Santa was nice enough to bring me a CAI WebControl for Christmas. Now I need to get some sensors. The lot of ten DS18B20 sensors from the Chinese sellers on eBay is pretty cheap, but I am worried about quality. It seems that the only other direct option would cost about $5 per sensor. What are people's experiences with the quality and accuracy of these from eBay? Is there is a particular seller that is known to come through?
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
DS18B20 Sensors from eBay
- Thread starter mstarks01
- Start date
apostolakisl
Senior Member
I have some of the cheapo ebay ones and believe them to be accurate. I haven't tested them against any kind of reference, but the temp it says seems correct. Of course, if you are looking for accuracy down to 10ths of a degree, then you would need to disregard my endorsement as I have never tested them to that level.
Thanks. Which firmware do you mean? It is new, so maybe it already has it.
I understand it is available 1-1-2012. Not sure of the revision number. I doubt you have it.
You should email support at CAI.
Michael McSharry
Active Member
The WebControl is fussy about the DS18B20. I obtained a lot of 10 a year or two ago that did not work with it, but works fine with my other 1-wire networks. Also be careful to not get the parasitic version of the DS18B20 as the WebControl has no means to use these. I did buy DS18B20 stainless probes from ebay and they worked quite well with WebControl. Mine had molex connector at the end of the pigtail. I do not see them on ebay now. The issue with the DS18B20 parts is not accuracy, but compatibility. All have resolution of 0.125 degree C and I think the accuracy is 0.5 degree C, but this may be in incorrect memory recall.
While CAI may not change for a firmware update, you still need to pay the postage in both directions and this is a very significant for a $35 product. Look at the firmware release notes and see if the update is worth the cost to you. Another option if you need the later firmware it to order a "spare" unit with shipping requirements that it have the 2012 firmware.
While CAI may not change for a firmware update, you still need to pay the postage in both directions and this is a very significant for a $35 product. Look at the firmware release notes and see if the update is worth the cost to you. Another option if you need the later firmware it to order a "spare" unit with shipping requirements that it have the 2012 firmware.
I have a WebControl with hw 2.2.2, which works well with 1-wire sensor. I believe that hardware revision added active pull-up resistors. I was able to run the DS18B20 sensors over 300ft cat6 cables with the 3.2.09 firmware. Anybody interested in making a 3rd party firmware for WebControl board? Since the board is cheaper than buying a Ardiuno Ethernet shied board without software.
Also be careful to not get the parasitic version of the DS18B20 as the WebControl has no means to use these.
How can I tell if they are the parasitic versions? The eBay listings I have seen do not mention this.
How can I tell if they are the parasitic versions? The eBay listings I have seen do not mention this.
They should say so. Otherwise, that aren't, in my experience.
Michael McSharry
Active Member
How can I tell if they are the parasitic versions? The eBay listings I have seen do not mention this.
When you order from manufacturer then you specify by part number/package etc. When you order via ebay lots you receive whatever they have. Typically the seller does not know or does not want to disclose. While I have always ordered the versions that would imply a power input, I Have received units in cases where they behaved as if the power input was not connected internally, but as if it was a parasitic device. I usually use them in parasitic mode so usually not a problem for me, other than for WebControl.
I have sent a DS18B20 that acted like parasitic even though it was not labeled as such to CAI in UK where there deveopment site is located. I believe they arrived at the same conclusion. This was a few years ago. Back in those days I was hacking the boards per CAI guidance to get the proper power to the sensors. I collected much data that I provided them and I eventually lost interest. I could be that the latest hardware did resolve some inherent design problems.
I cannot answer for others, but my automation environment is distributed and it uses xAP (or xPL) to intercommunicate. The HTTP layer used by CAI is unique to the CAI products (and not very well documented) so to achieve interoperabilty there needs to be another computer on the network to translate the CAI protocol to the automation/telemetry standard xAP. For those that have a centralize architecture they need WebControl "driver" or "plugin" to be able to use the device. While this is quite achievable it does illustrate where automation standards using HTTP have not been rolled out.What would you try to do differently in your firmware than CAI has done?
The real power of the WebControl is its abilty to be a self-contained inner loop control system with sensor input and actuator control and ethernet monitoring. Most hobbiest have computer(s) to implement their automation logic so there are limited places where this class of user can effectively apply WebControl. Most of what I have seen from this community is that it is used a dumb I/O device. The migration from BRE to PLA gives the unit a better abilty to act as a controller, but most users do not use this feature.
I'm going to be using mine as a controller for a home built solar hot air collector. It is ambitious, but doable. I'm currently waiting on my 2.2.2 HW revision board to come back with the 2012 firmware.
The code would also need to be able to support parallel processing of temperature data and relay control for a solar hot water collector and pump.
My logic requirements are:
If the temp at the top of the hot air collector (T2) is at least twenty degrees (variable 1) higher than the temp of the ambient inside air to be heated (T1), the fan will turn on.
As long as the air temp coming into the house (T3) is at least ten degrees (variable 2) higher than what is going into the collector, the fan will stay on, independent of the temp in the collector. Optionally, a second relay will turn on and off the whole house HVAC fan, in order to distribute the hot air throughout the house.
When the difference in temp between the input air and output air is less than ten degrees (variable 3), the fan will turn off.
In order to avoid turning the fan off and then on again too quickly, the code must check that the fan has been off for at least ten minutes (variable 4) before turning it back on again. By the same token, if the fan comes on, it will not turn off again until (variable 5) has passed.
If the temp inside the collector exceeds 150 degrees (variable 5), a trouble condition, such as fan failure, will be assumed, which will trigger the sending of an email alert. Also, if the controller determines that any of the temp sensors are no longer reporting, an email will be sent, and the fan will be forced on, as long as the hour is greater than 8 and less than 17.
The variables will be established separately from the logic of the code, so that the code itself doesn't need to be changed as the system is tweaked for real world performance. I understand the Webcontrol board allows variables to be easily changed via a web browser. I also understand it allows relays to be forced on independent of program logic via the web browser.
The code would also need to be able to support parallel processing of temperature data and relay control for a solar hot water collector and pump.
My logic requirements are:
If the temp at the top of the hot air collector (T2) is at least twenty degrees (variable 1) higher than the temp of the ambient inside air to be heated (T1), the fan will turn on.
As long as the air temp coming into the house (T3) is at least ten degrees (variable 2) higher than what is going into the collector, the fan will stay on, independent of the temp in the collector. Optionally, a second relay will turn on and off the whole house HVAC fan, in order to distribute the hot air throughout the house.
When the difference in temp between the input air and output air is less than ten degrees (variable 3), the fan will turn off.
In order to avoid turning the fan off and then on again too quickly, the code must check that the fan has been off for at least ten minutes (variable 4) before turning it back on again. By the same token, if the fan comes on, it will not turn off again until (variable 5) has passed.
If the temp inside the collector exceeds 150 degrees (variable 5), a trouble condition, such as fan failure, will be assumed, which will trigger the sending of an email alert. Also, if the controller determines that any of the temp sensors are no longer reporting, an email will be sent, and the fan will be forced on, as long as the hour is greater than 8 and less than 17.
The variables will be established separately from the logic of the code, so that the code itself doesn't need to be changed as the system is tweaked for real world performance. I understand the Webcontrol board allows variables to be easily changed via a web browser. I also understand it allows relays to be forced on independent of program logic via the web browser.
I'm going to be using mine as a controller for a home built solar hot air collector. It is ambitious, but doable. I'm currently waiting on my 2.2.2 HW revision board to come back with the 2012 firmware.
I have some hot air collectors that are mostly complete. My plan for fan control is to have the Elk read the inside thermostat temp and activate an X10 unit that the fan is plugged into if the house needs heat. The snap switch in the collector will complete the circuit if the collector is hot enough. When the house has enough heat, I'll send an off command to the X10 unit. Of course, I'll also needs some kind of delay to avoid too much cycling. It would be easy to add additional logic to run the routine only between sunrise and sunset.
Michael McSharry
Active Member
The sensors can be painted. You will not notice any insulating thermal effects. If you are using WebControl for any part of fan control, then why no do 100% of it with WebControl? Do you have other ELK variables that go into the control logic?
I have only worked with BRE logic, but I supect the PLC is similiar in that control logic that changes a relay with BRE/PLA cannot also be controlled by browser unless logic is setup to allow dual source of control. In essence if you control a relay with browser then the next PLA logic cycle will recompute a desired relay state and could change it back.I also understand it allows relays to be forced on independent of program logic via the web browser.
Similar threads
- Replies
- 3
- Views
- 798
- Replies
- 20
- Views
- 7K
