Just arrived: ZX Spectrum 48K

[carlsson]

I've joined the Dark Side. I just traded some random Commodore stuff for an unboxed ZX Spectrum 48K. This is my first (well, second) Z80 computer. The main reason I acquired one is because I already have built my own divIDE interface, and in combination with a CF adapter I will transfer Spectrum software to it.

My collection is slowly growing...

[hawk]

Cool...I'd love one, as they're a classic, but I can't justify the price they sell for here in Oz.

[MacbthPSW]

I'd like one too, but so far not enough to chase one down and presumably pay overseas shipping. For now my ZX81 and Timex-Sinclair 1000s will have to do.

[Stinky]

Cool...I'd love one, as they're a classic, but I can't justify the price they sell for here in Oz.

They're cheap enough in the UK that it's worthwhile shipping one to Australia. Good fun.

[carlsson]

Actually, I was watching a handful of eBay UK auctions the past weekend. Neither the loose nor boxed ZX Spectrums went quite as cheap as I had expected. It is all about timing though.

[carlsson]

Now I have introduced Mr. Soldering Iron to my ZX Spectrum. I modified it from RF to composite video output. It is very easy, should be doable in 10-30 minutes depending on how skilled one is.

http://www.neoncherry.com/retrorepairsh ... ticle_id=8

Then I bought a joystick interface with two through ports for the expansion bus, so I can connect the IDE/CF interface as well as a joystick.

[6502dude]

Modifcation looks like a neat project!

[carlsson]

Yes, in particular as there is an unused hole in the RF modulator to pull a wire through. Perhaps Sinclair technicians themselves did this mod to run development units onto a monochrome monitor.

In the coming weekend, I'm planning to build myself a mmbeeb MMC interface. I didn't quite source all the components, namely I didn't find a 3.3V 1086, but I found another 3.3V voltage regulator that I hope will do the trick. The main issue is just how many mA the MMC card will draw in operation. I believe the 1086 can generate up to 1.5A while the regulator I bought only does 500 mA. And yes, I have bought some odd surplus resistors that are 4k42 and 46k4 instead of 4k7 and 47k. I also have a bunch of 300 ohm resistors, and I suppose I can put them into ugly series to generate the exact resistance. On the other hand, the resistors are there to reduce 5V to approx 3.3V, so maybe the "odd" values will be within reasonable range on their own?

Speaking of which, I wonder if this rather simple (crude?) design could be implemented on e.g. VIC-20 as well. Certainly we'd need driver software, but hopefully parts of the BBC code (also 6502 based, if not written in Basic) could be reused. It looks almost too good to be true, even if the interface is read only.

[Victragic]

Nooooooooooooooooo.....................................!!!!

Faster processors... extra registers.. a Jedi craves not these things...

[Bacon]

Nooooooooooooooooo.....................................!!!!

Faster processors... extra registers.. a Jedi craves not these things...

The 3.5MHz Z80 in the Spectrum actually isn't that much faster than the 1MHz 6502 in the VIC due to the Z80 using many more cycles per instruction than the 6502. I read somewhere the rule of thumb among programmers in the 70s was that you needed twice the clockrate for a Z80 to perform as fast as a 6502. Also, see Stephen Judd's posts from the great spring 1997 Flamefest for some actual calculations and technical details. IIRC, he concluded that the Spectrum is ~1.5 times the speed of the C64.

And about those registers, with the memory speed on these machines and the 6502's use of zero page, you could argue that the 6502 has 256 external registers in addition to the ones in the processor.

[Victragic]

Seriously, I've always wondered about the speed differential, so that is interesting..