the ROMs but should the ST not be able to boot from a cartridge even if the ROM is invalid or missing? I think that I have tried all four combinations for jumper GLUE/1.02+2.06 near the CPU and the HI/LO TOS at the right hand side but then again not with a byte swapped ROM.

[olivier.jan]

The Diagnostic cartridge is called from the ROM, very early in the ROM code, but still the ROM needs to be present.

[exxos]

Regarding jumpering the two middle pins of the clockport header: I did jumper the pads to the left of that socket with a solder blob and that seems to be the same as jumping these pins at the header, right?

ah yes I forgot I added that on the later boards.

I will try byteswapping the ROMs but should the ST not be able to boot from a cartridge even if the ROM is invalid or missing? I think that I have tried all four combinations for jumper GLUE/1.02+2.06 near the CPU and the HI/LO TOS at the right hand side but then again not with a byte swapped ROM.

While that is often a issue, you using the wrong setting on the ROM jumper, you need to first use GLUE decoding and make sure you use TOS104 first. You selected TOS206 decoding which needs another jumper to enable the extra logic..

viewtopic.php?p=125032#p125032

[donix]

The Diagnostic cartridge is called from the ROM, very early in the ROM code, but still the ROM needs to be present.

Regarding jumpering the two middle pins of the clockport header: I did jumper the pads to the left of that socket with a solder blob and that seems to be the same as jumping these pins at the header, right?

ah yes I forgot I added that on the later boards.

I will try byteswapping the ROMs but should the ST not be able to boot from a cartridge even if the ROM is invalid or missing? I think that I have tried all four combinations for jumper GLUE/1.02+2.06 near the CPU and the HI/LO TOS at the right hand side but then again not with a byte swapped ROM.

While that is often a issue, you using the wrong setting on the ROM jumper, you need to first use GLUE decoding and make sure you use TOS104 first. You selected TOS206 decoding which needs another jumper to enable the extra logic..

viewtopic.php?p=125032#p125032

Thanks for the support to you both! Will switch back to GLUE decoding but I definitely had it set to GLUE decoding. I had set it to that at first actually when using the 1.04 ROMs that I programmed first and only set it to the DualTOS mode when actually programmig a DualTOS ROM and I did that just in an attempt to provoke some kind of response.

Also very good to know that the diag cartridge doesn't work if the ROMs are bad or incorrect.

I think that I will proceed by establishing a baseline of working components as far as I can take that. The idea being to fit sockets to one of the other working STs so I can verify that all custom chips are OK. I also have a kit for an Exxos dualTOS board for the STFM that I want to fit to the same machine so I can verify 27C4096 ROMs in the same machine.

I will be back shortly - or so I hope.

[donix]

I spent most of the day working on this but unfortunately I have not gotten anything different from the H5. Still, just a white screen with the ST2VGA in colour mode and a black screen when set to mono.

It was time well spent though. I have up a test rig in the shape of a 520STFM with a C070789-001 REV F PCB by desoldering all the custom IC except the MFP and replacing them with sockets. Using that I can determine that both sets of GLUE,MMU,SHIFTER,ACIA*2 and WD1772 seem to work fine, or at least they do not prevent the ST from starting up.

I also soldered up and fitted the Exxos dualTOS board for the STFM and the last 27C4096 I programmed with just a single image of TOS 1.04 with some padding also works just fine on the 520ST.

Both 68HC000 chips that I have hand also test fine in an STE so the CPU seems OK as well.

The only thing that I cannot test from the custom IC's is the MFP since I only have one of those. I also only have the one 4MB 72 pin SIMM should that one be faulty.

I have examined my soldering on the various connectors that I have fitted to look for bridges and was overjoyed to find what I thought was one on the cartridge connector but it was just a hair.

I also scoured the forum for information on jumper settings and found one for an earlier revision of the board which I hopefully have managed to follow. That led me to discover that I had missed the jumper linked on the RTC chip that was supposed to be jumpered with no RTC. Is there a comprehensive list of the jumpers for the C5B revision?

It has been a fun day working on the old STs to support the diagnostic work but I am itching to see signs of life from the H5. Any ideas on where to go next?

[exxos]

This is my build to get things running in a minimal config..

viewtopic.php?t=7532

You should really be working.. Bit those Hitachi eproms I've had no end of trouble with.. If you know the image is good I'd suggest burning a atmel one as they make much better contact in the sockets.

I'd suggest working though the troubleshooting page..

viewtopic.php?f=78&t=2362

[donix]

I have four AT27C4096-55 on order in addition to the HM27C4096 that are also on the way - both should be arriving in a few days. I only had two of these and one of them has started to refuse programming - it manages to set a few bits and then it fails. Erasing and blank checking works though.

I have followed some of the troubleshooting post already up until the checking of the first bus cycle. I just need to get myself another probe for the scope to connect the external trigger.