So, on both one of Richard’s Nanjg-092 drivers, as well as one of CK’s BLF-17-DD boards [that harvests the resistors/bits/mcu from a standard 105C], I have gotten at best, 3.7 amps from batteries that, direct-driving with 22ga wire, put 4.2 amps through a XM-L2 [laptop pull Sanyo 2250’s and Samsung –28A’s]
I just tested CK’s 105c/BLF-DD with [2P in my T08] Sanyo blue-tops [from Richard awhile ago] and it does 4.2 amps, the Nanjg-092 does 3.8 amps with one.
This still feels like much less than I should be seeing from these drivers. Am I doing something wrong, or do I need an Efest or a 20R to see 5ish amps?
Not enough info. What are you using to measure the current? What LED is it, domed or dedomed, and on Al or Cu?
There will be losses through the driver traces and spring, even if braided. To check if it's components on the driver or just the current path across the driver, short LED- to ground and see how much current goes up. If you're using a normal ammeter the best you can hope for is relative (X is more than Y), you can't compare your numbers to anybody else's. Well, you can, but you'll be using invalid info and any conclusions based on invalid info are bound to lead you astray.
If you want to test with your meter and only use it to compare to your own numbers using that specific meter, you need a baseline. Test current with your cell wired straight to the LED with nothing inbetween. That's the maximum possible. Each flashlight part you add after that will reduce current by some amount. Only knowing what to expect with no driver, driver with LED- shorted to ground, driver operating normally, etc., will you be able to find out if something is misbehaving.
I understand that the DMM has internal inaccuracy. But I have a Convoy M1 with a 2800mA Nanjg in it, and give or take fifty milliamps, that’s what the DMM says is flowing through the tailcap on high. Is the difference in error between 2.8 amps and 3.7 amps really that high?
It can be, but since I don't have your meter and your parts I can't say for sure. But generally, using a typical DVOM inline to measure current will have greater inaccuracy as the current increases. And there's no reason that inaccuracy will be predictably linear.
Best bet would be for you to pick up at least one high drain cell (20R/25R/VTC5), even if you just use it to get a better handle on how your meter reacts to different currents.
With my meters, the highest I have seen with single XML2/Cu, 17DD, & 25R (that's a fairly 'well known' combo) is around 5.65A - but with your meter and your leads, that exact same combo could give very different readings.
My meter reads about 8.5A with a triple XPG2, 17DD, & 25R, but IIRC Cereal_Killer's numbers with his shunt-modified ammeters is around 12A with the exact same components. That's lot of error added just by placing an ammeter inline - and I'm even using homemade leads with short triple 18AWG leads and the internal fuse holder replaced with a fat copper bus bar.
Certainly I am getting a shunt resistor modded DMM for nothing else that tonight I finally modded my Trustfire X6 such that its drawing 5.3A tailcap with the DMM [14ga automotive primary leads]. I’m afraid that its really over 6 amps on the cells when the light is closed back together, and that would be either 9 amps to the emitter or cooking the driver, neither of which are good. Backed the trimpot down 1/8 a turn [now reading 4.4A] and will wait for CK to start making more shunt resistors before using it to walk the dog again.
Not very. You're not going to get an accurate reading with any inline meter after you get up around 3-4 amps, there's just too much stuff for the current to pass through. Measuring voltage drop across a very very low resistance external shunt is light years ahead of trying to get it all to go through the meter, but a clamp meter is the only way to measure it that truly will have no effect on the circuit it's measuring, because you're not adding anything extra into the circuit.
But these meters work on shunt resistors themselves down in the unit. Soldered leads, plugs firm in sockets, and then 1/2” of pcb between the socket and the resistor on each end. Surely on an American made meter like the Fluke that costs $300 new, they have accounted for that. This $7 Cen-Tech, yeah, I buy that no question, but on a professional grade, high-dollar meter, if it says 5.5A on the screen, its at least close to that. Why the heck else would a $300 meter have a “10A” range if it cannot measure it?
I did consider a clamp-on ammeter, HF even had 3 different models, but none measured DC current.
Voltage drop. The current vs. voltage graph for an LED is exponential so the current will fluctuate greatly even if there voltage difference is 0.1V. Stock DMM leads drop enough voltage to affect this, it’s not the problem with the multi-meter. You should get similar results with a $300 fluke or a $5 HF multimeter with the same quality leads.
You’ll need custom made extra short leads if you want to accurately measure tailcap current. Some informative threads:
I’ve been using this Mastech Clamp Meter and love it. $40 shipped from Amazon.
It’ll read up to 40A DC current. The most accurate measurements come from a loop placed into the negative lead off the star. I typically use just enough 20ga silicone wire to get the clamp comfortably around, then I know exactly what the emitter is seeing.
For a tail amp measurement, I use a 6” long 12Ga Tenergy silicone wire with tinned ends for the connection. This is looped off the cell and to the battery tube with the clamp meter reading it. I’ve been seeing about 300mA difference with the FET drivers. (emitter amps slightly less than tail amps)
If you have more than one meter, hook one up inline like normal to measure current, then use the other meter to measure voltage drop between the ammeter leads. Do it with two different drivers, say a 1-2A driver and your DD driver (or no driver at all, just battery to LED), and you'll see how much more is lost as the current increases.
@ DBC, thanks for the suggestion. I think I’m going to be borrowing someone’s Prime acct. soon.
@ Comfy, I see your point. Testing the 2800mA Convoy, I’m seeing .15v drop through the Fluke 77. In the Trustfire X6 turned up, I’m seeing a .31v drop. Assuming that the Nanjg in the Convoy is doing truly 2800mA, then the X6 is sucking down 6200mA while the Fluke reads 5.78A. So the Fluke is…7% low then? Fair estimate?
Either way, I’ll be turning the trimpot a little bit back down.
