I want to make a starter tester, but I want to use a solid state resistor instead of an incandescent bulb.

What will be the value and the power capacity of the resistor for ballast an S10 starter at 220V?

I want to make a starter tester, but I want to use a solid state resistor instead of an incandescent bulb.

What will be the value and the power capacity of the resistor for ballast an S10 starter at 220V?

Just replace a 40W incandescent:
So at 230V it should pass around 170mA, so that means 1.35 kOhm / 40 W total. I would use higher wattage, as the formal rating assumes the resistor is in a free space and it can safely operate at its maximum temperature (could be above 500degC with some models)
So e.g. 5x 270 Ohm / 10 W each in series.
Or 5x 6800 Ohm / 10W each in parallel.

Personally I see the incandescent as a more practical for a test load - it is way easier to dissipate the heat from it (just radiates by IR)...

A 10k resistor should be ok, for pass 5w, most starters can work with 5w load. For a current flow of 0.022A at 220V a 5w dissipation is created. I should use a 5w resistor?

The resistor power rating means, that the resistor itself can withstand the rated power at the specified conditions. But it may not be what you like :

 - Aluminum body resistor rated at 10's of Watts, but this rating applies when it is attached to a large heatsink (which you gotta provide of your own)

 - Wirewound resistor rated at 10's or 100's of Watts, but which is bare wire in the air with bare connections, and can get to 100's deg C temperature on the surface of the wire. so needs some mounting (that you gotta provide) so it won't touch and short to or burn something else

 - Potted Ceramic resistor rated at few Watts, but which can get well over 100 deg C temperature, so even when mounted on a PCB needs some measure to stand off the PCB + it have a specified "burning position" (i.e. not below the PCB)

You could ballast a starter with way higher value resistors, which will greatly lower the current and so the dissipated power. The starter will glow as long as it is intact, but won't heat up enough to close the contact



Alternatively, look into inductive or capacitive ballast solutions :

Inductive - 2 or more low power (PL, ...) ballasts in series, or the primary winding of an unloaded small isolation transformer, etc

Capacitive - using high voltage and transient rated, low value film capacitors (typically EMI filter capacitors). A series resistor of a minimal value still must be added to prevent "infinite" current spikes in response to line transients

Can a bare "D" size carbon rod drive the starter? I don't have a multimiter for check its tesistance and see if it will work.

Can a bare "D" size carbon rod drive the starter? I don't have a multimiter for check its tesistance and see if it will work.

The battery rods are on purpose designed with minimum resistance possible (so the resistance is in 10's of mOhm range), so I do not see how that could in any way limit the current from a mains voltage to any reasonable level. And even when it would have the kOhm resistance, it will still have to dissipate the 40W of power in heat. Even a D-cell rod size heated up uniformly, it would lead to orange heat, what woud ignite the carbon, at least...
Or Ico pletely missed the idea...


For the dapacitor: It would stress the starters contacts to the large current spikes, heavily damaging the contacts. So I wont do that, the tester should not cause any damage to the device ubder test in the first place...
Using an inductive ballast coil wou.d be a way better idea - it is what the starters are designed for in the first place... To keep the current under control, you may connect two ballasts in series, so they wont saturate (they are designed to saturate slightly during preheat to boost the preheating current, wth an assumption that load ks just for few seconds till the lamp starts)

A 5W resistor obviously is rated to dissipate 5W continuously. However, if your duty cycle is lower, you may use a lower power resistor. A 1W resistor would likely be fine at 5W for just a few seconds because of its thermal mass.

It may surprise many to know that a humble 1N4007 (1 amp rated general purpose diode) is rated to happily withstand 30 amps, so long as the current is applied only for 8 milliseconds.

I like to build redundancy into things. A 5W resistor will dissipate 5W obviously, but if the current were to increase for any reason, there's nothing to take up the excess. Same with wiring; a 20A circuit will obviously carry 20A all day, but one day you might want to put 30A on it. It will handle that for a while, but not for ever and things will get a bit warm.

Better to use a higher rating than you need at the start, say 7 to 8W or whatever the next rating up happens to  be.

That's kind-of my point. The redundancy is already built into the rating. A medium size window air conditioner might draw 10 amps nominally. During startup, however, the inrush current is closer to 40 amps. No one plugs these into a 40A circuit though. You might think you're being "safer" by doing so, but the opposite is true. If there is a fault that draws 35 amps for instance, the circuit breaker doesn't care, and the AC catches fire in short order.

A 15 amp circuit and breaker is rated to (and regularly does) handle 20-40A for short periods just fine. It's not as simple as "apply safety margin of 3" to every problem you see. That can get unnecessarily expensive in the best case scenarios, and deadly in the worst cases.

I will try the starter in series with a 10w C7 incandescent bulb. Making the test device with a "solid state" ballast will make the setup bulky and heavy, if I instead use a little incandescent bulb the thing will be smaller, but with the disadvantage that the incandescent bulb is delicated.

For the resistor ratings:
As Ash wrote, the resistor itself will survive its full rating, but it will reach its rated temperature during that.
And this temperature is then problem not for the resistor, but for the things where it is mounted, because unless it is high temperature porcelain, it will start to degrade at way lower resistor load than the resistor itself is rated for.
With power resistors it is quite common nowadays to reach nearly red hot temperatures (500..600 degC), the modern ceramic materials used for the resistor construction are able to handle that easily, but e.g. the FR4 is still imited to 125 degC long term. So if you mount a 20W rated resistor directly on the PCB surface, its maximum load becomes barely 5W (unrealistically assuming the mounting wont affect the thermal resistance to ambient, which is highly questionable assumption, when at 500degC majority of the heat transfer goes via radiation, so it counts on the PCB being colder than the resistor itself).

For the starter the 5W is way too low: The starter needs to dissipate few W in order to close and for the timing to be ballpark where it is supposed to be, 10kOhm 5W load will limit that, so you wont be able to asses if the starter timing (preheat time,...) is correct.
Plus in case the starter develops high contact resistance failure (and consequently refuses to open during lamp start), you need certain current to make this defect present itself. Same for its tendency to weld contacts together. These faults may get undetected by the tester when used with too small current.

As I said before I will move a bit from the solid state resistors for a while, and I will experiment with 10w and 25w incs. I'm planning to mount it in a little plastic box, all connected with bare wire.

As I said before I will move a bit from the solid state resistors for a while, and I will experiment with 10w and 25w incs. I'm planning to mount it in a little plastic box, all connected with bare wire.

That should be fine, indeed...