First Heatermeter Build - It works..... almost


 
D

D Blake

New member
This is the first time I have attempted anything like this but the directions were great. Super cool effort going on here. Thanks to Bryan and everyone else that contributes to the development.

Plugged power into the Rpi and everything has gone great. Configured the wireless. Played with the settings a bit just to see what happens. Finally installed the whole unit in the case........

I guess I should have tested using the 12 volts. The unit powers up for a second or so and the display will indicate no pit probe but then it shuts off. It just keeps turning on and off. I think I hear a faint clicking noise when it switches on or off. Any idea where I should start looking for the problem?

Thanks,
Dax
 
Definitely take it out of the case and remove the rpi. Does the heatermeter work without the rpi attached ? And does the rpi still work using the mini usb cord?
 
Yeah you've got a short somewhere and it is overheating the regulator and it is shutting down. I'd check your capacitor polarity and the diodes to make sure they're facing the right direction as these are easy to put in backwards and make a short.
 
Pulled it out of the case this morning.
1) Plugged the USB power back into the Rpi and it worked fine again.
2) Checked the diodes and capacitors. They look like the photos in the wiki to me. I must admit I am colorblind so I have a bit more trouble making sure the resistors are in the correct locations but i used the part numbers during assembly.
3) Tried 12v power supply and same issue as last night
4) Unplugged Rpi and then tried 12v power supply. The heatermeter appears to work fine without the Rpi connected. Of course I could not access via wifi but it did say No Pit Probe until I plugged the thermocouple in then it read temp too.

I took a bunch of photos and uploaded them to flickr.
https://www.flickr.com/photos/daxnsuz/sets/72157650025490495/

I hope the link works. Could also try www.flickr.com/daxnsuz

Still scratching my head?

Looking at the photos, I only noticed that a couple of pit probe jacks are bridged with the nearest resistor. The pads are so close that I would assume that is where they connect anyhow. Do I need to get some soldering braid to clean them up?
 
Scratch what I said in number 4 above. It seems to take a little longer but it does the same thing on 12v power and having the Rpi disconnected.
 
From your pics, I can see a lot of solder pads are not completely covered in solder, Also, you need to remove the 10k resister that is behind the TC jack. Its not needed when you have the TC components soldered. Still looking, I have not been through the pics but some more with better lighting of the solder side of the board would be useful
 
I found the pics that helped. But I don't see anything that pops out as bad. that could be good or bad. Anyways I recommend that you first take you soldering iron and hit each solder joint again, add some where you still see solder pads. Also, run your soldering iron in-between the joints at a somewhat fast movement with very little solder on the iron to remove any stray solder. Then give it another try without the RPI. If you still have problems then take your solder wick and apply it to the main components like the DC/DC converter, transistors, Atmega328, shift register, and any where something just does not look like a small pyramid shaped solder joint, Shiny joints Equal good joints.

I have not seen any parts that look wrong except the 10K resister that is not needed
 
Looking at the photos, I only noticed that a couple of pit probe jacks are bridged with the nearest resistor. The pads are so close that I would assume that is where they connect anyhow. Do I need to get some soldering braid to clean them up?
Click to expand...

Those that are close are fine, as they are connected
 
Thank you John. I saw your first post today and took more photos with several flashlights pointed at the board. They are uploaded now. Same album but I tagged them with "more light"

Based on your feedback, i.e. "I don't see anything that pops out as bad", I thought that I should check the power supply. I bought it from Amazon and got the 2amp version rather than 1 amp. Just wanting to make sure I had juice for the fan and the wifi to work reliably. Well I checked the voltage on the 2.1mm jack. It was bouncing between 17 and 18 volts. Certainly not 12 volts. I will dig around and see if I can find a different power supply at home.

I'll need to go to radio shack to get some soldering wick. Never needed the stuff before.
 
Forgot to ask, would the high voltage be a likely cause?

edit: Found a power supply (12v 1.5a) that goes to an old Vonage device. It actually has a grounded plug too. Checked the voltage and it was rock steady around 12v. Next step, I plugged it into the heatermeter and it has been reading consistently for five to ten minutes.

Looks like my brand new power supply is a dud. I am going to let it run for a bit longer on the vonage wall wart and see if it holds.

Next, I need to complete John's recommendations for the resistor. I may avoid doing any other solder work for fear of messing up the now working meter.

After that, I need to find the details so I can access it from the internet when away from the house.

THANK YOU for taking the time to look at my work.
 
Last edited:
Byran wiuld need to chime in to answer that, as i had wondered about using an adapter that was rated at 13.5v. That could be the problem. As the dc/dc gets hot from over voltage, its thermo protection triggers and shuts off. The pi just makes it trigger even faster.


Edited....

Glad it was something simple and since its running, i would worry to much about the solder joints, but i would definitely cut the pins on the lcd or any other pins that were not cut at are a tad long, if they are near the pi.
 
Last edited:
I'll still chime in! I've tested the input up to 24V and the regulator works better and better the higher the voltage goes (at least to 24V). The fact that the power supply was putting out such a high voltage may indicate it was just bad to begin with. I've seen some power supplies that put out 12.5-13V with no load but 17-18V is pretty high. Glad you worked it out though, such a shame when the one piece you didn't make yourself is the piece that ruins the whole thing!
 
Well I let it run for about 18 hours. It was just logging my basement temperature and the effects of our programmable thermostat. Worked like a charm. I had the set point at 65 degrees and when the furnace thermostat went to a lower set point around 11pm, the heatermeter started ramping up the fan to warm the fire. Of course there was not fan or fire in the basement.

Bryan, I agree. A real frustration that a brand new item such as a power supply was the source of the problem. Interesting that you have run them at a higher voltage without a problem. The other aspect when I put the multimeter on the defective supply was how it bounced around so much. I never got a stable reading. The power supply I dug up in my box of old electronics was very stable by comparison.
 
You must log in or register to reply here.

 

Back
Top