Hi all, I’m working on a fairly low power set of builds where I have to tune drivers for for the exact current I need to the LED. The target is 2.5a.
My clamp meter has never been perfectly repeatable, but its fairly so on these builds. I get around 2.29-2.68 amps over several tests. Not great, but ok I guess. I wanted to double check this so I tried using my standard DMM running the power thought the leads on the 10 amp side, and I get a much lower reading 1.71amps. This reading is consistent, but I suspect consistently wrong. I’ve always assumed DMMs “drop” current as a result of using long leads and adding resistance to the circuit. I tried using a very thick heavy set of wires and that reads about 1.88 amps.
I really need to get these into a tight tolerance, but I’m not sure what device to trust since they all spout random gibberish.
I assume you’ve got a Uni-t ut210. Those are good to about 200 mA woth decent accuracy. The most accurate way is with a high precision bench power supply or bench DMM/hand held DMM with a current clamp. The only other way is measuring the voltage drop across the load to calculate the current.
If your current changes significantly by adding a tiny additional resistance, the constant current source doesn’t seem to be doing its job in the first place. A working constant current source keeps the current constant, no matter if output resistance rises or falls, unless your resistance is high enough that the input voltage of the CC source isn’t enough to reach the desired current anymore.
Or the wires for the DMM are broken and have really high resistance.
With DMMs there is the ”burden voltage” which is the voltage drop across the sense resistor and traces inside the DMM, this burden voltage may be specified by the manufacturer, or you can find that info on HKJ’s tests for example, you can also simple measure it with another voltmeter and a known current, and additionally the resistance of the leads you are using as you mentioned, The total is usually pretty significant so it may affect the output of your driver depending of the type, and the input/output voltage.
For example if you have a DMM+wires 200mV/A of burden voltage (=200mR resitance) and a 2A linear driver, led Vf of 3V, you need an Supply voltage of at least 3.4V + the drop of the linear driver to not affect the output current of the driver.
For precise current measurement with minimal voltage drop I use low resistance 0.1% current shunts, ranging from 1mR to 50mR depending on the current I want to measure, with short thick wires for minimal added resistance, and I measure the voltage drop with a relatively precise DMM. Of course there is still a voltage drop as small it may be so one need to know what they’re measuring and if it can affect the output current or not.
I was testing my newest light with three different batteries a few days ago. I was surprised at the different numbers and I was only dealing with 5 amps and under. So I rotated through the same three batteries a second time and the amps matched nearly identical on each battery as the first round. So I would say it was very repeatable. That is with the UT210E.
But I realized at the beginning you have to hit the zero button before every measurement. I’m used to my analog AC AMProbe where I don’t have to do that.
The ubiquitous UNI-T UT210E has DC current spec of ±(2%+3). I did a quick test using a linear driver, bench power supply, DMM in current mode, and UT210E clamp meter. On various levels I’m seeing around 3% max difference compared to my DMM (and power supply).
Maybe not accurate enough for precision driver/LED testing but good enough for most other flashlight measuring where you want to measure the tail cap amps.
Accounting for losses in the leads, burden voltage, not bad. Of course, as you get higher currents, this changes a lot, woth the clamps generay getting more accurate readings than the DMMs.
As long as the power supply is a proper constant current supply, and it has enough voltage reserve (e.g. not a 3.05 V LiIon battery on a 3.0V_f led) it won’t matter. The CC supply will increase voltage accordingly to compensate for the higher wire resistance.
Had a long PM conversation with Sidney_Stratton, who tried to educate me about using shunts… unfortunately, I just didn’t really understand it, and had no idea where to get, and what to get
Perhaps he, or someone else who really dived deep into this could help out.
At least, I learned that I don’t really trust my Amp readings …
At low currents: DMMs, shunts are way more precise than current trafos there
At high currents: clamps, flashlights do not have the necessary reserves to keep in CC regime (battery sags, V_f of LEDs rises) and the added resistance of DMM leads + shunt leads to the current dropping.
Same issue as DMM, in the case of a flashlight, really. You add resistance to the circuit.
It’s hard measuring off a battery that is already close to LED V_f when full, and will drop even closer after load. You either need a very low reistance shunt, with short and thick wires, so you do not add much resistance, or clamps (which will in return be inaccurate at lower modes).
If you power the flashlight via an external power supply that stays fixed at 4.2V it might be easier to read even with a DMM.
Speaking of buck and linear drivers here. Boost drivers should not suffer from the added resistance of a DMM. Neither should 3V bucks with 2S input.
Then you’ll never going to get a fixed current. The Vf can vary quite a bit with heat, and of course any resistance added to the circuit will reduce the current.
For the measurements yes, but I mean that in the first place the current will not be fixed because you are using a CV supply, if you want a constant current then use a CC supply.
