SDRAM on the lower ram

[KWF]

Thanks for sharing the schematic diagram and good work.

These issues, you are facing with A0 and A15 on SRAM, are driven by SCLD and the way how VRAM address lines (MAx) are multiplexed by SCLD internally.

Good luck with 128K upgrade

Re SRAM chip: have a look at any 62256 (e.g. AS6C62256) or 628128 (e.g. AS6C1008) types in SMD, they are 5V compatible, still in productions and available. Hitachi SRAMs have been obsolete for long time.

[Powervr]

Thanks KWF, now that  I have acess to the bus and lower ram, I could make something in the CPLD (that is the objective) for yours "Timex 128k" to work with the other screen 7? of  the 128, thus improving compatibility on my timex 2068 for128 titles. Could  also  change  stuff on your's  timmy...

I don't  understant much, this was one of my first schematics.

One of my first "works" was to remove the not black on my 2068, with  the inclusion of some gates logic on the bright signal... but I did that not inside the timex computer 2068 but on the tmx 128k expansion you made ... from  the work of Don Superflo. I was hooked when this worked. 

[Powervr]

the boot sequence it's faster by maybe 2 seconds, I find this weird... do you know why KWF? this is not faster, it uses the same RAS  and CAS values.

[KWF]

Are you sure that is 2 sec. faster? This is a lot of time. Could you record and share the boot up sequence of your modified TS2048?

[Powervr]

Are you sure that is 2 sec. faster? This is a lot of time. Could you record and share the boot up sequence of your modified TS2048?

Maybe just a sec... Will do it when at home today

[trojacek]

Boot-up is visibly faster when there is 16K of RAM only

[Powervr]

Boot-up is visibly faster when there is 16K of RAM only

I didn't removed the upper ram.

[trojacek]

On your first video upper RAM is removed.
On your second video I cannot see the moment of boot-up, just the diag software running.

[Maryjan]

It's hard to get SRAM to work as the best RAM in a ZX Spectrum or TIMEX.
It seems simple, but it is not quite so.

[Powervr]

It's hard to get SRAM to work as the best RAM in a ZX Spectrum or TIMEX.
It seems simple, but it is not quite so.

if it's hard the more reasons I have to do it. if it was easy... why I would lose time to learn VHDL ... nah...


The hard work is to make the cpld several times quicker than the ram/speccy..  buffers, and first arrive,  first served... 
Will make a video why not...
hehehe

I will even try to make the timex more compatible... don't know if I will arrive where I want... but who cares... it's the trip. speccy as some  resistances for isolation of the bus... I will have my CPLD orquestrating this... I can even remove some IC's...  so when as z80 asking for some missing port... why not implement some of the "wrong doings" of the spectrum... when even the Z80 can know what the lower ram is doing.

this tricks  prevent games like arkanoid the original working on the timex.

This timex, was not working for decades...  it lacked several components, and some were bad... I  took more than 1 week to repair  it... got a new clock oscillator that measures 4.0001 mhz... when the older one didn't work and were more lax... like 3.97 mhz...  also the capacitor that goes with.

The last person that bought it, bought it not working,  gave it to me, after I repaired his timex 2068, removing and implementing the sound from the speaker instead of the TV via RF out....like some last models, I repaired also the ram... like usual.

He is the first person to build an spectrum emulator in MSDOS, Rui Ribeiro. long time ago in the 90's...

I am not an electronic master, I am mathematician... but I love this stuff... what I am good at is soldering and resoldering stuff... and I love it. I already builded several Reamiga (amiags from scratch) so the soldering of the SCLD is... easy ... as pie.
Did the same to this one...

The weird stuff is after I remove and cleaned oxidized soldering pins  then the SCLD stops to work... and I have the same result in two of my working timex.

that is if you remove the SCLD...  resoldered it... it stops to work... the last time it was working when I put my oscilloscope reading one of the clock...  if I was not doing that it didn't started...
matrix glitches...

nah... two of the times after I cleaned it with my ultrasound cleaner... both were ok after... so even non conductive flux, make some  capacitance preventing the clock to work... unless you touch some pins with your hand... or... using the osciloscope... hahaha
matrix glitches...

Allways due to some reality check...

I make mistakes but I learn with them. I make stupid mistakes... and this one is one of  them. he...hope I have learned from it.

nop...  it's not faster ... it's the same thing... I found that weird... and I was right.

This booted  faster, and I HAD not  removed the upper ram (yes on the first video I think... don't remember).
It was faster before I put the last 8 kb in  service...(I am a little ashamed)  I had an bad soldered adress ... so this rom must bypass the rest of the ram checking or something. just like the upper ram, but in the lower ram.
Then I thought it was allways like that...  before retesting it again today... IT's possible  to have malfunction lower RAM, with an good boot image.

So the same speed...  I love this glitches on the matrix...  hahahah

[Powervr]

https://www.youtube.com/watch?v=cfzyJYazWGQ