The HP meter added a surprisingly large amount of resistance.
Spec says
DC Current Shunt Resistance:
5 Ω for 10 mA, 100 mA
0.1 Ω for 1 A, 3 A
Was it in Auto-Range-mode ?
Maybe that screwed something up…
-S
Nope, it was set to high range - even in auto-range mode, it will physically select the alternate shunt once current rises above it’s low-range threshold. Looking at the numbers though, you’re right - it acted like it was a .3 or .4 ohm shunt instead of a .1
I’ll have to look into that… Interesting indeed. it certainly wasn’t in low range mode - a 5 ohm shunt at 3 amps would drop 15 volts!
I agree for the current measurement.
What about the voltage measurement at the LED ?
Let’s say we have an XM-L, that is optimally drived at 3.35v, 3A.
While running, should I measure 3.35v at the LED, or about 3.15v (because of the voltage drop at the diode (I am saying 0.20v but could be more could be less, I have no idea)) ?
The LED IS the Diode (Light Emitting Diode) If you measure 3.35v at the LED, then that is the voltage that the LED is dropping (Vf)
Just for the record, the voltage at the LED is a pretty worthless number. It doesn’t consistently relate to current.
Two XM-L emitters sitting right next to each other could both be running at 3.35V, one of them might be drawing 3A of current while the other draws 4A - it’s literally that inconsistent.
PPtk
Welcome to BLF, Silicon!
Thank you 
The obvious point here is that to use your average clamp meter for an led current reading, you would at best have to treble the length of the cable to the led just to fit the clamp……
I use a 4mm square cable to clamp meter my tail cap currents if I’m not happy with my avo reading both tend to be within .1 of an amp so far, but all that precludes driver and emitter losses. All it tells me is I can trust my meter for the traditional reading at the tailcap.
My clamp meter is generally used to check start current of 12v dc fan motors (transport fridges) but my meter tends to read pretty close since I started cross checking the two, but we’re talking upwards of 12a generally, I’m not sure there’s even a multi emitter array that takes that as standard but I’m happy to be proved wrong.
Aloha and welcome to BLF Silicon!
Another possible way to measure LED current is to measure the LED voltage, after letting the temperature stabilize, and then replacing the driver with a power supply that has an adjustable current limit. Again the LED voltage is measured (using the same connection points) and the current through the LED is increased till the LED voltage is the same as it was with the driver. A current meter in series with the external supply current displays the same current that the driver produced to get the same LED voltage. The meter and leads doesn’t cause any affect on the reading because we’re only interested in the current it takes to create the same voltage drop at the LED.
Obviously this method requires a current regulated power supply ($200+), and very good resolution of the LED voltage, but it should be reasonably accurate. Other drawbacks are: it won’t work for PWMed drivers (except at 100% PWM), the LED temperature for each of the measurements needs to be the same, and there are others.
Just some food for thought
Regards, Crux
That certainly is one way to do it
You don’t actually need a current controlled supply, either. A voltage controlled supply will be fine - you just slowly adjust voltage up until voltage at the LED matches what you saw when it was in-host. Current will flow according to this voltage; no need to regulate it…
The problem is that I think you have somewhat underestimated how much measurement resolution you’ll need for this to be anything near accurate. millivolts of voltage can induce significant changes in current in LED emitters. The voltage/current curve is nowhere near linear, and with the way it plots on a graph, VERY small voltage changes create VERY large current changes.
It’s a cool theory, and if you had enough measurement resolution, it would work… Unfortunately, I think you’d need 5 or 6 digits after the decimal of voltage measurement for the numbers to be any kind of meaningful.
PPtk
Agreed! A high resolution voltmeter would get more accurate readings here, it all depends on how much is necessary.
Perhaps equally important is insuring the LED is at the same temperature for both readings. Although, given the same heatsinking, the same ambient temperature and the same current, the temperatures ought to be very close.
This is a case where the a few ohms of resistance in the leads of the meter and external power supply is actually helpful, this becomes the series resistor to limit the current as the power supply voltage is changed. It makes it less twitchy, and removes the absolute need for a current regulated supply. Keep in mind that this resistance will generate heat at high currents.
Regards, Crux
Edit: This setup doesn’t take into account the heat added to the LED by the driver… that will skew things a bit.
The idea does also have a problem with pwm, but most methods has that (A oscilloscope is needed)
In many lights it is possible to measure the tailcap current at different voltages and then estimate the led current, the exception is lights with boost converters.
Need to know?
If you really need to know, you likely know how.
If you don’t know how, you likely don’t really need to know.
Disclaimer: Not intended as a reason not to ask/learn. Just a thought.
The video above states that tailcap readings are not accurate. However, both measurements on DMM seem within 10% of expected. Add this to the fact that the tailcap itself doesn’t have dissimilar resistance to meters and results should be within the ballpark. Also, though LED’s are nonlinear, even adding a simple correction for this should make the readings fairly accurate (enough for our purposes at least).
Thanks for making the video though. The Tek probe looks to be pretty accurate; I wonder if some cheapo meter can replicate its measurement accuracy. HKJ seems to think no.
I love it when you talk dirty. Thanks guys, great thread. Many(most?) diyer’s seem to just measure tailcap current which is battery current and doesn’t include driver requirements. Good for DD though. I’m not really all that concerned with pinpoint accuracy anyway and since I’ve decided not to budget what I would need to do so. I do use 14awg stranded for current measurements as that is a free upgrade to my el cheapo DMM but there are so many places I’m introducing inaccuracies that I’m really just “ball parking” it. I’m glad there are members that have the knowledge, the technical skill, and quality equipment to let me know what’s really going on. I do actually read and try to understand.