As if this hasn't been enough of an ADVENTURE!

I have a distinct recollection telling myself I'd help my chemistry studies by programming up molecule simulations using the terribly primitive character graphics on the uk101. Of course I never did, and I got poor grades in chemistry. But these facts are not related.

And 2MHz would not have been fast enough, although I think my uk101 was overclocked by a similar simple rewiring.

Sorry...



Well it is working again @2Mhz. Just the one crash, might have been a glitch or something.. I think I'll fit a turbo button somewhere on the board. There is already a power light next to the reset button. Still need a power switch too, but Maplin are rip off merchants for these little things. I want to run it off 3xAA batteries so it's portable (it does work at 4.5v, I tested it) because I can't get into the usb-rs232 dongle for the +5v as it is entirely cast in tough, translucent plastic. It is bloody indestructible, they should use this stuff for tank armour or something!

Might try a 4Mhz run later

I gutted a USB extension lead once to get at the 5V line. I made a mess of it, but that's optional. It should still work as an extension lead.

Some new pictures. I found a Duracell promotional pack with the perfect plastic box to use as an enclosure. The board fitted in with a little trimming. I also added an on board power supply (3xAA calls; it seems to run fine on 4.5v at 2Mhz), a reset button, power switch and indicator LED.

Better watch out for static electricity with a plastic box like that.

Any idea how long it will run on 3 cells? That's amazing, considering the hefty PSU for the original, and the temperature of the heatsink. All that TTL and the RAM chips add up to something.

Sure? If the ICs are all on the board there ought not be any problem. None observed so far, at any rate, and having it in a neat slimline case (you should see the size of the project boxes I looked at so far.. yuk too tall) will afford it some physical protection. Let's face it, those point to point wires are fragile!



Hi Ed

Ah well, the original UK101 has lots and lots of ICs on it that this one doesn't. I'll wager that using a 32K ram / ROM chip instead of all those tiddlers on the original board saves loads of watts. To say nothing of the video circuitry, tape circuit, etc, etc. No wonder the power supply got toasted regularly.

This Micro UK101 has a 200mA draw (measured) and with alkaline 2100mAh cells I'd estimate 10 hours, maybe a little less. Good enough. Of course, it isn't on all the time, only when I want to muck about and run a proggie. I'm experimenting with VT100 control codes at the moment, and how to position the cursor as fast as possible in BASIC. Came up with this:



CL$ is a two dimensional array of cursor placement strings, where printing CL$(x,y) sets the cursor to an absolute position on the terminal. The idea is to pre calculate all the values for speed (now don't say "6502 machine code!", I know, but it's to much of a learning curve for the time being.

Unfortunately, and for some unknown reason, it crashes whilst populating CL$() (maybe it is like machine code). Which is odd, because the adventure program works just fine, and it is stuffing a load more arrays with data than this. The program runs as-is in another BASIC I am using (on the Mac, called Chipmunk BASIC), so I think there's nowt wrong with it.

I'm convinced that this is the reason I built the Micro UK101...

Unless there is ESD protection on the board, having the ICs all connected doesn't offer a whole lot of protection. A piece of cardboard, or paper, between the board and the case should help. A piece of anti-static foam would be even better.

ESD doesn't always manifest itself as a immediate problem. Sometimes it can slowly damage a component, and reduce the life span.

Hmmm... ok, thanks for the explanation. Anti static foam? That black stuff that ICs are delivered on? Is it not conductive?

Usually that black (or pink) foam has a very high resistance. You can verify with an Ohmmeter. The conductivity is enough to dissipate the static charge, but usually not enough to cause problems. I've used it on a number of occasions to put boards on, when they're on my desk.

Could this be the well-known(!) bug in the garbage collector?

There's supposedly a fix here:
http://www.osiweb.org/osiweb/software.html

See also here: http://www.pagetable.com/?p=46

Yes, Ed, it could. i thought that too, but I didn't know where to get a fix (thanks for that). Might be a good idea to test it on an emulator.

Need to patch the ROM & reburn... duhh...

Like I said, it's why I built the thing. endless tinkering..

Cheers

JonB

Now that's interesting, I always thought those black anti-stat mats they used to puts IC's on, were very low resistance to keep all the pins at the same potential.

I just measured that black "foam" at between 5K-7K, and some trays containing FPGA's from Avnet at infinity, thereby proving your point.

EDIT: Added "...thereby proving your point."

They will get down to about 1V per cell too, so it's not like they will stay at 1.5V per cell through their lifetime. I did alkaline battery-life tests last year for a product and found that a 9V battery wasn't really done until it got down nearly to 5V. The product got about 12hours on a battery, but used a switching regulator, meaning the battery actually had to put out more current as the voltage dropped.

You could add a switching regulator to run your computer at 5V until the three AA batteries are dead, and the switching regulator won't take much room. It looks like you have plenty of room in the box. Power trends is the only brand of integrated switching regulator modules I've used, and they worked very nicely. Mouser carries them. I would not recommend trying to breadboard a switching regulator circuit yourself though, as layout, even on a real PC board, is extremely critical to good operation, related to the extremely high dI/dt's.

We have had plenty of CMOS stuff in plastic boxes with no problems at all after many years in the field, but it was ABS plastic, not polystyrene which it looks like you have.

OK, I applied the fix (basically hack the 2k BASIC-3 rom into the corresponding address of the 32k eeprom buffer in overwrite mode - using HexFiend it is a simple cut & paste job) and burned a new PROM.

The results are disappointing. It is still hanging, but it manages 11 iterations of the outer loop (Y) before crashing. This means one of three things; the fix isn't a full fix, or something went wrong in the burning process, or there is something wrong with my board.

I'm running it on the 5v switching supply now as the batteries have run out. Interestingly, this is evidenced by the serial interface dropping characters - the MAX232 circuit must be sensitive to voltage levels.

Would anyone with an 8k BASIC installation care to try this out?



It should print 24 rows of dots on the screen, each 71 chars long.