Thank you for your help Nicolas,
but even with your setup i can not recieve any answer from the sensor. I did that with a small breadboard, shortend the legs of resistor and cap and put them directly near the sensor…
no chance 
i must believe it’s broken.
yes, di7 and ethernet shield isn’t a good choice, i realized that too - tried it without anything but the sensor setup.
Mark,
My setup was using a FEZ firmware at least 2 releases behing. Just to make sure, I will upgrade to the latest GHI SDK and check if it is still working. I will let you know, I must have one or two spare DHT22 arround. Unlike DHT11, I never had fiability problem with them (DHT11 would work for a few months, and then its reading would become very unstable).
Yes, of course it is also possible that your sensor is dead…
By Chance, would you happend to have some logic analyser (or memory scope) just to check visualy that the sensor is really sending data on the line ? Those are really aluable tools when troubleshooting electronics 
Hi Mark,
I had a chance get a hand on a spare DHT22 and test it this morning with latest GHI firmware.
It worked straight out of the box with the code provided on code.tinyclr.com
I used DI40 and DI42 of the pandaII. See below a picture of the setup I made for the test - it’s quick and dirty but that works.
So I think there is a chance your DHT22 might be defective or damaged ?
(sorry I was too lazy to remove my ethernet shield. The extra wiring comes from that I usualy debug using serial those days / busy developping USB applications 
)
Nicolas3 -
I tried your driver as well with no luck. Nothing back from the sensor. I ran the sensor on an Arduino and it works. Used Di5 and Di7 on a Domino. Also tried 5V.
What is the black wire on the Panda2 (looks like its in Gnd and D33)?
Hi Eric !
The black wire is only here between “MOD” and “GND” to use serial debugging. Those days I am working on applications using the USB client port, so it’s not available for debugging.
Did you try other pins ? if you have a fez connect shield on your panda, most likely it is already used by the shield. Best thing is to try to use it with a simple fez board with nothing else connected.
I have 3 DHT22 here, and I tried them on several boards domino/panda/pandaII. The only trouble once I had was because there was a bad contact on a pandaII connector (I just switched pins)
Good luck !
Nicolas
Nothing else connected to the Domino. Will try other pins and a PandaII. Also my jumper wires are kind of flimsy. I’ve been busted by a bad one before.
Victory! I removed the resistor close to the sensor and changed pins to Di2/3, which both have 2.2K pullup resistors on the Domino.
[edit] @ nicolas3 - also thanks again for your work on NetConfig.
Shortly (this weekend) I’ll have my son’s hermit crab tank logging temp and humidity to a Sql Azure database. Next up I’ll add 3 more turtle tanks, once of which is a 200 gallon aquatic tank. End game is to be able to monitor and adjust temp/lighting on all of these from a silverlight app.
Great ! It’s nice to hear that I am not the only one able to make it work. There have been several threads of peoples having difficulties to make it work, so it is a little reassuring !
I just wonder why you can’t have it working with a regular external pull-up resistor. What value did you try ? I used a 1k because that’s what I had sitting on my desk, but you may try values a little higher (2.2K, 4.7K).
BTW I several extension projects for netconfig, but unfortunatly my professional days have been very busy during the last 6 months at least ! I was working porting a uPNP class ( NAT Traversal with UPnP in C# - CodeProject ) but I decided to wait for netmf 4.2 to have regexp available. And also a remote software upgrading. But days have only 24 hours !
Is this for breeding or are your reptiles just spoiled?
Breeding?! No way I would agree to that. It’s already a freakin’ zoo in this house. It’s pure geek factor. My wife is the crazed critter-ologist, and I swear she’s fast tracking my son for a career in zoology. I’m leveraging her love of turtles to my advantage! Light and heat for these critters adds a good chunk to our monthly electric bill, particularly in the winter. Automating climate control will be more energy efficient and also keeps them closer to their optimal environment. It’s a perfect project for all this uC hardware I can’t seem to stop buying 
Now all I’ve gotta say when she sees another package come through the door is “It’s for the turtle project, dear”…
I was quite a reptile (mostly lizards) fan in my youth also. If the cages are in close proximity, you can probably control just about all of them off a single Panda-II. Hmmm… A “zoo shield” that has an array of common sockets (similar to an e-block shield) where each socket has everything you need to control a single aquarium. Sounds like that would be a fun project!
Currently the critters are scattered around the house, so I’m going with Xbees and a central bridge to ethernet. I scored several of the v1 Pandas when they were on sale. The Xbees and PowerSwitch Tails cost more than the controllers!
Sounds like a good plan. XBees & PowerSwitch Tails are not cheap indeed. You could probably cut the cost of the PowerSwitch in half by using your own SS relays. Watch eBay for XBees. You can find them for half what they cost at SF.
Merci & thanks guys!
ransomhalls trick did it…
use di2 or di3 as output pin and get rid of the pullup resistor.
But why it’s not possible to get that work on any other interrupt port WITH external pull up?
That’s a mystery. An electronics mysteries need to be solved, that when you get to learn something : there are no ghosts in the electron world. Internal or external pull up should be the same. Don’t give up !
Gus?
Output without pullup on di2&di3, but why not on other ports and external pullup? :
Those pins are open drain, the I2C pins
So, what’s the magic to the other panda pins according to the dht22?
i don’t understand your question, others may not either.
Sorry Brett,
i don’t understand why it’s only possible with di2/di3 to drive the dht22. Why is an open drain output nessesary - why couldn’t i use other pins with an external pullup (as it was done by others, mentioned earlier in this thread) to make the sensor sending a response?