Erratic Harrison clock

[Blitzen]

Hi all,

 What could make one of the Mike Harrison Nixie clocks run erratically? It's basically running fast, but it also does this erratically- sometimes it looks like the seconds are counting at the proper speed, but then they will speed up...?

 I'm using the pre-made board with an added seconds driver board, and it's a new build; just testing it at this time.

 Thanks for any ideas,

-Brian

[Mark Moulding]

 What could make one of the Mike Harrison Nixie clocks run erratically? It's basically running fast, but it also does this erratically- sometimes it looks like the seconds are counting at the proper speed, but then they will speed up...?

I build clocks using an 8051, synchronized to the power line.  (I use a similar circuit - a resistance and voltage clamp diodes, but I pull the AC off the power transformer, and use external clamp diodes - still, there's no reason that Mike's circuit isn't OK.  Microchip recommends just this type of circuit for direct off-the-line embedded products.)

Regardless, I had the same problem with my clocks (running several minutes, sometimes as much as an hour) erratically fast each day.  It turned out that the problem was false triggering from line noise.  I solved it in firmware by ignoring the power-line pulse for 16.6 mSec after each previous pulse.  Statistically, there could still be a noise event in that last 67 uSec, but in practice this seems to have completely solved the problem.  After running for months, my bedroom clock is still within 1 minute of my (presumably always correct) cell phone.
~~
Mark Moulding

[Nick]

On Monday, 17 March 2014 17:29:15 UTC, Mark Moulding wrote:

 Microchip recommends just this type of circuit for direct off-the-line embedded products.)

Just because Microchip discuss it (AN521 & TB3013) doesn't make it an ideal solution or right for this application - those app notes don't discuss line noise, which is a serious issue - they also relate to connecting to the inputs of a uProcessor, not discrete logic - digitally filtering in a uP is easy - not so easy using standard logic...

Nick

[greg...@hotmail.com]

If you suspect line-noise, plug a noisy device into the socket next to the suspect clock and turn it on/off rapidly for a few minutes to see if the clock gains time. If your clock is susceptible to line-noise, it will gain time. I found that my Weller WTCPN soldering station is REALLY noisy; far more noise than my circular saw or garage vac.

My first-generation nixie-tube clocks use the 60Hz line signal for keeping time. I have 2 stages of low-pass filtering. The first-stage is for high-frequency/high-energy noise on the AC line, right where the AC line comes onto the PC board. The second-stage is a simple RC filter for the 60Hz reference signal to the CMOS counters with a cutoff frequency around 300Hz. Incidentally, my clock uses 4000-series CMOS gates, running at 10VDC. It has no transformer, and the "ground" of the clock is tied directly to the AC line. Yes, it's perfectly safe to run a CMOS gate tied directly to the AC line; no worries of surges, ESD, noise, etc because the other side of the AC line is well-protected and filtered.