|
Post by DCish on Mar 15, 2019 5:28:38 GMT -8
Definitely exciting what Peter's documenting! I'm not sure my first airtight build will get into the sub-1000ppm range right away, but it is fantastic to know that it is reliably achievable. Stoked to hear how your experimenting goes! My big project for today is getting a clothes line up now that the tree that used to hold it is stacked in small pieces in my yard waiting to heat my house.
|
|
|
Post by esbjornaneer on Mar 16, 2019 14:34:14 GMT -8
You bet Peter's dsr2 development is great, but so are these posts. I hope you get on well! It would be great to have a components list and connection diagrams if it goes well for you. That is unless you could send one to Calgary in July... I am Europe based but go to Calgary in July this year. I look forward to your developments.
|
|
|
Post by hallinen on Mar 23, 2019 15:47:36 GMT -8
Update on flue gas project...
Carbon monoxide: The MQ7 sensor circuit appears functional. I'm waiting to get the rest of the sensors before I rent a commercial CO tester for calibration. I ordered a few from different manufacturers to see what seems most stable.
Particulates: The Nova sds011 arrived. It now has a functional circuit. I'm waiting for the two versions of the plantower counter so I can compare them all. This circuit was easy, and I was able to use the super cheap, nodemcu, microcontroller with it. I've downloaded the literature that I could find on it. The scientific testing comparing the Nova sds011 to high quality sensors was good. I wrote to a research group in Waterloo, Ontario that does wildfire smoke testing to see if they would be interested in testing it. I've not heard back. I'll dig around to see if I can find any other research groups who might want to mess with it.
Lastly, and most frustratingly, is Oxygen. The good news is I found an Italian brand, Figaro, that makes a $75 sensor that is said to be okay for flue gas. I researched some of the other commercial sensors and I'm pretty sure flue gas would be bad for them, especially the ones that use photoluminescent technology. I have not given up on the super cheap zinc air battery circuit. I found three different (Rice, UC Irvine and Hopkins) college biomedical engineering groups that got a zinc air battery to work for an oxygen concentrator meter. They were interested in differentiating oxygen concentrations between 40%-60%, 61%-80% and above 81%. The group from Rice said they got to within 1% accuracy and they were kind enough to send me a print out of their circuit yesterday. One of the students was sceptical that I would get within 3% accuracy at the low oxygen concentrations that we are interested in. I told the group if the knew of any electrical engineering students who are interested in the project to let me know. They told me they spent 3 years optimizing the circuit for the high oxygen levels.
I also reached out to a professor at University of Michigan. He said he could offer the project to his students in September. I'm going to dig around for a student who wants a summer project since I hate waiting! If anyone knows of any tech loving college kids interested in forestry or electrical engineering, let them know about this please!
So, progress is happening, slowly.
|
|
|
Post by DCish on Mar 23, 2019 17:23:10 GMT -8
I'm curious, where do you go to rent a commercial CO tester?
|
|
serg247
Junior Member
The mountain can not be conquered, it can allow it to ascend...
Posts: 111
|
Post by serg247 on Mar 24, 2019 2:47:26 GMT -8
|
|
|
Post by hallinen on Apr 2, 2019 14:13:44 GMT -8
Ahhh, oxygen, my good friend. I've learned it is smarter to change only one variable at a time and test it before blundering on to the next variable. I think I finally have some good news on the cheap PR44 zinc air hearing aid battery oxygen sensor. It can reliably tell me if my ziplock bag is full of room air or my exhaled breath! The current protocol is to short the zinc air battery for 30 seconds, then watching the voltage recover to various arbitrary numbers. For example it takes the battery to recover to a voltage of about 0.5 in 70 seconds when the bag is full of room air (21%) and 92 seconds when full of exhaled air (16%). I was using the oxygen depleter packets yesterday and was getting inconsistent results. I was changing the recovery time on the battery, not realizing that was critical. So today I just used my breath and was careful about controlling for the recovery time, and boom, consistent results! I still have a bunch of tests to run, and have to figure out a program for the little arduino run so it will automate what I've been doing. From the looks of the response time of the zinc air battery and the MQ 7 CO sensor, I think it makes sense to have at least two testing chambers running. The current configuration for the zinc air oxygen sensor costs... zinc air battery $4.99 for four of them. I ran 57 tests on one battery today and it was still giving consistant results. www.amazon.com/Kodak-Hearing-Battery-K675ZA-1-45V/dp/B00JG30ZGG/ref=sr_1_2?keywords=pr44+battery&qid=1554241752&refinements=p_89%3AKodak&rnid=2528832011&s=gateway&sr=8-2 I think they came from England so it took a few weeks to show up. arduino nano $8.99 (I bought mine directly from China for $2.36) www.amazon.com/ATmega328P-Microcontroller-Board-Cable-Arduino/dp/B00NLAMS9C/ref=pd_day0_hl_147_3/155-0427600-6989261?_encoding=UTF8&pd_rd_i=B00NLAMS9C&pd_rd_r=713522ec-5590-11e9-8348-f91a7e0ecc75&pd_rd_w=7vUX1&pd_rd_wg=873MD&pf_rd_p=ad07871c-e646-4161-82c7-5ed0d4c85b07&pf_rd_r=FNQVWQZWQ9KHCF2W8BVF&psc=1&refRID=FNQVWQZWQ9KHCF2W8BVF2 resistors and 6 wires, one transistor. Each costs pennies when ordered in bulk. cloths pin, sandwich size zip lock bag. I was at my local library yesterday and noticed that they are now lending out arduino kits that have resistors, wires, and it even had a transistor. I am STILL waiting for the particulate sensors to arrive from China. I also emailed a professor at University of Dayton. They have a really fancy combustion analysis lab so I thought this would be a good project for them. I'm sure with a few op amps (whatever the heck they are) and young brains, this sensor project would be pretty solid. In the meantime, I slog away. In fact, I just now I realized I didn't control for humidity, my breath has more moisture than room air, dang. Tomorrow I hook up the humidity and temp sensor and try again. Maybe desiccant will work, or maybe a wet sponge sealed up in the room air bag. Hmmm.
|
|
|
Post by hallinen on May 3, 2019 5:53:32 GMT -8
It has been awhile since I posted an update of the flue gas analysis project. It took many tedious experiments, but I think the zinc air battery is going to work. The issue is accurately calibrating the battery in a user friendly, automated fashion. Right now, I'm covering the holes of the + size of the battery, where the oxygen flows in, and shorting it out so most of the left over oxygen is used up. Then, I watch the voltage rise after the short is stopped. The variables I'm trying to manage include how long the battery sat uncovered, how fresh the battery is and of course variables I have not figured out yet. For the room air calibration, the holes are uncovered, the battery is shorted, and the voltage recovery is observed. I've noticed the first calibration run can be way off, but by run 3 it settles down and becomes very consistent. Yesterday, I used an oxygen absorber in a 100 cc sealed mason jar and watched the oxygen level drop form 21% (well actually a little lower, but I like integers) to 5%, which is what is was suppose to do. What improved the sensor was using a 10 ohm resistor between the + and - sides of the battery during the short. I checked a 0.7 ohm and 100 ohm resistors to narrow it down to 10. Now I should run through resistors in 5-50 range to see what works best. I also need to figure out the best calibration run voltage curves, and write the program that calibrates with minimal user input. I also plan on using some LED lights to communicate with the user and save the data to an SD card so the user doesn't have to bother with installing software to run the arduino serial port. I discovered that the arduino drivers don't like old, dying laptops. They also do not work on my chromebook. This is a solvable problem, but since I make every possible mistake that can be made, it will take awhile. My particulate sensors finally arrived! They work, but I didn't realize they needed a particular connector. I managed to solder the hair thick wires to a connector, but boy, that challenged my eyeballs! I emailed with the author of a paper, Nanoparticle emissions from residential wood combustion, www.sciencedirect.com/science/article/pii/S1364032119300012. She didn't know of any researchers doing flue gas analysis with the laser sensors, but she thought it would be useful. One issue she brought up is right now the regulations do not take into account particle size and count, they only care about the mass (weight) of the particles. From a human health perspective, count and size is critical. After I finish soldering the rest of the sensors, I'm going to do some runs to see how consistent the sensors are between each other and on repeat measurements. My CO sensors are sitting in a box, patiently waiting for me stop fooling around with those zinc air batteries.
|
|