3K Rampack issues

Discussions about Sinclair ZX80 and ZX81 Hardware
User avatar
1024MAK
Posts: 5101
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: 3K Rampack issues

Post by 1024MAK »

siggi wrote: Mon Nov 26, 2018 6:42 pm The first new board worked, so number 2 and 3 also "should" work ;)
The issue 1 board above has a big ground plane next to the connector and maybe below the ram chips. Does the issue 3 board also have such a ground plane?
Yes, Siggi, I agree that good decoupling, big ground plane, and wide and short/directly routed power rails are good practice. Also as you indicate, just because the first board works without apparent problems does not mean that it is a good working design, or that other boards built using the same layout will work.

But one thing I do know, is that using SRAM chips that are too slow are very much likely to cause problems. Sinclair used 2114L-1 (150ns) type for a reason. Otherwise Sinclair would have used the cheaper 2114L (450ns) SRAM chips throughout.

Mark
ZX81 Variations
ZX81 Chip Pin-outs
ZX81 Video Transistor Buffer Amp

:!: Standby alert :!:
There are four lights!
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :!:
Looking forward to summer later in the year.
User avatar
mhudson52
Posts: 141
Joined: Tue Sep 27, 2016 2:17 am
Location: Decatur, Alabama (US)

Re: 3K Rampack issues

Post by mhudson52 »

Thanks for the info, Mark. I had seen the extract of the instructions before, but somehow failed to save that particular page.

Just received a fresh batch of 200ns 2114's yesterday, so I'm fixin' to go see what I can do using the info you guys have posted.

BTW, I am attaching photos of both sides of Charlie's Issue 3 board. Issue #2 corrected a couple of flaws in the original Issue #1 from Brenner, and Issue #3 just "fine-tunes" Issue #2.

With report the results. Thanks, guys!

Mike
3K RAM Pack Rear bw.jpg
3K RAM Pack Front bw.jpg
TS 1000 bought in 1982 and brought back to life by Sir Ian
User avatar
mhudson52
Posts: 141
Joined: Tue Sep 27, 2016 2:17 am
Location: Decatur, Alabama (US)

Re: 3K Rampack issues

Post by mhudson52 »

Mark, I just received a fresh batch of 2114N-2 chips and tried them on the board both fully populated and in pairs as shown in the drawing you posted. Still no joy.

Also tried having only the timer chips installed. Got "K" prompt with one and nothing on the other. So I went back over all the connections once more with my trusty low-tech continuity meter and once again found no issues. +5 and +9 volts are indeed present at the edge connector. About to lose my mind.

Makes me wonder if it is my TS1000 that is not playing nice with these packs and that it is just "almost" compatible with these RAM packs due to differences in addressing schemes between the 1K and 2K machines. It has been refurbed by Ian so I have full confidence in it.

Regards and thanks,

Mike
TS 1000 bought in 1982 and brought back to life by Sir Ian
User avatar
1024MAK
Posts: 5101
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: 3K Rampack issues

Post by 1024MAK »

mhudson52 wrote: Sat Dec 01, 2018 2:21 pm Also tried having only the timer chips installed. Got "K" prompt with one and nothing on the other.
Hi Mike

Timer chips? Confused :|

There are no differences in the addressing schemes between the ZX81 and TS1000. In both, the ULA partly decodes both the ROM and RAM areas as follows (regardless of the size and how much memory is actually fitted):
All ROM access decodes to 0x0000 to 0x3FFF, and 0x8000 to 0xBFFF
All RAM access decodes to 0x4000 to 0x7FFF, and 0xC000 to 0xFFFF
In each case, if the actual fitted memory is less than the decoded area, you will see repeats (echos, duplicates) of the same physical memory chip(s) throughout the decoded area.

So for an unexpanded ZX81 or TS1000, the 8K byte ROM chip will completely fill the 0x0000 to 0x3FFF (16k byte) area, but will repeat after 0x2000... Same for 0x8000 to 0xBFFF.

The same happens with RAM. For the ZX81, the 1k RAM chip can be accessed at 0x4000, 0x4400, 0x4800, 0x4C00 etc.
The only difference with a TS1000 is that it has 2k bytes of RAM, so instead of the ZX81 1k RAM being present 32 times in the memory map, the TS1000 2k RAM is only present 16 times at 0x4000, 0x4800, 0x5000, 0x5800, etc.

You can prove this for yourself by using PEEK to examine the memory one byte at a time...

Memory expansions either ignore the ULA’s decoding and have their own memory decoding circuitry (e.g. 16k byte RAM packs and greater) or combine the ULA’s decoded output with their own circuitry.

Does your TS1000 work okay with other expansions? If so, can you list them please.
And I am assuming that you have cleaned and inspected the edge-connector on it...

One last try. If you fit RAM chips in only positions 3A and 3B, do you get the inverse K? And if you do, can you POKE numbers into this memory (0, 85, 170 and 255 are good numbers to try) at the relevant addresses and read back the same number with PRINT PEEK?

The only other thing that I can think of, is for you to ask someone else to test one of your boards.

Mark
ZX81 Variations
ZX81 Chip Pin-outs
ZX81 Video Transistor Buffer Amp

:!: Standby alert :!:
There are four lights!
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :!:
Looking forward to summer later in the year.
User avatar
mhudson52
Posts: 141
Joined: Tue Sep 27, 2016 2:17 am
Location: Decatur, Alabama (US)

Re: 3K Rampack issues

Post by mhudson52 »

Hey Mark. Yeah I probably mis-named the 74LS138 1 of 8 Decoder / Demultiplexer IC as a "timer" chip out of my abundant ignorance.

Thanks for the info on addressing schemes on the machines. That's good stuff I am trying to digest.

The TS edge connector has little wear and is clean. Plugging my ZXpand+ in brings up a nice "K" prompt and short memory tests run without a hitch.

Alas, RAM chips in 3A and 3B produce nothing, same as 1A/1B and 2A/2B. Board with only 74LS138 plugged in does produce "K" prompt.

Yeah, having someone else to look at them is where I am headed next.

Thanks again, Mark, for your great help and patience.

Mike
TS 1000 bought in 1982 and brought back to life by Sir Ian
User avatar
1024MAK
Posts: 5101
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: 3K Rampack issues

Post by 1024MAK »

One thing I forgot to make clear, was your test with the 74LS138 fitted?

If not, try again.

If yes, fit the 74LS138 and either ONE RAM chip in 3A or in 3B and see what happens. Then try ONE RAM chip in the other position (3A or 3B, whichever you left empty previously).

Mark
ZX81 Variations
ZX81 Chip Pin-outs
ZX81 Video Transistor Buffer Amp

:!: Standby alert :!:
There are four lights!
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :!:
Looking forward to summer later in the year.
User avatar
mhudson52
Posts: 141
Joined: Tue Sep 27, 2016 2:17 am
Location: Decatur, Alabama (US)

Re: 3K Rampack issues

Post by mhudson52 »

Yes, the 74LS138 was fitted. I'l try your suggestion using only one RAM chip at a time in those locations.

Then I'll probably just sob quietly to myself over in the corner.

Mike
TS 1000 bought in 1982 and brought back to life by Sir Ian
Post Reply