Print

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

[Noam]

Depending on the frame, I might even be able to place them BEHIND the picture. I'll have to see if I have something that wouldn't look out of place on that wall.

I was already using the DS18B20 and DS18S20 sensors for temperature, before I got the DHT11's.
The specifications certainly indicate the DHT11 is less accurate, and It isn't very hard to run both of them through one CAT5 cable.

I was thinking of running the wire up into a shallow plastic wall box, and then all I'd need is some sort of vented cover-plate to allow the sensors to read the air temperature in the hallway. If I insulate the box (haven't figured out the best way to do that yet), then I have a better chance of getting a reading from the hallway, and not the interior of the wall. Finding the right type of cover for the box might prove to be very tricky, though.

I wonder if I could mount the two sensors right behind the grille of the return duct, which is also in that hallway. It certainly gets airflow (especially when the HVAC system is running), and nobody would notice it (unless they look for it). I'd probably have to place the DHT11 with the "holes" facing the inside of the duct, to try and minimize dust buildup.
What do you think of that?

[bkenobi]

That sounds like a good solution to me!  The air return duct would get you a pretty accurate sense of what conditions in the hallway really are.  If you put the sensor on the wall, as you mentioned, you run the risk of sensing wall temperature.  Since there's a pretty slow flow rate there, the boundary layer at the wall will probably be somewhat elevated.  I don't know how that would affect humidity since that's driven by diffusion.  But, the air return would solve the problem fairly effectively.

As for dust, you could probably just wrap a layer of speaker foam or fabric around the sensor if you noticed a build up.  I'd probably leave it in the open for a while to see if it's a real issue as you may slow readings significantly by installing such a filter.

[Noam]

Another advantage of installing it behind the return duct grille is that I don't need to worry about punching a hole in the wall, and cleaning it up afterward.
The question is, could placing the sensor in the stream of the airflow in the duct cause an inaccurate reading, as compared to what the "real" thermostat is sensing on the opposite wall? I suppose I'd have to try it to be sure.
I used up my spool of CAT5, so I'll have to go buy another one before I can wire this up.

[bkenobi]

The question is, could placing the sensor in the stream of the airflow in the duct cause an inaccurate reading, as compared to what the "real" thermostat is sensing on the opposite wall?

I'd actually ask that question the other way around. 

Could placing the sensor in the stream of airflow in the duct create a more accurate reading than what the "real" thermostat is sensing.

I'd say that the sensor in the air stream is going to be much more accurate than the thermostat.  It won't precisely match the value of the thermostat, but that's primarily because the thermostat is correctly reading what it sees at it's installed location.  That will not match the average conditions in your house.  If you want the average conditions, the best thing you can do is to use the air return duct...assuming you have enough spread throughout the home to provide correct central heating performance.

In my last house, we had a tri-level with a single return duct in the lowest floor.  It returned air from the entire house and caused a very large delta between upstairs and downstairs conditions.  In my current home, there are 5+ return ducts.  If I wanted to find the most accurate average temperature/humidity in the home, I'd put the thermohygrometer in the return duct where all the ducts meet at the furnace itself.

If, on the other hand, what you want is to see what the conditions are in a single room, you might be better placing the sensor(s) somewhere near the center of the room volume.  If there's any air flow, the distance from the center becomes less critical.

[Noam]

The question is, could placing the sensor in the stream of the airflow in the duct cause an inaccurate reading, as compared to what the "real" thermostat is sensing on the opposite wall?

I'd actually ask that question the other way around. 

Could placing the sensor in the stream of airflow in the duct create a more accurate reading than what the "real" thermostat is sensing.

I'd say that the sensor in the air stream is going to be much more accurate than the thermostat. 

What I was getting at, is would the moving airflow throw off the temperature sensors, the way you feel cooler when there is a fan blowing on you?

We have a two-story house (main floor and basement), with one return on each floor. We have a forced hot air gas furnace, with central air. My one thermostat controls both systems (and can actually switch between the two to keep the house within a predefined temperature range). My idea was to put temperature sensors behind each return grille, and an additional one just before the air enters the furnace (which should give me a mix of the two temperatures). There isn't much I can do to fix airflow issues without a major reworking of the ductwork, but I'm curious to chart the difference between the upstairs and downstairs, especially as we move from winter into summer.
I also plan to put a sensor in the plenum just above the furnace, so I can compare the inlet and output temperatures  and make sure the A/C and furnace are doing their jobs properly.

I'm having an energy audit done next week, and I'm hoping the auditor will find the data I already have to be somewhat helpful.
I can already look at how fast the temperature drops when the system is turned down to "away" mode, and chart that against (slightly inaccurate) outdoor temperature.

I have to figure out how to get the sensor into the return duct now. The grille is painted to the wall, and I really don't want to mess that up by removing the grille. My dad suggested dropping a string with a weight on it down through the grille, and pulling the wire with the sensors attached back up. If I make a hook out of a paper clip, I should be able to carefully hook it onto the inside of the grille. I can get the wire out of the return duct at the bottom pretty easily.

[bkenobi]

Ahhh, I see what you are asking now.  No, measuring the temperature in still versus moving air should give you the same results.  I work in aerospace with fluid flow (CFD primarily), and the differences between moving and static air are important and looked at using total and static air temperatures.  In your case, you are looking at flow speeds so small that there will be no significant pressure difference thus there should be no issues.

I just replaced my furnace last month because it was really old and did not provide sufficient heat.  The furnace would provide air through the registers at up to 70°F, but because the air was moving fairly quickly it felt more like 50°.  The obvious issue comes down to convection.  Our bodies are 99.5°F nominally at the core and 98.6° measured orally.  Skin temperature is lower and depend on location on the body as well as time of day relative to activity level (and lots of other things).  In any case, the skin temperature will certainly be much higher than 70°.  In order for the air to feel warm, the air must be warmer than skin temperature such that the delta temperature between the air and skin drives heat into the skin.  70° air moving at any speed will feel cold.  90°+ air will feel warm.  Oil furnaces can provide air at over 100° easily and are considered the most comfortable for that reason.

In any case, the point is that the sensors that you are using do not take this convection into account and simply report the air temperature itself.  In short...you should be fine.


EDIT:  I just found this site that has a cool plot with different body temperatures at different ambient temperatures.  I'm a geek engineer, so I'm gonna post it!   

http://www.healthyheating.com/Definitions/facts_about_skin.htm

[Noam]

Thanks for the info.
That explanation made a lot of sense to me (and also better explains why I could heat up a sensor by pinching it between my two fingers (which are warmer than room temperature), but I couldn't get it below room temperature by aiming a fan at it. I WAS able to cool it slightly below room temperature by clipping on a copper clip, that had been cooled in a cup of ice water. Of course, the effect only lasted a few seconds.

I need to find the time (and more wire) to wire up the sensors on a long enough wire to run up through the return duct.

[bkenobi]

One thing I forgot to mention about temperature measurement that may be important here depending on your furnace's design.  There are 2 types of thermocouples generally used for measuring temperatures (shielded and unshielded).

Thermocouples can be used to measure both surface and fluid temperatures.  When the temperature of the surface/fluid is around the same as the surrounding, it is acceptable to use unshielded sensors.  If, on the other hand, there are large variances in temperature between the area of interest and it's surrounding surfaces, the sensor can (and likely will) see a radiation component.  Think about standing in the sun on a cold day with no wind.  If you face the sun, your face might feel warm, but your back will be cold...radiation!  I can go on and show the physics, but basically here's the gist...make sure the sensor is not facing anything hot (like the burner or heating elements).  If the installation location can see this location, just make sure you have something blocking the device's view.  The DHT11/22 has a plastic case that probably works fine for this since the sensor is on a PCB inside the shell.  The DS18B20 and DS18S20 do not have a shield, so it might be advisable to consider where they are installed.

Since radiative heat transfer is 4th order (Tsurf1^4 - Tsurf2^4), the hot surface needs to be very warm to make a significant difference.  You could probably just install both the DHT11 and DS18S20/DS18B20 and see if they report different values.  If they are, consider making a simple change.

[Noam]

Thanks for the tip.
I found one more long scrap of CAT5 in my wiring box last night.
I cut up a DC-ROM audio cable, to get the 4-pin sockets from the ends. These are perfect for both the DSB1820 and the DHT11, allowing me to swap out sensors if need be, without having to solder more.
I put both connectors onto a small piece of perfboard (I still have to hot-glue them in place or something), and wired them to the end of the cable. A little white electrical tape will make it less noticeable once it is inside the vent (of course I can't tape over the sensors themselves).
I hope to have time this weekend to pull the wire up through the duct, and hang it from the inside of the return grille. Then, I have to wire it up to the arduino and test it.

I'm still trying to figure out the best way to interface all this with the Arduino.
I have a prototyping shield, that already has my status LEDs, and all the needed resistors for the inputs and outputs. I have a set of 4 wires that go to a terminal strip where I connect the wires from the furnace/thermostat interface. I have another set of wires going to a second terminal strip, where I have the wires from most of the DS18B20 and DS18S20 sensors coming in. I also have wires going from there to a small breadboard where I have two more DS18B20's. It is a good place to add stuff for testing, without disturbing the screw terminals on the terminal strip. Another set of jumpers temporarily goes from the Arduino to that same breadboard for the DHT11 at the moment. I just have to pull the shield off the Arduino, and solder in a more permanent set of interface wires for the DHT11 (I might add a second data channel wire while I'm at it, for future expansion).
Eventually, I'll replace the Arduino Mega with the Uno that I bought. I have a few of the "Really Useful Boxes" (from Office Depot), which are the right size to make an easy enclosure. I'm not sure how to do the external connections, though. I want to be able to disconnect the wires coming in relatively easily, if I need to take the system apart to move it or make changes. I was thinking of attaching a screw terminal strip to the box, and connecting everything that way. I can join the DS18B20/DS18S20 wires together at a point closer to where they diverge to go all over the house, and have that be a bit neater, too.

[bkenobi]

Is this going to be a permanent setup?  Since you are having a HVAC person come into your home and he will know about this setup, he is going to tell you that you need to use plenum grade wires if they are installed in the HVAC.  If it's temporary, I wouldn't worry about it.

Otherwise, this looks like it will work out nicely!  I'm itching to get my DHT11's and temp sensor samples!

[Noam]

Is this going to be a permanent setup?  Since you are having a HVAC person come into your home and he will know about this setup, he is going to tell you that you need to use plenum grade wires if they are installed in the HVAC.  If it's temporary, I wouldn't worry about it.

Otherwise, this looks like it will work out nicely!  I'm itching to get my DHT11's and temp sensor samples!

I don't know about permanent, but certainly long-term.
I don't have an HVAC person coming - I'm having a Home Energy Audit done.

[bkenobi]

Ahhh, ok.  Plenum grade wires are designed to go inside plenums (HVAC ducts).  The difference is that these wires have insulation that won't create a toxic cloud if burned.  It's something that needs to be considered when installing in buildings and would definitely be better in your case.  However, since no one will ever check, it's not really a huge deal.  I can't say for sure how much better it is to breath smoke from plenum grade wire bundles vs standard.  I personally would avoid breathing in either one!

[bkenobi]

My DHT11's showed up yesterday.  WOHOOO!  I dug out some CDROM audio cables, but they generally seem too loose fitting on these very thin connectors.  I found a couple power y-adapters (standard molex to standard molex + disk drive molex) and they fit very snugly.  I bet this is the type of cable i can order from Hong Kong for $0.30USD shipped.   

[Noam]

My DHT11's showed up yesterday.  WOHOOO!  I dug out some CDROM audio cables, but they generally seem too loose fitting on these very thin connectors.  I found a couple power y-adapters (standard molex to standard molex + disk drive molex) and they fit very snugly.  I bet this is the type of cable i can order from Hong Kong for $0.30USD shipped.   

I found a few that were too loose, but I also found a few that seemed perfect, when I pushed the DHT11 all the way in.
I don't know if it helps, but they were the kind with the same type of "CD-ROM" connector (with the little latch) at both ends.

I pulled the wire up through the return duct last night, but I pulled it out after about an hour. I wasn't seeing a real difference between the upstairs and downstairs temperatures (there probably isn't much, AND the DS18B20 probably isn't sensitive enough to matter for this application). Also, the DHT11 was reading way off (about 10% high). I think one of mine was off my more than the others, and I'll bet that was it.

I think I can do without measuring both upstairs and downstairs temps, certainly for now.

[bkenobi]

Interesting details.  I just placed an order at allelectronics.com last week for a few supplies to tinker with (potentiometers, push buttons, etc).  I wish I had realized that the cdrom cables I have won't work that well.  Those guys have 4-pin connectors that look perfect and have a 10" lead on the end for soldering/whatever.  $1.35 each, but shipping is a flat $7.  DOH!