Release 303
10 February, 2012 - 15:12
Whats Changed
- Support for the EDF current cost monitor
- issue 31 - Graph web page did not handle multiple feeds
- issue 32 - Plugboard not instance aware
- issue 33 - bscmsg not instance aware
- issue 36 - Support ID=0 when using xap-pachube as a service
- Note on instance web page about unique MAC addresses
- LUA frame:getValue() api call should be case agnostic
Thanks goes out to Karl for his code contributions to this release.
Brett
15 February, 2012 - 14:15
#2
false start
ECO manager - it sorta works. I have flat graphs on Pachube. It takes the first reading and repeats.
XAP messages show first reading.
what do you need? debug/ output raw data?
updated as I wasn't very specific.
have rebooted & hit with hammer.
15 February, 2012 - 13:03
#3
only certain feeds
Currentcost feeds are fine as are jeenodes.
1wire seem to be flatligning at one level.#
https://pachube.com/feeds/15114
Dean
15 February, 2012 - 13:17
#4
Any Other Changes?
Dean,
Are you seeing all the 1wire devices in the web GUI, try resetting the 1wire bus, or power off/on the HAH. It looks like they went off at 12:47, did anything else change at around that time? Your alarm, hot water tank and pump also stopped at this time, could it be a cable issue?
15 February, 2012 - 13:28
#5
will try a power off/on when
will try a power off/on when i get home (might phone wife who will groan) also noticed this
303/1.0
the 1.0 is worrying as i have a 328 chip on 3.4 (memory serves correctly)
I am not at home so nothing has changed, note to self do not upgrade firmware when not at home (250 miles away)
Dean
UPDATE.
power off/on resolved the issue.
15 February, 2012 - 13:22
#6
That's your problem. The
That's your problem. The 1wire is on the PCB, and the HAH software is no-longer talking to the PCB, that's why it shows "/1.0", a power reset might sort it out, but you need to get that resolved for the 1wire to work,
Karl
15 February, 2012 - 13:47
#7
Yep that fixed it
will be putting in a remote reboot system.
wife is not impressed.
Dean
18 February, 2012 - 19:31
#9
Maybe quick to fix?
In looking at the data coming from the ecomanager, the xml is sometimes pretty cr*p!
From (briefly) looking at the source code in currentcost.c I think that what is happening is the SAX parser is failing when the xml is malformed e.g. in the following capture that contains the historic data there is no closing </msg>, opening <d18m02> closes with </d00m02> etc etc
<msg><src>EMC2-v1.79</src><date>18/02/2012</date><time>18:40:00</time><d18m02><kwh>0</kwh></d00m02><d19m02><kwh>0</kwh></d01m02><d20m02><kwh>0</kwh></d02m02><d21m02><kwh>0</kwh></d03m02><d22m02><kwh>0</kwh></d04m02>
<d23m02><kwh>0</kwh></d05m02><d24m02><kwh>0</kwh></d06m02><d25m02><kwh>0</kwh></d07m02><d26m02><kwh>0</kwh></d08m02><d27m02><kwh>0</kwh></d09m02><d28m02><kwh>0</kwh></d10m02><d01m03><kwh>0</kwh></d11m03><d02m03><kwh>0</kwh></d12m03><d
03m03><kwh>0</kwh></d13m03><d04m03><kwh>0</kwh></d14m03><d05m03><kwh>0</kwh></d15m03><d06m03><kwh>0</kwh></d16m03><d07m03><kwh>0</kwh></d17m03><d08m03><kwh>0</kwh></d18m03><d09m03><kwh>0</kwh></d19m03><d10m03><kwh>0</kwh></d20m03><d11
m03><kwh>0</kwh></d21m03><d12m03><kwh>0</kwh></d22m03><d13m03><kwh>0</kwh></d23m03><d14m03><kwh>0</kwh></d24m03><d15m03><kwh>0</kwh></d25m03><d16m03><kwh>0</kwh></d26m03><d17m03><kwh>0</kwh></d27m03><d18m03><kwh>0</kwh></d28m03><d19m0
3><kwh>0</kwh></d29m03><d20m03><kwh>0</kwh></d30m03><msg><src>EMC2-v1.79</src><date>18/02/2012</date><time>18:40:00</time><m02y12><kwh>0</kwh></m02y12><m02y12><kwh>0</kwh></m02y12><m02y12><kwh>0</kwh></m02y12><m02y12><kwh>0</kwh></m02
y12><m02y12><kwh>0</kwh></m02y12><m02y12><kwh>0</kwh></m02y12><m02y12><kwh>0</kwh></m02y12><m02y12><kwh>0</kwh></m02y12><m02y12><kwh>0</kwh></m02y12><m02y12><kwh>0</kwh></m02y12><m02y12><kwh>0</kwh></m02y12><m02y12><kwh>0</kwh></m02y1
2><msg><src>EMC2-v1.79</src><date>18/02/2012</date><time>18:41:00</time><sensor>S</sensor><id>0</id><type>1</type><ch01><watts>0</watts></ch01><ch02><watts>0</watts></ch02><ch03><watts>1</watts></ch03><ch04><watts>0</watts></ch04><ch0
5><watts>0</watts></ch05><ch06><watts>0</watts></ch06><ch07><watts>0</watts></ch07><ch08><watts>0</watts></ch08><ch09><watts>0</watts></ch09><ch10><watts>0</watts></ch10><ch11><watts>0</watts></ch11><ch12><watts>0</watts></ch12><ch13>
<watts>0</watts></ch13><ch14><watts>0</watts></ch14><chH><watts>0</watts></chH></msg>The xml block we're interested in the above capture is actually well formed:
<msg><src>EMC2-v1.79</src><date>18/02/2012</date><time>18:41:00</time><sensor>S</sensor><id>0</id><type>1</type><ch01><watts>0</watts></ch01><ch02><watts>0</watts></ch02><ch03><watts>1</watts></ch03><ch04><watts>0</watts></ch04><ch0
5><watts>0</watts></ch05><ch06><watts>0</watts></ch06><ch07><watts>0</watts></ch07><ch08><watts>0</watts></ch08><ch09><watts>0</watts></ch09><ch10><watts>0</watts></ch10><ch11><watts>0</watts></ch11><ch12><watts>0</watts></ch12><ch13>
<watts>0</watts></ch13><ch14><watts>0</watts></ch14><chH><watts>0</watts></chH></msg>But all the rubbish before it causes the parser to fail.
The logic in the source code in "serialInputHandler" looks for </msg> and then searches for <msg> from the start of the buffer, which then includes all the rubbish xml and passes it to "parseXml".
If the logic could be changed to "look for </msg> and then for the first <msg> working backwards" and then pass *just* that block of xml to "parseXml" I think it would solve the problem.
My knowledge of C is limited, so I've not been able to fix it myself (other than I added some warning, error and fatalError handlers for SAX) but in perl (for example) it would be relatively trivial to use a regex to grab just the section of xml that is needed.
Hope this helps!
M
18 February, 2012 - 20:20
#11
Hi Brett,I've got 2 of the
Hi Brett,
I've got 2 of the ecomanager units, they both do the same :-( Not sure if it's bad programming on EDF / CurrentCost part, or something wrong with how they're squirting the data out over serial....
The capture above was direct from minicom - I noticed the data in the other thread about the ecomanager is the same.
I grabbed a debug capture from your app and attached it. Note that I added an "errorSAX" handler so I could see the SAX parser throwing errors.
You can see that when presented with the badly formed XML the SAX parser throws errors and therefore the "good" bit of the XML is never seen.
Martyn
| Attachment | Size |
|---|---|
| debug.zip | 82.51 KB |
19 February, 2012 - 11:48
#13
Hi Brett,I did try all sorts
Hi Brett,
I did try all sorts of combinations of flow control settings and such like on a PC but the captured xml was always malformed where the historic data is concerned...
Your idea of re-setting the input buffer gave me some inspiration and I came up with some (likely badly written) code that fixes things for me:
diff currentcost.c currentcost.c.old
402,406d401
< static void errorSAX(void *user_data, const char *msg, ...)
< {
< debug("</%s>", msg);
< }
<
424,426d418
< handler.warning = errorSAX;
< handler.error = errorSAX;
< handler.fatalError = errorSAX;
463,473d454
<
< if(strstr(serial_xml, "<msg>")) {
< debug("BOM");
< //reset serial_xml
< serial_cursor = 0;
< serial_xml[0] = 0;
< //set a new tag
< strcpy(serial_xml, "<xml>");
< serial_cursor = 5;
< debug("xmlmw1:%s", serial_xml);
< }
475,478c456
< char * found;
< found = strstr(serial_xml,"</msg>");
<
< if(found) {
---
> if(strstr(serial_xml,"</msg>")) {
480,486c458
< debug("xmlmw2:%s", serial_xml);
< //set new end tag
< strncpy(found, "</xml>", 6);
< debug("xmlmw3:%s", serial_xml);
<
< if(strncmp(serial_xml,"<xml>",5) == 0) {
< debug("FULL XML");
---
> if(strncmp(serial_xml,"<msg>",5) == 0) {
491d462
<
496d466
<
I also attached the new currentcost.c file.
So far it seems to work well for me!
Martyn
| Attachment | Size |
|---|---|
| currentcost.zip | 5.65 KB |
19 February, 2012 - 22:40
#15
looks a lot better
Brett
Just grabbed the beta and my pachube graphs are now updating a treat!
https://pachube.com/feeds/41007
Thanks.
The local graph seems stuck but I guess thats just impatience? -yep impatience
Alex
19 February, 2012 - 19:36
#16
all good here
Hi Brett,
I checked out the latest version, compiled and have it working - much more efficient than my feeble effort ;-)
Thanks,
Martyn
22 February, 2012 - 10:33
#17
catching up....
"I believe Kevin9 has the EDF meter working using this code base. Kevin can you confirm that?"
After the capture I sent you, I did run the beta for a short while, but my xap-livebox.ini became corrupt, so I reverted to standard release as I had to go away for a bit.
Clearly the edf xml is a mess, never seems to stop sending history, always seems to append the 'realtime' message to the end of a history message, etc
Probably worth pointing out, for the record
The edf device does support an optional, extra cost, whole house transmitter/clamp, you can use a Currentcost one.
The edf device only supports a single clamp on the whole house transmitter (3 on the Envi)
There is no on board temperature sensor on the eco manager
The edf IAMs (Individual Appliance Monitors) are different to the CC IAMs, are capable of switching on and off and do not use the proprietry currentcost C2 protocol at all
The eco manager does not support the use of any of the currentcost development boards, or transmitters or mini transmitters ( other than the whole house one)
There are 14 IAM channels on the eco manager compared to 9 on the CC Envi
To me functionality of these devices are sufficiently different to suggest that having interfaces for both on the HAH would be useful.
The similarities of the xml interfaces means this could be an easy "win".
For example I could see that it would be useful to have an EDF IAM on the TV's and media centres etc, so I could monitor the consumption and switch them off. On the other hand, I would like to be able to monitor the consumption of my fridge and freezer but would not want there to be any possibility of them being turned off inadvertantly.
Any volunteers for decoding the RF commands to the edf IAMs?
kevin9
23 February, 2012 - 00:00
#18
Its very cool.
I was blown away by the holy grail of energy management switching off sockets & measuring them as well. Having a cool interface like HAH with XAP flash (which I had already bought) as well was the icing.
£50 later and I have 6 EDF IAMs and 2 displays, gotta love eBay.
Mine seems to be working fine now, apart from the rather poor rf range. I think this weekend a tear down is called for and making the controllers arial a bit meatier. Derek & Co would you mind if I leant on you for advice?
Agree with Kevin multiple Current cost devices on HAH would be really useful. I have already allocated most of the 14 channels mentally. Would like to put the read only stuff on the CC128.
Agree would love to see a C2 TX, however from what I have read on a few blogs Currentcost don't seem that keen on people developing it publicly. Maybe worth running it past them first?
Thanks Kevin - that info was really useful, came to the same conclusion last night about needing measurement only.
The EDF IAM's I have are labelled as
Model=IAM
manufacturer=Current Cost ltd
serial numbers =IAM0076827a - 29a etc.
So not sure if its a lack of the EDF manager that stops them being switchable or the Current cost supplied ones are truly different?
Or is it only the MK1 IAMs that are read only? They did one with an external arial according to some of their press releases. They have announced a new on off controller so maybe the EDF ones are MKII IAMs and Current cost will be supplying those from now on? In which case I need to get on eBay/Amazon.
Ordered another transmitter tonight for the oven & hob. Now I won't just moan about the tumble dryer I can winge about the cooker as well! Should make me popular not much in the way of WAF.
I was originally thinking I would need to buy normal transmitters and create breakout plugs but if the current cost IAM's can be set measure only then the Freezer & fridge are getting those. I can take my 1 wire on/off sensors out or use them just for lights/ fans.
23 February, 2012 - 11:14
#19
Antenna boost
I found that moving from the '6.48" piece of wire' antenna to an external magmount unit made a big diffference. As much as anything else, it is a load easier to experiment with various antenna positions.
I currently use the one that is in the shop & can reliably control an RF socket in my garage (external to the house).
RF is such a fickle thing .. lots of variables to play with. Best to allow time to experiment.
Cheers,
Derek.
23 February, 2012 - 23:48
#20
EDF Ecomanager, EDF IAM's/RF Plugs CC IAMS
AlexS
As Derek says, RF range with module like these can be problematic and depends very much on your environment. Certainly adding an aerial to the head unit is a viable way of improving things. I have done much as Derek suggested with a CC128 Envi, the only difffernce is the type of plug I used:-
http://www.flickr.com/photos/netcompsys/sets/72157625058490399/
I have done a similar mod on my weather station with success
The original CC IAMs did have an 'aerial' as seen here
http://www.flickr.com/photos/netcompsys/sets/72157625168084766/
The IAMs were a long time announced so I made my own from Mini transmitters
http://www.flickr.com/photos/netcompsys/sets/72157624901424769/
and
http://www.flickr.com/photos/netcompsys/sets/72157624985430957/
but don't be fooled by the smart looking aerial in any of these devices, it is just the wire inside a plastic tube!
I would not advise modifying the IAM itself, for safety and insurance reaons
I still think there is a place for using mini transmitters in place of IAMS on those high current items like hobs and cookers. Not sure I want to trust these things to an in line device made in china even if they are rated to 16A
From my brief dialog with CC tech support
CC IAMS use the C2 protocol, are purchaseable from CC, are incapable of switching and are not compatible with the EDF Eco manager
EDF IAMS do not use the C2 protocol at all , are not purchasable from CC who made them for EDF, you buy them from EDF and they are not comptible with the Envi etc
The newer CC IAMs use the same molding as the EDF IAMs, that is all
The C2 protocol will only be released to partners under non disclosure agreements
Since the Ecomanager and its IAMs do not use C2 then it should be possible to decode the commands sent to switch the plugs. I have read somewhere that the RF module in the EcoManager is an RFM12B variant
Could you pass on the link to the announcement of the CC on/off controller?
HTH
kevint
24 February, 2012 - 01:33
#21
Thanks Kevin.So the CC IAMs
Thanks Kevin.
So the CC IAMs are non switchable. Thats where some of this months 'given up the smokes' cash goes! Freezer & Fridge.
announcement:
http://www.ceesquared.com/?p=118
http://www.ceesquared.com/?m=201201
and a picture, Looks actually quite stylish, might even get that into the lounge if SWMBO agrees.
http://www.currentcost.com/media-centre-downloads/catalogue.php
From the looks of that switchable IAM's are coming.
Agree no playing with the IAM. I have a Trec & transmitter coming for the hob - ebay again.
Kevin & Derek cheers for the Aerial modification, Look forward to a 'Franken ECO Manager' variant.
I found this today, RC switch coding for Kilk un Klak looks interesting.
http://code.google.com/p/rc-switch/wiki/KnowHow_LineCoding
and it sort of provides a raw output.
http://code.google.com/p/rc-switch/source/browse/trunk/examples/ReceiveDemo/ReceiveDemo.pde
On Ken Sheriffs Blog IR & Arduino.
http://www.arcfn.com/2009/08/multi-protocol-infrared-remote-library.html
One of his users came up with this:
right bed now, watched Rising Sun - again!
Thanks Alex
24 February, 2012 - 13:00
#22
"From the looks of that
"From the looks of that switchable IAM's are coming"
On past performance (IAM's and Optismart) it takes 12 months plus for CC to get from published artwork to available product, so don't hold your breath!
Of course you probably know the Trec has no RJ45, so no serial output, but also that you can sync transmitters/mini-transmitters/dev boards and IAMS (all of the CC variety as opposed to EDF) to multiple receivers at the same time. After a bit it can get confusing if you don't keep track of what is where though unless you write it down.
kevint
24 February, 2012 - 13:04
#23
Some more
Pictures of shiny new EMC/c2 Touch
http://hotzone.org.uk/edf-ecomanager/
may not materialise
and the Aerial, looks like its sane as envi
http://www.flickr.com/photos/robomotic/6413419547/in/set-72157628179034525/
and details on protocol (not much)
http://forum.jeelabs.net/node/661
http://www.silly-science.co.uk/2010/03/04/current-cost-envi-cc128-teardown/
24 February, 2012 - 13:13
#24
Ah yes the more I read, the
Ah yes the more I read, the more I realise they are a marketing led company.
Now if only they did a Bridge with XAP or similar. Keeps the C2 protocol private but gives them a way to keep us happy.
Maybe we should ask?
The Trec has the option of daisy chain CT's I can use it to sense Hob/oven.
The display will be popped in the Hallway on the wall and show top 9 offenders.
So it ticks 2 boxes! For less than 1 transmitter.
24 February, 2012 - 14:18
#25
"Pictures of shiny new EMC/c2
"Pictures of shiny new EMC/c2 Touch
http://hotzone.org.uk/edf-ecomanager/"
Of course, the top picture is of the Eco Manager, then the EDF IAM
the EMC/c2 device circled was only in public in limited numbers and was withdrawn, at about the same time as the early edf IAMs were recalled after a few went up in smoke. Stuarts post is dated September 2010
BTW the xAP Bridge for currentcost is otherwise called the HAH lIvebox(!) no point asking about opening up the C2 protocol, plenty have asked but are refused or just ignored
On the daisy chain clamps to a single transmitter, they show up as one value on the display but are available as 3 vaules in the xml. When the transmitter is linked to an Envi channel it also reports 3 values in the xml but i don't know if the HAH parser supports this confiuration
If you check out my Flickr Photostream you will see a series of Photo Sets covering CC device unboxing, teardowns, comparisons of different devices, mods (handheld battery aerial), devboards, IAMs, Pseudo IAM's, appliance monitoring, etc. Ecomanager photos will probably be next
http://www.flickr.com/photos/netcompsys/sets/ and
http://www.flickr.com/photos/netcompsys/sets/?&page=2
kevint
18 September, 2012 - 21:05
#26
Have you seen this ? EDF ECO receive code..
Cheers Brett
Ecomanager manager seems to work.Now have the tumble dryer on an IAM.
£25 to monitor the Washing machine/ tumble dryer, Home theatre & my PCs. Cool.
Now if people stopped bidding the ECO managers up on eBAY I can drive the wife & family completely mad.
Have you seen this?
http://www.homeofthefuture.tv/
some interesting stuff and EON running a competition, maybe a way to get some publicity?