Partially working Burroughs B5094, possible recover?

[Paolo Cravero]

Hello from a newcomer on the list, looking for help and hopefully soon going to show a completed build.

I might have used wrong search terms, but I found no previously written text on what follows.

I have a Burroughs B5094 Nixie where only 2 out of 10 symbols light up. "u" and "A" FWIW. All other pins show no activity when pulled low (one at a time). Is there a procedure I could try to revive other symbols? Is there any chance they could spring back to life?

Thanks,
Paolo

[gregebert]

Double-check the datasheet; maybe you're connecting the wrong pins ? See http://www.tube-tester.com/sites/nixie/data/B-5094/b-5094.htm

How much voltage are you applying ? If it's too low, some cathodes wont fire. 150V-170V is usually enough, but a used tube may need more.

As long as you limit the current to a safe value (0.8mA for the 5094), you can apply higher voltage to a tube to fire it. As an example, my first nixie clock uses 340V for the anode supply, even though the tubes only require about 150V to run.

If you have a 300V supply, you can take a guess at a safe series-resistor. Just assume the tube needs 150V to fire. That means you need to select a series resistor that will drop (300V supply-150V tube) at 0.8mA. By ohms law, your resistor will be 150/0.0008 = 187K. The exact value isn't critical, so 180K to 200K should be fine. You can use other supply voltages and calculate the series resistor accordingly.

Fortunately, the current is quite low so 1/2watt resistors will work fine.

If you're fortunate enough to have a current-limited supply, just set the current to 0.8mA and crank-up the voltage.

[Paolo Cravero]

Hello.

Connections look fine, since two symbols light up I probably got the anode right. Also if I plug a B5092 all digits light up.

 

How much voltage are you applying ? If it's too low, some cathodes wont fire. 150V-170V is usually enough, but a used tube may need more.

My lumos.sk supply is currently set at 164V. I had tried at 170V and still got the same result.
 

As long as you limit the current to a safe value (0.8mA for the 5094), you can apply higher voltage to a tube to fire it. As an example, my first nixie clock uses 340V for the anode supply, even though the tubes only require about 150V to run.

Fine. I will crank up the voltage to the maximum, probably 180V and adjust the resistor accordingly (37kR). If that's not enough I will begin stacking up batteries ...

I will report my findings. Glad to be learning something new (but old).

Paolo

[Paolo Cravero]

 

How much voltage are you applying ? If it's too low, some cathodes wont fire. 150V-170V is usually enough, but a used tube may need more.

Gregebert, I followed your tip and I could light up all filaments in the B5094 at 198 V.

I had to pay a small price for the test, though: I fried a K155 driver chip. This probably happened when I plugged in the socket a 5092 that works at 170V leaving the supply voltage at 198V and a 100k current limiting resistor.

The circuit firmware simply cycles through all symols in the Nixie and all of a sudden all symbols would light up together. Once 170 V were restored and the K155DI replaced, everything returned normal.

Too many wires in there (and too hot weather) to do a proper troubleshooting, but I kept the chip in case further investigation can reveal useful information and avoid similar faiulre in the future.

Paolo

[gregebert]

I'm glad your tube is working. It's unfortunate your driver got fried, but that is a risk when using a supply-voltage significantly higher than the max-rated voltage of the driver. The reasoning behind using the lower-rated driver is that the nixie tube will provide a voltage drop around 130-150 volts, so a 170V supply should only result in 40 volts at the driver. That's only true under ideal conditions. In reality, hot-swapping a nixie tube will momentarily expose the driver to the entire HV supply voltage while the nixie tube's parasitic capacitance charges-up (I've measured a few tubes around 10pF). I'm sure there are other scenarios that expose the driver to unsafe voltages.

If you cant change your driver, you may consider placing a bleeder resistor from each cathode (ie, driver output), to GND to shunt any pre-ionization current. You will want to use the smallest resistance that doesn't cause any cathode-glowing, which means it will be several megohms. This wont eliminate the risk of zapping, but it should reduce it. The only way to eliminate the risk is to clamp the driver pins to the maximum-allowable voltage of the K155. That would require a diode+resistor on every driver pin to a clamp voltage, which could be a shared zener diode. At that point, for size and cost reasons you would use a different driver.

BTW, even a higher-voltage driver such as the HV5530 has some risks because there probably isn't full ESD protection on the output pins, therefore you have to rely on the offchip circuitry (which likely includes the nixie tube) for protection. As long as the driver is rated higher than the maximum possible voltage on your HV supply, there aren't any operational risks even if you swap tubes while your clock is running.

[Paolo Cravero]

On Thu, Jul 30, 2015 at 12:27 AM, gregebert wrote:

driver. That's only true under ideal conditions. In reality, hot-swapping a nixie tube will momentarily expose the driver to the entire HV supply voltage while the nixie tube's parasitic capacitance charges-up (I've measured a few tubes around 10pF). I'm sure there are other scenarios that expose the driver to unsafe voltages.

Right on spot gregebert: I did some hotswapping or I had not fully inserted the Nixie in the socket. I am always afraid it breaks.

Also thank you for the brainstorming about drivers and how to protect them.

Paolo