How to measure power consumption of SMPS?

[alphacat]

Hello,

I have a board which is driven by a half-wave rectifier SMPS, meaning its configuration is like this:
- The Input voltage is AC mains (110VAC/220VAC, 50Hz/60Hz).
- The mains voltage is rectified by a single diode (Anode's connected to the Live wire) and a single (electrolyte) capacitor that comes after the diode, and connects to the Diode's Cathode and Neutral wire.
- The rectified voltage (~311VDC in the case of 220VAC mains) goes into a buck converter which I assume that works in discontinuous mode since the current that is drawn from the converter is 350mA maximum, normally the drawn current is around 100mA.

I own a dual channel digital oscilloscope.

I'd like to measure the active (real) power that my board consumes from the main, using my scope.
I'd be happy to receive guidance on how to perform such measurement.

Thank you very much.
Any help is much appreciated.

 

[Nigel Goodwin]

You need a true RMS meter, or better still a true RMS power consumption meter.

It's very difficult to do it with a scope, and the current consumption is a series of small pulses, at the peak of the mains cycle, as it tops up the capacitor.

But you should really use a bridge rectifier anyway, half wave is a bad idea.

 

[alphacat]

Hi,

I dont have a true RMS power consumption meter in possession.
The only measuring tools I have here are Fluke and digital oscilloscope.
I heard by an engineer that used to work around here that it is possible to measure the true power of an SMPS using a 'scope.

I fully understand the waveform of the current drawn from the mains, which you described - current spikes that charge the capacitor to the voltage peak (~310VDC in the case of 220VAC mains).

So how is it possible measuring the ture power with an oscilloscope?

 

[Nigel Goodwin]

Hi,

I dont have a true RMS power consumption meter in possession.
The only measuring tools I have here are Fluke and digital oscilloscope.
I heard by an engineer that used to work around here that it is possible to measure the true power of an SMPS using a 'scope.

I fully understand the waveform of the current drawn from the mains, which you described - current spikes that charge the capacitor to the voltage peak (~310VDC in the case of 220VAC mains).

So how is it possible measuring the ture power with an oscilloscope?

You could measure both current and voltage, measure the surface area under the curves, and multiply them together - but it's not likely to be very accurate, and a lot of hard work.

 

[blueroomelectronics]

True RMS meters aren't that expensive for a company purchase. Get the right tool for the job.

Is the two channel scope a DSO?

 

[alphacat]

Hey,

I guess that the hard work and lack of accuracy derive from measuring the area under the current waveform.
Dont you think that in order to measure the power, you actually need to first multiply both voltage and current curves, then measure the area under the resulted curve, and at last divide the result by T (time period)?
That is since P = 1/T * ∫v(t)*i(t) dt .

I dont mind doing this hard work, but do you know any practical way to calculate the area under v(t)*i(t)?

To Bill,
Our budget is very limitted at this stage,
and besides, since I mostly deal with SW issues on these days, I'd love to utilize any HW issue that comes into my hands.
The oscilloscope isnt a DSO.

 

[blueroomelectronics]

eBay has oodles of true RMS meters, many under $200 not even a capital expense.

 

[alphacat]

As I said, i'd rather at this stage learn how to calculate the power consumption by myself, than purchasing a tool that would do this job for me.

Perhaps after i measure this, i'll have an extra motivation to purchase such a meter to verify my result.

 

[alphacat]

anyone please.

 

[blueroomelectronics]

Remind me never to buy anything from the company you work for, they sound like major cheapskates that won't even spring for tools or qualified personal.

 

[alphacat]

you remind me that when I was a kid, I connected two cans to a long thread and placed one can at my home and the second can at a friend's home, and all other kids called me a cheapskate.

Its much more thrilling to acheive your goal with your own hands than pressing some buttons and let it do it for you.

Anyhow, I'd really like to learn how is it possible to measure the true power of a SMPS.
I'm sure people do it using only an oscilloscope, and i'd like to be capable of doing it myself.

 

[Nigel Goodwin]

you remind me that when I was a kid, I connected two cans to a long thread and placed one can at my home and the second can at a friend's home, and all other kids called me a cheapskate.

Its much more thrilling to acheive your goal with your own hands than pressing some buttons and let it do it for you.

Anyhow, I'd really like to learn how is it possible to measure the true power of a SMPS.
I'm sure people do it using only an oscilloscope, and i'd like to be capable of doing it myself.

Why not ask the guy who told you it was easy?.

 

[dknguyen]

you remind me that when I was a kid, I connected two cans to a long thread and placed one can at my home and the second can at a friend's home, and all other kids called me a cheapskate.

Its much more thrilling to acheive your goal with your own hands than pressing some buttons and let it do it for you.

Anyhow, I'd really like to learn how is it possible to measure the true power of a SMPS.
I'm sure people do it using only an oscilloscope, and i'd like to be capable of doing it myself.

While you are sitting there figuring out how to get the most thrill, your company is losing money. THis is a job, not a hobby so efficiency outweights thrill-factor.

If you have a current probe for your meter, you might connect one channel as current, the other channel for voltage and have the meter multiply the two for you and then sit there and work out all the kinks. I doubt they have a oscilloscope current probe though...those things costs even more than power quality analyzers.

 

[blueroomelectronics]

Or use the meters that are part of electronic workbench. It's free for 30 days and has all sorts of simulated hardware I'm sure an RMS meter could be found.

 

[sheldonstv]

before you go measuring anything signal wise from your smps connect it up via an isolation transformer first or as soon as you connect yr scope there will be a bang!!!

 

[The Electrician]

Hi,

I dont have a true RMS power consumption meter in possession.
The only measuring tools I have here are Fluke and digital oscilloscope.
I heard by an engineer that used to work around here that it is possible to measure the true power of an SMPS using a 'scope.

I fully understand the waveform of the current drawn from the mains, which you described - current spikes that charge the capacitor to the voltage peak (~310VDC in the case of 220VAC mains).

So how is it possible measuring the ture power with an oscilloscope?

What is the brand name and model number of your oscilloscope?

 

[alphacat]

Hey, The oscilloscope is
Tektronix TDS 360, Two channel digital real-time oscilloscope, 200MHz, 1GS/s.
Thats whats written on it.

 

[Mr RB]

Connect a second, large electro cap in parallel with your first electro cap, then put a 1 ohm resistor in series between them. Measure the current by the voltage drop on the 1 ohm resistor and the voltage by the voltage on the second cap.

Assuming the 2 caps are large enough you will be able to measure input power within a few percent accuracy as the AC ripple on the 1 ohm resistor is very small and is averaged on your multimeter. You can even add a volt for the voltage drop of the rectifier diode, that might get you within 1 or 2 percent of the true input power.

 

[alphacat]

Hello Mr RB,

Thank you for the great advice.
I have a few questions please.

1. You're actually saying that I can treat the current that comes from the capacitor as DC current, and the voltage on that capacitor as DC voltage; and then you get P = I * V .
Did i get you right?

2. I just dont quite understand why do you suggest to connect a large capacitor in parallel with the existing one?
Is it only to decrease the ripple current and voltage, and therefore being able to measure the true power as the product of two DC components (I and V) with better proximity?

3. Moreover, adding this large capacitor will only increase the apparent power that the SMPS consumes from the power plant, but wont change the true power the SMPS consumes from the power plant, right?

4. Why did you say the connect the resistor between the two capacitors, and not after them? we dont care about the current thats going to the capacitors, but only the current thats going to the system which the PSU drives.

If I understood your method right, then its a really good one, it relays on the fact that Pin = Pout (where Pin is the power that the power plant provides, and Pout is the power that the capacitor provides).
The capacitor itself also requires power from the power plant, but its reactive power therefore we dont need to measure it.
So I wonder why i was suggested to measure areas of curves which is such a hard work.

 

[alphacat]

anyone guys?