DataMod integration with ELK M1

robolo

Active Member
Is anyone familiar with this new product for temperature and humidity sensors? This was posted on the Charmed Quark web site. I was wondering if this item could be integrated with the ELK M1. It would be an alternative to the 1-wire integration many of us have been waiting for. Plus I really like the unobtrusive tempurature probes. High WAF. Spanky, any thoughts about this?

"DataMod Series I/O Modules and Temperature Sensors. These Modules can also be used for Contact Closures, Humidity, CO2, Pressure, Flow, Voltage, Current... The first module to be supported by the driver is the MD01 which has 32 Analog Inputs. Other modules include the MD02 providing 8 Analog Inputs and 8 Analog Outputs rated 0 to 10 volts and the MD03 providing 8 Analog High Speed Pulse Inputs and 13 Relay 1Amp @ 24VAC Outputs.

These products will hopefully provide users with an economical way of monitoring and or controlling of Temperature, Humidity, CO2, Pressure, Flow, Voltage, Current, Dry Contacts, Dampers, Valves VFDs, Pumps and the list goes on.

More information on the Modules and Temperature Sensors can be found at:
www.kanak.com/datamod.asp and www.kanak.com/tempsensors.asp"
 
I am sure that if Elk did some programming on their end it could but right out of the box, No.
Kirk,
I'm glad you posted here.

I figured it would need some programming on Elk's end but maybe it would be easily deployed through the Elk M1SXP?
 
The M1XSP is capable of RS485 so that would not be an issue. I know Elk has a lot on their plate so it just may be a matter of time.
 
Is anyone familiar with this new product for temperature and humidity sensors? This was posted on the Charmed Quark web site. I was wondering if this item could be integrated with the ELK M1. It would be an alternative to the 1-wire integration many of us have been waiting for. Plus I really like the unobtrusive tempurature probes. High WAF. Spanky, any thoughts about this?

"DataMod Series I/O Modules and Temperature Sensors. These Modules can also be used for Contact Closures, Humidity, CO2, Pressure, Flow, Voltage, Current... The first module to be supported by the driver is the MD01 which has 32 Analog Inputs. Other modules include the MD02 providing 8 Analog Inputs and 8 Analog Outputs rated 0 to 10 volts and the MD03 providing 8 Analog High Speed Pulse Inputs and 13 Relay 1Amp @ 24VAC Outputs.
These products will hopefully provide users with an economical way of monitoring and or controlling of Temperature, Humidity, CO2, Pressure, Flow, Voltage, Current, Dry Contacts, Dampers, Valves VFDs, Pumps and the list goes on.

More information on the Modules and Temperature Sensors can be found at:
www.kanak.com/datamod.asp and www.kanak.com/tempsensors.asp"

The breadth of the services offered by Kanak through the INTERSCIENCE INTERNATIONAL web site reminds me of Dr Bonner's original soap label (Wars quelled, Revolutions started ... ;-)

I also have several more serious observations:

1) Perhaps the most important step forward would be if Dean/CQC did indeed support a Modbus-over-RSxxx interface. To the best of my knowledge, this would be the first readily available HA application of that venerable and highly appropriate industry standard. Visit www.modbus.org/

2) The modules, sensors and acccessories themselves all appear to be rebranded/debranded items from other sources of the sort that have been around for decades. I see nothing "new". The sensors themselves are generic and not specific to the input modules. In the realm of scientific/industrial sensors, $15 for a thermistor assembly may indeed be a good price . But what is offered apears to be a $1-$2 thermistor attached to a $2 stainless steel plate.

3) The 10-bit analog input resolution of the analog-to digital converters cited is a modest step up from the 8-bit resolution of existing HA devices such as Elk, Ocelot &etc. but is no better than that built into a $1.50 AVR microcontroller. This is not a Modbus limitation, but of the particular modules offered. For many/most purposes 10-bit resolution and accuracy is sufficent, but folks wanting to make full simultaneous use of the accuracy, range, and resolution available from (eg) the thermistors they tout will/may be disappointed. As may folks wanting to (eg) make light measurements that begins to approach the actual range of environmentally significant illumination levels (eg 0.1 to 10,000 lux).

4) The modest ADC resolution helps explain the relatively low price. IC's with sixteen analog inputs with 24-bit analog resolution and eight DIO wholesale for $8.00 http://focus.ti.com/docs/prod/folders/print/ads1258.html

5) Assuming for the moment that Elk did offer a way to communicate with Modbus over RS-xxx, the M1G might be challenged with respect to providing calibration and conversion from a voltage to engineering units. Unlinearized themistors, for example, are typically converted to temperature using equations that may be outside ithe panels' capabilites ( eg Steinhart-Hart uses natural logs).

6) There are reasons why HA "panels" historically have not provided analog inputs and some of the reasons revolve around the facts that 1) "panels" typically have no/negligible data storage capabilites so most anything other than a real-time reaction to the data is not practical/posssible) and 2) it greatly simplifies the manufacturers task if the analog world is converter to an 'Event" with binary outcome (ON/OFF; TooHot/NotTooHot; etc) before the data arrives at the panel.

7) It follows (at least in my mind) that making full use of analog inputs requires addressing points 1) and 2) in the paragraph above. That would be the most time-consuming part of building a Modbus 'interface' for CQC, not downloading the measured voltages from the devices.

8) Compared to other 'dumb' devices that have been available for DIY HA, these are relatively expensive ( but not when compared to scientific/industrial devices). For example, the LCD+ from Netmedia has been around for 7-8 years, provides eight 10-bit analog inputs, eight DIO with built-in relay driver IC, a small multi-line ASCII LCD display, and a 16-input keypad interface -- all for $60. And a USB data acquisition module with eight *12-bit* analog inputs, 16 digital I/O lines, two programmable
pulse-width output channels, and one 16-bit counter channel cost costs ~$100 shipped (Seach eBay for " iUSBDAQ")

Reading what I just wrote, the remarks seem to be more negative than intended ...

Having a Modbus interface, sensor calibration and conversion front-end, and historical data storage and display would be a dynamite addition to CQC !

... Marc
Marc_F_Hult
www.ECOntrol.org
 
2) The modules, sensors and acccessories themselves all appear to be rebranded/debranded items from other sources of the sort that have been around for decades. I see nothing "new". The sensors themselves are generic and not specific to the input modules. In the realm of scientific/industrial sensors, $15 for a thermistor assembly may indeed be a good price . But what is offered apears to be a $1-$2 thermistor attached to a $2 stainless steel plate.
Of course anyone can buy a $1-$2 thermistor and tape it to a steel plate. But the solution will not be a quality one, and it will cost you much more in time than it’s worth unless you are making a few hundred of them at a time. The number & variety of styles available for a very competitive price (especially the flush mount version). These are high quality products and are made by the thousands… Not something that you patched together in your basement

3) The 10-bit analog input resolution of the analog-to digital converters cited is a modest step up from the 8-bit resolution of existing HA devices such as Elk, Ocelot &etc. but is no better than that built into a $1.50 AVR microcontroller. This is not a Modbus limitation, but of the particular modules offered. For many/most purposes 10-bit resolution and accuracy is sufficent, but folks wanting to make full simultaneous use of the accuracy, range, and resolution available from (eg) the thermistors they tout will/may be disappointed. As may folks wanting to (eg) make light measurements that begins to approach the actual range of environmentally significant illumination levels (eg 0.1 to 10,000 lux).

4) The modest ADC resolution helps explain the relatively low price. IC's with sixteen analog inputs with 24-bit analog resolution and eight DIO wholesale for $8.00 http://focus.ti.com/docs/prod/folders/print/ads1258.html
Both the AVR & TI solutions are typically bought in quantities of 1000 or more to get this price. How many end users are going to do that, design, assemble, & solder the PCB, package it nicely with terminal blocks and a case that looks good, add the 485 driver & modbus protocol, put on the cute LEDs, power supply etc, test it through 5-10 life cycles, and then feel comfortable using it to control their home?
The solution for temperature monitoring using the DM01 & the 10K thermistor sensors is actually quite accurate by space temp and outdoor air temp standards. Each increment will be about 0.15 DegF (a 12-bit 1-wire sensor measures 0.1125 DegF per increment)

5) Assuming for the moment that Elk did offer a way to communicate with Modbus over RS-xxx, the M1G might be challenged with respect to providing calibration and conversion from a voltage to engineering units. Unlinearized themistors, for example, are typically converted to temperature using equations that may be outside ithe panels' capabilites ( eg Steinhart-Hart uses natural logs).
In microprocessor applications we are able to achieve around +/- 0.5F (better than 1-wire) by simply using lookup tables. The thermistor element that we are using is actually quite linear between 55F & 100F anyway, but when you use lookup tables, you can achieve very high accuracies with a small amount of code. This can be done without any messy equations and even with using a cpu with FPU capabilities.

6) There are reasons why HA "panels" historically have not provided analog inputs and some of the reasons revolve around the facts that 1) "panels" typically have no/negligible data storage capabilites so most anything other than a real-time reaction to the data is not practical/posssible) and 2) it greatly simplifies the manufacturers task if the analog world is converter to an 'Event" with binary outcome (ON/OFF; TooHot/NotTooHot; etc) before the data arrives at the panel.
This is why I am offering this as a solution to CQC. Why not allow the HA ‘panel’ then to view the real data and actually do some things with it rather than reacting to simple and dumb ‘events’ that it needs to trust from a ultra cheap 60’s era reed switch? We have all the processor power in the world to play with now. No need to offload the intelligence and limit the amount of information we can use. We are in the information age!

7) It follows (at least in my mind) that making full use of analog inputs requires addressing points 1) and 2) in the paragraph above. That would be the most time-consuming part of building a Modbus 'interface' for CQC, not downloading the measured voltages from the devices.
It is simple to convert the 10-bit value into the corresponding engineering units with less than 15 lines of code. You need a scale, offset, & min in order to do this. Not difficult at all.

8) Compared to other 'dumb' devices that have been available for DIY HA, these are relatively expensive ( but not when compared to scientific/industrial devices). For example, the LCD+ from Netmedia has been around for 7-8 years, provides eight 10-bit analog inputs, eight DIO with built-in relay driver IC, a small multi-line ASCII LCD display, and a 16-input keypad interface -- all for $60. And a USB data acquisition module with eight *12-bit* analog inputs, 16 digital I/O lines, two programmable
pulse-width output channels, and one 16-bit counter channel cost costs ~$100 shipped (Seach eBay for " iUSBDAQ")
Comparing this to the LCD+ is like comparing Apples to Oranges. This solution is a finished product ready for installation where as the LCD+ still requires time to finsh. Also, USB controllers are very limited by the distance of the network and by the number of nodes you can drop off of the network. Not a good solution unless you want to have all of your wires coming back to your computer.
 
Kirk,

Do you think that system would work with the PC-12 sensor? It is also a 10k thermistor and it looks like the temperature chart is about the same. I am still looking for a way to use the PC-12 sensors. I wanted to bring them into the Elk but I don't think that is happening. I can bring them into CQC and would be happy with that, although not too happy to have to buy another $200 controller to do that. Are there any other options to use a 10K thermistor with Elk or CQC?
 
Kirk,

Do you think that system would work with the PC-12 sensor? It is also a 10k thermistor and it looks like the temperature chart is about the same. I am still looking for a way to use the PC-12 sensors. I wanted to bring them into the Elk but I don't think that is happening. I can bring them into CQC and would be happy with that, although not too happy to have to buy another $200 controller to do that. Are there any other options to use a 10K thermistor with Elk or CQC?
Yes it would work. The Curve Chart would have to be added to the CQC driver to support them but I see do reason why it would not work.
 
The latest M1KAM update allows for use of one (1), 1-wire temperature or humidity module. Hobby boards has these available.

The current temperature module and probe that connects to the M1's zone input uses a very slow manchester encoded protocol. If any one wants to make their own temperature probe, the protocol is quite simple.
 
Back
Top