24 December 2016

My first Nixie based instrument at work

Since the Philips PM2422A instrument works, it has been put into use. Here it is measuring a 1% (vintage) 10.7 Mohm resistor.

The reading has been jumping between 10.71 and 10.70, and finally settled on the latter value. I wonder if there is a warm up time for these instruments?!

The picture also shows the emulated "1" leftmost symbol.

22 December 2016

My first Nixie-based instrument

At the ARI Torino HAM rally few Sunday ago (Nov, 27th 2016) during my third or fourth tour of exhibitors I spotted what looked like a box with Nixies. The box read "Philips PM2422". A quick picture search with the smartphone confirmed my suspicions: a 3 and half digit multimeter, with Nixies. I was even able to power it up right there, and the tubes happily glew.
I paid 30€ and I think it was the right price. The seller then realized he could have asked more, I nodded and I replied that I did not negotiate the price on purpose.
Once home I opened it, and everything looked fine. No fat electrolytics, no sign of burns. But a surprise: there are only three tubes inside, not 4 or 4+1 as in the PM2421. The most significant digit is a long neon lamp, so there's implicit leading zero blanking. And the minus sign is another small neon lamp. That's it. Cheaper than full feature tubes but effective. I think more recent versions of the same instrument carry four tubes. I got a PM2422/A2, while they went up to /A5.
Finally I own a bench multimeter!
Oh, by the way, can it be a clock? Well, it can count from 0 to 1999, so as a 24h clock it would work up to 19:59. It is suitable as an office desktop clock since most people leave way before 8 pm. Or simply make it a 12h clock. I want to explore if a simple DAC produces a stable enough reading, soon.

12 December 2016

Nixie Tube Ciapapuer - 2 - video

If you are curious how the Nixie Tube Ciapapuer looks, here's a short video:

In the end I dropped the fading effect descried in the previous post and opted to include an "off" step in the sequence.

The object is now sitting on my office desk and pretends to be reading visitors' minds. J One funny moment was when a colleague picked up the box and the "off" moment kicked in, that caused a bit of surprise and worry. I should include an accelerometer and intentionally switch off the tube when the box is moved.

06 December 2016

Nixie Tube Ciapapuer - 1

Or, in plain English, "Nixie tube home decor".

Literally: "Nixie tube dust collector."

Why, why would I ever want to build a device with just one Nixie? First of all, to use symbol Nixies like ZM1001 (side view) and B5094 (top view). Then because: housing is simpler than using several tubes, firmware is pretty simple (no RTC or strict timings), circuit is simple, it will result in a compact object.

My first attempt is with a ZM1001 that has 6 symbols, so there are 6 cathodes to switch. The typical 74141 driver has 10 outputs, nevertheless since I am lazy I choose it for this project, thus reducing the connection to uC to 3 data lines.

In order to differentiate from my previous projects and learn something new, this ornament firmware will also:
  • randomize transitions and next symbol
  • fade-in/fade-out
Fading requires one more control line. One way is to PWM the anode voltage through a PMOS or PNP: this requires a component with high Vds/Vce (that is missing in my junk box). Another way that is doable in this project is to switch the BCD combination to unused 74141 outputs. Either way uses a 4th output from the microcontroller, but the latter is "low voltage" and implemented in the basic circuit "by design."
So, going for fading through switching to unused outputs, a look at the 74141 truth table indicates that I can use the MSB (Most Significant Bit) to switch over to unused outputs (8, 9). Indicentally I had not planned to use those outputs, so now it is a matter of writing the proper software routine.