Help!

[jmmcg]

A beginner's advice:

• Practise assembly!!
  When everything is built and the end-caps finished and the perspex cylinder washed: doing a first assembly then is a disaster: the boards are a very tight fit and could so easily scratch the perspex cylinder. Yes: the cylinder is extremely sensitive to scratches and gouges: leave the wrapping on for as long as possible and avoid gouging the inside with the PCBs! The end-caps aren't dry enough so nasty finger marks and scrapes begin to appear. The lacquer has filled the channels so the end-caps don't seat nicely on the cylinder so the PCBs don't mount in the end-caps properly. Argh! A mess has been left on the ends of the cylinder which would just love to spread down the inside of the cylinder... Avoid this nightmare! Sand the end-caps nicely, scraping out the routings in them of wood-swarf. The test fit with no PCBs. Then before any soldering has been done, put in the PCBs. Does the cylinder need shortening? Gently sand as necessary. Add the nuts to the display PCB. (see below re soldering) Repeat re-assembly noting well how the bolts fit and feel: don't over-tighten and damage the display PCB! Note the tightness of fit and whilst holding a nixie in the cylinder. Make sure the power cable has been test-fitted and familiarisation with how it should be bent. *Familiarisation* is the key to a happy final assembly and a clock that looks worthy of your effort!
• Mounting the nuts onto the display PCB.
  The stresses from the bolt can rip the nut, solder and pad right off the PCB. This is a nightmare scenario. I tried to avoid this by soldering de-soldering braid (nicely squashed in pliers to keep it flat) on both sides of the PCB then soldering the nut onto the braid. Note that the holes in the cedar end-caps might need tweaking due to the slight increase in width, in 3 cases I have not found this to be an impact, but best check! Recall: familiarisation is the key here!
• Forming the nixie holders.
  Experts, forgive me: but for the beginners (like myself): push the pins into the holes in the PCB. Once they are all in, then before soldering, push the desired nixie into it (making sure it is nicely aligned) then solder the pins onto the motherboard. This will ensure the pins are all nicely aligned and the nixie is guaranteed to fit the socket.
• Spacing the PCBs on the jumper wires.
  Test-mount the PCBs onto the end-caps (with the paxolin insulator), fitting the bolts & washers. Once this is done, the spacing of the PCBs is now known and some of the jumpers may be tacked to set the correct gap. De-mount and solder all the jumpers appropriately.
• How I finished the cedar end-caps.
  I used "Wurth Lacquer Spray Special" for the cedar end-caps. After sanding, but before lacquering, I heated the end-caps in an oven at about 180degC for two hours to dry them and this also enhances the grain. Longer & hotter would scorch the wood & also enhance the grain, but may let the teak wood turn greyish (I had heard this may happen to over-seasoned teak)... Be very very careful - if I were braver, I'd have toasted for longer for a better grain. Wait for the caps to almost completely cool - below about 40degC then apply the first coat of lacquer: this will get absorbed into the wood, so is the foundation of the grain, so the finish. It ensures the stability of the lacquer in hot weather. If applied with the wood too hot, it will bubble. Let it dry & re-sand. DO NOT put the lacquered wood into an oven, no matter how dry, otherwise you'll get a horrific "rice crispie" effect (the air in the wood will expand, forcing out the lacquer....) & a nightmare of sanding. Make sure all the lacquering is done on a dry, hot day or in a dry, hot room. Over 25degC if you can. This will make the lacquer dry better & look better. Use tons of very thin coats (avoiding tiny bubbles in the lacquer - it can "fizz"; I don't think experts would ever use spray lacquer - if I were doing it again I would not use spray either but it is oh so much easier for the ham-fisted such as I...) & piles of sanding to build up smooth layers for that "piano finish". Make sure you leave it to dry very thoroughly afterwards - I'd leave them at least a week to avoid fingerprints in the lacquer. When drying, make absolutely sure the end-caps can't slide together - they have a nice habit of doing this (I think there is a Van-der-Waals force in play here...) and once stuck together, after all that sanding to make them smooth, they part very very messily - think lots of sanding again....

[jmmcg]

One of the cylinder clocks I built has a curious issue: the IR transmitter/receiver circuit for the menu system seems to "be too sensitive". It can trigger if sunlight flashes on it, I pass by, or even it would look like it happens spontaneously! I have tried turning it off, waiting a few hours until the backup capacitor discharges, then turning it on to let the IR reception circuit "settle" and this had not effect. I have tried different houses & rooms too (again repeating the off-wait-on process). Likewise no effect. Note that I have verified that the random display setting (7-4) is not set.

Can you suggest what I might be doing wrong? I have the heat-shrink on LED8 as required in your instructions. I note that the receiver IC4 is not black plastic, but a silver-clad chip that looks like IC3 in your picture [1]. Is there something I can do to reduce the sensitivity of IC4 (temporarily) to verify if that sensitivity may be the issue? Or might the PIC chip, IC1, have a setting to reduce the sensitivity of IC4?

[1] https://www.lasermad.co.uk/forums/viewt ... 7&start=10

[Tony]

The replacement sensors we have found for IC4 (as Sharp have discontinued production of the original and there are very few alternatives with no internal AGC) is a bit more senstitve than the original. Try covering about 3/4 of the window with heatshrink for now. A software mod will probably be needed though in future kits.

[jmmcg]

... Try covering about 3/4 of the window with heatshrink for now. A software mod will probably be needed though in future kits.

[SOLVED] As Tony suggested, the addition of the heat-shrink to reduce the sensitivity of IC4 fixed it. Some experimentation was need regarding how much to cover up, but reducing it by about 50% seemed to do the trick! Thank you Tony!

[jmmcg]

When the clock does the "slot-machine mode" I notice that it stops at either 1047, 1048 or 1049 roughly evenly then holds that display for a few seconds before returning to the normal time display. What do these numbers mean?

[Tony]

If you had 2 clocks running in series they would display '1.048596'. Probably.

It made more sense as a reference back when I designed it.

[jmmcg]

If you had 2 clocks running in series they would display '1.048596'. Probably.

It made more sense as a reference back when I designed it.

So it is simply was test code to verify the communication between them?

[Tony]

Try a google for 1.048596

[jmmcg]