Help Wanted: Sydney, AU. Oscilloscope + Current Probe Measurement

[foetusmachine]

Hi All,
I hope this is the correct forum to post this.
I'm looking for someone who can measure a current waveform for me using a decent oscilloscope and current probe (clamp type).
I have a LED camera flash unit that has a two-lead output to the LED.
I've been able to measure the output voltage (4v), but would like someone to measure the current waveform to the LED for me. Access to the LED cable is easy.
I have had no success measuring the current myself using a shunt (wildly varying results).

I can bring the camera to you (i'm in Canterbury (Sydney) Australia) once we're out of lockdown or the restrictions ease.
Any help would be greatly appreciated.

[julianhigginson]

What's the actual experimental setup for your shunt resistance measurement?
And what order of magnitude of current are you expecting?

[foetusmachine]

I tried a 10m ohm shunt placed into the +ve wire feeding the LED. Unfortunately, I don't have a spec on the LED (other than it's likely to be a discrete 3535 style) and I don't have any idea of the driving circuit as it's in a sealed enclosure.

The 10m ohm shunt provided a voltage drop on my scope of 0.145 volts equating to 14.5Amps.

Maybe it's right, but I would have expected something less than 1A.

[julianhigginson]

how long is the flash for?  My understanding is LEDs are power rated more or less by limits in energy dissipation before they are thermally destroyed, and you can get huge currents flowing in lots of things for very short periods of time.. so.... possible you're seeing that amount of current?  but have never personally looked into running a LED like that and we are way off datasheet territory here.

It at least seems possible you're measuring what you see. 15A x 4V = 60W so that's a fair bit of light energy being emitted. A bright 1500 lumen LED lightbulb is around 15W constant draw (including conversion losses, so less in the actual LEDs) but maybe 3-4 times that is the kind of momentary light power you want to blast out for a well lit flash photo? 

Anyway, I'm curious to know and I have a clip-on current probe and a 200MHz scope at my office in western Sydney that you're welcome to come and look at it with, if you don't get it sorted in the next few months before lockdown ends.

In the meantime, different shunt resistances could be interesting to try. like, 5mR or 20mR if you can get your hands on more shunt resistances?

[foetusmachine]

The LED flash is something like 5-10ms. It's actually dual flash TTL, so the first flash doesn't really plateau, but the second main flash does.

The LED itself is simply a trigger mechanism to get light through an underwater camera housing to the outside where it is picked up by fibre optic leads that go off to the underwater strobes that do the real lighting work. However, some underwater strobes have notoriously poor light sensitivity in their light receivers, so maximising light levels is important for the LED trigger.

It is possible that the current is high, but a few things lead me to think the current is not that high:

1. the trigger unit itself is very small - a lot of SMD-only circuitry only on the inside I believe and powered from a rechargeable lithium coin cell of some sort (hence the 4v power reading i'm getting on the LED). I'm by no means an expert in all of this, but that order of current seems to be outside the realms of what would reasonably be achievable by such a small package. You could probably design some kind of small circuit to do it, I just not think it's what's going on here.

2. The LED connector is a micro-JST and the two power leads to the LED are ~0.9mm OD on the insulation. I'd only be guessing what the copper area is, but it would be extremely small (28 guage was the order of size when you search online for 0.9mm OD cable). That order of current again sounds very high for such a small conductor. Yes, they only activate once every 1/2 second, but still, it doesn't feel right.

But, I think it's true that there may be a degree of over-drive going on. And also...

3. The only spec I have on the LED is that it's capable of 150 candella. Taking that back to lumens by assuming a typical 110 degree 50% beam spread, that's about 400 lumens. That kind of output is possible on an XP-L2 led being driven to about 1A.

That's kind-of where i'm at. I really would just like to verify my gut feel.

[Brumby]

I tried a 10m ohm shunt placed into the +ve wire feeding the LED.

Just to clear this question.... Is your LED circuit floating with respect to your measurement setup?  (until connected, of course)

If there are any doubts, then I'd suggest trying your shunt on the low side.

[foetusmachine]

Thank you @Brumby for the suggestion.

I rewired my test setup using soldered connections and properly screwing the small cables to the shunt resistor and also had it on the low side. Previously I was holding cables together with the scope clips (my bad). I think I have a much better set of results now.

Using my 10mR shunt, i'm consistently getting 0.01v across the shunt. Therefore 0.01v/0.01R = 1A. Unfortunately, my only other proper shunt is 1mR, so the voltage there is way too low for me to pick up using that one.

Much more like the result I was expecting.

One thing I do notice is that the open circuit voltage on the output of the LED trigger is 4v (probably the voltage of the internal rechargeable lithium cell). When the LED is in-circuit and triggered, the voltage drops to 1v. I presume this is becasue the tiny driver is working as hard as it can to deliver the required LED drive current.

The LED pre flash and main flash run for 0.2ms and 50ms respectively (as metered by the camera at the time of the test), so I feel like such small activation times equate to a very small duty cycle such that the driver is not over-burdened.

[Brumby]

Previously I was holding cables together with the scope clips (my bad).

Aha!

I think I have a much better set of results now.

To be expected.  Good to hear.

[bob91343]

I wonder how well the venerable HP 428B would measure this.  It is primarily for measuring steady dc but has a signal output that has a decent frequency response.  If you can find one, give it a try.

[foetusmachine]

Thanks Bob,

So you're basically saying that you ignore the panel meter and maybe the HP428B would act as a current probe?

[julianhigginson]

aaah, I didn't have a chance to look back till now, but seems like you sorted it out...

Yeah a micro JST connector (and the wires used) really limits the expectation of crazy current. obviously that connector and wires would take that kind of current for a short period of time, but would need a lot of volts to do that, and take up most of the energy being supplied.

And  yeah... piping light as a trigger source to a bunch of light synchronised flashes would give you way less need for light energy than filling a whole room with enough light to illuminate it directly.