Ive had my Tx connected to the pulse output on my water meter for a couple of days now.
Data is transmitted from the meter cupboard in the kitchen, to where my monitoring PC and Nanode is, about 20M diagonally through a couple of floors and several walls. Radio reception works well.
However, having been successfully logging to EmonCMS for that time, Ive noticed some 'spurious' pulses. Regularly, I get 1-2 pulses, when no water is being used. I have confirmed there are no leaks, and a commercial logger also attached, does not record them.
Any thoughts? Glyn - how accurate/susceptible is the pulse monitoring?
Atatched CSV export of the SQL DB (its 15000 lines, but small in size).
EDIT:Just seen I cant attach a file, so a cut n' paste is at the bottom.
If you look between 01:00 - 07:00, when no one is awake, so you can the examples.
| 01/10/2011 01:26 | 2 |
| 01/10/2011 01:26 | 0 |
| 01/10/2011 01:26 | 0 |
| 01/10/2011 01:26 | 0 |
| 01/10/2011 01:26 | 0 |
| 01/10/2011 01:27 | 0 |
| 01/10/2011 01:27 | 0 |
| 01/10/2011 01:27 | 0 |
| 01/10/2011 01:27 | 0 |
| 01/10/2011 01:27 | 0 |
| 01/10/2011 01:27 | 2 |
| 01/10/2011 01:27 | 0 |
| 01/10/2011 01:27 | 0 |
| 01/10/2011 01:27 | 0 |
| 01/10/2011 01:27 | 0 |
| 01/10/2011 01:27 | 0 |
| 01/10/2011 01:27 | 0 |
| 01/10/2011 01:28 | 0 |
| 01/10/2011 01:28 | 0 |
| 01/10/2011 01:28 | 0 |
| 01/10/2011 01:28 | 0 |
| 01/10/2011 01:28 | 0 |
| 01/10/2011 01:28 | 0 |
| 01/10/2011 01:28 | 0 |
| 01/10/2011 01:28 | 0 |
| 01/10/2011 01:28 | 0 |
| 01/10/2011 01:28 | 0 |
| 01/10/2011 01:28 | 0 |
| 01/10/2011 01:28 | 0 |
| 01/10/2011 01:29 | 0 |
| 01/10/2011 01:29 | 0 |
| 01/10/2011 01:29 | 0 |
| 01/10/2011 01:29 | 0 |
| 01/10/2011 01:29 | 0 |
| 01/10/2011 01:29 | 0 |
| 01/10/2011 01:29 | 0 |
| 01/10/2011 01:29 | 0 |
| 01/10/2011 01:29 | 0 |
| 01/10/2011 01:29 | 0 |
| 01/10/2011 01:29 | 0 |
| 01/10/2011 01:29 | 0 |
| 01/10/2011 01:30 | 0 |
| 01/10/2011 01:30 | 0 |
| 01/10/2011 01:30 | 0 |
| 01/10/2011 01:30 | 2 |
| 01/10/2011 01:30 | 0 |
| 01/10/2011 01:30 | 0 |
| 01/10/2011 01:30 | 0 |
| 01/10/2011 01:30 | 0 |
| 01/10/2011 01:30 | 0 |
| 01/10/2011 01:30 | 0 |
| 01/10/2011 01:30 | 0 |
| 01/10/2011 01:30 | 0 |
| 01/10/2011 01:31 | 0 |
| 01/10/2011 01:31 | 0 |
| 01/10/2011 01:31 | 0 |
| 01/10/2011 01:31 | 0 |
| 01/10/2011 01:31 | 0 |
| 01/10/2011 01:31 | 0 |
| 01/10/2011 01:31 | 0 |
| 01/10/2011 01:31 | 0 |
| 01/10/2011 01:31 | 0 |
| 01/10/2011 01:31 | 0 |
| 01/10/2011 01:31 | 0 |
| 01/10/2011 01:31 | 0 |
| 01/10/2011 01:32 | 0 |
| 01/10/2011 01:32 | 0 |
| 01/10/2011 01:32 | 0 |
| 01/10/2011 01:32 | 0 |
| 01/10/2011 01:32 | 0 |
| 01/10/2011 01:32 | 2 |
| 01/10/2011 01:32 | 0 |
| 01/10/2011 01:32 | 0 |
Re: Spurious Pulse generation on TX
Hi Andy,
What method are you using to interface the emonTx with your meter. Are you using an optical sensor or wired? If your using an optical sensor it could be another light source is effecting it.
I'm yet to run any long term tests using the emonTx as a pulse counter. I don't have easy access to a pulse-output utility meter so I'm not too sure how reliable it is. Your experiences will be very valuable to the project.
Good luck, Glyn.
Re: Spurious Pulse generation on TX
The Tx is connected to a wired water meter (purchased specifically for monitoring, as the fitted meter only seems to work with the utility companys touchpad). The meter pulses every 1 litre, and Tx seems to be quite reliable in detecting and transmitting.
I originally ran the logging with the default firmware interval of 10 seconds, and thats where i saw the spurious pulses. I changed the Tx firmware to collect every 60 seconds, and the Base radio timeout check to 90 seconds. With this configuration, it looks like the extra pulses arent recorded.....could the radio transmission generate a spike on the Tx which is triggering a pulse detection?
Re: Spurious Pulse generation on TX
Mmm thats strange. I guess the only way to find out would be to put on scope on the interrupt pin while the RF was being transmitted. Have you tried running the emonTx (sending RF's) without your utility meter connected? This would determine if the spurious pulses were coming from the RF or the utility meter.