Sinclair ZX80 / ZX81 / Z88 Forums

Claudius,

May Jupiter ACE have its 32k or 64k RAM chips? How about a switching mode voltage regulator instead of LM7805? Direct composite video connector? SD slot or MicroSD slot?

Thank you very much in advance.

DesSony

Yes, it’s possible for an updated design to use a 32k byte SRAM chip for the main memory. There are pictures of the kit board that I bought and built up here. This includes a composite video output and a DC/DC switching regulator (in this case a RECOM part).

On board mass storage would be nice in the form of a SD card slot, but is rather more complex to implement. A bit bashed I/O port is not too hard to do, but it would take up more space on the PCB. Then there is the matter of developing the code and integrating it in a ROM - that would be the more difficult bit.

Mark

dessony: have a look into the Jupiter's memory map. It looks like a Swiss cheese due to a partial address decoding. Any user memory extension can sit between 0x4000h and 0xFFFF addresses. So, the easiest way to extend the user memory from 1K to 17K (+16K) is employing any 16K RAMpack for ZX81. In this case my "Twister" is very handy. The other possibility would be using a 128K SRAM chip like 628128 type and a simple decoder. In this case you will have 1K + 48K (bank 0) + 48K (bank 1). Banks are independent to each other and switchable. The hole idea was invented by Edwin Blink in 2006 as a part of Jupiter Ace Team developments.

Yes, you can use any +5Vdc switching regulators. I installed one in my Jupiter and it have worked without any issues since then.

Another "Yes", the composite signal can be made easily. There are some minor issues with signals levels on some modern LCD TV sets, but I worked around the issue with additional diodes and resistors. In my replica I use an emitter follows and it is sufficient for my cheap 4.3" monitor.

Regarding a modern storage media, well it would be tricky but feasible. The ACF project stoped a long time ago, and I was not able to find any creditable information how it had been done. I think that ZXMMC could be adopted for Jupiter but new ROM/OS should be created by a clever software geek. We (my much smarter than me colleague and I) checked the ROM content and there is no single byte left spare. We have some ideas about squeezing characters area and free up 100 bytes there for an other purpose. This would not be enough for data storage routines. Anyway some works in progress.

I don't have the Jupiter's memory map chart. Sorry. As for the SD or MicroSD, we need another voltage regulator, cut to 3.3 volts and do something about a data serial streaming and clock. Maybe we can use some "empty" upper section of the EPROM/EEPROM for these SD or MicroSD functions. Can we consolidate the Jupiter ACE's ROM since we now have enough EPROMs and EEPROMs laying around? It means reading 2 assembly language listings and merging them into one listing.

Jupiter ACE memory map can be found there: http://k1.spdns.de/Vintage/Sinclair/80/ ... memory.txt

Please note that the 8K ROM is full. You can use larger EPROM chip but the address deciding is simplified and you will not gain much. The new address decoder is needed.

See also this linked document.

Mark

Thank 1024MAK and McKlaud for the Jupiter ACE's memory map chart!

Have we tested a ZXpand or ZXpand+ with the Jupiter ACE yet? Can we hook up a CF to the computer?

Which new address decoder is needed for the "combined " 8k ROM?

DesSony

A ZXpand or ZXpand+ designed for a ZX81 will not work, as it is very closely designed to work with a ZX81 including having a custom ROM (or equivalent) for that machine.

In theory an equivalent device for the ACE could be developed, but it is a lot of work.

Mark

1024MAK,

We can try to beg Sir Morris (aka Charlie Robson) to create a ZXpand+ version for the Jupiter ACE.

DesSony

Jupiter ACE is not popular like ZX81. Please bear in mind that there was sold no more 10,000 units of Jupiter ACE in 80-ties. A few more replica PCBs in recent years do not make any difference. I would not be surprised if SirMorris said 'no'.

Regarding the memory map, the partial address decoding of RAM result in shadows in different places. ROM area is limited to first 8K of addressing space (0x0000 to 0x2000). Keep in mind that the existing ROM is full, no spare space for any additional routines. Any new interface with the ROM would be very challenging exercise.