Programmable Logic DesignLine Blog
Men versus Women : Turn the knob on your side...
Clive Maxfield
2/6/2009 1:25 PM EST
As fate would have it, all sorts of things have been happening. First of all, in order to start prototyping in earnest, I had to purchase a bunch of large solderless breadboards, which I've mounted on a common baseboard to make my life easier (I wired all of the power and ground connections yesterday evening and I'll provide a photo of this setup in my next related blog).
Another consideration is that I'm going to have a lot of input switches (toggle, rotary, etc.) and – potentially – other input devices; similarly, I'm planning on a lot of LEDs and meters and other output devices. The problem is that the PICAXE microcontroller I'm using has only 8 input pins and 8 output pins (along with 4 analog inputs). The solution I've decided to go with is to use shift register chains.

The input chain will consist of a series of 8-bit 74HCT597 chips (high-speed CMOS with CMOS/TTL compatible inputs and outputs, where each chip contains an 8-bit storage register and an 8-bit shift register). This consumes one of my uC's input pins (for the data-in signal) and three output pins (one to trigger a massive parallel load of all of the inputs into their corresponding storage registers, another to copy the storage registers into the shift registers, and a third to clock the shift registers).
Similarly, the output chain will consist of a series of 8-bit 74HCT595 chips (each of which contains a shift register and a storage register). These consume three more of my uC's output pins (a data-out signal, a clock signal to drive the shift registers, and a third signal to copy the contents of the shift registers into the storage registers, which then present the signals to the outside world).
I already have a bunch of these chips, plus a pile of 10-bit DIL switches and 10-bit DIL "bar-graph" LED displays, so my next task is to wire these all up and start playing around to make sure I'm on top of things... meanwhile...
... there are several things I need when it comes to the full-up "Display-O-Meter". These include – but are not limited to – antique (or antique-looking) toggle switches, antique knobs (to be connected to modern rotary switches), and antique (or antique-looking) meters. Well, I'm still looking for the toggle switches, but I have good news on several of the other items...
A couple of days ago I was completely taken by surprise when a package arrived from my friend Douglas in California (Douglas is making the Steampunk TARDIS Console I was waffling on about a few days ago). Amongst other "goodies", this package included a mega-cool antique Phasetron Meter with which to start my collection:

I wish you could see this in real life, because it is absolutely gorgeous (much nicer than my poor photo conveys). Also, another reader introduced me (via email) to one of his friends called Martin who has a business refurbishing antique audio equipment (and creating things like vacuum-tube Theremins and suchlike – I'll be doing a future blog on Martin once he has his revamped website up-and-running). Martin and I ended up having a long chat on the phone, with the result that I now have a number of amazing antique meters winging their way to me as we speak.
But wait, there's more... Martin told me that one of the real pains was going to be finding affordable-priced antique knobs. (I'm reminded of a sketch from an old English radio comedy program called The Goon Show, in which one character is trapped in a cellar and another shouts "Turn the knob on your side." The first guy replies "I haven't got a knob on my side," to which the second responds "...of the door, you idiot!" ...you had to be there...)
But we digress... it seems that antique knobs can cost an "arm-and-a-leg". One solution that I was considering was to purchase reproduction knobs (which also aren't cheap)... but then ... while looking for something completely different on eBay, what should I run across than a job-lot of knobs that were acquired from an old electrical repair shop (these may not be mega-antique, but they are certainly well-aged – possibly older than myself, which is saying something these days):

So I hung around in stealth mode for a day or so until the auction was poised to end ... and then with only 30 seconds to go I pounced ... and won the little scamps. Once again this was an occasion for me to perform my "happy dance" (be thankful you weren't forced to watch).
And, as one last thing for us to ponder, I was originally planning on using conventional (uni-color) LEDs for my displays. But say I wanted to reflect the state of a toggle switch, for example, this means that I would either have a single LED that's ON or OFF, or I would need two LEDs (say a red one and a green one) associated with the same switch.
Hmmm... I started looking at tri-state LEDs (4 terminals; a common anode and three cathodes, one each for red, green, and blue LEDs). These are pretty cool, because you can use a single device to represent a range of colors. Using three of my shift-register chain's output bits; I could make each device display red, green, blue, yellow, cyan, magenta, and white...
"...but wait just a moment," I said to myself (I find I'm talking to myself more often these days), "this technique means that each LED has to be hard ON or hard OFF. In turn, this means that a secondary color formed from two primaries will be brighter because there are two LEDs on at once, and white will be brighter yet..."
...on the other hand, if I could pulse-width-modulate the control signals, I could (a) ensure that all of the colors were the same overall intensity and (b) actually achieve a much wider range of colors. My initial thought was that you could find an off-the-shelf chip to do this sort of thing. Indeed, there are devices available like the LTC3212, but these don't have all of the capabilities I'm looking for. But turn that frown upside down into a smile, because my friend Joe in the UK has come up with a FANTASTIC solution that will blow your socks off ... however, I think we'll leave that as the topic for a future blog :-)
2009 New Year Resolution (Goal: Walk 1000 miles at ~3 miles a day)
[A=Actual, C=Current, P=Plan-to-Date, R=Remaining, T=Total]
| Days: | T=365, C=37, R=328 |
| Miles: | T=1000, P=111.00, A=130.56, Δ=+19.56, R=869.44 |
| Note: | It's not getting any easier... |
Questions? Comments? Feel free to email me – Clive "Max" Maxfield – at max@techbites.com). And, of course, if you haven't already done so, don't forget to Sign Up




