A4000 Acill - Issues booting with Diagrom

[amigasquirrel]

I've compared all the signals on U450 to those on a working A1200, and they are all the same.
Clock inputs and outputs are all identical, and so is _CCCAS.
The data and address bus shows the same activity and "noise" as the A4000. My probing setup is probably less than ideal so that might not be helping.

Think it might be time to swap U450 out and see if anything changes.

What's unusual is when comparing the DiagROM display sequence on startup between the A1200 and A4000, the A4000 shows the same sequence, up until the point when the menu should be displayed. I wonder what DiagROM is doing just before it displays that screen. I'm using the same DiagROM in the A1200 and A4000 so it's not an issue with the ROMs.

[Castellen]

Made some progress with the FastRAM, it is all now being detected correctly and error free over multiple passes.

Great, what was the problem with that?

The IC's are all supposed to be working, either new old stock or tested used.
U700 is not installed at the moment until this issue is solved.

What's unusual is when comparing the DiagROM display sequence on startup between the A1200 and A4000, the A4000 shows the same sequence, up until the point when the menu should be displayed. I wonder what DiagROM is doing just before it displays that screen. I'm using the same DiagROM in the A1200 and A4000 so it's not an issue with the ROMs.

As mentioned, I don't know exactly how DiagROM might behave with a partially populated board.  You mentioned earlier that U700 and U890 were installed, I gather you've since removed U700?

I've previously seen DiagROM behave as you've described where there's sometimes no menu displayed on the monitor, but some of the video tests (as operated from the serial terminal) do show the expected image on the monitor.  I've no idea if this was due to a hardware problem at the time, or some kind of quirk with DiagROM.  I don't usually do much low-level fault finding with the monitor connected, so never looked into it any further.  I typically either use DiagROM or my own design in-circuit debugger to diagnose more complex hardware faults, and once these are resolved, the hardware is typically in a state where it can run the standard OS.

I'm not fully convinced there's a problem with U450 since you're mostly seeing the expected video output.

My suggestion would be to get the board more complete (i.e. U700 fitted) and try to get it running with the standard OS3.1/v40.68 ROMs.  If you get nothing but yellow screen exception errors, I can send you a specially modified ROM that sends out debugging detail when it boots to better understand where the software is having a problem.  I can't share that publicly since it contains parts of the original Commodore code.

[amigasquirrel]

The issue with FastRAM was a bad connection to one of the RAS/CAS lines on U890, after fixing that it all started working. The finish on the legs of some of the NOS IC's isn't very good (presumably due to age) so it appears what looks like a perfect solder joint isn't necessarily making a good connection at high frequency. Had to clean up the legs so they look nice and clean, and re-install that made a big difference.

I decided to swap out U450, and got a display as a test I put the suspect U450 in my A1200 as a test and got exactly the same display so we know that is definitely faulty. Installed U700 (DiagROM works fine without it installed) and managed to get it booting properly into Kickstart and workbench, and was able to run some software.

After playing around with that for a few hours, powered it on and was occasionally getting no display. Went back to DiagROM and when there was a display, started to get weird issues where I couldn't get DiagROM output on the serial port, but the serial tests would pass with characters getting sent and received.

The the system started misbehaving, not wanting to boot at all. Power supply started making a chirping noise (no idea why as I hadn't changed anything) as if it was shutting down and restarting - never a good sign. Checked all the voltages, everything looks correct. Tried another power supply, no difference.

Without changing anything, DiagROM then started to output in the serial port again but started to crash at "Set all DMA enablebits" and the system has got steadily worse since.

I'm now getting no H sync / V sync, no activity on the serial port even with all RAM removed.
Occasionally I'll get a burst of activity on the CPU bus, but it's mostly dead and stuck at 5v.
Clock inputs to U450 show little/no activity as if they are being dragged down, so something has died. 

All I can think of doing now is removing U450 and see if I start to get some activity, it felt like it was getting a little warmer than usual.

[amigasquirrel]

I removed the replacement U450 (which was borrowed from my A1200) and the master clock input now looks correct, so the IC is definitely faulty and was pulling the clock down. Why that's happened I have no idea, maybe it didn't like the stress of the hot air removal.

As test I installed the original U450 (that was displaying the vertical lines) and the system is working again as before, so at least that's something. Was getting worried that something catastrophic has happened.

Still wasn't getting a serial output and discovered that the +12v on U304 was nearer 5v. Measured D175 which supplies +12v_USER and it was 12v on one side, and 5v on the other. Reflowed the solder joint (which looked perfect) and then magically got the correct +12v on the output, and now serial is working again.

So I'll have to get a new U450, and continue on from there. At least I know the system has been fully working, so it should again.

My suggestion would be to get the board more complete (i.e. U700 fitted) and try to get it running with the standard OS3.1/v40.68 ROMs.  If you get nothing but yellow screen exception errors, I can send you a specially modified ROM that sends out debugging detail when it boots to better understand where the software is having a problem.  I can't share that publicly since it contains parts of the original Commodore code.

That would be much appreciated, when the system was fully working I was occasionally getting yellow screens until I removed the RTC. There is no crystal fitted at the moment because I don't want to put it through the ultrasonic cleaner.

You are absolutely right about sockets, when it was running the full OS I found that applying gentle pressure to the board near sockets would make the system crash/glitch, so all the PLCC's are soldered now. I don't think sockets are any use with these old IC's and less than perfect finish on the legs. I also know next time to apply new solder to the legs and remove with solder wick to refresh the coating before soldering to the PCB.

[Castellen]

OK, so in summary:
   - The ICs which were all "supposed to be working" weren't necessarily working
   - You were still trying to persevere with the IC sockets, and these were continuing to introduce intermittent problems
   - There were multiple bad solder joints
   - The oxide wasn't cleaned from IC leads before soldering or trying to fit them into sockets

I don't think sockets are any use with these old IC's and less than perfect finish on the legs. I also know next time to apply new solder to the legs and remove with solder wick to refresh the coating before soldering to the PCB.

Next you'll be telling me you used 100% tin solder instead of 60/40 tin/lead with a decent flux   If anyone cares, I'd recommend one of the Kester leaded solders with 3.3% 245 no-clean flux, which is a great all-rounder for this type of work.

If you're trying to solder to heavily oxidised IC leads, I'm not surprised there are so many soldering issues.  You can do the solder wick thing, or a much faster/easier way is a couple of wipes with a fibreglass scratch brush to expose a clean surface, then brush IC leads and PCB solder pads with a no-clean liquid flux pen immediately before soldering.

Sounds as though you may still have intermittent solder joints somewhere based on some of the occasional problems you're seeing.  With the board running normally, you should be able to gently flex/twist it and it should continue to run normally.  If it freezes, resets, crashes when you flex the board, then you've likely got other solder joint issues to track down.

Regarding the system not booting with the RTC/U178 fitted, than can happen if U178 doesn't have 5V at the supply pin (18).  The system will run normally without the 32.768kHz crystal fitted, only the stored time won't increment, it'll stay 'frozen'.  While powered on, the system RTC will increment normally since that uses one of the CIA hardware timers, but it'll always load the frozen time from U178 at every boot.