Sinclair ZX80 / ZX81 / Z88 Forums

I just picked up a Lambda 8300, It's a tad incomplete, and as I'd rather not destroying it the first day out, does anyone know:

• What is the DC voltage & amperage in? I've seen it listed between 10.8V 500ma and 12V
• Can I use a standard ZX-81 16k RAM pack?
• Can I just jumper the missing expansion bus lines from the CPU?
• If the missing lines can be reconnected, and I install the ZX-81 ROM molded for Lambda, can I then use a ZXpander?

Thanks!
-John-

There are more knowledgeable folks about than I, but I'll tell you what I found with my relatively recently acquired example.

questarian wrote:What is the DC voltage & amperage in? I've seen it listed between 10.8V 500ma and 12V

Mine came with a monster 12V 1.8A combination transformer and cosmic ray deflector. That scared me a little, so I substituted a 9V 1A transformerless unit. The Lambda seems happy enough to be on the same diet as a ZX81. If you're using a RAM pack I'd guess at 800ma optimal, but 500ma may suffice.

questarian wrote:Can I use a standard ZX-81 16k RAM pack?

I slapped on a standard ZX81 RAM pack and was off to the races immediately. Worked just fine.

The other questions I'm even less qualified to answer. If you haven't seen it already, you'd find this thread interesting: viewtopic.php?f=7&t=1006

Ian

TMAOne wrote:Mine came with a monster 12V 1.8A combination transformer and cosmic ray deflector. That scared me a little, so I substituted a 9V 1A transformerless unit. The Lambda seems happy enough to be on the same diet as a ZX81. If you're using a RAM pack I'd guess at 800ma optimal, but 500ma may suffice.

I was a little surprised when I saw 12v listings as well, good to know a 9v works.

TMAOne wrote:I slapped on a standard ZX81 RAM pack and was off to the races immediately. Worked just fine.

Great, have a bit of room to play in then.

TMAOne wrote:The other questions I'm even less qualified to answer. If you haven't seen it already, you'd find this thread interesting: viewtopic.php?f=7&t=1006

Thanks, I saw that thread and It was helpful. I haven't decided on how daring I'm going to be with it... My knowledge of electronics is just enough to get me in trouble if I don't check my assumptions. On the one hand I didn't pay alltogether that much for the Lambda so I'm not risking loosing much if the modding goes badly, and the ZXpander makes things soooooo much easier. On the other hand Its an interesting ZX-81 clone, with it's own uniqueness, which might be worth preserving, and I don't think there's really anything unique enough about the Lambda, if any thing at all, that would justify the effort of turning it into a less-then-compatable ZX-81.

-John-

Picked up a second working Lambda so I've decided to be a bit daring and mod it. One of the changes I want to do is to bring the missing CPU lines out to the edge connector. The question is, electrically, what is the best way to handle the lines? My inclination is to run them directly from CPU socket to the connector pins.

-John-

I am not sure for the Lambda, but the Sinclair ZX80 and ZX81 have the dataline (databus) not directly wired from the CPU but from the RAMs. There are resistors decoupling the (external) RAM content when the ULA becomes active during video display and takes the opcode from the (external) RAM and serves a NOP to the CPU instead. So there are resistors between RAM and ROM databus to the CPU databus. Only the ULA is connected directly to the databus.
The address bus is taken directly from the CPU for ZX80 and ZX81.

As far as I know, the Lambda is compatible to ZX80 and ZX81. So if you want to have a compatible expansion slot it is maybe recommandable to use the datalines or RAM, decoubled via resistors from CPU instead of CPU lines directly.

these may be of use...(many apologies if I am breaching any copyright, but I found these freely on the 'net) . I know it says 'Power 3000' but it's supposedly identical to the Lambda!

PokeMon wrote:I am not sure for the Lambda, but the Sinclair ZX80 and ZX81 have the dataline (databus) not directly wired from the CPU but from the RAMs. There are resistors decoupling the (external) RAM content when the ULA becomes active during video display and takes the opcode from the (external) RAM and serves a NOP to the CPU instead. So there are resistors between RAM and ROM databus to the CPU databus. Only the ULA is connected directly to the databus. The address bus is taken directly from the CPU for ZX80 and ZX81.

As far as I know, the Lambda is compatible to ZX80 and ZX81. So if you want to have a compatible expansion slot it is maybe recommandable to use the datalines or RAM, decoubled via resistors from CPU instead of CPU lines directly.

Thanks for the explanation, from the schematics I've seen I couldn't make out where the expansion connects into the lines. I'll have to open up a lambda, trace out the expansion and see how things are laid out, but a total ZX compatible rework of the expansion is probably beyond me. At very least I'm looking to return the missing INT, NMI, HALT, BUSACK, WAIT, BUSRQ, RESET & MI lines to the connector.

gozzo wrote:these may be of use...(many apologies if I am breaching any copyright, but I found these freely on the 'net) . I know it says 'Power 3000' but it's supposedly identical to the Lambda!

Yes, it's the closest thing I've been able to locate as well... at some point I should spend some time poking around inside the Lambda and see.

-John-

BUSRQ and BUSACK are not really important.
I don't know any hardware using these signals (for DMA access only).

PokeMon wrote:BUSRQ and BUSACK are not really important.
I don't know any hardware using these signals (for DMA access only).

Maybe you don't know Wilf's High Speed Interface?
http://www.user.dccnet.com/wrigter/index_files/HS4.htm

Siggi

siggi wrote:

PokeMon wrote:BUSRQ and BUSACK are not really important.
I don't know any hardware using these signals (for DMA access only).

Maybe you don't know Wilf's High Speed Interface?
http://www.user.dccnet.com/wrigter/index_files/HS4.htm

Siggi

No, I don't know. Do you have it at home ?`

I think this is more a theoretic study or case model to show what's possible.
By the way you need 2 TS1000 and have one TS1000 running approx. 2 or 2.5 times faster than in SLOW mode.
And need additional hardware.

If you want it faster and can live with a smaller display it would be more economic to reduce (maybe half) the screenheight to have nearly the same effect with a simple software driver.

It would be more interesting to balance the program execution between these two computers like a multi core or multi threaded application.