Wattcher consistency

Xbee projects like the adapter, xbee tutorials, tweetawatt/wattcher, etc.

Moderators: arduwino, adafruit

Wattcher consistency

Postby Kal » Tue Apr 07, 2009 8:17 pm

Nice project, just what I was looking for.

So we built 6 transmitters and three receivers. They all work, but no two transmitters give the same reading for the same load.

With no load the output ranges from about 5 watts to about 60 watts. By modifying the VREF constant we were able to bring this down considerably on the most eratic ones. But none came all the way to zero.

Also, I mde a change in the python code that seemed to help:
As downloaded:
Code: Select all
# sum up power drawn over one 1/60hz cycle
        avgwatt = 0
        # 16.6 samples per second, one cycle = ~17 samples
        for i in range(17):         
            avgwatt += abs(wattdata[i])
        avgwatt /= 17.0

As modified:
Code: Select all
# sum up power drawn over one 1/60hz cycle
        self.aveWatt = 0
        # 16.6 samples per second, one cycle = ~17 samples
        for i in range(17):         
            self.aveWatt += wattdata[i]
        self.aveWatt /= 17.0
        self.aveWatt = abs(self.aveWatt)


With those two modifications it is better, but the output jumps around a lot, pretty much at random within maybe a 10 watt range.
Kal
 
Posts: 4
Joined: Tue Apr 07, 2009 7:41 pm

Re: Wattcher consistency

Postby adafruit » Tue Apr 07, 2009 10:42 pm

its...not that consistant. any help you can give would be really nice. its only meant to be good to about 10% any more would require a microcontroller (or, alternately, not having Power factor correction) :(
User avatar
adafruit
 
Posts: 3863
Joined: Thu Apr 06, 2006 3:21 pm
Location: nyc

Re: Wattcher consistency

Postby jckohler » Wed Apr 08, 2009 5:46 pm

>its only meant to be good to about 10% any more would require a microcontroller

Which of course would add costs and complexity, but for those of us obsessed with accuracy, that might be well worth it. What rate would you have to sample at? 5 kHz, 10 kHz or higher?
jckohler
 
Posts: 3
Joined: Wed Apr 08, 2009 5:40 pm

Re: Wattcher consistency

Postby Kal » Wed Apr 08, 2009 8:04 pm

Anyone know of an implementation of this.
Kal
 
Posts: 4
Joined: Tue Apr 07, 2009 7:41 pm

Re: Wattcher consistency

Postby adafruit » Wed Apr 08, 2009 11:11 pm

even 10khz would be just fine
User avatar
adafruit
 
Posts: 3863
Joined: Thu Apr 06, 2006 3:21 pm
Location: nyc

Re: Wattcher consistency

Postby franklin97355 » Thu Apr 09, 2009 8:56 am

Anyone know of an implementation of this.

What do you mean by 'this'? (sounds like a quote by Bill Clinton)
User avatar
franklin97355
 
Posts: 787
Joined: Mon Apr 21, 2008 1:33 pm

Re: Wattcher consistency

Postby Kal » Thu Apr 09, 2009 10:24 am

franklin97355 wrote:
Anyone know of an implementation of this.

What do you mean by 'this'? (sounds like a quote by Bill Clinton)


"any more would require a microcontroller", as applied to a Kill-a-watt and with a radio link similar to the Wattcher's.

I know, I should have included a quote. But I was expecting a more explicit attachment to the immediately preceding post.
Kal
 
Posts: 4
Joined: Tue Apr 07, 2009 7:41 pm

Re: Wattcher consistency

Postby franklin97355 » Thu Apr 09, 2009 3:01 pm

something like this? http://enerjar.net/
User avatar
franklin97355
 
Posts: 787
Joined: Mon Apr 21, 2008 1:33 pm

Re: Wattcher consistency

Postby Kal » Thu Apr 09, 2009 3:43 pm

franklin97355 wrote:something like this? http://enerjar.net/


Thanks for the link. It is interesting, but I need something that would be safe around middle schoolers. Too bad the link on that page to the Cornell project is dead.
Kal
 
Posts: 4
Joined: Tue Apr 07, 2009 7:41 pm

Re: Wattcher consistency

Postby fricna » Tue Nov 10, 2009 12:53 pm

sometimes the power factor on my tweet a watt fluctuates wildly, even if there is no load connected. has anyone else experienced this and do you know a solution? is it possible to read the power factor from the quad op amp?
fricna
 
Posts: 4
Joined: Fri Jul 31, 2009 2:18 pm

Re: Wattcher consistency

Postby adafruit » Tue Nov 10, 2009 1:07 pm

note that the power factor is calculated -from- the data
did you check that you have the tweetawatt wired up? most people who have had big fluctuation had a miswiring/assembly
User avatar
adafruit
 
Posts: 3863
Joined: Thu Apr 06, 2006 3:21 pm
Location: nyc

Re: Wattcher consistency

Postby fricna » Tue Nov 10, 2009 6:51 pm

thanks fro the note. the wiring appears to be correct and it seems to be working most of the time...

i actually meant at times ive observed that the power factor that is displayed on the kill-a-watt fluctuates wildly, even when there is no load. typically when i then disconnect the xbee radio and super cap, the power factor settles to 1.00 but ive also noticed that sometimes the kill-a-watt by itself with no xbee radio and no load connected will have a fluctuating power factor. could my kill-a-watt be damaged?

this doesn't happen all the time but im not sure what causes it and id like to find out the cause so i can use the tweet a watt reliably.
fricna
 
Posts: 4
Joined: Fri Jul 31, 2009 2:18 pm

Re: Wattcher consistency

Postby adafruit » Tue Nov 10, 2009 10:46 pm

the PF fluctuates or the Amp draw? how much do each fluc. by?
User avatar
adafruit
 
Posts: 3863
Joined: Thu Apr 06, 2006 3:21 pm
Location: nyc

Re: Wattcher consistency

Postby fricna » Thu Nov 12, 2009 2:27 pm

when there is no load connected, the reported current is as high as 0.03A. the power factor varies anywhere from 0.00 to 1.00. sometimes the variation is rapid, other times it settles to some value and varies by about +/- 0.05. at other times, these same devices will work perfectly and maintain a power factor of 1.00 with no load or a simple resistive load or maybe drop to about 0.96 with a switching power supply like a laptop power supply.

the main problem is when there is no load connected or a very small load connected, then i will get very incorrect readings because the power factor becomes so low. usually when there is some significant load connected, the power factor is at or near 1.00 and the readings become reliable again.
fricna
 
Posts: 4
Joined: Fri Jul 31, 2009 2:18 pm

Re: Wattcher consistency

Postby adafruit » Thu Nov 12, 2009 3:19 pm

30 mA isnt that much. with no load, the PF will vary. it sounds like its working fine, you have to adapt to the data
User avatar
adafruit
 
Posts: 3863
Joined: Thu Apr 06, 2006 3:21 pm
Location: nyc


Return to XBee

Who is online

Users browsing this forum: No registered users and 1 guest

Stuff to buy from the Adafruit store and links to product documentation!