Well I'll go to the end, with the solution ... Clean your Boards !!!!

Back to the beginning. I'd been selling some multplexed nixie kits for 
close to a year and a half, now. It uses a unique form of multiplexing 
called "Charlieplexing", invented at Microchip. There it was used for 
driving LEDs. A clever member of this forum (when it was back at Yahoo), 
named Jason Harper saw that it could also be used to drive the base-emitter 
junctions of transistors, back ~2003. For our pleasure, base-emitter 
junctions, of HV transistors. In short, charlieplexing nixie tubes. That's 
what I incorporated in my kit.

One important issue with charlieplexing, is that it can only be done with 
uC tri-state I/O bits. It needs the I/O bit to not just output a one (1) or 
zero (0), but also present a high-impedance (Hi-Z). With charlieplexing, 
only one "node" (p-n, or base-emitter junction) of its matrix can be active 
at any time. One output issuing a "0", while another issuing a "1", with 
the rest "inactive" and presenting their Hi-Z state (DDR bit set to 
"input"). This becomes a key to the problem later.

Multiplexing, in general, and not just charlieplexing, can exhibit a 
problem called "ghosting". This is when more than one numeral appears to be 
ON at any time. Usually, the intended numeral is on brightly, while a 
second numeral is only slightly glowing in the background. Hence the term, 
ghosting. It can have multiple causes, that usually boil down to something 
being ON at the wrong time. A timing issue. This isn't an issue with direct 
drive displays, since they don't rely on timing.

Now to the problem at hand, and its inadvertent solution. A customer 
returns a board (for evaluation) on what's causing his ghosting problem. I 
turn it on to confirm the problem ... yep, ghosting. Pretty severe, too. I 
examine the board visually, but all the parts are in the right place, and 
the solder joints look okay. So, before probing around further, I decide to 
clean off the flux residue. When I'm done, I let it dry for a half hour, 
before applying power. Turn it On ... no ghosting. Apparently, the flux 
residue (and whatever contaminants it trapped) was conductive enough, to 
turn ON, inactive "nodes" (transistors). 

In conclusion, clean the flux residue off the board !

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send an email to [email protected].
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/7959ab9a-0e9c-4795-b37f-798f4ae822bb%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to