Temp Sensor Hooked to Zone 1-16 ELK M1

[ddennerline]

I agree a wireless (GE Crystal or Z-Wave) temperature and humidity sensor would be nice especially if device could be flush mounted. In addition, if ElkRP was updated to support humidity then I could start working on trying to keeping inside RH at optimum balance.

The only concern would be batteries and freezing tempatures. Batteries tend not to perform optimally at extreme ends of the weather spectrum. I have been using Oregon Scientific MTR101. It’s extremely flaky – especially when it’s cold outside. When weather drops to freezing, I have to move the receiver to inside window that is closest to outside tempature sensor to force status to be updated.

I suspect the problem is related to batteries and age because if I swap out all the batteries, then problem goes away for a short period of time.

I do not have same receiver/transmitter problem with GE wireless door/window sensors though, so it’s probably specific to Oregon Scientific wireless protocol.

 

[fleetz]

A wireless Temp. sensor would be great!, especially for a retrofit situation. One of the big limitations now (besides the appearance) is that they have to be directly connected to zones 1-16 on the controller.

I think it would be fair to say that there would be a very good number of wireless temp sensors retro fitted should they become available. I believe a lot of folks would not have gone the the M1TZS because of its looks. I wired and waited and waited (nearly 2 years) until I really needed to get temps into the M1 then I modified them so they looked presentable.

As for the limitations of having to connect to 1-16 on the controller given the cost of an input expander shifting the zone inputs to an input expander is a very reasonable cost to pay to have the flexibilty of good looking wireless TZS.

My 2 cents....sorry dime!

Fleetz

 

[Lagerhead]

Would there be any interest in a wireless temperature sensor that would work into the ELK M1XRF RF receiver?

Unless wireless is imperative (can't get a wire to the location), reliability and price would be key considerations for me.

The M1ZTS makes sense in critical control situations (e.g. freezing, thawing, heat control, ventilation) where the M1 itself is involved in the control loop or must raise an alarm.

But I do not use Elk temperature sensors in "informational" situations (where I just want to know the temperature, or log a record over time) because there are alternatives that are far less expensive, like thermistors and 1-Wire sensors (which of course attach to other automation controllers and not to the M1 [with the M1KAM being the one exception]).

So the attractiveness of a wireless temperature sensor for the M1 depends on its price, the application, on what other choices you have, and -- not incidentally -- on whether you already have a wireless receiver in the M1 or would have to add it.

The only concern would be batteries and freezing tempatures. . .

Here are the reliability considerations -- the suitability of battery-powered temperature sensing in the application, and significantly, both remembering to change the batteries, and having the ability and opportunity to do so.

Dave

 

[sic0048]

Fleetz - would you mind giving a "how to" on your modifications? It looks great.

Thanks,

 

[TXFlatLander]

Fleetz - would you mind giving a "how to" on your modifications? It looks great.

I agree with sic0048. A how-to would be awesome given how great they look! Thanks

 

[signal15]

IMO, the best solution would be a true one-wire interface that supported all of the one-wire sensors. Someone said the M1KAM only supported a single sensor. I need to monitor the temp in the following locations:
- 3 freezers
- 1 storage fridge
- Possibly the kitchen fridge
- Wine fridge
- beer conditioning fridge (brewery)
- Attic
- Garage
- Outside temp

That's 10 sensors. If they cut into the max number of keypads I can have, well, that's a problem. Because I have 4 keypads right now, and I am going to add two M1KAM's for access control at two locations that are far from my existing keypads. So I'd be using all 16 possible keypad designations.

It seems to me that one of the major obstacles in adding functionality is the limited memory in the M1G panel. If a true one-wire interface were developed, couldn't it store all of the values for the sensors, and then the panel could just have code that polled that one-wire interface and walk the values on a periodic basis? Maybe I'm wrong, but it looks like there is limited room to store dynamic values in the device.

 

[fleetz]

Fleetz - would you mind giving a "how to" on your modifications? It looks great.

Thanks,

No problems. It is easily done.

I have chosen a Clipsal (Square D) mounting plate which matches my Clipsal C-Bus keypads. Any plate that takes a RJ45 will do the job so you can match what you have in your homes.

I removed the RJ45 connector on the M1TZS. Then I wire the new RJ45 in the whizzo plate of your choice with the following convention.

EP- > to pin 5 of the RJ45
EPD > to pin 6 of the RJ45
EP+ > to pin 7 of the RJ45

I used about 20cm (8") of Cat 5 cable so I could punch down the RJ45 connections

Then simply wire the M1TZS per the ELK instructions back to the M1.

I drilled a hole where the RJ45 was on the M1TZS to allow thw cable to pass thru for the connection to new RJ45. this allow the lid to clip back on. I then passed the M1TZS into the wall cavity which happens to sort out the blinking led and then fit the new plate in postion.

Plug the RJ45 temp sensor into the new RJ45 plate, set the M1 zone up to temperature sensor and there you have it.

In my case the cost of the bits worked out about the same as the M1TZS as the glass style plate are exy but they look as though they are meant to be there.

Hopefully it all makes sense if there are any questions please ask.

Regards,

Fleetz







The control unit

 

[robolo]

Spanky,
Could you have a wireless and a wired solution? My home is large and old with THICK plaster walls that make wireless almost worthless. I would really like a wired option. And I agree with signal15 that a true one-wired solution would be the optimal new add-on. I remember being very excited when the M1KAM was introduced thinking that would solve the 1-wire vacuum but, alas, not to be.

 

[me2]

Would there be any interest in a wireless temperature sensor that would work into the ELK M1XRF RF receiver?

You could have up to (16 minus the number of keypads) temperature sensors.

Yes!

 

[JayH]

Short answer: Yes, I'm interested in RF but there are some other options & issues.

I have 3 M1ZTS's installed, 2 attics and outside, but none inside because of accuracy and stability:

1) Accuracy. Outside it ranges from ~15 to ~108 here in Dallas but they only appear close in the 70-95 range. We were at 21 last week but I never read below 34 on the outside sensor (but the transmitter next to it for a wireless weather unit did). The attics are similar. I've heard the remote is much better but never tried it.

2) Stability. From monitoring the ST messages the sensor appears to be "raw" with no ballistics or averaging built in to keep it from bouncing. There are many circumstances where it will bounce between 2 temperatures (ie, 74/75) for several minutes before stabilizing. I would like to be able to control this and know how it handles fractions/rounding.

Some other options:

1) The tr/TR (Thermostat) can have it's temperature SET via a message (newer firmware) so you can completely virtualize a thermostat and this information is available within the rules, so there is no reason unused "thermostats" couldn't have values plugged in from an external source if you already have temperature sensors. Note that when you feed it via the "tr" update there is no automatic ST generated for a temperature change (bug?) so other external systems won't see it. My RCS stats are handled this way so I can get to the rest of the details that are stripped/not available via Elk.

2) Could the other temperature sensors (keypad, zone) be opened to setting via a command (st or lw?) (and have the corresponding ST sent) for virtual temperature sensors? In the case of zones, still allow an independent temperature even if the zone is in use and not a temp zone.

Spanky - Glad to hear you are back in Engineering!

My 2 cents.

Jay

 

[ddennerline]

I have not yet purchased a M1TZ, but if previous post is true and accuracy is questionable, then I would address this feature first. Whatever form or function, a thermometer must be accurate. I burned an expensive roast this holiday due to defective “backyard griller†digital (~$32) thermometer. I tested my cheaper backup thermometer using http://www.allfoodbusiness.com/calibrating_thermometers.php; it failed too. I sort of implicitly believed that thermometers are always accurate.

I was planning on installing four temperature sensors (downstairs, upstairs, outside, and attic). At $220, this is not a top priority.

In terms of wiring, wireless would be nice only if accuracy/reliability was not compromised. Having to run wires is not a deal breaker because the sensors will be positioned in walls that are accessible through basement or attic.

 

[Spanky]

How to use Zone Temperature sensor in a Rule:

Using the M1ZTS or a M1KP temperature sensor define the zone as a temperature zone.

Write a Rule using a timed event -
WHENEVER EVERY 30 SECONDS
AND ZONE 9 TEMPERATURE IS GREATER THAN 68 DEGREES F
THEN do something


Because temperature and voltage levels are analog levels and do not have event trip points, a timed event and the AND statement will evaluate the Rule for firing.

 

[Lagerhead]

[M1ZTS Accuracy] . . . We were at 21 last week but I never read below 34 on the outside sensor (but the transmitter next to it for a wireless weather unit did). The attics are similar. . .

Are you sure you are at the correct firmware level for your M1 (per the M1ZTS manual)?

This is a rather serious indictment of the M1ZTS if true. While ELK makes no accuracy claim that I can find, I think anyone would expect it to be within a degree or two across the specified temperature range.

I have only one of these but have never noticed any readings that are out of line.

[M1ZTS Stability] . . . There are many circumstances where it will bounce between 2 temperatures (ie, 74/75) for several minutes before stabilizing. . .

I have not noticed this either. Bobble has not been an issue.

 

[signal15]

I have two of the ELK sensors. One was off by several degrees. On the unit, you can calibrate it by shorting the ground pad to either the + or - pad with a flathead screwdriver or other metal object. It will increment one degree each time you short it.

What would really rock? ELK support for the RFXCom device.

 

[Monk]

[M1ZTS Stability] . . . There are many circumstances where it will bounce between 2 temperatures (ie, 74/75) for several minutes before stabilizing. . .

I have not noticed this either. Bobble has not been an issue.

I think Spanky's post was speaking to the possibility that JayH's setup was being incorrectly monitored - - Unless I got it wrong...
Anyhow -
I have 2 external Elk sensors and they are always close to what I think they should be.. No bobble here either.