1
\$\begingroup\$

I'm using an optoisolator for high side switching a Nixie tube. For some reason, the optoisolator is not fully turning off, allowing the a very minimal amount of current to flow through the Nixie tube.

How can I make sure the optoisolator is fully off, so no current will flow through the Nixie?

I'm switching the optoisolator with an Atmega8515 microcontroller.

I've attached a circuit diagram.

Thanks for any help you can provide!

enter image description here

\$\endgroup\$
1
  • \$\begingroup\$ What type of opto isolators are they? Given that very little current is required by the Nixie you might be seeing a small leakage current through the transistor. Take a look at web.jfet.org/nixie-1/NixieTransistors.pdf (page 8 or 8) which uses high voltage opto isolators to perform the same task. \$\endgroup\$ Commented Jan 22, 2015 at 18:56

2 Answers 2

2
\$\begingroup\$

First check the voltage rating of the opto outputs. Can they really withstand 200 V? What leakage are the optos specified for. Is that maybe what you're seeing?

Second, how do you know the opto is really leaking? Is it just because you see some of the segments dimly lit when they should be off? This might be due to turn off time, not leakage. How fast are you turning on the segment drivers after having switched the previous anode off? Perhaps you just have to wait longer.

If neither of the above are the cause, then put a pulldown resitor on each anode. However, you have to be careful about the value since you probably don't want to dissipate too much power. For example, a 100 kΩ would dissipate 400 mW when that tube is on. Of course each tube will only be on 1/6 of the time at most, so that's only 67 mW on average. Also make sure that you use a resistor rated for the voltage, not just the power.

\$\endgroup\$
1
  • \$\begingroup\$ Yes, I confirmed that the opto is leaking by switching on a cathode transistor, but leaving the anode optoisolator off. This resulted in the digit being very dimly lit, to the point where it only displays half of the actual digit value. I'll try the anode resistor, thanks! \$\endgroup\$
    – 331GT
    Commented Jan 22, 2015 at 20:01
2
\$\begingroup\$

Quite possibly you can't, there will always be some leakage current, especially if the optoisolator becomes warm. (I assume the optoisolator is rated for Vce=200V : expect trouble otherwise).

One thing you can do is to connect a high value resistor from each anode to ground. This will provide an alternate path for the leakage current. What value of resistor? First measure the leakage current : or equivalently the voltage across R31 etc when off and calculate the leakage from that. Then set the new anode bypass resistor to drop much less than the Nixie strike voltage (say 25V or less) at that current.

\$\endgroup\$
1
  • \$\begingroup\$ Ah yes, this makes total sense. I've seen this idea being thrown around on other websites, but I did not fully understand the reason behind it. I'll try this tonight and report back. \$\endgroup\$
    – 331GT
    Commented Jan 22, 2015 at 19:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.