Measuring LED power with math function on scope

Hello,
We wish to buy the cheapest scope that we can use to measure the LED power in our offline LED lamps. The LED current and voltage is as attached. As you can see, we need a scope which can multiply v(t) x i(t). Then the scope must be able to take the average of this, over a full number of periods.
Do you know which is the cheapest scope that can do this?

Our Tenma 72-8705A scope can't do this.
https://uk.farnell.com/tenma/72-8705a/oscilloscope-2-ch-50mhz-1gsps/dp/2499521

Why don't you just buy a true RMS meter? It would be a lot cheaper than a scope.

The LED current and voltage is as attached....

Click to expand...

Both V and I are clean sine waves with DC offsets and are in phase. It is very simple to do analytically OR even graphically by taking 6-8 points within a quarter cycle. The question is basic: how much accuracy you are looking for?

When simple just won't do.

Both V and I are clean sine waves with DC offsets and are in phase. It is very simple to do analytically OR even graphically by taking 6-8 points within a quarter cycle. The question is basic: how much accuracy you are looking for?

Click to expand...

Thanks, yes, they are sines, but thats just kind of an example, as you know, sometimes it is not sine.....the led current goes to zero for some 500us near the mains zero crossing....so we need the scope facility for that kind of thing.

- - - Updated - - -

Why don't you just buy a true RMS meter? It would be a lot cheaper than a scope.

Click to expand...

Thanks, as you can check in the attached LTspice simulation.......the power is not the same as either v(average) * i(average).....and is not the same as v(rms) * i(rms)
...so a true RMS meter will not help us.

the power is not the same as either v(average) * i(average).....and is not the same as v(rms) * i(rms)

Click to expand...

You are right. But they are approximations and that the accuracy depends on the individual waveforms and their displacements. If the waveforms are not sinusoidal, they can be pretty bad approximations.

I would download the I(t) and V(t) curves on a PC and do the calculation on the PC rather than on the scope. You can get 1% (or so, just a guess) or so accuracy.

One observation: the large fluctuation in the current shows that the LED is really flickering. It is not visible under normal conditions (50Hz is too fast for the naked eye) but can cause other problems (moving fan, for example) and can cause headache.

You should bias the LED to a slightly lower voltage (but that is a different story anyway) to reduce the flickering.

I do not know commercial power meters that can work at low voltage and current AND can take arbitrary waveforms (and by the way I know very little) and give reliable results.

Hi,

Why don't you just buy a true RMS meter? It would be a lot cheaper than a scope.

Click to expand...

It won't give useful results:

Some values:
Let's assume you use a LED, specified with 1A, 10V. It will be completely OFF (no current) with 8V

Now let's assume the LED current is DC 1A.
--> the RMS meter shows 1A, and 10V
--> an average meter shows 1A, 10V, too.
LED power is 10W.

Now let's assume the LED voltage is 8V/10V and thus the current is 0A/1A, 50 % duty cycle.
The LED power is 5W
--> the RMS meter shows 0.707A, 9.05V, ...how do you calculate the 5W?
--> the average meter shows 0.50A, 9.00V, ...how do you calculate the 5W?

Now let's assume the LED voltage is 0V/10V and thus the current is 0A/1A, 50 % duty cycle.
The LED power is 5W
--> the RMS meter shows 0.707A, 7.07V, ...here you can calculate 5W
--> the average meter shows 0.50A, 5.00V, ...how do you calculate the 5W?

--> you need a true V(t) x I(t) multiplying power meter = active power meter

Klaus

Added:
Too late...others recognized the same during me writing the post.

--> you need a true V(t) x I(t) multiplying power meter = active power meter

Click to expand...

Thanks, i agree, and also i used to previously do this measurement using a scope with the math function described in the top post....it was an expensive one....but now we want to buy something cheap.....either scope or power meter...but nothing is less than 1000 Euros.

either scope or power meter...but nothing is less than 1000 Euros...

Click to expand...

But the same can be done on a laptop (even the cheapest one can work wonders): in three simple steps

1. download the I(t) and v(t) on a worksheet from the scope

2. Select a half cycle or quarter cycle range

3. Multiply the two columns and sum the result.

First time it may take some time to figure out the details but repetitive tasks should not take more than 5 mins for each measurement.

Whatever scope you get you should also get a laptop for max flexibility...

(and the laptop can be used to design professional quality graphs and plots)- both are multipurpose tools unlike a power meter...

Don't post a picture of a sinewave and then later on state that well, its not a sinewave at all. What was even the point of posting that graph if it has nothing to do with the problem at hand?

thanks, it would apply if it was a sinewave, (or offset sinewave) and sometimes it pretty much is a sine wave, but sometimes it isnt.
The v and i are pretty much sines if a large-ish cap is used on the output....but not if a small cap is used.
The waveform of the top post exactly presents the situation which we have.......ie, the power can't be gotten by going either v(average) * i(average).....and is not the same as v(rms) * i(rms)

Hi,

I assume the best approximation is about : P_LED = I(average) × V (LED_typical)

In either case for a LED: lumenous_intensity is proportional to LED_current.
And lumenous_power is proportional to lumenous_intensity.

I wonder what's the aim to calculate the LED power.
* If you want to calculate brightness or lumenous_intensity, then measure the current only.
* But if you want to calculate LED efficiency, LED power dissipation or other LED power parameters you additionally need to take care of LED voltage.

Klaus

KlausST said:

I wonder what's the aim to calculate the LED power.

Click to expand...

Yes, I think that matters.

In "my" home automation community, we have users who measure the supplied power by sophisticated, precise instruments. No doubt, their numbers are precise, accounting for distorted waveforms etc. But when they check the data against the numbers measured by the "official" meter installed by the electricity supplier, there are significant differences. Their power consuption is charged based on the simple "official" instrument.