8-bit acorn hardware • Re: Issue 7 wont boot - now stuck for ideas

after a marathon beeb diagnostic session at the weekend we've run through the suggestions from the forum and checked out a few more of our own. There was no improvement nor any smoking guns we could see. All tests were run using Tricky's ROM. I had previously stated we saw no difference when the jumper was moved but there are small differences now. Not sure if we missed it before or something else has improved with the testing / socketing done so far. I've grabbed some more videos and they are shared on google drive as they are too big to attach:

https://drive.google.com/drive/folders ... p=sharing

Tricky's ROM was also used on the working machine for like-for-like comparison.

What does the memory test screen look like with the diagonal lines? Looking through my fixes for TTXVDU corruption, it is worth checking IC38 (74LS86) and IC15 (74LS273) it is also a good idea to check continuity from the legs of IC2 (6845) through to ICs 8/9/10/11 (81LS95s)

we checked IC38 and IC15 again as advised - in fact ic38 was removed, tested and replaced with continuity checks afterwards.
Continuity and clock signals were checked with a scope and all seemed fine. Continuity of IC2 through to IC8/9/10/11 was checked and no issue found.

I suggest during fault finding you go with a minimal configuration, but don’t worry about removing any soldered in chips and leave IC3 (system VIA) fitted IF you have a known good keyboard connected. Otherwise do remove IC3. If IC3 is not fitted, you will get a continuous tone regardless.

We spent all day with the minimal config suggested, apart from IC4 as it's not socketed on this board. There is no DFS, no Econet, no Speech, on this machine.

S9 was previously missing and dealt with a few weeks ago - S9 pins are bridged correctly for lack of disk controller.
S21 checked and correct
S25 is set "north" for 32k but has been tried in all positions and with inverter circuit applied with no real visible change to Tricky's test rom result.
The machine still doesn't boot. ** see note that there are some differences visible, perhaps we missed them before?

Do you have access to a thermal camera? You can look for things that are hotter/colder than expected.

Thermal checks were repeated across the board and compared to a working model. No chips were unusually hot (we had used this technique a few weeks ago to locate and replace the broken IC20) IC15 was the coldest chip in both machines. None of the DRAM was more than about 30 degrees and nothing stood out as unusual.

We also ran Andre's ROM as OS ROM for a few additional tests but that always failed after a single LED flash reporting an error very early on.
We removed, socketed and replaced IC53 and IC61 in case it improved the errors from Andre's ROM but it made no difference.

Looking at the MODE0 video, it looks like memory failure almost as if one RAM IC is still outputting when it should be high impedance. I have pulled all the DRAM ICs to cure a similar problem I had with some BBC boards and found a few dodgy DRAM ICs. The worst DRAMs are Hitachi that were made in West Germany or Mostek. The Japanese Hitachi DRAM ICs are, on the whole, very good.

It looks like I have West German RAM. We also wondered whether one or more DRAM could be outputting when not selected and causing such random behviour s but is it possible to detect that without removal and testing each DRAM individually?


thanks again in advance, Ed.

Statistics: Posted by eddawes — Thu Apr 10, 2025 6:45 pm

---