Nixie Tube Driver using HV5222

Driving a nixie tube with the HV5222 32 bit shift register high voltage driver
PCBWay Supports High Precision Advanced PCBs: www.pcbway.com/
PCBs at PCBWay www.pcbway.com/project/sharep...
Design files github.com/GadgetReboot/HV522...
Nixie Tube Info hackaday.io/project/1940-modu...
High Voltage Conductor Spacing www.smps.us/pcbtracespacing.html
If you'd like to support the channel, you can:
Join a membership on this channel: / @gadgetreboot
Support me on Patreon: / gadgetreboot
Become a channel member (click the Join button): / gadgetreboot
Shop through my affiliate links (I may earn a small commission, costs you nothing)!
Banggood www.banggood.com/custlink/3vK...
Aliexpress s.click.aliexpress.com/e/_AFeHwB
Amazon.com www.amazon.com/?&_encoding=UT...
Amazon.ca www.amazon.ca/?&_encoding=UTF...
Videos are mirrored at LBRY/Odysee: odysee.com/@GadgetReboot:4
#ElectronicsCreators

[@sootikins]

That power-up random output glitch is odd. I've done quite a bit with those Supertex/Microchip HV-series drivers and have never seen that. I'm not doing any sequential power-up tricks. The only thing i can think of is that I've always used the versions of the drivers that have a blanking pin. I connect it to an output on the controller for various reasons. I suppose that output is floating for the first few milliseconds as the program gets going. Since the blanking pin wants to be driven high to not-blank maybe the glitch is happening but being blanked.

[@GadgetReboot]

Yeah I assume I'm catching the output glitch due to my board pulling OE high until the controller can turn it back off and blank the outputs, so I think controlling the 12v rail is the next best thing if the controller can boot, configure OE for blanking, then turn on 12v so the driver can power up with OE set to blank it. Next time I may just use a proper level shifter so OE is less janky.

[@TheEmbeddedHobbyist]

There is one issue with your design i think quote from the datasheet " HV5122 / HV5222 are designed to be used in systems which either switch off the high voltage supply before changing the state of the high voltage outputs or which limit the current through each output."
Normally I would hold the OE input in a state to disable the outputs until I'm happy with the data to be displayed, but with this chip the OE enable in the disabled state pulls all the the outputs low, so turning on all the segments at once. This could lead to one unhappy hot chip and nixie tube.
looks like the only option is to send a set of segment off's through the chip at startup.
The problem with most Arduino designs is the bootloader will get in the way of good circuit design. At the time when you want to clear the display so you won't see random segments lit, the Arduino is sitting waiting to see if it's required to go into load mode before running your code. So you always get an issue with resetting displays.
The way round this is to remove the bootloader code and programme the chip directly, then the first code that runs is your clear display code. problem solved.
One Nixie fan boy here, waiting to see if Dalibor Farný can bring down the cost of his in design 'F' series. czcams.com/video/3nsg0q2Pz6g/video.htmlsi=IeKoDgNgpQmEIk9B
Edit was going to add, look for an old copy of EN 61010 safety standard this will provide great information on all aspects of circuit design requirements. great late night reading 🙂

[@GadgetReboot]

I found the datasheet somewhat confusing to interpret and lacking in some details, but for the high voltage supply switching or current limited outputs, I assume the fact that the nixie tube has a current limit resistor ( a few mA) would cover that part, but otherwise I was leaving it up to the broader project implementation to have another circuit potentially control the HV supply (the supply I'm using has a shutdown pin. I just need to re-learn what kind of voltage it wants)
I'm sure that when OE is set to disable, it keeps all outputs off rather than on (and confirmed where I have only one LED or nixie digit coming on at a time, and I use OE when updating the shift register). The datasheet pin description "When OE is LOW, all HV outputs are forced into a LOW state" was one of the confusing things. I think they are saying the output driver FET gates are low, so it is open drain mode.
I think Strobe is the pin that would force all outputs on at once (and I have it tied to vcc to never do that). When I look at the chip logic block diagram, the way it has OE and strobe going through the And and OR gates to control the FET, that all seems to make more sense.
If someone could start producing new nixie (and similar types of) displays again that would be good! I've been having fun using various old vacuum display calculators around the workbench rather than letting them sit in a box doing nothing.

[@sootikins]

@@GadgetReboot Okay... now you are starting to rememind why (oh so many years ago) I decided not to use the (slightly) cheaper HV5222 and went with the HV5530. I designed and built one clock using the 5530, was quite happy with the chip's performance, ease of use, lack of weird restrictions and to a lesser degree its data sheet (which could be better!). After that I never looked back - I just buy 5530s (if it ain't broke don't fix it) and so have no hands-on experience with the 5222.
So my advice would be to just get some 5530s and perhaps some legit level shifters. Oh, and ditch Arduino IDE and go bare metal if at all possible! There are packages out there that allow you to use gcc with AVR chips. You can use an UNO as a programmer to flash a bootloader-free Nano if you don't want to buy a hardware programmer. You'll have to Google how to use the UNO as I've completely forgotten how to do it since I get a hardware programmer a while back but I do remember it wasn't real difficult.
If you get some 5530s I can even give you the C code I use to drive them, if you want it.