keypad progress

Nice...

Though eventually the conducting matter in the button pod will wear off and then contacting becomes gradually tougher. Like with the FFWD and PLAY buttons on VCR remotes to fast-forward through commercials. They wear out while all others work just fine.

I'm toying with an idea of examining the HDMI stream and see if I can identify commercials... so I can automate the fast forward process ;-)

The other annoying thing, at least around here, are those damned storm warnings (*) that pop up with an annoying, and loud, klaxon. At least here with CenturyLink I can quickly press "Exit to TV" and terminate it... I want to automate that as well ;-)

(*) Winds and floods which don't impact those of us out in "boonie land"... only concreted-over city regions get street floods anymore.

Sometime, when I feel the urge, I'll go out and photograph all the flood controls we have here in backward AZ. (I live adjacent to a dredged section of the Queen Creek Wash... nice piece of park... sometimes the bottom gets puddles ;-) ...Jim Thompson

And then often white letters on a bright red stripe making it all but unreadable. Sometimes I wonder what the designers of such system were thinking or whether they ever field-test stuff.

I'm seeing between 5 and maybe 18 ohms switch resistance, and I'm working against a 10K resistor. So some resistance increase will be OK. I don't expect a huge amount of finger action; serious instrument control will be Ethernet. There's also a spinner knob to keep the customers amused.

The switches are independent, not wired in a matrix, which helps a bit too. Those carbon buttons work in parallel, not series.

OT, it's officially 106F now in SF. I measure 81 indoors (no a/c),

Humidity is 18%, and I'm pretty undressed, and now and then Mo blasts my surfaces with a water sprayer, which really helps. Last night, there were no stars and a deep orange moon. Maybe we should move north.

We're nice and cool here...

...Jim Thompson

If the button isn't pressed dozens of times per day every day it's probably ok.

After a while one becomes used to the heat. I often do 30-40mi bike rides when it is above 100F. Some people think it's crazy and on the MTB trails I am usually all by myself on those days.

If we ever move again it has to be to an area with pristine mountain biking turf and, for my wife, ideally the word winter should not exist in the people's vocabulary. Southern Utah looks and Northern Arizona look nice.

Around here the messages are so important that they overlay the broadcast with quickly scrolling white print on a light green background so you can't read it anyway.

Ditto, and I posted before reading yours.

Jim-out-of-touch-with-reality-Thomspon hasn't quite got his head around the concept of 100-year floods,

You set up your flood controls so that they wouldn't have been over-whelmed by the heaviest rain that you'd had for the last 100 years - or the last 1000 years and you live in the Netherlands where they've got records going that far back.

If the area doesn't get more than occasional puddles over a decade or more, nobody with any sense would be the least surprised.

The catch with anthropogenic global warming is that big floods are becoming more likely than they used to be, even in Arizona.

I would put a carbon printing on the PCB. This lasts much longer under the abrasive stress from the carbon pills in the keypad. I have done many such designs and this really helps to increase the lifetime. It also does not corrode.

We had a VCR - 20 years ago - that did that. IIRC, which I probably don't, the commercials were preceded by a black frame, or 2, or n. Broadcasters probably changed that 'cause it was too easy to detect.

Nice. I briefly considered using WS2812 5050 LEDs in a project but decided long term availabilty was a bit of a worry. They have RGB LEDs, a controller chip with 8 bits for each color brightness and daisy-chain so you only need a single port pin on your micro to control a whole whack of them. Cheap too.

--sp

I'm using 1206 sized upside-down LEDs (firing through a hole in the PCB) to backlight the buttons. There are three sources for the LEDs, Dialight and Kingbright and LiteOn. Only single colors.

I'm using shift registers to pick up the pushbuttons and drive the LEDs. TPIC6595s drive the LEDs; the G input is handy for PWM control of LED brightness.

There's also a New Haven color LCD, controlled by an FT800 graphics controller, in turn driven SPI from a Xilinx ZYNQ. Shockingly, so far everything seems to work. We just got the ZYNQ to configure from a micro-SD and run ARM code and run a DRAM test; what a relief that was.

You might try a black solder mask on the button side of the board to help block the light from illuminating adjacent buttons.

Might try plating through the holes under the LEDs, too.

There's already a dark blue solder mask, which should block the red and green pretty well.

The LED holes are un-plated, so maybe a little light can get inside the PCB and come out another hole. Clearances discouraged plating. If I disassemble the whole thing, PCB bleed should be visible, and we could darken the walls of the holes with paint or a sharpie to see if that makes a difference. I'm guessing the bleed is mostly scattering inside the milky silicone. Possibly a bit of internal surface reflection too. Several tedious experiments are possible.

Gotta earn your money somehow. :)

Today, they are all digital, so the switching is much tighter.

Prior to the video tape, commercials were spliced into the 16mm film before it was loaded onto he film chain. There was a blank frame or two at each end, so the commercial could be replaced, but each time you lost an additional frame for each splice.