Resistance and load in non-ohmic circuits

[Sunder]

Hi all,

Not sure where to put this, but this is probably the closest match...

I have a technical issue that I'm sure the more technical guys here will be able to answer. I have a LED headlamp that from 8v-32v, draws exactly 40w. So if I lower the voltage to 8v, it draws 5A. If I raise it to 20v, it draws 2A. So far so good right?

Problem is, my motorcycle controller complains if it detects greater than a 35W draw.

If this was an ohmic device, it would be easy to put a resistor in series, and just drop the entire circuit to 35w, and the bulb will get a little less.

But is there any way I can do this with a device that draws a fixed load regardless of voltage??

Thanks.

 

[Alan B]

Since it is an LED, it likely uses current on the output side for the regulating sense value. This is likely based on a series shunt resistor that it samples the voltage on. Slightly increasing the value of this shunt resistor would reduce the current to the LEDs, and they are a fairly constant voltage load, so that should control the power the system is regulating to.

 

[Sunder]

Thanks Alan. That makes a lot of sense, but considering the size of this, I dont think I will be able to modify it.

Back to the drawing board.

 

[Tom L]

I wonder what would happen if you fed the light from a simple one transistor constant current source designed to provide only 35W at your maximum supply voltage?
Would the light be dimmer or not work at all?

 

[d8veh]

I would say that the light has a dc to dc converter stuffed inside. That's how mine have been that accept a wide range of voltages. The LEDs have their own driver, which is separate to the buck converter..

 

[amberwolf]

I wonder what would happen if you fed the light from a simple one transistor constant current source designed to provide only 35W at your maximum supply voltage?
Would the light be dimmer or not work at all?

One problem is you'd be wasting power as heat in that current-source circuit, so there'd be even less for the light to use.

 

[Sunder]

Okay. I have an even curlier question now...

I hooked up a load simulator, and found that the bike doesn't complain between 2.2 amp draw and 3.5 amp draw. So far so good right? Get a bulb, measure its draw, and add resistors until it falls into that range.

Nope. Even getting a 2.0a bulb, and cranking up the load using the load generator causes a CANBus warning.

Now, I know very little about electronics, but I read that some CANBus computers sends a high frequency AC current through the bulb to look for impedance as a proxy for resistance. Is it somehow possible that my LED drivers have ultra high or ultra low impedance, while having nearly the right resistance?

One of the posts I found suggested using a large capacity (>1000uF) to lower impedance to fool the sensor, while not impacting the LEDs.

Does this make sense to anyone, or am I really going up the garden path?

Thanks.

 

[amberwolf]

Do you have an oscilloscope? If so, you can check what the system is actually doing pretty easily.

If not, do you have a set of speakers with a line-in? Put the ground of the line in to system ground, and a small capacitor from the signal of the line in to the point you want to check (blocks DC, allows AC). Then you can *hear* anything up to around 20khz, if you turn the speakers up enough, if there is any kind of AC on there it's probably not any higher frequency than that anyway.

Some mulitmeters also have a frequency counter that would also tell you if there is an AC component.

Even measuring with a multimeter on AC (lowest voltage range) might show something.

(also, you can get oscilloscope software that uses the line in of a soundcard as an oscilloscope to display on screen).

 

[Sunder]

Thanks Amberwolf. Will give those two options a try. Maybe borrow an oscilloscope from a mate of mine.

Cheers.