Topic: Tower sensor addresses, Aculink bridge max sensors?

[tandy1000]

I've begun the process of capturing the Aculink bridge data, and noticed a couple of interesting things. I have a 5-in-1 and 2 Tower sensors.

For the moment I'm sending the data to a network monitoring system, so I had concerns about distinguishing the 2 sensors. (Oregon Scientific sensors for example, generate a random address upon boot - but some have a channel selector, and that channel is part of the data packet)

However I've noticed the Tower sensors always send the same address (meaning each sensor has its own unique address), even if I remove/reinsert the batteries, or change the A/B/C switch. Does anyone know if it's hardwired? Or maybe I'm reinserting the batteries too quickly, and it has a capacitor to retain the address temporarily?

Secondly, Acurite indicates that the Internet bridge is limited to 3 sensors. I'm wondering how this limit is enforced, as I randomly picked up my neighbor's 3-in-1 sensor a few times. (My 3 sensors consist of the Two Towers  and the 5-in-1)

Has anyone actually tested the 3-sensor limit?

[nincehelser]

I think the tower sensor IDs are hard-wired.  I haven't been able to reliably detect any change by using the ABC switch, but I haven't tested this enough to say for sure.

From what I've been able to tell, the bridge just forwards what it hears.  Any "enforcement" seems like it would have to occur at the MBW website, which makes it easy to ignore.

The unanswered question is "why three?"  Is that just an arbitrary limit, or a practical limit? 

One concern I have is if there is enough information to reliably tell sensors apart.  For example, the tower sensor ID just doesn't seem long enough for it to be unique across all tower sensors. 

Then there's the issue of telling sensors apart by sight...the tower sensors seem to all look alike visually, so how is the user going to know which sensor is which?  Other than the ABC switch setting, how can you visually tell them apart?  I haven't noticed a unique serial number or anything like that, but it's kind of hard to tell when you only personally have 1.

So it looks like the "three" limit can be exceeded (at least locally) if you can just figure out a way to uniquely identify each sensor and keep it straight which is which (both in a program and in physical management).

[tandy1000]

Indeed, there is no physical distinction between the tower sensors. I even opened them up to see if there was any unique identifier - nothing. As far as distinguishing, I think the only thing I can do is put a physical label, so I know that "Tower 08936" is actually in my living room.

As I haven't looked into customization of weewx or any other weather software, I don't know how easy it would be to manage a lot of sensors. Since I'm just capturing/graphing the data in my monitoring software (Zabbix) it's not an issue since I create a unique object for each sensor.

So as far as the ABC switch, could it be a subtle frequency change to align with your console, and not necessarily part of the packet? Maybe I'll try to move my bridge to see if I can consistently pick up my neighbor's 3-in-1 and all of my sensors..

[vreihen]

I believe that the only reason for the three sensor limit in MBW is to limit the bandwidth consumed by each user of their free service.  (Same reason used for only having 15 minute updates to Weather Underground.)  Ryan from AcuRite probably has dozens of sensors on his personal bridge, reading between the lines in some of his public posts..... 

[nincehelser]

The ABC switch doesn't change the frequency.  It does something in the data packet.  Or at least that's what Acurite has confirmed before. 

I haven't seen this reflected in the bridge output, but then I again I haven't really looked.  I suppose it might show up as a change in the sensor ID number.

[SamH]

Today, I tried four sensors (2ea 5-in-1s, 3-in-1, and Tower)  with the Internet Bridge.  It looks like the Internet Bridge sees the four sensors, but MBW will only allow you to use three.
SamH

[tandy1000]

I'm also reliably capturing a 3-in-1, a 5-in-1, and 2 tower sensors. I just ordered a few more tower sensors, for a really good test. Anyone ever peruse the FCC testing documents? Here's the search link, where you can look for Chaney in the Applicant Name field:

https://apps.fcc.gov/oetcf/eas/reports/GenericSearch.cfm

And here's the 5-in-1:

https://apps.fcc.gov/oetcf/eas/reports/ViewExhibitReport.cfm?mode=Exhibits&RequestTimeout=500&calledFromFrame=N&application_id=280903&fcc_id=RNEVN1TX

At least you can see the inside of the 5N1 without taking it apart!

[tandy1000]

Well I can confirm I'm receiving data from a 3N1, a 5N1, and five Tower sensors.    So far they have unique addresses. I labeled each one on the outside with its sensor address, and in my monitoring system I give it a friendly name, like Living Room.

I'm also tracking packet loss, so I'll know if something drops off.

[tongsli]

Tandy, I just ordered the aculink bridge and 6 sensors.

[tandy1000]

Nice, tongsli! You can't go wrong with the Acurite price. I've moved my bridge to the middle of my house, the signal strength improved somewhat. I only seem to lose packets from the neighbor's 3N1 on occasion, which I can't complain about too much. I may try a few more spots in the house.

[nincehelser]

Well I can confirm I'm receiving data from a 3N1, a 5N1, and five Tower sensors.    So far they have unique addresses. I labeled each one on the outside with its sensor address, and in my monitoring system I give it a friendly name, like Living Room.

I'm also tracking packet loss, so I'll know if something drops off.

Can you verify what happens if you change the A-B-C switch?  I'm wondering if that changes the sensor ID or not.

Thanks!

[tandy1000]

I haven't been able to determine what the ABC switch changes. The sensor ID always stays the same, regardless of the position. In one aspect I suppose that's good - my TV room tower is always sensor ID 10201. So as long as it's labeled and my monitoring system maps 10201 to "TV Room" - it will be a constant.

I'm wondering how the sensor IDs are distributed. Did I get lucky by having 5 unique sensors? Or will they simply go beyond 99,999? So far my towers are 01359, 08936, 09088, 10201, and 14167.

[tongsli]

Thanks Tandy,

I'm trying to use the sensor to determine how efficient(or not) my house heating system and insulation is performing.  I know from observation that it's warmer upstairs at night, but I want to log and record the temperature delta. If I can set the thermostat even lower at night and still be comfortable upstairs then I'll do it.

Since I don't have them yet I don't fully understand the sensor ID's and how to configure them to the bridge.  But hopefully it'll show up this week.

[tandy1000]

Hey Tongsli, I'm going to do something similar. My old house only has one heating zone, and window ACs. I want to start calculating how long it will take to heat/cool a particular room, given an outside temperature. The upstairs rooms get blazing hot in the summer, so I have to start the ACs well in advance.

As far as the sensors go, note that My Backyard Weather will still only show 3 sensors, you will need to implement something custom to get all of that data. What's your computer setup/expertise at home?

I'm using a combination of things that seems to evolve as I poke away.

1. Zabbix to collect and graph all of the data. http://zabbix.com  I decided that weewx wasn't really for me. While Zabbix isnt' particularly "pretty" its powerful and easily accepts arbitrary data you define. Eventually I'm hoping to use http://highcharts.com to make things look nice.
2. A couple of things from EAS that he did:http://geekfun.com/hackulink and modified for my setup. Initially I started with a bridged Linux adapter that Nincehelser http://nincehelser.com/ipwx/ defined for the ncat/grep commands. However now I'm running those on my router with Tomato firmware. That way the Acurite can be on any Ethernet port in the house.
3. A python script (based on EAS) that captures the data and sends it all to Zabbix.
4. Separate from the Acurite, I have an Arduino that's capturing data from some Oregon Scientific sensors, and also sending it to Zabbix.

There are other solutions out there, such as Kevin Key's app http://kevin-key.blogspot.com/ so I think there are plenty of options!

[tongsli]

Tandy, OMG.  Well.  I have some reading to do.  I've got a NEST thermostat version 1 which is controlling our 1 zone hot water heat.  We have separate split ductless AC units down stairs which have their own remote controls that act as thermostats.

Upstairs I will soon deploy NEST thermostat version 2 which supports two stage cooling.  We had a conventional AC compressor and air handler installed in the attic.  I can monitor my NEST from my iPhone since it's WIFI connected.

Right now, I have a few acurite weather forecasters located in the kitchen and our bedroom.  I have a bunch of accurate 00380's that I can move from place to place.    I want to place one of the new sensors in a few places: attic (non-conditioned space), 2nd floor and basement.  I'd like to log the temperature in the basement and see if I can find a correlation when the furnace is running to the effective temperature on the 1st and 2nd floor.

I already know there is a 4-6 degree delta between the 1st floor and 2nd floor. And, even between our bedroom and the hallway upstairs.  Based upon this, I've already modified our thermostat program schedule.

I'd like to be able to see what the temps are when I am away from the house.