+- Discussion Forum for all things Microbee (https://microbeetechnology.com.au/forum)
+-- Forum: Microbee Forum (https://microbeetechnology.com.au/forum/forum-1.html)
+--- Forum: Microbee Hardware (https://microbeetechnology.com.au/forum/forum-6.html)
+--- Thread: DRAM 128K system has stopped working (/thread-481.html)
RE: DRAM 128K system has stopped working - Ernest - 21-05-2022
[quote pid='1611' dateline='1653053036']
Quote:Thanks again for all the help!
Hope these symptoms mean something to someone. I also have some photos of the screen if anyone thinks they'll help? Scope is on the bench, ready to go.
Cheers!
Chesh
[/quote]
Glad you haven't given up yet, photos are always a help, as they say a picture speaks a thousand words, but in case with a blank screen it won't tell us much at all.
Ironically, I have started having similar issues whit my set-up that I any trying to emulate your problems, mostly I have the MB8342 on Premium mainboards. The one I'm trying to use has the Colour board option, which for anyone who has deal with these adds a whole extra level of problems with additional connectors and sockets all potential points of trouble.
I have been able to confirm that the BN55 will boot to Monitor if the M key is held down just the same as the BN5443.
Anyway, the big difference between running the TEST ROM / BASIC and the Boot ROMs is how the Memory is used.
The TEST ROM and Basic are run from the ROMs themselves in the same way as the SRAM ROM based Microbees do.
With BN5443 or BN55 ROMs they are copied from their 8000h location to (I believe Bank 1) location 6000h. Then the ROM is switched out and Banks 0 & 1 RAM switched in. Control is now passed to the ROM image which is now located at E000h.
Any issues with the RAM and / or bank switching will corrupt the ROM image and prevent booting OK.
One of the first routines after passing the control to the ROM image in RAM is to initialize the Video driver 6545, which looks like this is not happening. Without an initialize 6545 there is no keyboard scanning and no output to the screen.
Port 50 latch IC16 is responsible for the RAM bank switching via the PAL IC26, however a problem with any of the DRAM chips could cause a multitude of issue.
Ewan asked about the jumper / board changes to the Mainboard as they are set up different between the SRAM models and the DRAM models. I assume as your system was working at one point you have the changes correctly set up for the DRAM. As Ewan said the screen RAM addressing at F000h is one of the big differences. This comes from the Main board for the SRAM models but from the Coreboard for the DRAM Coreboards. Another big difference is the !JLATCH and the RESET functions.
Hope this helps a little, I haven't given up on my experimenting to see if I can throw any extra light your current problem.
Ernest
RE: DRAM 128K system has stopped working - CheshireNoir - 21-05-2022
Hmmm. Always check your work.
Went back and checked my rework on the main board, and fixed up two problems that I'd missed on my re-rework pass. Multimeter in resistance mode for the win! (A link I thought was cut through wasn't and a jumper turned out to not be as conductive as I thought)
Can I just clarify something in the Hardware Notebook?
Section B Pg 13 under DRAM MAIN BOARD is the following line:
LINK SET 4. cut K. to L. join K. to J. (only change if DRAM core board IC48 is used).
I interpreted that as "Don't make any changes if you don't have IC48" but it could mean "only cut K. to L. if IC48 is not present"
BTW, fixing those patches and verifying them, things seem to have gone backwards. The RAM tests no longer work.
I'll do some more tests.
Cheers!
Chesh
OK Photos time.
Here are some screenshots of the 'Bee as she currently is.
https://photos.app.goo.gl/TMW9CWSKPyfAYxDh8
Check the notations for more info (Click the (i) in the right hand corner of each image)
If people want photos or captures, I'll add them to this gallery.
Cheers! Chesh.
RE: DRAM 128K system has stopped working - CheshireNoir - 21-05-2022
BTW I have also now tried a new Z80. (It was a longshot)
I have also tried disconnecting pin 18 on the X4/X5 connector (Thanks for the suggestion, Ernest) but that just left me with a board where even RESET no longer works. (Thankfully I tried it in a nice, reversible way so it was 2 seconds work to reinstate it. I simply removed the pin by desoldering it from the core board with my desoldering gun. When that didn't help I just rinserted the pin and resoldered it)
Chesh
RE: DRAM 128K system has stopped working - CheshireNoir - 04-06-2022
I seem to have fixed it :-D
Yesterday, in a fit of enthusiasm, I fitted my core board risers and started probing around the CPU, taking careful notes of all the "important" lines. After checking all the "inverted" pins (which tend to be the important ones) I found NMI was being held low.
OK why? To the schematics Batman! (Thank god for the iPad Pro work lent me)
Tracing back, NMI quickly jumps from the Main board to the Core board and through to IC 49 (pins 5&6) which was correctly inverting.
IC41 (pins 4,5 output 6) looked like a culprit. It was getting a high on pin 5 and a "low" on pin 4 so should be putting out a low (to be inverted by IC49).
I tested IC41 in my TL866 (Not a guarantee I know) and it passed.
Looking closer at the signal on pin 4 and it's hovering up at around 1.5-1.6v which is right in the "undefined" voltage level for an LS part, from memory.
Further back in the chain, we're back to IC 49 again (Pins 12&13). I popped it out and stuck it in the TL866.
Guess what? It fails on Pins 12 and 13!
Went out and bought a replacement and initially it didn't work (While NMI was now high, RESET was being held low) until I remembered to bend out pin 8.
After that I went through some more troubleshooting until I finally realised my video connector was being a bit "pants" and twiddling it I got the solid white "looking for floppy" cursor.
Plug in the Gotek and it boots. It actually works again.
Whew! I got there in the end :-D
Thank you to everyone who helped with suggestions, things to check and offers of help. I'd nearly got to the point of giving up.
(If I hadn't run out of my "mad" money this fortnight, I probably would have already posted it to Ewan. Don't worry though, Ewan. You'll be getting ALL my $$$ when the new Prem Plus gets released :-D)
Cheers all!
Chesh
RE: DRAM 128K system has stopped working - MbeeTech - 04-06-2022
Great to hear you solved it Chesh. Nice work.