Electronics > Projects, Designs, and Technical Stuff
Custom Nixie Tube Clock Review
Evangelopoulos Panagiotis:
--- Quote from: Macbeth on June 22, 2019, 08:29:06 pm ---Nice! Have you tested the DP's? You may want to add some resistors to them to limit current. I recall Dave used 100k on his Nixie subscriber counter. :-+
--- End quote ---
Truth to be told I didn't pay too much attention to that. :o I thought the 22k for each tube would be sufficient. I'll check the current consumption and see if they draw too much power.
So far the code does not allow a DP to be on if there is no digit displayed thus I hope that the current is somehow distributed based on the area of the cathode, creating a physical current limiter.
Sadly enough I don't think there is enough space to fit 8 100k resistors without making it look horrible.
So far they seem to work fine... Don't know if the extra current will burn them out quicker though.
Radiohead:
--- Quote ---The HV power supply could use a bit of work but it works well enough
--- End quote ---
What problems do you have with the HV supply? It looks like you try to power quite a few nixies with the MC34063, eventhough you've included a darlington PNP transistor to boost the efficiency.
I like the fact that you've outlined most IC's and added most component values in the silkscreen. It makes the PCB look quite nice without the components on the black silkscreen. (I'm used too getting PCB's without most silkscreen or designators due to size contraints, the massive amount of vias or designers lazyness)
Evangelopoulos Panagiotis:
--- Quote from: Radiohead on June 22, 2019, 09:33:51 pm ---What problems do you have with the HV supply? It looks like you try to power quite a few nixies with the MC34063, eventhough you've included a darlington PNP transistor to boost the efficiency.
I like the fact that you've outlined most IC's and added most component values in the silkscreen. It makes the PCB look quite nice without the components on the black silkscreen. (I'm used too getting PCB's without most silkscreen or designators due to size contraints, the massive amount of vias or designers lazyness)
--- End quote ---
Yeah since I power a lot of tubes the ripple voltage is quite significant and the display flickers. Not so much that's annoying but a flicker non the less. Maybe the switching frequency is a bit low? The MOSFET runs quite cool but if you remove the output cap the MC34063 freaks out and the MOSFET overheats in seconds.
In a future version I'm thinking of changing the HV PSU from a boost to a transformer based one. Either flyback topology or something else.
For version one I'm surprised of how good it works. :P
Detailed silckscreen is the best. I like seeing the values on the board as it also helps when debugging. :)
schmitt trigger:
What you could do, to require a lower step up ratio and a correspondingly lower duty cycle, is a diode-capacitor voltage tripler.
For instance, if your Nixies require +180v, then you design your supply for +60v and allow the tripler to boost it.
This tripler boost is not my idea, I picked it up from the website below. I had a circuit with both Nixies and Dekatrons, which require +510v anode voltage. It works flawlessly.
Scroll down towards the bottom.
https://threeneurons.wordpress.com/dekatron-stuff/
David Hess:
I would use common and inexpensive low voltage 1-of-10 decoders or the 74HC595 outputs without decoding to drive discrete cascode output transistors instead of rare 7441s or other integrated Nixie drivers. This would mean more wiring and parts though.
Navigation
[0] Message Index
[*] Previous page
Go to full version