5 Replies Latest reply on Mar 12, 2015 3:38 PM by JPMontero_Intel

    Latest DHT sensor wisdom


      The is conflicting forum information on using the DHT sensors as follows:

      On 11/13/14 AlexT_Intel  stated there is a 2 pin solution for DHT sensor DHT11 Temperature Sensor Issue


      I found DHT workaround for Galileo and Galileo Gen2 using 2 pins instead of one  Where SpiderKenny trashes the solution by stating - the only proper way to use the DHT sensor is with an interface.

      He seems to be a bright guy, and claims the interface costs pennies.  I was wonder if he would sell me one of his pic programmers for pennies because his interfaces requires an expensive pic programmer. Based on these forum conflicts please advise on the latest wisdom on the above solutions before I invest in many wasted dead end hours. I am using  a GEN 2.


      Being many of us and I were duped by Intel's  representation of "Arduino compatibility" , where Galileo's very slow  I/O speed  killed much of this compatibility for One Wire and DHT type sensors.   SpiderKenny has an eloquent explanation of why the speed is so slow, but Intel's per-purchase  datasheets daftly  failed to point this out.  Because of this deception I have a second question relating to the the new development environment which added Dallas one wire example programs into the examples menus.  This was not in earlier versions.  Do these sketches now work on GEN 2 or is someone screwing with us?  If they don't work why were they added?



        • 1. Re: Latest DHT sensor wisdom

          Well, I hope I don't regret weighing in here, but I want to provide my perspective on your first question.


          For background I want to state a few things up front:


          I am not an engineer.


          I bought a Galileo Generation 1 because it seemed like something I would like playing with and it was on sale and it had a cool box - it cost me US$34.99 and they sent me a replacement power supply for free although the original is still working fine (some potential interference as per the design I guess).


          I interfaced a DHT22 and DHT11 to the Gen 1 without too much difficulty using the serial port and I wrote it up here Gal 1 to DHT-22/11 interface through the serial port. I bought 2 DHT-22s and 1 DHT-11 - can't remember but they were something like $5-$10 each.


          I also interfaced an HH-5030 humidity sensor and wrote it up here Using an HIH-5030 humidity sensor . It didn't require an interface as it's output is an analog voltage and connects to the Gen 1 pretty much as is. I think it was $8.


          For temperature, I have used an LM35DZ and wrote it up here Easy temperature measurement with the LM35 . I think it was maybe $1.50 . Again, no special interface required.


          I also spent some time learning about thermistors and maybe I will post that as well as they seem like a good economical choice for temperature measurement.


          I don't personally know Spider Kenny, but I have read his opinion on this issue and I think he is more right than wrong (I would take issue with the term "nondeterministic", but for other philosophical reasons). Also, he had a complimentary response to my DHT post, so he obviously has good taste and knows his stuff


          Currently, I am working on a general PIC chip interface for use with my Gal Gen 1 that functions as a slave I2C interface (as seen by the Gal) but will be capable of operating with several sensors (like a DS18B20 and a DHT-22 and also as a port expander). The I2C interface part seems to work well (thanks in large part to some freely available app notes) - at least it does so far and I am curious to see how far I will go with this project.


          The programmer for the PIC that I bought (a PICkit 3) cost $18.80, delivered (slow boat style). It has worked flawless and there was nothing else to buy (assembler etc... downloaded for free) - not really what I would call "an expensive pic programmer". Cheaper and DIY alternatives are abundant.


          The Gen 1 *is* Arduino-compatible...sometimes... and sometimes not. I can't argue with you there, but I did not know what an Arduino was until after I bought my Galileo. I do know that I have used some LCD displays with Arduino libraries without problems. Recently, I interfaced a BH1750 using the Arduino libraries without modification. They (this site) put out a shield testing document that is pretty straightforward and that should count for something.


          If you look at the timing requirements for the sensors you cited, it does not surprise me that they might not work with a Galileo without an interface - for reasons that have already been stated having to do with what else is on the Galileo that is not on an off-the-shelf Arduino UNO, for example. Not every solution will fit every need.


          I should also add that I have done these projects because I can, not because I have to...that is, I know that I might feel differently if my livelihood depended on doing so.


          My point of all this is that, in my opinion, inevitably it is up to the person making the end device to know the power and limitations of the equipment they are putting together. If a "workaround" does not cut it (and some of them are down right clever), look into alternatives. I just accept that.





          • 2. Re: Latest DHT sensor wisdom

            Thanks RG,

            I appreciate your offering alternatives and links.  However all my analog inputs are already allocated to other functions.  I was looking for Intel to step up and address my stated issues.  I remain very interested as to why the One Wire examples were added to the the Development Environment


            • 3. Re: Latest DHT sensor wisdom

              Hi Bsessa,


              The reason we kept all the example sketches in the Arduino IDE 1.5.3 is because we plan to merge the Intel customized Arduino IDE with the mainstream Arduino IDE that will be available from arduino.cc, which will include all the Arduino boards and all the examples, but this does not mean that all the examples will work on each and every board.



              • 4. Re: Latest DHT sensor wisdom


                Thanks for your reply. Although the response seems logical, it reeks of unethical marketing tricks.  For example if Intel has an objective as stated and is actively modifying release versions of software, should they not be making some attempt to, at a minimum, add remarks in the associated Arduino sketches referring that they don’t function on Galileo?  Until that is done they should not be displayed in a Galileo menu under the guise of a future merge with  Arduino.   What would it take, about a minute,  to add a comment?


                Simple step as such, would probably have saved unsuspecting Galileo owners, Intel support staff, this forum and me, countless thousands of hours .  It would certainly add to my confidence in future Intel product releases not providing either misleading Vaporware or deliberate false representations of anticipated functions.


                Please forward this comment to the proper Intel management in the hopes that the next release is considerate of all the thousands of customer hours  wasted. I like the potential of the Galileo product and offer this dialog in the spirit of improvement.  I am using my pain to hopefully motivate change.


                Did you have anything to add on the official Intel position of  DHT support?



                • 5. Re: Latest DHT sensor wisdom



                  Intel followed the Arduino IDE module, the current Arduino has about 23 boards and soon Edison and Galileo will be added to the list. You have a good point and I suggest to post it to http://forum.arduino.cc/ to request enhance to the Arduino IDE to disable (gray-out) non-applicable examples for the selected board.