I'm attempting my first multiplexed nixie project, in this case a 13-character / 9-segment display. Direct-drive is always my preference, but that is not an option due to the way the display is pinned-out.
I'm using an FPGA to control the multiplexing, so I've put a lot of software-controllable registers to set the time when anodes and cathodes are turned on and off to prevent ghosting. For my display, the blanking time is 20-100usec per the datasheet; I believe this is the time required to dissipate the charge from the ionized gas. I measured less than 10pF of capacitance between the various terminals, but I'm certain this is an underestimate of the actual charge that an ionized segment holds, so you really cant calculate the correct value of the bias resistor based on the time-constant from the blanking time and the measurable capacitance. The anode-driver is a PMOS current-limiter that can be switched on or off, and the anode will be biased around +100V thru a 5meg resistor (two 10-meg resistors dividing the anode supply in-half). Cathodes are driven by NPN pulldowns that are current-limiting (3mA per datasheet). If I'm unable to eliminate ghosting by changing timer values, I will reduce the anode bias resistors. Unfortunately I have 3 projects going right now (too many!!), so I wont have any progress update on this for awhile. On Monday, May 24, 2021 at 10:25:44 PM UTC-7 Richard Scales wrote: > I think I am getting a little closer to understanding this all a little > bit more. I have almost grasped the concept of clamping the cathodes at an > intermediate voltage to aid switching them on and off in a timely manner in > an effort to minimise ghosting. > > Now it's all down to driver choice and chip count. > > How about HV5812's? i have seen these used in multiplexing circuits - > especially those where a group of say 6 nixies are split into two groups > then the 20 bits from the driver are split into two groups of 10 bits, > each group for one set of three nixies. that works just fine for nixies, > not so good for a 16 digit device with cathodes common to all! > > Regardless, it would be slightly more efficient than using a 32 bit device > though ultimately an 8 bit device would suffice. Using the HV5812 I can see > that I can just connect Vpp to 80V, job done. If I wanted to use HV513 (on > the grounds that I only need 8 bits) would I then just clamp each cathode > to 80V via a suitable resistor? > > I guess I could simply not worry about it and waste the bits, I'm only > ever going to do this once or twice and one of my design goals is to make > this thing as small as possible. > > For the high side i definitely need to switch 16 anodes so I need to > employ a push/pull device - HV5523 will do that or I could perhaps use a > couple of HV513's - again, the easiest thing to do would be to use a single > HV5523 and waste half of the bits. > > In an effort to home in on a solution - a single HV5523 for the 16 anodes > and a single HV5812 for the cathodes (clamped to 80V) - have I missed > anything crucial here? > > I just feel that I should get a handle on this multiplexing thing. I now > have a couple of different multi digit panaplex displays coming so I need > to come up with a plan. > > I have some ZM1500 (12 digits) on hand and some IGP-17 (16 digits) on > their way. > > ... and I've not even started to think about the timing of all the various > switching operations yet! > > Thank you all for helping me with my education > > - Richard > > On Friday, 21 May 2021 at 19:13:35 UTC+1 nixiebunny wrote: > >> Richard, >> A single push-pull HV driver chip will not be able to drive a multiplexed >> display. >> >> The reason is that it only connects each tube element to either 0V or >> 180V. It is not capable of leaving an element disconnected. >> >> >> On Thu, May 20, 2021, 8:45 PM Richard Scales <[email protected]> >> wrote: >> >>> Sorry, I mean using the same type of device to control the anodes AND >>> the cathodes but but the actual same piece. >>> So, using an HV5523 (or perhaps multiple HV513, multiple because there >>> are probably 8 cathodes (seven segment panaplex displays) and 16 Anodes (16 >>> digits)) could be a way forward then. >>> Please could you let me know more about the need to clamp the cathodes >>> to a certain voltage? (Exposing my real lack of understanding here!) >>> - Richard >>> >>> >>> -- 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 view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/0c9ff752-19c1-46de-87bd-39b03273c956n%40googlegroups.com.
