Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[bkenobi]

I've been monitoring the 4 DHT11 sensors hooked up to my Arduino for the last day and a half.  It appears that they all follow a trend pretty well, but they certainly aren't calibrated particularly well.  Here's the output I'm seeing from these 4 all lined up in a breadboard.

DHT11 test!
T1 (F)   T2 (F)   T3 (F)   T4 (F)   H1 (%)   H2 (%)   H3 (%)   H4 (%)
66.20   62.60   66.20   66.20   43.00   42.00   47.00   45.00

As you can see, T1, T3, and T4 are pretty good.  T2 seems to have an offset that's always around 4°F lower than the rest.  For reference, my DMM reports that the actual temperature is 67.5°F.  H1, H2, H4 are reasonably close with a slight offset that's consistent.  H3 is always off by a couple %RH.  These modules are spec'ed to ±2°C and within 5%RH.  At that rate, I guess these are within margins for the manufacturing specs.  If they are simply offset consistently, I think I'd be pretty well off as I could just label them and make sure they get installed in the slot where I apply a similar offset.

If this error remains consistent, I doubt I'll be investing in the DHT22 or SHT75 until these DHT11's start to die.

[bkenobi]

I just found this thread regarding SHTxx and found the results very interesting.  If the results on typical units is really that close, I may switch over after all!  Granted, mine cost ~$2.50USD each and the best price I can find the HST-75 for is $15.

http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1223311726/15

[bkenobi]

I wrote some code for the Arduino that should do what I need, but unfortunately the gap in readings may cause me enough issues that this won't work.  As reported before, the offset between sensors for both temp and humidity are relatively large.  The only way I can be sure that the delta is consistent is to vary the temperature and record the differences.  I guess I'll have to get a long extension for my USB and move the sensor to different parts of the house and outside (perhaps in the refrigerator and/or freezer as well.

[bkenobi]

I'm reviving my old topic to ask a higher level question directly related to this implementation.  I live in the pacific northwest which means highly wet winters.  I currently have a temperature and humidistat wired up in parallel. If the temperature or humidity exceed the set point, the fan turns on until both set points are relieved.  The problem is that 60%RH is what I've heard is good for an attic to avoid mold issues. If I set the humidistat to 60 and it's 60+ outside, I'm pulling in air that's at least as wet as is already inside AND I have the fan running indefinitely.

I'm going to use a sensor inside and outside to only pull in cooler air (for summer) or less humid air if under 60.  I was wondering if 60 is a good value though?  Also, is it better to have moving humid air or just let it stagnant since I can't reduce humidity anyway?

Any thoughts?

[bkenobi]

I completed the project and thought everything was going to work great.  Well, I have an issue that I never expected and am not sure how to resolve.  For reference, my full CM15A setup can be viewed in this thread:

http://forums.x10.com/index.php?topic=22657.msg143881#msg143881

Basically:

CM15A (f-connector modded) -> amplifier power inserter -> 3ft RG6 coax -> gold 90° connector -> Wall plate (zinc? coupler) -> RG6 coax to basement structured media box -> in-line amp (ebay) -> RG6 coax to attic -> in-line amp (RS) -> Coax-to-SO239 adapter -> Antenna (RS 20-176)

Now, I have my attic fan controller which consists of an Arduino Uno, 3 DHT11 sensors wired remotely (1 in each attic + 1 outside), CM17A.  The controller is located in the same attic as the CM15A antenna so I assumed even if there's some problem with the Arduino's power output to the CM17A, it should still be able to transmit 5-10ft.  Well, it appears that there is a range problem...it's too close!  I have tested with a palm pad remote (HR12A).  I can get inconsistent results if I place the remote in various locations, but I need much better accuracy than what I'm seeing.

The options I'm seeing currently are:

• Find a better antenna for CM17A (passive repeater perhaps)
• Switch to using TM523 or equivalent
• Move controller to a different section of attic
• Install ethernet shield on Arduino

I don't want to move the controller since the location it's currently in is much easier to access and obviously I'd have to redo the sensor wiring to some degree (small inconvenience).  I would prefer not to install an ethernet shield since I'm not positive on the wiring availability of network cable in the attic.  Also, I don't want to rely on the HA computer for the fan.  That's the whole point of the Arduino.  If I was going to use the PC, I could have just purchased some wireless Oregon sensors and moved on.  I'm hoping someone here will have some advice on the antenna situation.  If not, I guess I'll be buying a wired transmitter.

[bkenobi]

A bit more info:

I did a bit of testing with both location of the fan controller (CM17A) as well augmenting with a TM751.  When I used a TM751 in the attic, no matter where the controller was placed, the fan would work.  Without the TM751, I was able to find one location that I thought was very reliable, but it turned out not to be (~5ft above and 8ft away from the RS antenna).  The cheapest and easiest thing I could do would be to just install the CM751 in the attic, but that's dumb to have an antenna right next to an antenna.  It will likely cause signal collisions which would reduce the overall system reliability.  I also thought about adding a second antenna for the CM15A in the other attic to increase overall range but I'm concerned how that would react to seeing the same signal fractions of a second apart.

[dhouston]

I suspect you have a frequency alignment problem. The CM17A is LC tuned, the CM15A has a superheterodyne receiver which is very narrow-band but also LC controlled. The TM751 is LC controlled but is a very wide-band superregenerative design so it may hear signals to which the CM15A is deaf.

A solution might be to use a SAW controlled transmitter (would require recoding your Arduino app) and to replace the RF receiver in the CM15A with a ceramic resonator controlled wideband superheterodyne design.

• http://davehouston.org/RX310SH-WB.htm
• http://davehouston.org/ImproveCM15A.htm

If you'd like to try it, I'll give you a 310MHz SAW controlled transmitter (4-pin module like shown in the URL below) and RX310SH-WB receiver. You pay Priority Mail Postage.

• http://davehouston.org/modules.htm

[bkenobi]

I have some receivers on order on a slow boat, but no transmitter.  I was thinking of modifying the cm15a already, so now might be a good time to try.  Yeah, if you have the recommended pair already, I'd gladly pay postage.

[dhouston]

Check your Private Messages.

BTW, RF travels at near lightspeed so you'd be hard pressed to measure any difference in reception time for two receivers within range of any X10 RF transmitter.

[bkenobi]

I guess I was thinking the RF transmission would be the same 60Hz signal as the PLC.  Now that I think about it, that would be silly.  I suppose I could consider an extra antenna, but that's going to be a last resort.

[bkenobi]

I moved the fan controller down to my work bench in the basement to check wiring etc.  While I was down there, I realized that I have a XTBM sitting there that could have been used in the attic to see if ANYTHING was being seen by the CM15A and converted to PLC.  When the fan controller sends a command, it appears that sometimes it's not being received as expected.

For instance, I send a C3 Bright command and the XTBM displays "Cxx BRT".  I need to check the log in AHP to see what was actually received.  I wonder if the commands are actually being sent but are being broken by something in the setup or by some other interference.  This will take some testing!

FWIW, I pulled the controller and disassembled as I found that with no DHT11's plugged in, the controller worked perfectly.  When I plugged 2 or 3 in, it did not.  I was given the suggestion on the arduino.cc forum that perhaps the voltage was dipping for some reason on the 5v rail.  It does not change at all no matter what the controller is doing and stays at a consistent 5.08v the entire time.  I don't know how to adjust the averaging time on my DMM, but I would have expected a flicker if there was an issue...and there is none.

[dhouston]

I assume your Arduino sends X10 RF codes via the Firecracker. This is rather goldbergian.

If you look at the X10 RF Protocol (http://davehouston.org/rf.htm), the data pulses are only 0.005 seconds wide. Missing one of them would invalidate or change the code, causing the CM15A to ignore it. And, while the Firecracker is very forgiving in terms of timing, it is critical that one of the handshaking lines remain high at all times - if not, the Firecracker resets. And, the PIC in the Firecracker needs a certain voltage level on its pins to be seen as "logic high".

IOW, you have a very complicated data chain and you really need much more sophisticated diagnostic gear to see what's going on. You need to be able to record the voltage on the Firecracker handshaking lines and also monitor the RF transmissions directly using an independent RF receiver.   

My own preference would be to ditch the Firecracker and rewrite the Arduino code to utilize a standalone 310MHz RF transmitter. That greatly simplifies things, making it much easier to troubleshoot.

[dhouston]

When I used a TM751 in the attic, no matter where the controller was placed, the fan would work.

Can you clarify this a bit? Were you using your Arduino/Firecracker setup to send RF to the TM751? That's how I interpreted your post but now I want to be sure.

[bkenobi]

By chance, I do have a standalone 310MHz RF transmitter  

I was considering doing that, but since I already have the proto shield wired up and seemingly good to go, I was hoping to use it that way.  If it doesn't look like that will work out, my choices are either use a different RF device (like the one you sent), or just purchase a PSC04 or PSC05 and communicate via PLC.

I wasn't able to get the CM17A->CM15A to work reliably so I dug out a TM751.  I was able to use the CM17A->TM751 and it worked correctly as I recall (it was a while ago).  The CM15A was still connected at the time of that testing.  I simply added a TM751 to see if that would work better.

[Noam]

In my own setup, I'm connecting the Arduino right to the TW523 port of my XTB-IIR. It works just fine for me every time. However, it isn't using RF, so there is no chance of RF reception problems.