Cool, thanks. I did find that I had a bad CPU earlier, I grabbed my complete stock of 6502's and used my OSI 600D to test them all, and the one that came from the Electron is dead believe it or not, and it wasn't the original either, but one I bought from somebody on here. Luck of the draw I guess, but all the signals are looking good on the ULA at least.I had issues with pins making bad contact in my previous PGA socket - I put it down to a cheapo ebay no-name and damage due to an old ULA having some excess solder on the pins, but maybe you're experiencing similar issues? Probing the Address and Data lines to see any weird signals is a good place to hunt those poor connections down, and good old continuity testing - the JamSoftULA schematics will show what signals to expect on the sandwiching header pins are - it's just a tedious process...OK, so I am using a low profile turned pin socket (I don't really like them but it needed to be low profile) and I gave the socket a squirt of Electrolube and I am seeing address and data activity, plus , but now the clock has changed and instead of steady at 2Mhz I get flashes of 1Mhz.
Getting 1MHz signal is good as it should just be 2MHz when accessing RAM then 1MHz (or less) when talking to RAM/IO/keyboard...
Probe-probe-probe I think... Might be the RAM too as you say, if you have an Issue 6 board with the SIP TI board, you can add sockets and use that instead... you'll get there!
So that leaves me with a 'what the hack is going on?'
I'm getting address line activity and data line activity now, so that is something and I am getting 2Mhz/1Mhz clock, so that seems ok. I'm also seeing pulses on pin 7 of the CPU (SYNC) that suggests the CPU is running too.
What is odd is that my working machine that I'm using for cross checking (An issue 4) seems to behave exactly the same when I remove the keyboard. I mean it's exact. No beeps, no sound, similar address and data line activity and a black screen.
So could it be a problem with the keyboard not not being recognised? I need to explore that circuit some more.
Also, one of the RAM ICs is a 3764, which is ostensibly the same as the 4864's in the other 3 RAM spots, but are there possible RAM timing issues with that one IC? I don't have any spare 4864's but I could fit all 3764's which are 150ns rather than the 200ns 4864's. It would be trivial to remove them and fit sockets for the 3764's.
On the subject of sockets, I think the CPU socket is going to get swapped for a more normal dual wipe socket. Given I a) don't really like turned pin sockets b) don't need a low profile socket since I have a very tall ULA replacement next to said socket, I'm just going to swap it out for a normal socket, I just never get decent results with turned pin unless what's plugging in is also turned pin.
Then as you say @mogwaay, probe probe probe. At least I have a working machine to compare to.
Statistics: Posted by Ronin47 — Tue Jun 25, 2024 11:15 pm