Glow drivers

This site was last updated on February 15, 2008.

Here's the glow driver Howard Rush built for Paul Walker's B-17 stunter. We don't need quite this much for combat. Indeed, for normal operations, Paul simply uses 3 D-cell alkalines, soldered together in parallel. This lasts him for an entire season of stunt flying (which is a lot of flying).

Electrically, a glow plug is just a resistor. A Nelson HD plug seems to have a resistance of about 1/3 ohm. A McCoy #9 is a little higher - about 0.4 ohms. When we put a voltage across one, power is converted to heat. How much?

     V^2
W = -----
      R

In words, the voltage squared divided by the resistance in ohms gives the power in watts. So, applying 1 volt to a Nelson plug will produce 3 watts. Furthermore, Ohm's Law

     V
I = ---
     R

says that the plug will draw 3 amps of current (where current is traditionally represent by the letter "I").

However! The resistance of a glow plug varies with its temperature: When the plug is cold, the resistance is lower (this is true for most conductors). For example, when I squirt fuel on a glowing Nelson plug, simulating a severely flooded engine, the resistance drops to about 0.2 ohms, or even lower. This is a lucky circumstance, since it allows our plugs to draw more power (about 5 watts in my little experiment) and work to burn off the excess fuel more quickly.

Adding an ammeter in series will show you how much current the plug is drawing (expect around 3 amps, normally). It's useful because it shows when the plug is blown (the current will drop to 0), or when there's a flood (the current will rise), or when there's a short (the current will be very high and the meter will go off the scale).

You can get a Shurite 8203Z panel meter for $13 from Allied Electronics. Beware that nitro will melt the plastic crystal, so cover it with Fascal or something similar. You may be able to do a lot better at your local surplus outlet.

The schematic above illustrates another consideration, the battery's internal resistance. The ideal battery would have an internal resistance of zero ohms. Alkaline D-cells are on the order of 0.1 ohms, and the old Carbon-Zinc are much higher. Why do we care? Well, the internal resistance tends to limit the available current. The higher the resistance, the less current we can draw. Let's do some examples.

First, let's consider shorting the battery, so that all the voltage is dropped across the internal resistance. With an alkaline D-cell, we see 1.5 volts dropped across 0.126 ohms, implying 12 amps (and 27 watts, ouch!). Adding a glow plug with a resistance of 1/3 ohm, we'll see 1.5 volts across 0.46 ohms total, implying 3.26 amps. Measuring at the plug (across points X and Y), we'll only see 1.09 volts, which is fine (3.26A x 1.09V => 3.5W). Now let's flood the engine, lowering the resistance of the plug to (say) 0.15 ohms. The voltage across the plug drops to 0.82 volts, the current rises to 5.4 amps, and the power rises to 4.45 watts, which helps to burn off the flood.

Now let's consider a NiCad. A NiCad cell has a lower voltage (1.2 volts) but a much lower internal resistance (perhaps 0.01 ohms). If we short it, we'll see something like 120 amps, implying 172 watts (double ouch). Adding our plug, we'll see 1.17 volts across the plug, 3.5 amps of current, implying 4 watts. When we flood the engine, the voltage across the plug drops a little, to 1.125 volts, the current rises to 7.5 amps, and the power goes up to 8.4 watts. So! Even though NiCads nominally have a lower voltage, they're better for our application (driving a low resistance plug) because of their very low internal resistance.

Of course, the wires connecting the battery to the plug contribute a little more resistance to the circuit. For best results, we'd prefer those wire be short and heavy, which is why the little orange clip-on drivers work so well for ordinary applications. For combat though, we need the wires to be fairly long. These sorts of considerations led to the development of the Electronic NiCad.

As a final example, consider the GloBee Fireplug (no longer manufactured). This was a compact little package incorporating a 2-volt battery, an ammeter, and dropping resistor. The battery was a sealed lead-acid X-cell, with astonishing low internal resistance (0.0035 ohms, which implies a kilowatt of power if shorted). They can be charged quickly and easily, and they will retain their charge for a long time. In other words, great batteries all around and deservedly popular. But what about that dropping resistor? This is simply a resistor that appears in series with the glow plug to drop the voltage so the plug doesn't melt (2 volts across 1/3 ohm would be 12 watts). Some people use a 0.22 ohm resistor here. Let's see how it works.

If the dropping resistor is 0.22 ohms and the plug is 1/3 ohms, the voltage across the plug will be 1.2 volts and the current will be 3.6 amps. The plug will emit 4.4 watts, which is fine, and the dropping resistor will emit another 2.9 watts. Now let's flood the engine. The plug's resistance drops to 0.15 ohms, the voltage across the plug drops to .81 volts, the current rises to 5.4 amps, and the power drops to 4.37 watts. Hmmmm! It's particularly ironic to see people using this approach with 10-gauge wires and a battery the size of a beer can. To be fair, they're able to handle floods by using some sort of variable resistor -- when the engine floods, they turn the resistance down.

Of course, all my numbers are approximate, since individual glow plugs, batteries, and wires will vary. Indeed, there may be mistakes in the arithmetic. Nevertheless, the conclusion is accurate:

Strive for very low resistance for all parts of the circuit outside the plug.

And if you see the competition using a big battery with a dropping resistor, congratulate them on their foresight. You might even ask them where they bought their battery. Don't snicker!