Tweet-a-watt power issue?

XBee projects like the adapter, xBee tutorials, tweetawatt/wattcher, etc. from Adafruit

Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.
nullop
 
Posts: 22
Joined: Tue Feb 09, 2010 2:38 am

Tweet-a-watt power issue?

Post by nullop »

Greetings,

Built my first tweet-a-watt transmitter this evening, and sort of strange behavior.

First part was that the board didn't look like the photos in the instruction (the "new" board or the old board). In fact, the board looked more like the "EZ" board (ie, the zigbee chip doesn't even really fit in the case).

Anyhow, it had the same chipset in it, built the same.. just put xbee setup and capacitor outside of the housing. It powers up, but the LED flashes maybe every 5-6 seconds, and the display never fully recovers.. it'll show data, clear.. show more.. Seems like some of the low power issues others have reported.

It's a P4400 Kill A Watt, serial SH47933 (I had this a couple years prior to tweet-a-watt being announced, so perhaps this version is just fail).

Curious though if anyone else has run into this. Even with the every 6 second burst, the receiving xbee chip never shows the rssi indicator live. Display goes out everytime the LED blinks

THoughts?

nullop
 
Posts: 22
Joined: Tue Feb 09, 2010 2:38 am

Re: Tweet-a-watt power issue?

Post by nullop »

Here's a couple of photos of the actually Kill-a-watt for reference of this older model (Doesn't look like either P4400 listed on http://www.ladyada.net/make/tweetawatt/solder.html )
Board with LM2902N
Board with LM2902N
IMG_0738.JPG (276.1 KiB) Viewed 6193 times
Both sides of P4400
Both sides of P4400
IMG_0737.JPG (261.48 KiB) Viewed 6193 times
backside
backside
IMG_0740.JPG (192.05 KiB) Viewed 6193 times

nullop
 
Posts: 22
Joined: Tue Feb 09, 2010 2:38 am

Re: Tweet-a-watt power issue?

Post by nullop »

When I looked at the pins on the LM2902N, pin 4 measures about 2.4V and goes up to about 3.8 over a few second period, then drops back down to 2.5V. Never comes close to 5V. (Seems to coincide with the every 5-6 second burst of the LED on the zigbee)

W
Last edited by nullop on Sat Apr 16, 2011 7:40 pm, edited 1 time in total.

nullop
 
Posts: 22
Joined: Tue Feb 09, 2010 2:38 am

Re: Tweet-a-watt power issue?

Post by nullop »

One last piece of information.. Channel 1 (yellow) here is pin 4 on the LM2902N, and channel 2 is is the actual LED on the zigbee (used because it was trivial to clip on to, but to show that the voltage drop on pin 4 corresponds to the activation of the xbee.

When the capacitor is not hooked up, it has an even 5.8V coming off pin 4.

so it appears its acting normally, could there be something in the way I put it together to cause the capacitor to never quite get enough juice in it?
oscope
oscope
oscope.jpg (84.82 KiB) Viewed 6191 times

User avatar
adafruit_support_bill
 
Posts: 88086
Joined: Sat Feb 07, 2009 10:11 am

Re: Tweet-a-watt power issue?

Post by adafruit_support_bill »

The Kill-A-Watt units in general have minimal power to spare. Some models have less than others and it sounds like yours falls into that category. There just isn't a lot of spare power to scavange. You could try extending the time period between XBee transmits to allow more time for it to recover.

nullop
 
Posts: 22
Joined: Tue Feb 09, 2010 2:38 am

Re: Tweet-a-watt power issue?

Post by nullop »

As it turns out, works perfectly fine.. when you actually write your changes to the xbee chip like you should have when you started. So dumb!

Thanks, this old model works fine :)

mrlisk
 
Posts: 2
Joined: Tue May 03, 2011 6:52 pm

Re: Tweet-a-watt power issue?

Post by mrlisk »

I have that same configuration inside mine. Unfortunately I still cannot get mine to work, and its not an xbee configuration issue. the connections on the ribbon cable between the two boards broke when i opened it up, so i soldered in some spare wire i had laying around. It's odd, because when i plug the unit in, the keys will beep and if i push on the surface of the screen, it will show the digits as it should. The LED never flashes on my xbee though.

Is it possible that solid core wire in a heaver gauge than the original is causing me power issues? Here's a few pictures...
Image
Image
Image

adafruit
 
Posts: 12151
Joined: Thu Apr 06, 2006 4:21 pm

Re: Tweet-a-watt power issue?

Post by adafruit »

remove the xbee
does the kaw work

davebmiller
 
Posts: 2
Joined: Wed Mar 23, 2011 10:25 pm

Re: Tweet-a-watt power issue?

Post by davebmiller »

I have the 'conventional Kill-A-Watt with the surface mount chip, and have a similar problem--with the XBEE in place, the screen dims when the XBEE broadcasts. Will this cause inaccuracies in the readings? Is there a solution?

mrlisk
 
Posts: 2
Joined: Tue May 03, 2011 6:52 pm

Re: Tweet-a-watt power issue?

Post by mrlisk »

adafruit wrote:remove the xbee
does the kaw work
Yes, unit works fine with the xbee removed. Just tested it.
Also hooked up 5V through capacitor to test the xbee board. transmitter lights up just fine.

Maybe it's as you said previously - theres not enough juice coming from the KAW to power the xbee.

adafruit
 
Posts: 12151
Joined: Thu Apr 06, 2006 4:21 pm

Re: Tweet-a-watt power issue?

Post by adafruit »

try increasing the transmission to once every 5 seconds

Sima
 
Posts: 6
Joined: Wed May 18, 2011 1:35 am

Re: Tweet-a-watt power issue?

Post by Sima »

Hi everyone!

I want to make Tweet-a-watt with two zigbee modules.
I've already configured modules and installed Python 2.5 and also pyserial2.5.win32 on Windows XP machine.
I changed the wattched.py script for COM7 ("the one that my zigbee is plugged to"). But when I want to run the wattcher.py I receive this error:
---------------------------------------------
C:\watcher>c:\python25\python.exe wattcher.py
Traceback (most recent call last):
File "wattcher.py", line 56, in <module>
ser.open()
File "C:\Python25\Lib\site-packages\serial\serialwin32.py", line 56, in open
raise SerialException("could not open port %s: %s" % (self.portstr, ctypes.W
inError()))
serial.serialutil.SerialException: could not open port COM7: [Error 13] Access is denied.
----------------------------------------
Any body knows why is that so? My COM port is not occupied by other software and I have Administrator privilege on Windows.
PLEASE HELP, I must do this in only one week :'(

User avatar
adafruit_support_bill
 
Posts: 88086
Joined: Sat Feb 07, 2009 10:11 am

Re: Tweet-a-watt power issue?

Post by adafruit_support_bill »

It is possible that the COM port was opened and not closed properly. This is a common problem when debugging code. Re-starting the machine will reset it.

Sima
 
Posts: 6
Joined: Wed May 18, 2011 1:35 am

Re: Tweet-a-watt power issue?

Post by Sima »

Unfortunately I even tried this on two different PCs, but again I got this error.
when I disconnect my Zigbee module and run the wattcher.py I receive an error indicating that no device is detected, and when I connect it and run wattcher.py, I receive mentioned error.

User avatar
pegasus
 
Posts: 1
Joined: Mon Feb 21, 2011 11:57 am

Re: Tweet-a-watt power issue?

Post by pegasus »

I have put together three of the tweet-a-watt kits with the P4400s over several months (and models) of the P4400 and noticed that they all power up within about 30 seconds, and the display stabilizes but...

With nothing plugged in, the voltage display is off on all of them. One, for instance, is showing 125.5-126.9vac at idle. Using an unmodified P4400, I double-checked the outlet voltage and it was steady at 120vac. Something with the modified tweet-a-watss was going on...

Also, using a modified version of the python script as the basis for code on a Teensy++ 2.0, I noticed the readings seemed to be all over the place and never really stabilized although the P4400s acted perfectly normal to key presses.

Doing a little research, I extended the transmit frequency from 2 seconds to 4 and noticed that the displayed voltage got 'closer' to the actual target value. Extending the transmit time even further to 10 seconds steadied the voltage readout and now the data coming across is rock-steady on the current and voltage values.

It appears that although the XBee is only transmitting for 25ms (1/4 second, 19 samples at 1ms each and 3ms for the wake-up before sampling and 3ms for the power-down after sampling) that the capacitor never reached a full-enough charge to completely power the ADC circuits on the XBee. Extending the transmit interval to a minimum of 10 seconds seems to allow enough time for the capacitor to fully power the XBee for it's sampling run.

The inaccurate voltage display on the P4400 appears to result from the under-charged capacitor placing a load on the P400 ADC circuitry, resulting in bad samples. Extending the transmit interval appears to allow the capacitor time enough to recharge.

Bearing in mind that electrolytic capacitors 'age' over time, I'm going to extend the transmit interval to 15 seconds to ensure long-term reliability.

Just thought I'd pass along these observations since other appear to be having the same issue. The clue all along was the inaccurate display of voltage on the modified P4400 compared with an unmodified unit. If I only knew then what I know now it would have saved a lot of time tweeking code to compensate for an underpowered XBee adapter :roll:

Locked
Please be positive and constructive with your questions and comments.

Return to “XBee products (discontinued)”