Ignition delays

[newxito]

I noticed some weird ignition delays with B-5870 nixies while entering digits into the calculator.  1 is the most affected number. If the background LEDs are on, the delays completely disappear!?! Don’t know what’s going on.

I know that the drivers are ok, no delay with IN-17 and IN-16 versions. The anode resistors should be ok, 8.2k/170V.

[Dekatron42]

I hope I get this right, but from what I remember you get a few extra ions in the nixie from the background led's which helps the gas to ionize faster. Just like some trigger tubes contain an extra electrode that keeps ions present all the time in the gas for quicker turn on times, likethe GTE175M. You can try to use a resistor in the megohms range and connect it to one of the decimal points and to ground, high enough value will mean that some ions will be produced but the decimal point will not glow but this will result in quicker turn on times.

/Martin

[newxito]

I'm sure you are right because after some more testing, I can clearly see that there is a connection between light and delays.

Actually, the delay adds some charm to the calculator :-) I guess I'll keep it that way. Another solution could be to pad the display with leading zeroes, so all the nixies are always on.

Anyway, I will make the test with the resistor with a separate tube, thanks!

[Paul Andrews]

Also, I was reminded recently of a couple of things. Firstly that older nixies often have a higher maintenance voltage. Secondly that the data sheets always recommend a minimum voltage of 200V, and the higher the better. I have certainly had tubes that refuse to perform at these low modern voltages, but work fine at higher voltages with the resistance recommended by the manufacturer. So check the actual current they pull when they are on, and consider using a higher voltage anyway.

BTW, ZM1000 have a pin specifically intended to keep the gas ionized.

When and why did we all settle on 170V?

[Paul Andrews]

Oh. Are you also using multiplexing? In my experience that can exacerbate the problem.

On Oct 29, 2020, at 7:52 AM, Paul Andrews <pa...@nixies.us> wrote:



--
You received this message because you are subscribed to a topic in the Google Groups "neonixie-l" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/neonixie-l/2FkQM46B2j8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/f74e2c30-d163-47be-a1aa-c3314d63eef7n%40googlegroups.com.

[newxito]

It’s direct drive, 5 x HV5622.

According to the datasheet, for 200V the resistor should be 18k. I don’t want to change the 14 anode resistors because I don’t want to play around the tubes with the hot air gun if not absolutely necessary.  

I can try to increment the voltage to 175V, I think that should be still ok with the used 8.2k resistors.

 

When and why did we all settle on 170V?

Good question.. not my fault, I only started playing with nixies 4 years ago :-)

 

[Max DN]

I had the same issues with ignition delays of B5870, but only in complete darkness and not all the times. It would present itself only if dark and the tube had been off for a while, if dark but the nixie had just been on a few seconds to a minute before, than no issue. After a brief search online, I guessed that it must have been a small issue with ionisation and I read that some people add a small UV LED to help with that. I opted for background blue LEDs. After adding the LEDs, I realised that the top of digit 2 and 5 wouldn't completely ignite when both decimal points on the digit where on. I increased the voltage of the power supply so slightly by changing the bottom resistor of the resistive divider to compensate for the voltage drop from the LEDs (added in series on the anode of the nixie) and it works perfectly now. Perhaps slightly different issue than yours but the idea of adding a small UV LED may be worth thinking about (I wonder whether an infrared LED hidden somewhere works as well).

[GastonP]

Probably IR would work better as normal glass and borosilicate are opaque to UV light. In the normal case, the UV LEDs emit near-UV so they have a reasonable content of blue /violet light to pass through the nixie walls and help ionization. IIRC some sleepy Nixies took several seconds to trigger and light in complete darkness. Natural beta and gamma rays help too, and that is the reason to put Kr-85 into some of them that needed certain operation.

Regarding to the incomplete lighting of the digit, it is a totally unrelated issue, that has to do with low anode voltage.