USB / 18X woes

I wanted to do some wireless stuff to my shed & chicken coup - control of lights, remote temp sensing etc, so excitedly bought the XBee starter kit. Put them together yesterday and went through the XBee configuration over a Serial to USB cable. All that went OK. XBee's configured and passed the ping test.

Then tried to program the 18X. Serial chip removed, jumpers configured - programming editor spits out at "memory verification failed (byte 4)" Error. Tried the other board - same problem. Tried different power supplies, 4.5v (3 nice fresh out the packet AA's), freshly charged 7.2v nicad through the 9v input, 5V regulated via 7805, fresh 9v PP3 on the 9v input - all give the same problem. Swapped the 18X for another and went through it all again. Still no joy.

Tried the USB/Serial cable onto an old 18 in a breadboard setup. Programmed OK. Made up a new breadboard for the 18X which gave the now familiar error. Swapped in the old 18 and it works fine. So, tried another few 18x chips - some of which I bought some time back. The error byte varied - 3, 6, 251 but apart from one older 18x, all the others failed to program.

Now - I can't believe TS have just sent me 5 dud 18x's - but what else could it be? I've never had any problems in the past - except when I switched to the USB cable - after which I seem to have had problems downloading programs where none existed before.

Any ideas please?

Tht's been reported before. The 18X seems to have different input hardware from the 18A and does not work with certain USB->serial adapters.

Those who have reported the problem also report that the 18X is fine on a "real" serial port or with a better (closer to RS232 spec) USB->serial adapter.

Do a search for 18X* (need the * because 3 character groups are ignored) and USB.

John

Sounds familiar!
http://www.picaxeforum.co.uk/showthread.php?t=10772

Are you using USB010, AXE027 or something else?

BeanieBots said:

Sounds familiar!
http://www.picaxeforum.co.uk/showthread.php?t=10772

Are you using USB010, AXE027 or something else?

Click to expand...

Thanks for the pointer to that thread.

I'm using USB010. I'd switch to the 027 but I tend to breadboard using the Shrouded 3 pin header (CON038) - just because it's quick and easy. There is a com port on the motherboard - but it needs a backplane bit which wasn't supplied. I've ordered something which will hopefully fit (fingers crossed). I'd make my own but blowing up PICAxes I can handle - blowing up a whole motherboard ... dunno if I'm prepared to trust my skills that far!

Quick & Final update for anyone in the future searching with the same problem : I tried the USB cable on an IBM lenovo laptop and everything works fine.

That's interesting. I tried two HP machines and had the same result on both.
I've done a few more tests to try to find route cause.
At first, it looked like pressing F5 gave better results than using the mouse to click on "run". This was then whittled down to moving the mouse during download caused a problem.
As mentioned in my earlier thread, this suddenly happened in my case.
I can only conclude that some sort of windows update (or anti-virus software) has reduce the serial/USB driver priority and now timing can be out far enough to cause an error. Unfortunately I don't know windows well enough to investigate this theory any further.

How much "always on" stuff is in the system tray?

PDA sync, camera picture transfer, and similar "in the background" applications all eat up some resources.

Some updates definitely break things. There was a Flash update a couple of weeks ago that broke the "virtual card number" application for one bank's credit cards. The application couldn't read the Flash version number and complained that it was too old - that's an OOPS on somebody's part.

John