I/O Cartridge (Show and Tell)

[Rolo]

PCB would be enough. I can do the soldering myself. Thank you very much. Maybe we establish an email account for PM.

[Rink]

I'm a bit wary of posting any email address here just because of the amount of spam I get already. If you're a member of the Retro Computer Museum or Sinclair ZX World forums then you can find me on those with the same username and PM me there.

Since you're just asking for the PCB then we just need to find a safe, secure way for you to transfer your address across to me and I'll send that right out.

[tormod]

Until/unless we get PM activated, I can suggest some alternatives:

• Add contact info to your profile. Only forum members can see your profile. There are some suspicous-looking members though, but I think they are "sleepers" waiting to post spam and not address harvesters. Light obfuscation of e-mail addresses is probably sane.
• Send PM on IRC. The #coco-chat channel on Freenode is the right place to meet for Coco/Dragon fans!
• Use public key encryption 8-<:-) We'd better have a separate subforum for those conversations

[Rink]

[*]Add contact info to your profile. Only forum members can see your profile. There are some suspicous-looking members though, but I think they are "sleepers"

Good plan. I doubt the odd sleeper will contribute much to my email woes.

Rolo - I've put my email address in the "Interests" description that appears on my profile. Drop me a line and we'll see about getting that board off to you.

[Alastair]

I suspect that you have seen the good news that PMs are now enabled? All we had to do was ask

[Rolo]

I finally had some time to play around with Rink's fine I/O cartridge. I did not start any big project, just made some small tests.
Handling of the cartridge is really easy: output via &Hff40 and input via &Hff41.
I tried the fasted switching loop I could make (without speed poke):
- one-line basic program: ~20 ms (~50 Hz)
- machine language: ~23 us (~43500Hz)

speed test

DigitIO01.JPG (1.04 MiB) Viewed 4535 times

I connected some of Fischertechnic's components, a building set, that I liked very much, when I was a kid, to see, if that works as it should. The FT-components are specified for a supply of some 6...7 VDC. If you feed with the standard FT transformer, voltage is very weak and unstable. It drops with every light bulb, that is switched on. A 5VDC signals of the IO-cartridge's output often is not interpreted properly, i. e. relays easily start to oscillate. Putting a gate in between (if you have one (they were not cheap)) helps. But if it comes to operating a real model with motors, magnets, lights and so on, changing to a solid laboratory PSU probably is the best thing to do.

FT's electronic cmponents

DigitIO03.JPG (952.73 KiB) Viewed 4535 times

Another point, which we've already discussed somewhere above, is the choice of connectors. I haven't found anything that really convinces me. It surely depends on the hardware the IO-cartridge is connected to. In addition there always is a mechanical problem of pulling or pushing wires, which forces have to be sustained by the Dragon's edge connector. I'd propose a solution with a separate patch field, like that one on the picture. I built that decades ago (oh boy!) for the official "Fischertechnic computing interface". That 20pin connector is common, cheap and easily squeezed onto a flat cable. Well, we know that: Dragon users call that the printer connector .

custom patch field

DigitIO04.JPG (263.01 KiB) Viewed 4535 times

Summing up I really have to say, it's a pity, that I did not have that back in the day! The cartridge is easy to use and to understand, but still lets the computer be a real computer, other than some of the modern systems, which, in my opinion, underestimate the childrens' intelligence, and reduce computer and programming to a bunch of clickable coloured icons, which have to be arranged in the right order. Personally I do not like that kind of abstraction layer, which totally hides the computer and leads to users not knowing anymore, what they do and what happens. Dragon's Basic is easy to learn and quite sufficient for controlling all of those typical models, that we built back then, like elevators, assembly lines, constructing machines ...
Well done, Rink!
___________________________
PS: Is there a way to shrink/autosize the photos, so that they are fitting onto the page?

[Rink]

Lol. Cool stuff mate - I've never played with those blocks, but they look like lots of fun.

Strange choice of headers, but then the ones I sent you are horrible so I can't blame you for not using them And a patch box is a good idea - wish I'd thought of that when building the thing.

[Rolo]

Yeah, my headers really are crappy, too. I'll change them again. At the moment I have no good solution on my mind...

[devo]

Are these I/O boards still available, posting to Australia? This project to control the AY 3-8910 sound generator chip looks interesting, Rainbow Magazine Aug 1984.http://archive.org/stream/rainbowmagazi ... 6/mode/1up Has anyone tried similar with a Dragon as controller?

[Rink]

Are these I/O boards still available, posting to Australia?

Well, they were never really for sale. I'd thought about it early on but then I re-used the board shape from an earlier project which made the PCB larger than needed and pushed the manufacturing into a higher price bracket. I figured they came out too expensive for anyone to be interested in using.

That said, if a WoD user wanted one, I'd do my best to help out. As long as I wasn't losing money on the build, I'm more than happy to flog a few.

Can't see any reason why you couldn't hook up an AY chip. The AYs (and other similar chips) is a much more conventional way of providing I/O ports - I just did mine with basic latches etc. because I had them lying around.