Most Accurate Way to Measure Power

What do you know about how the scope does the analysis? Also, just how accurate is the scope in general? I don't remember scopes being all that accurate to start with, but that is from the analog days I expect.

Ignoring basic accuracy, any difference has to be involved in the ripple. Voltage goes up a little and current goes up a lot as a capacitance in the load is charged. That power is returned when the voltage goes down. Meanwhile an averaging meter is going to give you the product of the averages rather than the instantaneous product of the voltage and current, right? Or do I misunderstand?

Rick

So now you are also getting into the design of high-PF converters? Welcome to the club.

I never imagined I'd be doing lots of SMPS. Used to be 1-2 a year. Now that is >80% of my daily work.

Displaying PF in SPICE is a real pain. So far I am using product, RMS and CTRL-mouseclick.

Yeah, forget using a scope for this.

In the US it's done the American way, by renting the stuff :-)

Example:

--
Regards, Joerg

http://www.analogconsultants.com/

It isn't a "Spice" problem, it's a real-world measurement issue.

As for Spice issues, PF and such is trivial in PSpice.

As I get myself immersed into the device modeling world I'm finding other Spices are sorely limited in user-added features...

For example, the "public" version of LTspice can't do an adequate analysis of device-level CMOS unless you generate the netlist in a full-featured schematic capture... or laboriously hand-edit the LTspice-generated netlist. ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

But it's blazingly fast and most of all it does not crash. Simulating a converter with complicated switcher chips on PSpice can take forever. LTSpice has behavioral models. Ok, that has its shortcomings but at least I have a result after 10mins and not 1-2h.

--
Regards, Joerg

http://www.analogconsultants.com/

I can't remember when PSpice last crashed.

Oh, yes, Now I remember, you got suckered into using OrCAD :-)

That's where your crashes lay.

Behavioral models... extensive support by PSpice.

But, "sometimes", us chip designers must do it at the device level. ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

El 12-09-12 17:30, Jim Thompson escribió:

If you don't know the load (so it may be non-resistive), you need to determine (1/T)*integral(U(t)*I(t)*dt) over one or more periods. You need to multiply before integrating.

So you need a multiplying action and an integrating (or averaging) action. You can find chips for that, but depending on the required accuracy you could use an oscilloscope with multiplying action, or import measured values into a spreadsheet and do the math on a computer.

What is the highest frequency that may contribute to the power. Shunt resistors may cause non-negligible errors at several 100 kHz frequencies.

I would not use a current transformer unless you know for sure there is no DC current component.

--
Wim
PA3DJS
www.tetech.nl
Please remove abc first in case of PM

Load is a resistive load bank

There IS a DC component... 24VDC + 2.5VP-P/100Hz ripple.

You provoked an interesting thought. Couldn't you do this function via a PC sound card, and some post-processing? ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

They don't do DC well. You could chop it and then use the sound card.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

hobbs at electrooptical dot net
http://electrooptical.net

OK. That's easy enough to do. I'll build up something for my client. He needs help... he's a PhD ;-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

It's only the ones who didn't get soldering irons for their 10th birthday that are helpless like that. (Sending somebody useless to school for 20 years just makes him arrogant as well as useless.)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

hobbs at electrooptical dot net
http://electrooptical.net

Indeed :-)

I grew up in a Radio & TV repair shop. I had a Weller "Gun" in my hands as soon as I could lift it ;-)

Built a TRF toooob radio when I was in Cub Scouts... thrilled that I immediately picked up WBZ (Boston, ~750 air-miles away)! ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

How about using an incandescent lamp and a photocell? Get a reference point for the unknown DC+AC signal then substitute pure DC + measure current. Or measure the temperature?

It was the client's wish and the client is always king :-)

Well, Cadence now owns PSpice. So what can ya do?

Doesn't help at all if you can't get any switcher IC behavioral models that run in PSpice. In LTSpice there's tons, but of course only from LTC. Which is why I pretty much default to their ICs unless something is very cost sensitive or where they don't have something, like on an upcoming push-pull converter design.

I think their decision to provide LTSpice to the community was the best marketing strategy ever conceived in any semiconductor company.

Sure. But us circuit designers are of the impatient kind, always under schedule pressure to get things done.

--
Regards, Joerg

http://www.analogconsultants.com/

How is that relevant? My understanding is that there is a difference between RMS and average for a reason... You can't just take the average of each and multiply, can you?

Rick

El 12-09-12 20:31, Jim Thompson escribió:

If you exactly know the resistor, you could use the soundcard to measure the AC component and use a DC voltmeter (average responding) to measure the DC component

RMS(total waveform) = sqrt(DC^2 + ACrms^2)

You could also use an RMS responding AC meter for the AC component.

I also like the thermal based measurements as suggested by tm. I did this before with loooong mercury thermometers (and now with a thermocouple).

--
Wim
PA3DJS
www.tetech.nl
Please remove abc first in case of PM

Exactly. Use a decent DVM to measure the DC voltage and the true RMS AC voltage. As noted, some advanced math is required.

That same DVM can measure the average current and the average voltage to compute the live load resistance, with the usual cautions.

Most DVMs AC-couple the RMS AC measurement, and average the DC measurement, but some allow you to measure the RMS voltage all at once.

--

John Larkin         Highland Technology, Inc

jlarkin at highlandtechnology dot com
http://www.highlandtechnology.com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom laser drivers and controllers
Photonics and fiberoptic TTL data links
VME thermocouple, LVDT, synchro   acquisition and simulation

Not much. They've neutered the latest version, forcing you to use OrCAD Capture to print, even if you're entering via PSpice Schematics :-(

So I'm sitting tight at v15.7, and shoving needles in the asses of everyone I know at Cadence ;-)

Of course. Marvelous marketing strategy, except it kills sales to any company hanging their hat on Cadence Virtuoso, used by _many_ BIG companies... I've been called in to help them roll their own... as recently as last Spring.

Back in my breadboard days I've been known to run a discrete version along-side my breadboard and build-out one function at-a-time... that's how I got the Bosch motor controller up and running so quickly... used their motor and discrete controller as excitation for my breadboard.

Since no one wants to pay for modeling services I'm considering doing it on my own, creating models by that side-by-side technique, that run anywhere, but charging for the library... and for each "update" >:-}

I could run an LT converter in LTspice, then behavioral model it in generic Spice... allowing comparison to other brands... that ought to annoy a bunch of people >:-}

_Usually_, in a chip design, I'm not under _intense_ pressure, just moderate :-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

[snip]

"tm"? I'm not seeing him in the thread. Got a message-ID?

I thought that thermal was certainly the ultimate, but don't know how to implement it accurately ...Jim Thompson

-- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

| 1962 | I love to cook with wine. Sometimes I even put it in the food.

That sort of method is called "electrical substitution", and is often used in radiometry at wavelengths where there aren't good detectors. The best true RMS meters still use thermal sensing, as well. (I'm very partial to my HP 3403As, but they aren't the most recent developments, though their bandwidth and crest factor are still unequalled AFAIK.)

The difficulty with using a lamp is that the resistance isn't constant, and you can't readily set it. I'd be looking at good quality electronic loads.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

hobbs at electrooptical dot net
http://electrooptical.net