Spikes in Tweet-A-Watt

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.
Locked
User avatar
s-fighter
 
Posts: 9
Joined: Sun Feb 28, 2010 12:19 am

Spikes in Tweet-A-Watt

Post by s-fighter »

Hi,

I am trying to make the Tweet-A-Watt using MSP430 microcontroller. I am taking the current and voltage samples as explained in the website. I am getting spikes in voltage pin whenever I switch the Kill-A-Watt ON and OFF. There are also spikes on the current pin when I connect a device to the Kill-A-Watt. These spikes have blown two of my microcontrollers. Any one had the same problem before? Any remedies for the problem?

Thanks,
SF

User avatar
Franklin97355
 
Posts: 23911
Joined: Mon Apr 21, 2008 2:33 pm

Re: Spikes in Tweet-A-Watt

Post by Franklin97355 »

Circuit diagram? From your description I'd say you don't have it engineered correctly. You are trying to do something this device was not designed to do and as such you are responsible for the calculations theat make it work correctly. Since the Tweet-A-Watt does not use a MSP430 you are on your own with that.

User avatar
s-fighter
 
Posts: 9
Joined: Sun Feb 28, 2010 12:19 am

Re: Spikes in Tweet-A-Watt

Post by s-fighter »

Yes, thats true. But the spikes are created even when I don't connect any microcontroller or the XBee module, and observe the waveforms in an osciloscope.

I am using the same circuit as given in this link:
http://www.ladyada.net/make/tweetawatt/solder.html

User avatar
Franklin97355
 
Posts: 23911
Joined: Mon Apr 21, 2008 2:33 pm

Re: Spikes in Tweet-A-Watt

Post by Franklin97355 »

Where are you connecting the scope?

User avatar
s-fighter
 
Posts: 9
Joined: Sun Feb 28, 2010 12:19 am

Re: Spikes in Tweet-A-Watt

Post by s-fighter »

I am connecting the scope to pins from the opamp of the kill-a-watt, before the voltage dividers in the tweet-a-watt circuit diagram.

I blew my kill-a-watt while doing this yesterday :(
I do not know the reason yet.

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

Re: Spikes in Tweet-A-Watt

Post by adafruit »

"You should only attempt this project if you are comfortable and competant working with high voltage electricity, electronics and computers. Once the project is complete it is enclosed and there are no exposed high voltages. However, you must only work on the project when its not plugged in and never ever attempt to test, measure, open, or probe the circuitboards while they are attached to a wall socket. If something isnt working: stop, remove it from the wall power, then open it up and examine. Yes it takes a few more minutes but its a lot safer!"
is there for a reason! if you don't understand the difference between ground and earth ground in your circuit then you will only damage yourself or equipment.

User avatar
s-fighter
 
Posts: 9
Joined: Sun Feb 28, 2010 12:19 am

Re: Spikes in Tweet-A-Watt

Post by s-fighter »

I am sorry to contradict you, but I did not probe the Kill-A-Watt by opening it when it was connected to the wall socket. I had soldered wires to the pins, as shown in the website. I had closed the box, and then used the wires to connect to the scope. I have worked a lot with circuits before, and I found this the safest way to investigate the spikes.

I do not understand why earth ground came to the picture here. I am only using the ground from the LM2902. Is that wrong? Please let me know.

Thanks,
SF

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

Re: Spikes in Tweet-A-Watt

Post by adafruit »

It is not a contradiction:
However, you must only work on the project when its not plugged in and never ever attempt to test, measure, open, or probe the circuitboards while they are attached to a wall socket.
phil, adafruit safety sheriff here.

Is quite clear! You should never attempt to connect test equipment to something while it is plugged in to mains power unless you completely understand what is occurring and have proper safety procedure. Whether or not the box is open is irrelevant
Your scope is earth grounded. The killawatt 'ground' is not.

To reiterate: never ever attempt to test, measure, open, or probe the circuitboards while they are attached to a wall socket. Period!

At this point we cannot really help out and will likely not answer more questions if you're going to do things that are very very dangerous. what you're doing is something we do not suggest or support - while you're free to post here, we may remove the post(s) if we think they're not something that is "good" for the project forums. please email us "offline" if you'd like too.

be safe! k?

:)

User avatar
richms
 
Posts: 558
Joined: Tue Jan 20, 2009 3:05 am

Re: Spikes in Tweet-A-Watt

Post by richms »

You need an isolating transformer on the supply if you are to scope or meter or poke at a non isolated mains powered device like the kill-a-watt and the like.

You should already know that sort of thing before starting to work on high voltage powered stuff. I am suprised that your scope isnt blowed up from this misadventure.

Also you cant connect a load to the kill-a-watt that is grounded, without the possibility of introducing problems to the circuitry if you are interfacing the K-a-W circtuitry to a grounded thing like a scope or computer etc.

Adafruit specified a wireless connection because it means none of that happened, the end result was safe from an end user perspective as it was all sealed in the orig K-a-W case, and the PC was isolated by a radio link.

Otherwise you could whack an arduino or other micro on the sense lines and get a much better more accurate reading than the low res ADC's in the xbee could ever give. The problem with that is people would want to connect it to their computer while live and "blow BANNED up" as you have done.

Count yourself lucky you have only lost a few cheap micros and isolate before playing more, and think about how you are going to get the readings out of the micro to the real world (IR, cheaper RF modules, optocouplers with appropriate design of the board etc)

User avatar
s-fighter
 
Posts: 9
Joined: Sun Feb 28, 2010 12:19 am

Re: Spikes in Tweet-A-Watt

Post by s-fighter »

Well, okay. I apologize for my mistake.

So, you suggest I build the whole circuit with the XBee, put in the box, and get the values at the receiver; as explained in the ladyada site, and not play with the transmitter circuit?

I actually wanted more accurate readings, so was experimenting with the opamp pins, as the voltage levels were low enough. I guess the lack of isolation between the mains and my circuit is the cause of the problem then. But then, what is the isolation in the original tweet-a-watt transmitter? the 10000 uF cap?

I am sorry that I violated those rules, please keep replying.

Thanks!
SF

User avatar
richms
 
Posts: 558
Joined: Tue Jan 20, 2009 3:05 am

Re: Spikes in Tweet-A-Watt

Post by richms »

There is no isolation in the tweet-a-watt between the circuits and the xbee since it doesnt need it. It is sealed in the box so the fact that its circuit ground is sitting at some large voltage over protective ground doesnt matter.

Its only when you connect that circuit out to others that you have problems. I am suprised the current from the connections that you made via the scope didnt have your power protection tripping out, as it was probably leaking 10's of mA at least via the scope connection you did.

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

Return to “XBee products (discontinued)”