OpenEnergyMonitor Forums

Open-source wind monitoring options (1 reply)

Lomeli — Wed, 13 Apr 2011 14:09:27 +0000

I've run across two open-source wind monitoring options while researching equipment for a project to mount a small wind turbine to a building. [Yes, I know that mounting a wind turbine to a building is probably not a good idea! cf Warwick Wind Trials] Interestingly, neither one is explicitly open-source -- only by request and in consultation.

APRS World, Inc.
Their web site doesn't say that they are open-source, but I have confirmed that they are, over the phone. They are "...working on a new system called ReDAS (Renewable Energy Data Acquisition System). The new system does need a live internet connection on site, but it will be much more flexible and expandable than our current offerings. It allows you to simultaneously monitor multiple outputs from different inverters, along side a variety of other wind and weather sensors. We're still working on the new web interface and display for this new system..."

Etesian Technologies
The company owner was working on an open-source wind data logger. They also sell a nifty wireless, battery-less, anemometer-logger-display pair. The anemometer generates enough current to send the data packets down to the logger and display.

I'm not interested in creating the hardware and software, only in explaining the options to my client. In turn, I doubt they will want to go the open-source route, although I'll certainly present it as an option. At the same time, I can't imagine a technical college being happy with Fat Spaniel as a display solution. So I'm still looking... if you're interested, I can send my annotated compilation of web sites that I've found so far.

Best regards,

Alex D
Clean Energy Partners
Madison, Wisconsin
USA

--------------------------------------------------------------------------------------------------

ps3usercheat
SuperCard 3DS
R4 for 3DS

Arduino data throughput (no replies)

ksquared — Mon, 11 Apr 2011 18:12:35 +0000

I am trying to setup an energy monitor with the Arduino. But what I'm wondering is what is the maximum data throughput with either just serial output or using the ethernet shield. I have tried to research this, but can't find any definitive information that I can understand. Of course, I'm very new to this too, so hopefully people will be understanding.

How to set up multiple data steams with Pachube (1 reply)

Paul — Sun, 10 Apr 2011 20:13:16 +0000

I have downloaded the NUClient Pachube, however, I made changes to the code because I am using an ardiuno ethernet shield. I have one graph working fine but cannot figure out how to add another feed. Has anyone put up their data on pachube and have sample code?

Here is the code:
/*
Nuelectronics ethernet shield client example - sending data to pachube
Created by Trystan Lea, April 31 2010
GNU General Public Licence: [openenergymonitor.org]
*/
#include
#include

//Set these to your feed ID and pachube api key
#define SHARE_FEED_ID // this is your Pachube feed ID that you want to share to
#define PACHUBE_API_KEY "" // fill in your API key

byte mac[] = { 0x90, 0xA2, 0xDA, 0x00,0x3B, 0x8F};
byte ip[] = { 192,168,2,50 };
byte gateway[]={192,168,2,1};
byte subnet[] = {255,255,255,0};
byte server[] = { 173, 203, 98, 29 }; //Pachube.com

//Some variables to send
double valA = 10.23;
double valB = 22.34;

int content_length;

Client client(server, 80);

void setup() {
Ethernet.begin(mac, ip,gateway,subnet);
//analogReference(EXTERNAL);
Serial.begin(9600);

delay(1000);
Serial.println("Connecting...");

}

void loop()
{

//Do things that need to be done even when
//the ethernet shield is connecting etc here.

//If the client has finished sending.

//Increment values so that we can see something happening!
valA += 1.2;
valB += 0.2;

//Do things that only need to be done when
//the client has done sending.
delay(10000);
//Tell the client that it is not done and it needs to start sending again.

Serial.println("start");

//client.process opens and closes the client connection
//It returns a one when the connection is at the data
//sending stage.
//client process will also check for pings
if(client.connect())
{
Serial.println("connecting");
//This seems to work set as 1?
content_length = 1;

// 1) Send feed id, pachube api key
client.print("PUT /api/");
client.print(SHARE_FEED_ID);
client.print(".csv HTTP/1.1\nHost: pachube.com\nX-PachubeApiKey: ");
client.print(PACHUBE_API_KEY);
client.print("\nUser-Agent: ecotinker");
client.print("\nContent-Type: text/csv\nContent-Length: ");
client.print(content_length);
client.print("\nConnection: close\n\n");

// 2) Send the data - comma seperated
client.print(valA);
client.print(",");
client.print(valB);
client.println();
//sends all of the above
client.stop();
Serial.println("data sent");
}
else {
Serial.println("Failed to connect to client");
}

}

Thanks,

Paul.

Help using bridge rectifier instead of voltage divider (no replies)

BenFranske — Tue, 05 Apr 2011 06:09:41 +0000

Hello! I'm trying to create a circuit to measure current used by a motor with an Arduino using a current transformer and a bridge rectifier. I'm afraid it's been too long since I've done AC circuits or calculated RC values so I'm having some difficulty making sense of the numbers I'm getting, any help would be greatly appreciated.

My specific hardware, with links to the datasheets where applicable:
CR Magnetics CR8410-1000 Current Transformer
Fairchild DF02M DIP Bridge Rectifier
Currently using 200 ohms of burden/load resistance on the AC side of the rectifier for the CT
Currently using 100uF capacitor from the positive output of the rectifier to ground
Currently using 10K resistor in parallel with the capacitor
Measuring across the dc side of the rectifier

Based on the datasheet for the CT I thought I should be using a 100 ohm burden resistor but the voltages across it that I'm seeing when the motor runs (it's on a sump pump either 1/4 or 1/3 HP I think) are lower than I expected based on my calculations (all well below 1V) so I bumped the burden resistance up to 200 ohms to raise the voltages. With a multimeter I'm now measuring voltages around 1.4 Vdc when the motor is running, this still seems a little lower than I'd expect...

One of the things that's somewhat important to me is getting an accurate run duration (being able to detect when the motor starts and stops with reasonable accuracy, say 1/2 second or less) and I'm somewhat concerned that the capacitor is causing a delay in detecting the stop time. If I'm thinking correctly I could reduce this lag by reducing the capacitor size however the only other capacitors I have on hand are a couple non-polarized 1uF and some tiny ceramics. If I were to order another size capacitor what would be ideal.

The circuit itself is a dedicated 20A circuit so ideally I would be able to measure up to 20A but realistically I expect the motor draws somewhere around 10A. I'm located in the US so this is 110/120v. Again, any help is greatly appreciated!

Burden resistor sizing (no replies)

reh — Sun, 03 Apr 2011 19:02:53 +0000

Hi

I have a circuit board to take inputs from 4 CT sensors - it is basically 4 copies of the single CT sensor circuit. I want to have different inputs sensitive to different current ranges, but am having some trouble choosing the burden resistors for low current values (i.e. the most sensitive input).

using the same CT sensor and a test 8.6A purely resistive load, i get the following results (the first column is the burden resistor in Ohm as directly measured on the board; the second is the minimum-maximum reading from the Arduino when using an RMS 8.6A current; the third is the consequent calculated maximum RMS current that can be read; the fourth is this range time the burden resistor value):

burden		min-max A/D	max current range	range x burden
30.2		106		83			2509
51.4		157		56			2883
102.7		259		34			3492
997		340		26			25823

According to my reading of the detailed system design page, the last column should be a constant (related to the number of turns in the CT sensor and the bias voltage). The agreement is pretty good for the first couple of lines, poor for the third, and useless (out by a factor of 10) for the last.

Is there a more accurate way of calculating the burden resistor value, or should i just use trial and error?

Brand new (no replies)

Jeremy.grimm — Thu, 31 Mar 2011 19:10:02 +0000

Hello,

I am brand new to this and want to build a wind turbine. I have some questions that I can't seem to locate the answers to. All I want to start with is powering my little shed in the back yard with a turbine. This is something that I will not use every day for energy, but I will have an air compressor for airing up tires (should only need to run at about 10 minutes at a time) maybe a circular saw once in a while, an electric drill once in a while, and 2-4 lights once in a while. I want to know how do I keep my turbine from overcharging the battery bank. Is there a product that I can get that will shut off the power to the bank when it's at peak charge? or am I going to have to manually turn it on and off to charge it? I want to run power to my shed, and I though this would be a fun and inventive way to do it. Thank you.

Communication (1 reply)

pyro — Fri, 18 Feb 2011 12:54:46 +0000

From reading on the site it sounds like the general concept is to have a networked Arduino send data to a webserver somewhere. A few basic communications question on this approach:

When is a TCP socket opened and closed?
How often do you send data? Is it a single sample, or averaged?
What port? Most web servers have firewalls blocking ports from incoming connections

My approach has been to have the webserver in control. I use a bi-directional interface with a request from the webserver and the power meter transmitting a reply. Do you invision this type of setup would be compatible with your setup?

Nothing connected and Arduino shows reading !=0 :S (1 reply)

zayanimeister — Fri, 04 Feb 2011 09:07:11 +0000

Hello everyone,

I am having a problem with my setup. I am using a Blackwidow arduino board (with built-in wifi) [asynclabs.com]
with a current sensor and an AC-AC adapter connected to two analog-in pins as detailed on the websote: the same electronics, same 2.5v DC bias, and yet, I have some problems with the readin:

For a 1900w purely resistive load I have a real power lower than the apparent power, and the apparent power is the correct one. The resulting power factor is 0.93 instead of 1.

I am suspecting noise is the source of these problems, and tried to see what's the reading when everything is disconnected, so I connected the board to USB, launched the energy monitor latest sketch and read on the serial port some values for both current and voltage + a power factor of 0.56

I am on my ass, banging my head against the wall, can anyone confirm the same result with their setups?

My biggest worry is that the board is creating this noise, it is not FCC or CE certified :S

PLZ help :'(

Amin
=============
www.bouhablog.com

"Router" energy monitor (5 replies)

angelvalorreed — Sat, 19 Mar 2011 17:04:30 +0000

Hello Trystan and everybody,

I have reduced the number of modules of the system, from two to just one, by replacing xbees by a Linksys wrt54gl router.

The final system is depicted here:

[picasaweb.google.com]

I have substituted the two xbees, plus the two xbee adapter boards (LadyAda) plus the Ethernet Arduino, with the router. This allowed to cut costs by approximately 40 € ( $ 54).

I hope it can be of interest for you guys, as I think it makes the system easier to setup and to mantain (of course you need to have a wireless modem, but I think these are common nowadays).

I can supply further details on upgrading the router firmware to the open operating system dd-wrt (which I did) as well as supply the new arduino sketch (which now does everything).

I would like to thank Trystan and the rest of the team once again for sharing ideas and knowledge... thanks to you I am monitoring my house and reducing consumption!

Best regards, Angel Valor Reed

Problem with USB port and web port (2 replies)

binooetomo — Sat, 29 Jan 2011 04:39:23 +0000

Dear All ..

I just try to use the software.

Well , I try to simplified the "MasterEthernetTemplate" , since I just want it to read one Analog pin (Pin A3) and send the value to web:
- Strip out the NewSoftSerial call
- Strip out all line that have "laptop" in it.

The ethernet-send is work well.
But I have a problem with USB port.

Whenever I plug the ethernetShield, the USB port become unuseable.
I look to your code and there is ---> Serial.begin(9600);
I knew that you use the default serial to call another unit via readFromPVunit() function.

Is it true that the ethernetShield will always make USB port disabled ?
What about the UART port behind FT232 chip ... is it still fuctioning ?

Problem number 2 : How to make this sketch to send the data on port 8888 instead of 80 ?

Sincerely
-bino-

Multiple current sensors (2 replies)

reh — Tue, 22 Mar 2011 18:41:14 +0000

Hi

apologies if this is the wrong forum - it isn't really a hardware question, but there isn't a software forum.

I'm following the open energy monitor v3.0 design, but with multiple current sensors. I have a question about how to implement the phase calibration in the sketch.

Currently this is done by modifying the filteredV (only), but I guess that the phase calibration takes into account both the voltage and current sensor behaviour, and with more than one current sensor the calibration constant might be different for each.

Is it just a matter of switching the phase calibration to the filteredI measurements and going through the power factor calibration once for each sensor?

Efergy CT appears not to work - what am I doing wrong? (2 replies)

edjones — Sun, 16 Jan 2011 20:17:48 +0000

Hi all,

I'm struggling to get the non-invasive energy monitor (referenced here) working. I have the AC adaptor measuring voltage fine, and it's tuned to within 0.5V of my multimeter's reading. However, I'm struggling with the current sensing. I've tried two different CTs - one from CurrentCost, and one from Efergy as referenced on the page above. Neither appear to report sensible readings when using the 'basic sketch with calibration' (I've also tried the emon sketch without the tuning parameters).

I'm at a loss to know how to debug this. I've tried a few high-current things with the CT clamped around (hairdryer, toaster etc.), but I appear not to get a current reading at all. When the appliance is switched on, a typical line in the Arduino serial monitor is:

0.06 0.82 0.08 245.56 0.00

It's almost identical when the appliance is off :-(

I'm using a 100 ohm burden resistor.

Can you suggest how I might work out what's wrong? (I'm confident the circuit is breadboarded as per the diagram)

Thanks in advance.

Ed

Kirikou: an open electricity monitor (2 replies)

zayanimeister — Fri, 07 Jan 2011 23:59:50 +0000

Hello folks!

I have been following this project for quite some time now, and I finally decided to take on the challenge: launch a commercial product based on the open energy monitor.

I applied to a fellowship at the Zurich HUB that aims to incubate entrepreneurs with promising ideas that serve sustainability issues, but in order to do my pitch, i'll need to have a prototype to show.

My idea is to build an energy monitor based on the Blackwidow board (native Wifi capabilities), but I am a poor intern now, and I need support.

So I created this Ulule page, to raise 120 euros and build my Kirikou:

[www.ulule.com]

Your help is much appreciated! Please spread the word and comment with your opinions and sugestions.

Amin

mini JQuery AJAX tutorial (no replies)

trystanlea — Wed, 08 Dec 2010 17:06:34 +0000

I've written a mini JQuery AJAX php mysql tutorial here:
Using JQuery AJAX and php to fetch data from a mysql database

Demo of tree structure backend (1 reply)

trystanlea — Sat, 04 Dec 2010 10:41:34 +0000

Suneil and I have been working here on getting a demo of the tree structure navigator running

Here's a link to it:

[dev.openenergymonitor.org]

Gives a bit of a feel for how it could be..

How the different modules link up (no replies)

trystanlea — Fri, 03 Dec 2010 20:52:42 +0000

Hello Guys

Thought I would start a thread to explore how we connect up the different modules we have outlined in the system design diagram, as we have been discussing on the phone.

So how does this work?...

The emonCore is php scripting interfaced with a mysql database. The main interface communicates with the core via restful http requests.

The interface is php scripting communicating via curl to the core?

The backend could be javascript + jquery as I have been developing it, communicating with the interface via AJAX http requests. This would make the backend mainly client side. Or it alternatively it could be done via server side php...

The interface consists of: emon or wattDepot client, does the interface consist of any other modules?

Does the backend communicate with the interface via http requests and then the interface with the core again with http requests?
Would the reason to do it this way be principally to make it possible to interface with other cores such as watt depot? as we have discussed is a good thing.

This would work but I liked the idea of the client being a library class that could be included in the backend.

Do we need a second client class for this?

What do you think?

Trystan

Statistics calculations (no replies)

cagabi — Mon, 22 Nov 2010 11:38:55 +0000

Hi,

After reading the basic actions (object methods) proposed by Trystan in design ideas.odt i was wondering about calculating statistics for a given object.

"Who" do you think should calculate them? The emonBackEnd application, the interface or could it be a plugin?. I think that maybe the best option is the plugin, it ask for the raw data throw the interface and then calculates everything.

emoncms dev : start here (1 reply)

trystanlea — Mon, 22 Nov 2010 11:28:49 +0000

Hello!

Welcome to the emoncms forum, thought I would add this here, in-case you have just stumbled across the forum and are wondering what we are talking about. So here's a very brief description of what we are doing:

We are building an open source web app (called preliminary emoncms) for storing, structuring, visualising and sharing sensor data from sustainable energy setups.

Here's a brief walkthrough of how the site could be setup and operated to give an idea of the steps involved and the site features:

Concept walkthrough

install on your own server

user

(I0 objects)

visualisation

statistics

Arrange

tree structure

Share

custom frontend dashboard

interactive maps

Scalable
We want to make it useful for home scale up to CAT microgrid and community energy scale. On the home scale there may just be a single energy monitor or maybe a solar hot water or PV system. On the CAT microgrid scale you have 12+ different monitored loads and generators. On a community/group energy scale you might have 20+ houses each with their own energy monitors plus maybe a community hydro or solar array.

Modular
We are going to try to design it to be as modular as possible, we would like it to be able to interface with other initiatives like wattdepot via their API.

Design Overview

The diagram below describes the system layout.

The core + api module contains the database + database queries a core API to call those queries.

The interface is a module is there to provide the ability to swap out our core+api for say the wattdepot server.

The backend is like an admin panel it is the GUI for setting up and operating the system.

The system can be expanded with plugins.

Concept screens

Backend GUI
Interactive map
Custom dashboard page

Hope this gives an overview glimpse of the idea, for more detail on each part check out the other topics here.

Thanks for reading and feel free to join in!

Trystan

Tree datastructure (2 replies)

Suneil — Mon, 22 Nov 2010 18:49:12 +0000

(see thread on 'words used to describe things' for some terminology used!)

So far have investigated two hierarchical datastructures in MySQL, these are:

1. The Adjacency List Model

2. The Nested Set Model

The first has some limitations such as:
-to see the full path of a node we have to know the level at which it resides
-care must be taken when deleting nodes because of potential for orphaning an entire sub-tree

The second (nested) seems to be more robust, but further reading is needed. It would appear that the first method might be adequate, so long as we put in place good code to deal with all eventualities regarding insertion, deletion, updating, searching and returning path/structure of the tree.

The main things I have been considering at this stage is where to store data within the tree, so as not to replicate data, and also to allow fast searching/retrieval of the required data. Some issues considered so far:

1. Priviledges and how these are inherited through the tree - I think it would be best to have priviledges associated with the IO_Object i.e at the leaf level. This is so that the owner of the leaf can have control of who can view that leaf.
The User object would contain the list of nodes that a particular user can view, thus e.g. if userA shares her IO_Object with nodeX, userB would be able to see it if they belong to nodeX. For a user to become a member of a node, the node (group/location) would be either publicly viewable/joinable, or an admintrator would need to approve joining that particular node. Nodes may also contain lists of users, but this is repetition of data which could be derived from the User Object. As priviledges will be associated with the leaf objects, the owner of the leaf will determine what is viewable by other nodes, however by default, the IO_Object 'name' will be publicly viewable by all nodes in that branch, but not necessarily the data within that IO_Object.

2. Granualarity/statistics of data: for security reasons, certain real time data (such as a users power consumption) should not be made public, however, cumulative totals over longer time periods would be safe to share with the wider community. There may be tradeoffs between deriving statistics from raw data, and execution time to do this. Statistics derived from real-time streams of data (i.e. IO_Objects) can be stored in other Objects also of type IO_Object (leaf). E.g. the IO_Object 'DailyEnergy' could derived from the IO_Object 'InstantaneousPower'.

3. The tree structure will need to follow strict rules to maintain its integrity. E.g. should we allow a leaf to share its data with two parents? Also, a group node may also belong to many physical location nodes, but in order to preserve the tree integrity, social groups that belong to more than one physical location may just have to be contained in one physical location. These exceptions will be rare, but may need to be enforced to maintain correct cumulative totals of energy further up the tree. So perhaps leafs can only belong to one branch of the tree, as they should in real trees! ;)

4. Lastly, when a node is created, I would assume it is up to be administrator/creator of that node to decide which other parent nodes to connect to. Ultimately IO_Objects data is protected by the owners of those objects, however if for example two community energy groups were to merge, it would be good to combine their energy generated. Thus I guess, individual alerts would have to be sent to each leaf, and owners of those leafs can decide whether the share their data with the newly merged node/group.

---------------

Some initial thoughts on core data for each object:

Users: userID, name, nodeID(list)

Nodes: nodeID, name, adminID

Leafs: leafID, name, typeID, adminID, sharedIDs(list of userIDs, nodeIDs)

LeafType: typeID, name, Load_Gen

Tree: nodeID, leafID

-----------------
LeafType 'name' might be: SolarGen, WindGen, CHP, HydroGen, Temperature, WindSpeed etc.
LeafType 'Load_Gen' would be: Load, Generator, neither.

LeafType will allow users to query the tree for all generators/loads..

Well done if you managed to read this ramble far! Hope you get the jist/angle/scope at which I am approaching.. Could be that many of these things will be better dealt with later on in the project, as plugins /add ons to the core functionality.. anyway, the more we discuss these ideas, the clearer things will become!

words used to describe things.. (3 replies)

Suneil — Mon, 22 Nov 2010 16:33:37 +0000

some definitions/terms used for our tree structure, these are fairly arbitrary, we can change them if we find better words to describe stuff.

Objects: Nodes or leafs
Nodes: (physical) locations, (physical) devices, (social) groups (Perhaps for now social groups basically reside in physical locations, social groups can also have 'child' nodes and leafs)
leafs: IO_Objects

Other definitions/terms: Users, Loads, Generators, WindGen, SolarGen, HydroGen, CHPGen, Temperature, WindSpeed etc..
Some fictional/real community energy groups/locations could be: Snowdon community, Mach, CAT, NorthWales, Wales, England etc.

The reason to call it IO_Object as oppose to a variable might be good so that we can include actuators (i.e. output objects, displays etc.) as well as input objects that get data from the outside world. We could just called them leafs too!

Backend Demo's (2 replies)

trystanlea — Mon, 22 Nov 2010 12:00:36 +0000

Here are my thoughts so far in demo form:

These demo's are only a start, they only demonstrate a flat structure at the moment.

Starting to get the hang of javascript and jquery, jquery has a very powerfull feature called
selectors its worth having a look at, I have attached the source so if any of you would like to have a look.

Backend Demo
Download: backend.tar.gz

Demo object list only
Download: objectList.tar.gz

Next Im going to try to further develop the GUI to allow objects to organise hierarchically.

Here is the design ideas document I sent via email, which adds a bit of context to the above: designIdeas.odt

emonClient (9 replies)

cagabi — Thu, 18 Nov 2010 23:45:33 +0000

What is emonClient?

emonClient is an experimental class to test different possibilities for the emon system

emonClient is a class which talks directly to the emon core API. This client would alow every developer to easily write programs that works with emon core , so that everybody can use it in their own projects.

Do we need emonClient?

Maybe it is not strictly necessary to have this class as once we have defined the emon core API it shoud be easy to talk to the emon core. Once we have the API defined and this class, other developers will be able to write their own functions/methods that follow the API language or they can include this class and just call to its methods

One positive thing about having this class is that we can write a very good client, for example introducing error handling using exceptions. If we do it in this client instead of in the backend (or interface) we can easily write, for example, a very good CLI backend to talk to emon core (and not have to write the exceptions again). Another posibility is to add extra functionality, for example i don´t think it is a good idea that the core calculates stadistics (as i see it, the core would only provide a set of data) but we can write a method in the client to calculate these statisctics (for example the average generation in a certain period of time)

Maybe you think i am adding to many "layers" to the system and that we don´t need this client, in that case please read the previous paragraph again think about it and then lets have a chat, of course i am open to new ideas suggestions.

How does emonClient fit in our system?

The basic structure of the system is: backend - interface - emon core API (see systemDiagram.png by trystan, email 16 Nov)

The "interface" is a class that implements all the methods that the backend calls. The interface for emon core API would include the emonClient class.

I am workin in a very basic example about this, that hopefully will be ready tomorrow.

IDs for locations, devices and variables (1 reply)

cagabi — Wed, 17 Nov 2010 23:31:12 +0000

Suneil, once again (as with the xml file for organizing the database) i think you were absoutely right: the IDs for locations and devices should be a number defined by the Core.

It can´t be something like "location_device_etc". If we do it this way, ID will change everytime a user reconfigures a device, for example when moving the device from one location to another. If the ID changes we would have to upload the new ID to the device and it would be a mess

Anyway, i still think it is good if every device and location has a "name" or any verbal form that makes the to be recognizable.

database structure organized in an xml file (no replies)

cagabi — Wed, 17 Nov 2010 21:58:15 +0000

Hi team,

The other day i saw clearly that it could be a good option to have the information about parents and childs of locations and devices in an xml file. Now i think it is a crazy idea.

Imagine that we have some users, all of them configuring their locations at the same time. Each time that one of them makes a change we have to write the whole xml file again, and if we have one hundred locations and thousands of devices it can be a big file. Then for sure two users will want to save their changes at the same time. This is the main reason to dismiss the idea: if we do the changes directly to the database we only have to change one field in one table.

I haven´t even tried to write any code or experimented anything, but i see it very clear, so it doesn´t make any sense.

Anyway i think that the tree structure of locations + devices suits really well in an xml file, so lets keep it in mind.

CT PCB Board (6 replies)

pyro — Wed, 02 Mar 2011 08:29:44 +0000

Hi, I currently have my own version of a house energy monitor using clamp on CTs on the mains coming into the main breaker panel. I am now interested in extending this to sub-circuits. Not sure exactly how many, but somewhere between 20-40. I have settled on solid core CTs for the sub-circuits. That means you'll have to disconnect the circuit temporarily and pass the 14AWG or 20AWG wire through it. This is the lowest cost option available, and while its a pain to install, I don't think it would take that long.

I was contemplating designing a PCB to mount as many CTs as I can on a board. My circuit panel breakers are exactly 1 inch wide. So far I found 14"x10" PCBs for fairly cheap. But they are expensive for low quantities. Each CT measurement board would be 1.5" x 14" meaning I could put about 12 CTs on each board with an external connector. I could get about 6 CT measurement boards out of a single panel. The CTs will transfer current to the microcontroller board. I'll have a burden resistor on that board near the uC A/Ds. However, I realized that since there are not many connections, its much easier to breadboard this on a piece of vector board. Especially since I'll start off with one or two for testing and then go up to a larger number.

I plan to use an arduino mega board to start, but using the GNU AVR C compiler to get maximum speed for A/D and RMS calculations. If I did create an entire PCB panel, I was thinking I could fit a shield on the end. The shield would have the burden resistor, some analog multiplexers to increase the quantity of analog inputs, and connectors to go to the various CTs.

But I wanted to run it by you guys to see if you had any interest in fabricating a board like this? Again, I'm most likely going to just do a prototype board since the connections aren't that numerous.

Advice on getting started. Looking forward to a great project (1 reply)

AnthongRedbeard — Mon, 29 Nov 2010 18:59:03 +0000

Hi, I'm looking to take on some alternative enegry generation monitoring as a project.

I want to mess around with various energy generating projects but I don;t want to do that until I have a nice setup for monitoring and storing that energy. Once I have a few things going I want some metrics to tell me which one is doing best, and when I do experiments I want to know the measurable results.

I have found many sites with info on building power monitors for consumption but not much in the way of generating it. I assume I'll want an arduino, I'm familiar with it somewhat from my reprap project. Beyond that can anyone point me in a good getting started direction?

Thanks, I look forward to being at the point where I can contribute :)

Open-source wind monitoring options (4 replies)

Alex DeMegaJoule — Tue, 04 May 2010 21:21:54 +0000

Welcome (2 replies)

trystanlea — Mon, 23 Aug 2010 16:31:48 +0000

Welcome to the open energy monitor forums. If you have any questions, want help with building or post up developments of your own please feel free to use the forums here :)