General Stuff(tm)

[PhilC]

Just fired the Lynx, and it boots with a singular beep and you get the Lynx logo and prompt.

Then the power supply has started making noises and it keeps losing sync. Typical old computers, know what I'm fixing this afternoon.

BTW, you can borrow the Lynx if its of use to getting the others working?

Edit: Wasn't the PSU making the noise, it was the speaker They were right next to each other.

Just pressed the socketed chips back in and off it went again.

[stephen_usher]

Thanks. Apparently in 48K mode it beeps once and in 96K mode it beeps twice.

The RAM refresh issue is due to the system using the Z80's refresh which only does 128 rows and not 256, which all the TI chips need. Older 4164 allowed 4116 compatible refresh cycles.

I'm still not getting anywhere with the keyboard issue. From the hardware point of view everything looks fine. I've put some pull-ups on the keyboard read lines which has helped what they look like when a key is pressed but it made no difference. It's almost as if an I/O read isn't happening now and again. Maybe it's a timing issue?

[PhilC]

Ah I see, I need a third eprom, will see if I can find and burn that.

[PhilC]

I've put the three 96K roms in and can confirm it does beep twice on boot.

[stephen_usher]

Thanks.

The machine seems to be running happily in 48K mode, other than the keyboard issue. I've managed to load the game "Star Trek", which is mostly BASIC, so the tape interface definitely works.

With respect of the keyboard, looking at the ROM source (https://github.com/w5Mike/Lynx-Rom-Anal ... 8k_Rom.ASM) I can't see how it can not work other than if an "in" is not returning the correct value.

[stephen_usher]

I've ordered a 74S138 from Mouser, along with 500 decoupling caps to make it half worthwhile.

I have a feeling that the 74LS138 is marginally too slow for the keyboard read timing and the CPU is reading zeros on random lines on the bus.

[stephen_usher]

Seeing as the keyboard issue is probably timing as I have some 74Sxx parts on the 128 board I thought I'd fit them where maybe they could do some good.

The two 74S04s worked in two places but caused video timing issues in another.

The 74S74 dual flip-flop caused all sorts of fun when I put it in the RAM timing circuit, only with the video RAM though. I moved it into another signal path and the machine is happy.

I also tried replacing the 74LS244s and 245s with AHCT variants, in case they would help as they switch slightly faster. The only difference it made was to the temperature. The AHCT parts run stone cold, unlike the LS ones, one of which had been running at over 40C.

After these changes the keyboard issue may be slightly less but not remarkably better.

[rubber_jonnie]

Seeing as the keyboard issue is probably timing as I have some 74Sxx parts on the 128 board I thought I'd fit them where maybe they could do some good.

The two 74S04s worked in two places but caused video timing issues in another.

The 74S74 dual flip-flop caused all sorts of fun when I put it in the RAM timing circuit, only with the video RAM though. I moved it into another signal path and the machine is happy.

I also tried replacing the 74LS244s and 245s with AHCT variants, in case they would help as they switch slightly faster. The only difference it made was to the temperature. The AHCT parts run stone cold, unlike the LS ones, one of which had been running at over 40C.

After these changes the keyboard issue may be slightly less but not remarkably better.

That's reminding me of all the timing issues I had with my OSI 600D and modern ICs...