# What do I measure to size a capacitor appropriately to soak up a pulse

As you may be aware, modern vehicles go to reasonable lengths to detect whether the various globes (indicator, headlamps, etc) have failed.

I have a car designed for non-LED globes that sends pulses upon startup and periodically to the headlamps to measure resistance. These pulses are small enough that for a halogen bulb, it never lights up, but for the LEDs that I have installed, the effect is like a strobe for up to 10 seconds upon starting.

I'm familiar with "CANBUS cancellers" and the like, in fact I've tried out a few different ones, but they all seem to just be a resistor and heatsink which puts an appropriate load on the circuit, but the pulse of current is still visible.

I was thinking that an appropriately sized capacitor across each bulb would soak up enough current to prevent the bulb flashing on-off with each pulse, but I need to know how to measure the pulses to size the capacitor appropriately.

I was thinking I essentially need to measure the total amount of current sent through the whole sequence of pulses and keep the capacity as close as possible to that so that I'm not causing a significant delay when the headlights are actually turned on, is that correct? How do I measure "current over time" using a typical multimeter and convert that into a capacity?

• If you multiply the pulse duration (in seconds) by the pulse current (in amps) you will get Coulombs. You can divide into that, the maximum voltage rise you can tolerate before the LEDs show something, to get a capacitance value: $C=\frac{\Delta Q}{\Delta V}$. Whether or not that's a useful value (and how to hook everything up, besides) is another issue. I'm not savvy about these auto circuits, and you've not posted something I could read to become educated about it either, so I'm not going to be much help beyond that.
– jonk
Aug 9, 2021 at 1:06
• A LC filter might work, but the required inductance could be excessive.
– Drew
Aug 9, 2021 at 1:16
• By globe, I would hazard a guess that you are from Australia? Aug 9, 2021 at 7:35

That would depend on how the circuit behaves with a halogen lamp. If it pulses and detects a cold 50W ~70W(2) lamps that draws 500W initially, it may stop quickly after the 1st pulse. Then you have to simulate that load with a halogen lamp to verify the pulse width and you don’t need a large cap for a 1/10 th of second but just long enough for the measurement to complete and exit the self test. This would be a cap with about 20 to 50 mOhm ESR which to draw expected cold current initially. For low ESR caps this would be around T=ESR*C=<20us so C would be20us/20mohm = 1mF @ >=16V. This assumes the car detects this current in < ~ 20 us which might be optimistic.

So I would verify the pulses on a halogen lamp stop after 1 or 2 pulses then choose a few of these and see if this simulates the cold halogen lamps and is detected fast enough to prevent the undesired action. Place the caps across the LED lamp circuit near the source but not output to ground.

Assume:
-. Test current = 500mA
-. LED headlight takes below 500mA to lit, and drops 10V.

Approximated voltage and time correlation is: dV = 1/C * I * dt
To hold voltage below 10V for 10 seconds: C = (500mA * 10sec) / 10V = 0.5F

So, combination of super capacitors with 0.5F/12V may do the job, for this hypothetical case. Meantime, I thought, LED headlights use inline converter(voltage boost/current regulation). If so, the story will be different.

Edit
I would feel sorry, if I don't tell this now; You will need a few tons of fuses, once you start testing with the capacitor.