Archive for August, 2009

working PhidgetLED64 6-channel lamp driver

I believe cockpit building and electronics is a lot like rug weaving. As legend has it, weavers deliberately introduce a small error into their designs because “only god is perfect”. That’s my story and I’m sticking to it.

Assembly of the board went beautifully, except for one small detail: a dodgy power connector. Once fixed, the first (Master Caution) channels worked a charm.

Proceeding with the remaining 5 channels, I gained a new appreciation for my Dremel. The flexible extension and drill chuck is absolutely brilliant for circuit boards – set to the lowest speed setting, I flashed through the boring bits (geddit) without breaking a single bit.

Testing the fully assembled board revealed my very own “god stitch” – channels 4 and 5 are swapped over, probably due to a cabling issue rather than a problem with the board. It has no effect on the final result, and I’m happy as larry that it works.

lamps_worknig (51k image)

how did I miss these guys?

CNC cockpit kits. One to watch.

slip rings

new toner transfer method

I was having inconsistent results with the Press’n'Peel Blue toner transfer film using an iron, so I thought I’d try a different tack. 2 G-clamps, a thick piece of nice flat aluminium, two solid pieces of wood, and make a clamped sandwich with the flat aluminium against the PnPBlue, and the wood spreading the clamp pressure evenly. The whole thing went into the oven at 150deg C for 30 minutes. The results:

only a couple of transfer problems, touched up with a PCB pen
board1 (51k image)

clean! Ammonium Persulphate, do your stuff
board2 (29k image)

teeny bit of dremelling where my PCB pen touchups were a little enthusiastic, and it tests well
board3 (56k image)

so here it is

cheesy enough for ya?
pit_aug09 (48k image)

Work’s progressing nicely on the pit for the expo. I’m deep into the PitManager software now, and it’s getting a good refactor as I’m building it to best OO practice.

Work on the HSI has stopped for now, but that doesn’t mean I’ve stopped thinking about it. After the expo I’m going to build a VSI and RMI/Compass, to practice the skills I’ll need to realise the HSI. The design is pretty tight now, I have access to gear cutters for the ring gears (if I ask Scott REAL nicely!) and all that’s left to do is prototype it.

MCSSOH working again

Well, this was weird.

After extending the wireless doorbell receiver away from the MCSSOH module via an ethernet cable, it didn’t work. When I took the cable out of the loop again, it still didn’t work.

Uh oh, I broke something.

So I did a bit of poking around and determined that there might be some sort of issue around the rectifier, that was intended to take the AC signal through the speaker wires and convert it to DC to drive the PIC pin. A bit of googling revealed that someone else had done something similar using a reed relay rather than a rectifier.

So, I pulled the rectifier and put the reed relay over the AC pins (using wires out of the existing DIL socket), leaving the DC pins unconnected, just to see if the relay would be closed by the doorbell.

Not only did it close the doorbell, it lit the SSOH lamp! With the PIC pin completely disconnected from everything! Wtf?

All I can think is that the inductive load on the doorbell has somehow compensated for a fault elsewhere in the circuit and by some coincidence two wrongs are making a right. I will be totally unsurprised if it just stops working sometime in the future, but for now I gaffer-taped down the reed relay, closed it up and re-mounted it in the pit :) The missus is happy that she doesn’t need to traipse down the stairs to get me late at night if I have my headphones on.

great source for cheap steppers

Thanks to Scott Hendry for the link.


Cool article on using light-dependent resistors to sense movement in small ranges (example uses an F-16 high-G HOTAS).

candidate PCB

Herewith the first cut of a 6-way PhidgetLED64 lamp driver PCB: {{popup pl64lad01_01_PCB.gif pl64lad01_01_PCB 677×720}}clicky

And here is apparently the best toner transfer paper for printing PCBs:

*edit* Luckily a very kind gent on the AllAboutCircuits forum picked up an error with the board – the jumper pads should have been connected to the MOSFET sources, not to the #6 pins. Doh.

{{popup pl64ldv2.gif pl64ldv2 660×699}}clicky for final version

converting a PhidgetLED64 to drive 12V lamps

My Master Caution and targeting Mode Control Panel illuminated switches have a 12V lamp in them, which can’t be driven directly by my PhidgetLED64 which is designed to have directly connected LEDs. So – after much mucking about, the following:

your mileage may vary depending on the qualities of your lamps
pl64ladr (3k image)

hoping the magic smoke will stay inside
pl64ld_01 (48k image)

it’s alive!
pl64ld_02 (48k image)

Return top