Get a string of VR tubes to go pop

You know, I like VR tubes (aka glow tubes, gaseous tubes, STV tubes), but one minor problem is to get two or more of them in series to—reliably—pop (aka start, strike, ignite).

One piece of feedback to the latest coppatone instalment was: ‘are you sure that string of three VR tubes is going to ignite?’ Well, I wasn’t. I had seen in the past 120k or 1M resistors in parallel with one (out of a string of two) VR tube, but no idea how that worked. So it was time to find out.

It is not that the internet is exactly bubbling with this information. If there is info out there, it is well hidden from google. Also a query on the Joelist showed that handling VR tubes seems to be a lost art. Luckily two ‘Joesters’ were able to help out: Bjørn Aaholm pointed out that the STV280/40 datasheet has some good info on this topic and Jeffrey Jackson answered my questions of how this stuff works. Here is what I learned.

The most important thing I had to get my head around is that the aim is to pop the VR tubes in a string one after another instead of all of them at the same time. A VR tube has an operating voltage (the one it regulates to, say 105V) and a starting voltage (say 115V for the same tube) it needs to see to go pop. Normally you would need 345V to get a string of said tubes started. With the aid of resistors, as shown below, you can get all three to pop, with the last one striking at 325V (2 × 105V + 115V). There are two methods to do this.

The parallel resistors method is shown in the STV280/40 datasheet:

In this example we see a string of three VR tubes, hooked up with a dropping resistor and a load. Some things do not matter: whether the + or - side of the supply is grounded or neither of them; also Rdrop can be on the + or - side. We see that the resistors that help with the starting are in parallel to each other, one to every VR tube junction. These resistors can be attached to the + side (left) or the - side (right) of the supply.

I have mixed a bunch of different VR tubes here, just to make the point that it can work for any combination: the VR-150 (0D3, 150V, 160V start), VR-105 (0C3, 105V, 115V start) and the VR-75 (0A3, 75V, 100V start).

Here is how the circuit on the left starts up, as the supply voltage ramps up:

1. as the supply voltage after Rdrop hits 100V, VR-75 will pop, dropping to 75 Volts and staying on with the tiny current (25V / 270k = 93µA) that runs through the 270K resistor;
2. the supply ramps up further and at 190V (75+115) the VR-105 will pop, fed by the 150k resistor;
3. at 340V (75+105+160) the VR-150 will pop and the string is on.

We see it took 340V to ignite the whole string, instead of 375V (100+115+160) as expected without the use of resistors. The circuit on the right starts differently:

1. at 160V, VR-150 will pop;
2. at 265V, VR-105 will pop;
3. and at 355V (no mistake: 150+105+100) the VR-75 will pop and the string is on.

Yes, the same string will be fully on at different voltages depending on how you rig the resistors. It is the last VR tube to strike whose starting voltage premium is the one of the whole string. I picked resistor values so that during normal operation about 1mA flows through them. Much higher values can be used, this datasheet shows 1M, with a start drop of 25V that means 25µV is enough to keep the tube running.

The series resistors method is the one Jeffrey explained to me:

We see resistors in series, one paralleled to every VR tube in the string. On purpose they are not proportioned like the voltages of the VR tubes. The aim is to get every step along the way a disproportionate part of the supply voltage over one of the VR tubes, which then will pop. Here is how the circuit above starts up, as the supply voltage ramps up:

1. as the supply voltage after Rdrop hits 220V, 330/630 × 220 = 115V will be over VR-105 and it will pop, staying on with the current trough both 150k resistors;
2. the supply ramps up further and at 305V, 150/300 × (305-105) = 100V will be over VR-75 and it will pop;
3. at 340V (75+105+160) the VR-150 will pop and the string is on.

No surprises, 340V is exactly the same voltage as with the parallel method and it is also determined by the start voltage of the last VR tube that strikes.

Thanks to Bjørn and Jeffrey I now know how to get a string of VR tubes to strike at a lower voltage, with less chance of the circuit locking up in the off state. I hope this post also helps you to achieve that.