Is this a legit way to measure the maximum amp draw of a light?
Following
I have used that method before. One thing to watch out for is that if you want to measure high currents (like the 5A on your meter), you should use very short and heavy gauge wire. The alligator clips shown in the OP likely have a decent drop across them, and the 4.2V out of your power supply is likely much lower at the light. If you have 0.1 ohm resistance in the leads, you are dropping 0.5V with the 5A shown. So instead you have 3.7V at the light. Ideally you would measure the voltage right at the light, and adjust the power source such that you have the desired voltage (4.2, or whatever you are trying to characterize) right there.
The highest currents I have measured in my lights is about 3A. Much more than that ( and likely even at those levels) you should be very careful about wire and contact resistance if you want accurate results.
TL;DR. Your setup is really only legitimate for measuring the impact of various changes in the head of the light because it doesn’t reproduce or account for the characteristics of the battery and sources of resistance in either the test setup, or the intact light. Further, translating between measured differences into differences in real world performance will be complicated by the slope and non-linearity of emitter performance (ie, relatively small changes in voltage at the emitter can result in large changes is current and output).
Longer version:
As sbslider said, there are issues with lead and contact resistance and resulting voltage drop.
More fundamentally though, you aren’t measuring the light. An intact light has its own sources of resistance, but they are quite different from those in your test setup.
The body and springs are, in effect, leads, but the body has low resistance, the springs, though, can be an issue. There is also contact resistance between springs and battery, between the body tube and the contacts at tail and head, and in the switch.
And, then there are the characteristics of the power sources. Most powersupplies are designed to have low impedance (~resistance). Their output voltage doesn’t change much as the load varies. If their open-circuit output voltage is 4.2v, then their voltage (at their output terminals) at full load wouldn’t be much less than that.
High-drain batteries are also designed to have low impedance, but there are limits. So, the workhorse Samsung 30Q, which HJK tested to have 40mOhm internal resistance (IR) at full charge would have a voltage drop of 200mV at 5A, 400mV at 10A, etc. And, of course, a battery is only at full charge for an instant once a load is attached.
On the flip side, with a low resistance load a high-drain battery could deliver a lot more than 10A (and cells exist that are specced for operation at >20A). At those drains, many/most bench power supplies will be overloaded, and either cut their output, or limit current.
There are power supplies designed to simulate other power sources, like batteries. There are power supplies that can have external “sense” lines to measure voltage close to the load (to eliminate the impact of lead resistance). One can carefully measure and account for the other sources of resistance in a light.
I think the easiest, fastest, and cheapest thing to do though, rather than trying to simulate the properties of the system, is to, as much as possible, test the actual system.
At a high level, the approach is simple: insert a sensor into the system while minimizing the impact on system performance. The implementation is also pretty simple. A clamp meter allows measurement of current through a conductor with minimal impact. The only real complication is finding a conductor in a flashlight that you can get a clamp meter around.
A typical flashlight only has two isolated conductors, the leads between the driver and emitter MCPCB, but they are typically short in order to reduce resistance and fit the space constraints of a flashlight, too short to allow access to a clamp meter.
So, some modification is necessary. Temporarily replacing one of the emitter leads with a longer, lead is an option. One should compensate for the added resistance of longer leads by choosing an appropriately thicker larger wire gauge to compensate iff this approach is taken in a FET driver light, or a light with a constant current driver that is being operated close to its limits.
Another, easier to reverse approach, is to do current sensing at the tail-end of the light. The simplest version is to remove the tailcap and use a short loop of thick wire to bridge between the battery negative and the bottom of the tail tube. This has the significant advantage of using a battery as the power source, but it doesn’t account for the contact resistance of the intact light. It also adds what could be considerable variability in the contact resistance at both ends of the wire, which will depend on pressure and position.
The approach I’ve taken is to make a dummy tail switch assembly. I use the existing tail cap and replace the switch PCB with my dummy unit, threading the test loop out the back. My dummy unit uses an actual tail switch PCB, with a bypassed spring and a loop of thick (13AWG) wire soldered across the pads intended for the switch. I’ve inserted a standard Omten 1288 switch into the circuit in order to reproduce switch resistance and allow easy cycling of modes.
This setup uses or reproduces the entire circuitry of the light under test, including spring-pressures, with the trivial additional resistance of a short length of thick wire. It may be overkill, but it didn’t take much to fabricate, and it only takes a couple of minutes to setup when testing a light.
I’ve also experimented with using predictable characteristics of the light itself. For example, I tried using the resistance of the body tube as a current shunt. I have a sensitive & precise “6.5 digit” bench multimeter, and I used some fine wire to make contact with the tailcap and the head end of the body tube and then measured the voltage drop on various modes. For a given light, I could reproducibly measure consistent voltages at various currents. I had trouble calibrating it though, because at low currents, it was hard to get a good signal, largely, I think, due to thermocouple effects between the lead wires and the light. In retrospect, I probably could have successfuly used the top one or two regulated modes of Fet+1 lights to calibrate.
It’s also possible to measure the forward voltage across the emitter. The V/I curves are published in emitter data-sheets, but we often run emitters far above their specified current, and extrapolation from the published data could introduce significant error. The resistance of the MCPCB could also be a significant source of error, particularly at high currents. Both of these sources of error could be corrected by taking some calibration measurements at known currents and your setup would be suitable for taking those measurements.
In general, your setup is suitable for measuring relative changes in the head of a given light (different emitters, leads, drivers & driver components. This may or may not provide good estimates of how those differences will be reflected in the real-world performance of the intact light.
not really,because a real battery isnt; going to supply the most total power at 4.2V
plus you don;t really know the voltage at the light head, some will drop across the 2 wire pieces and 4 alligator clips.
Interesting. Thank you for the detailed run-down. Voltage at the head is simple enough measure, but the rest… well, I’m not sure how serious I will get with this. I was mostly just messing around with my PS at work and playing with the SC26 comparing a single 26650 battery vs. two 26350 batteries.
I think you need to specify whether you want to measure the amp draw of the total light, meaning the driver circuit, springs, wires, etc… or if you want to measure just the amp draw at the emitter.
The majority of the people on BLF tend to measure the amp draw at the tail cap. So all you need is the light, the battery and then a heavy, but short loop of wire to bypass the tail cap and then use a clamp-on ammeter to measure the amperage. This is the simplest way and allows you to compare your results to most of the other members on BLF.
Check out this post: