Wattmeter? / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Wattmeter?

previous 1 2

Move Thread

LAN_403

twist2b

Sun May 03 2009, 11:05PM

Registered Member #2086 Joined: Tue Apr 21 2009, 02:33AM
Location:
Posts: 117

Klugesmith - AWESOME thanks!!! :)
That was exactly what i was looking for. :)

Well, why not this though:

Put a resistor inside the meter, measure in parallel, get the voltage, then use the formula I said earlier:
E^2/R
That would create a simpler method to measure, but it would need to be a logic meter....

That work?

...

Mon May 04 2009, 12:31AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

Klugesmith wrote ...

Electromechanical KWH meters in the USA, for split phase 120/240 volt service, commonly have four terminals and no instrument connection to neutral. They respond to the instantaneous product of (Line 1 current + Line 2 current) and (voltage between line 1 and line 2). Even if the meter is otherwise perfect, there are small errors when unbalanced loads (reactive ones in particular) cause unbalanced voltage at the meter.

It depends on what you want to measure, the meter your electric company provides is pretty dang good at measuring *power* delivered (as opposed to VA, balanced or otherwise, it doesn't matter how much current you draw on the neutral or how crappy your power factor is the power meter will read accurately). It is power that the power company sells you (surprisingly enough), so you actually don't pay anymore to run something which crappy power factor since its the power company that eats up the slack.

As a side note, this has been more and more of a problem recently now that everything is powered by cheep SMPS's which have terrible power factor--which has resulted in power companies needing to install better power factor correction.

Dr. Dark Current

Mon May 04 2009, 06:17PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Steve McConner wrote ...

Bob Pease of National Semiconductor has a simple wattmeter circuit:

I have seen that schematic, but I can't get the LM394 here. Also I'm not exactly sure what to mod to make it work at 230V.

EEYORE

Mon May 04 2009, 07:49PM

Registered Member #99 Joined: Thu Feb 09 2006, 06:10PM
Location: florida, usa
Posts: 637

There is a device on the market for $25.00 or so that plugs into the wall and measures all sorts of things such as power. I forget the name, but I could have sworn Steve Ward used one for his tesla coiling and found it to me reasonably accurate. I have one as well.
Kill-A-Watt meter or something like that. You just plug into the outlet, and then plug your device into the meter. It has a display to show you power, VA, voltage, ect.
EDIT: Bam!

Matt

aonomus

Tue May 05 2009, 08:17AM

Registered Member #1497 Joined: Thu May 22 2008, 05:24AM
Location: Toronto, Ontario, Canada
Posts: 801

Having seen the discussion gone towards the Kill-a-watt as a watt meter, I should bring up a few things...

1. It shows power factor, this can be useful if you seem to be tripping breakers with a low Arms. A really low power factor might be due to a bad SMPS or something... (residential power services rarely charge for power factor, but it still affects the instantaneous load for some breakers tripping).
2. The 'tweet-a-watt' mod, essentially hacking a xbee wireless module into the kill-a-watt, and using its on-board ADCs, transmits the voltage and current data to a separate computer. The recieving computer can log the raw serial data and process it, which could provide some useful automatic data acquisition (provided the RFI isn't too bad).

Link:

Steve Conner

Tue May 05 2009, 10:56AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

LM394 is just a dual transistor. You can substitute two ordinary transistors, the accuracy will be lower, but it'll still work. If you like, you can stick them together with thermal paste to keep them at the same temperature, but it still won't be as good as the LM394.

To modify the circuit for 230V, you need to double the value of the resistor chain at bottom left (two 49.9k in series plus a 50k pot, should change to two 100k plus a 100k pot) and change the 0.2uF 250V capacitor at top left to 0.1uF.

GeordieBoy

Tue May 05 2009, 10:57AM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Probably best to buy something ready made as it is likely digital in this day and age. In short, a cheap professionally manufactured digital wattmeter is likely to be way more accurate than an amateur constructed analogue one. Particularly as you said that you want it to measure with some accuracy when presented with difficult loads like rectifier-input SMPSUs with their wide-bandwidth distorted line-currents.

As Steve C said, the traditional analogue way is to use some circuitry that implements a four-quadrant multiplication. (Voltage and current can each be positive or negative at any point in time, so you can be in any of four quadrants.) Most analogue circuits that implement this multiplication are inherently non-linear so they only work well over a limited dynamic range before errors become large. The problem with SMPSU loads is that they draw very narrow but very intense current spikes. This high crest factor can easily drive an analogue multiplier well outside of its small-signal linear region. A second additional problem is that the current waveform also contains hamonics right up into the kHz region which must be preserved during the multiplication. Trying to make an analogue multiplier that is wide-band, has a large dynamic range, low-noise, and has low offset and feed-through will be very challanging. If you achieve this, it will be similarly challanging to calibrate it accurately and temperature compensate it, so that it stays accurate with variations in ambient temperature.

The digital solution with a two-channel ADC and a DSP or microcontroller is so much easier and is free from many of the technical challanges that hamper the analogue multiplier. Digital wattmaters typical sample V and I many times over the duration of one mains cycle. Each pair of measurements is multiplied together in the DSP and added to an accumulator. At the end of each mains cycle the total in the accumulator is read, multiplied by a magic scaling factor to calibrate the instrument and displayed on a screen as the average power. (Alternatively the average power over more than a single mains cycle can be displayed to reduce fluctuation in the reading.) Integrating over whole numbers of mains cycles eliminates 50Hz/100Hz ripple in the power measurement result.

-Richie,

previous 1 2

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob