"From the raw data, can you tell us if it is 3000 mV or closer to 3300 mV? It is hard to read from the graph."
Tracy,
The maximum millivoltage for the trial for the attached graph #1 was 2960 at 56 seconds, 2962 at 58 and 61 seconds, and 2964 at 63 seconds. At the point it dropped, maybe in response to turn on the vacuum pump (?)
It's important to note:
1) The trial for the data above was only 3:24 (minutes and seconds) and it was conducted inside a vacuum chamber.
2) In previous trials where we got 'clean' data, trials NOT done in the vacuum chamber, the average maximum millivoltage was typically between 3000 and 3500 mV after the heater had warmed up and the millivoltage had stabilized (see attached graph #2.)
Can you help me understand why this is significant?
Also, I have a nagging question, probably with a simple answer. When we solder resistors in-line with some of the wires, why do we use a 1 kilo-ohm resistor? Why not a 1/2 ohm, or 2 ohm resistor? What's the "magic" in using a 1 K-ohm resistor?
=============================
The Rocketeers are doing awesome work, aren't they? And we sure were glad, at 1:30 AM last night, that the glitch with the MAWD was simply two reversed wires. That's what happens when you work late, and when you're tired. It's not the best time to do quality work. I hope the Rocketeers will remember this lesson if they're thinking about staying up all night to study for an important exam the next day... not a wise decision!
"Question #1: What do you suppose the sensor would read if you seal it in the air space inside a bottle half full of soda water? Would that reading change with temperature?"
The sensor would read CO2 because like Justin and Andrew said, CO2 would come from the soda water and go into the air space and thats where the sensor is so it will read CO2.
TO SEAN: Sean, can you be more specific? Would the CO2 concentration increase, stay the same, or decrease?
Question #2: do you expect the mV to go up, or down, as the ASP-2 ascends to 10000 feet? (you looked at this question before, but it is time to think about it again!)
Like Mr.Kibler and Justin said,CO2 decreases as millivolts increases and CO2 increases as altitude increases. So as I understand it, mV would decrease as the ASP-2 ascends to 10,000 ft.
TO SEAN: In saying this, are you saying that CO2 increase with altitude? Is that your hypothesis? Also, does pressure affect CO2 concentration? If so, how? (*That's why Dr. Allen gave you a "CO2 sensor in a soda bottle" homework question. I think.)
Thanks for getting your homework questions answered promptly!
Sorry for my late answers. I had them written out since last night, but I was busy all day until now, and wasn't sure if we were going to answer all together in one document or alone. Here are my answers....
For Dr. Allen's question #1, " What do you suppose the sensor would read if you seal it in the air space inside a bottle half full of soda water? Would that reading change with temperature? "
**(I did not think of the pressure of the container being a factor, but more of evaporation being the main culprit)
I believe this has something to do with the common chemicals in the soda. Since CO2 is trapped in the soda by the carbonation, and the temperature is increased, causing evaporation, the CO2 trapped by the carbonation in the water, would evaporate causing the sensor in the air portion of the bottle to sense the CO2, causing mV to decrease...
I might be wrong on this one, I just researched Soda briefly to learn how it works, in both sweet, (coke) and water, (sparkling water) based solutions.
And for Dr. Allen's second question, "Do you expect the mV to go up, or down, as the ASP-2 ascends to 10000 feet? (you looked at this question before, but it is time to think about it again!)"
Well, looking back on our conversations, mV decrease as CO2 increase for the sensor's inner workings. And as I remember from our past conversations, CO2 concentrations decreased as altitude increases. So then in retrospect mV would increase as altitude increases.
The reason I asked about the maximum EMF is that the output from TP1 to your MCP3202 analog to digital converter cannot exceed 3.3 Volts. That is because of the fact that the Parallax #27929 CO2 sensor module has a 3.3 Volt power supply to run its amplifier. Therefore, if the expected signal would be greater than 3.3 Volts, the amplifier would "flat-top", or "saturate", kind of like turning up the volume control too far on your guitar amplifier--you get distortion and lose the true peaks of the signal. Unlike a guitar amplifier, distortion in a measurement system is not a desireable effect. If you have a maximum output of 2964 mV, then there is still head room before it hits the 3300 mV ceiling.
Also note that if you divide 2964 by 7.6667 (the gain of the amplifier on the #27929 module), the result is 386 mV. That is a bit higher than the value on the typical MG-811 graph (the one in Andrew's most recent post). The typical value would be around 320 mV at an ambient CO2 concentration of 390 ppm.
I have two MG-811 on running side by side, on on one BOE as I write this, and one of them has equilibrated to around 1100 mV and the other to 1000 mV after 1 hour of warmup, although they are still increasing gradually. I am puzzled, because that is quite low on the curve. Very far below the typical value. They do respond well to CO2.
By the way, it is possible to install the sensor backwards in the socket. The documentation does not really mention this, but there is a red dot on one side of the sensor, at least on the ones I have here. The red dot indicates the positive output side, and the red dot has to be on the side facing TP4 as you install the sensor in the socket. Remember that the sensor element is a little battery, and like any battery, it has a positive and a negative pole.
"Question #1: What do you suppose the sensor would read if you seal it in the air space inside a bottle half full of soda water? Would that reading change with temperature?"
The sensor would read CO2 because like Justin and Andrew said, CO2 would come from the soda water and go into the air space and thats where the sensor is so it will read CO2.
TO SEAN: Sean, can you be more specific? Would the CO2 concentration increase, stay the same, or decrease?
Question #2: do you expect the mV to go up, or down, as the ASP-2 ascends to 10000 feet? (you looked at this question before, but it is time to think about it again!)
Like Mr.Kibler and Justin said,CO2 decreases as millivolts increases and CO2 increases as altitude increases. So as I understand it, mV would decrease as the ASP-2 ascends to 10,000 ft.
TO SEAN: In saying this, are you saying that CO2 increase with altitude? Is that your hypothesis? Also, does pressure affect CO2 concentration? If so, how? (*That's why Dr. Allen gave you a "CO2 sensor in a soda bottle" homework question. I think.)
Thanks for getting your homework questions answered promptly!
Mr. Kibler
Mr.Kibler,
For Question 1: Pressure does affect CO2 concentrations but I can't seem to figure out how. I think that CO2 concentration would increase.
For Question 2: I was thinking that CO2 increases with altitude earlier, but now, I checked it , and CO2 decreases as altitude increases.
I won't be on the forum for the rest of the night, my mom is having a cookout for my birthday today . Just letting you all know.
"... the output from TP1 to your MCP3202 analog to digital converter cannot exceed 3.3 Volts... Also note that if you divide 2964 by 7.6667... the result is 386 mV... By the way, it is possible to install the sensor backwards in the socket."
Thanks for the helpful information Tracy. It will help us dial in the 'constant' we'll use in the final version of the program (if it really is a constant.) The data we got from this afternoon's trial showed the millivoltage stabilized at at 3315 (see attached graph... and EXCEL data!)
For some reason I knew someone was going to say that the sensor had a right and wrong pin orientation: "When in doubt, read the documentation and directions...!" I recalled that you asked about that before and then I remembered no one replied. Now we know. Thanks!
How do we 'calibrate' the sensor before flight? What do we use as a starting point, as a baseline? How do we use the data that the Rocketeers calculated for your previous homework assignment (below)? Where do we use the constant they calcualted previously? Is it valid since we're seeing variation between sensors? Could you explain how we change the program to make the potientiometer mV appear on the flsh drive too?
We've finished the hardware portion of the ASP for the most part. We still have a few minor things to do at Saturday's practice. I would like to spend most of it walking the Rocketeers through how to add program code for the (elusive!) constant, making a pre-flight checklist, and a post-flight data sheet and power-down sequence so we don't lose any data. This will be one of our last "build" sessions as a team.
The following weekend the Rocketeers are helping the Henniker Rotray Club serve at their annual "Chili Fest." They're one of our sponsors and the Chilli Fest is great fun! Thouands attend. There'e a chilli contest for amateurs and restaurants, a classic car show, a big cookout, lots of displays and crafts, etc. It's a small way to say 'thank you' to the rotary for their support, and an excellent way for the Rocketeers to be involved in a community service project. Rocketeers from (many) years gone by come each year to help. It's held at ski area... without snow! If it's like previous tears it'll be hot, especially parking car or behind the grill.
[The data we got from this afternoon's trial showed the millivoltage stabilized at at 3315 (see attached graph... and EXCEL data!)
Hmm. In the past your sensor was taking many minutes to settle in, and settled in at lower levels. Tracy's sensor acted like that as well, if I remember correctly.
But these data show it settling in at 3319 mV (that is, 3.319 V) in about 30 seconds (attached - a graph made from your Excel file). I suspect that it reached the full voltage available (which could certainly be a few 10s of mV over the nominal 3.3V level) and then stopped there, which means that you don't have that "overhead" that Tracy mentioned before.
There's no sign of any meaningful variation in these data either. Once it settles in, the values only vary from about 3309 mV to about 3317 mV, which sounds well within the margin of error of the system, and not a reflection of real CO2 level changes.
Does the sensor as set up still respond to CO2? Can you verify that?
QU: "In the past your sensor was taking many minutes to settle in, and settled in at lower levels. Tracy's sensor acted like that as well, if I remember correctly."
ANS: Sylvie, I didn't mention that the sensor had been running in another room-- warmed up-- before we put it in the vacuum chamber for the trial. It didn't really stabilize in 30 seconds; it was already warm.
QU: There's no sign of any meaningful variation in these data either. Once it settles in, the values only vary from about 3309 mV to about 3317 mV, which sounds well within the margin of error of the system, and not a reflection of real CO2 level changes.
ANS: Yes, there is still very little variation in the data... Hooray!
QU: Does the sensor as set up still respond to CO2? Can you verify that?
ANS: Yes, the sensor still responds very nicely to CO2 and the ASP responds well to temperature, humidity, and pressure altitude (see the attached graph.) WE didn't test for that while it was in the vacuum chamber though; we should do that. Susanne and I had the ASP running on the table on our patio yesterday when we ate breakfast outside. It was interesting to see the CO2 data fluctuate every time the wind blew even a slight breeze. That's a good sign, yes?!
MY QUESTION: How do we get the potientiometer voltage to read out on the flash drive? What do we add as far as program code? The data is there on the screen so we may as well record it as flight data, right?
QUESTION #2: So what about calibrating the sensor before flight, or adding the 'constant' into the program? Or should we simply record CO2 millivoltage and worry about that after the flights?
QUESTION #3: Are you and Tracy joining us in Nevada for the launch? We'll put you up in a hotel (you get to room with me!), feed you, and even chauffeur you around (we'll have a 15-passenger van.)
We might even speak to the students at Gerlach High School again this year. It would be really cool to have the whole project team there so everone has a full understand and appreciation of the scope of the project. Two of the Rocketeers' parents are coming on the trip, too...!
ANS: Yes, the sensor still responds very nicely to CO2 and the ASP responds well to temperature, humidity, and pressure altitude (see the attached graph.) WE didn't test for that while it was in the vacuum chamber though; we should do that. Susanne and I had the ASP running on the table on our patio yesterday when we ate breakfast outside. It was interesting to see the CO2 data fluctuate every time the wind blew even a slight breeze. That's a good sign, yes?!
Ok, good. Phew. You had me worried. I do still wonder about the level of those readings, though. 3319 mV is right up there at the top of the scale. Tracy?
MY QUESTION: How do we get the potientiometer voltage to read out on the flash drive? What do we add as far as program code? The data is there on the screen so we may as well record it as flight data, right?
I'm not sure why. The voltage on that point is simply a function of how you have the potentiometer set, and not of anything about the environment in which the rocket is flying. I suppose it might tell you something about how the sensor is working, but you have no reason to expect it to change in flight, right?
QUESTION #2: So what about calibrating the sensor before flight, or adding the 'constant' into the program? Or should we simply record CO2 millivoltage and worry about that after the flights?
Well, that's the $64,000 question. I think that Tracy is still working on that part. The issue is the fact that the three sensors we're looking at give such different output, and that they don't correspond neatly with what's in the datasheet. I suspect that means that each sensor has to be "characterized" separately, and calibrated to some known values, as you suggested in that long list of mV and ppm figures you posted a little further up. I could say more, but it'd probably just muddy the waters - this is really Tracy's specialty again.
QUESTION #3: Are you and Tracy joining us in Nevada for the launch? We'll put you up in a hotel (you get to room with me!), feed you, and even chauffeur you around (we'll have a 15-passenger van.)
We might even speak to the students at Gerlach High School again this year. It would be really cool to have the whole project team there so everone has a full understand and appreciation of the scope of the project. Two of the Rocketeers' parents are coming on the trip, too...!
Much as I'd love to, and much as I'd seriously considered it, I just don't think I can spare the time or the money right now (yes, even with how little it would cost). I've been away quite a bit lately, and I'm going to have to be a bit more responsible with respect to indulging my whims, at least for the fall.
With regard to the warmup issue, I am attaching two graphs where the heater control was set up to turn off and on in a cycle. Off for 50 seconds, and On for 950 seconds, then Off for 50 and On for 950, etc. etc. After a long rest, it does take a longer time to equilibrate, and it overshoots, but in subsequent Off-On cycles it comes pretty close to equilibration in one minute. The second graph is a detail from the first over a shorter time interval.
Note the difference between these two sensors, and the relatively low resting output level, around 1 Volt. I have confirmed the readings with a multimeter. Compare with the typical value which would be around 2.5 Volts or with the close to 3.3 Volts that Mark found in the recent ASP-2 experiment .
Yes, those resting readings from the altitude experiment are right up there near the limit at the top of the scale. The mV reading does not have any room to increase. My prediction is that the mV reading should increase substantially as altitude increases and pressure decreases. The vacuum chamber experiment cannot show if that happened or not. If it is already at 3.3 Volts, it has nowhere to go in the up direction.
How many MG811 sensors do you have now? Please plan to test each of them for a few minutes at or before your Saturday meeting. I'd like a few more data points for ambient mV levels. It might be a good idea to run them all in the same circuit with the same software.
Rocketeers, there is a distinction between parts per million (ppm) and partial pressure (pp). Do you understand that distinction? I think the mV output of the sensor is primarily a function of partial pressure, and secondarily a function of parts per million. Do you understand what I mean by that hypothesis and how it could apply to the ASP-2?
I agree with Sylvie, why would you want to record the potentiometer voltage? It has nothing at all to do with the reading from TP1. Is the mV reading currently recording okay on the ASP-2 flash drive?
My feeling is that we can make a stab at translating the mV reading to CO2 partial pressure, but if we do that in the data stream, we should also record the raw mV in a separate column. We need to work that out before your Saturday meeting. There are still sooooooooo many uncertainties in the sensor performance.
My feeling is that we can make a stab at translating the mV reading to CO2 partial pressure, but if we do that in the data stream, we should also record the raw mV in a separate column. We need to work that out before your Saturday meeting. There are still sooooooooo many uncertainties in the sensor performance.
Given that last point, I agree that you should plan to record the mV output of the sensor as part of your data stream. If you don't do that, you'll forever lose the raw data that might give you something to work with if the rest of what you do doesn't work out as planned. It's an easy safeguard. It also makes it far more likely that you'll have a unit ready to fly a month or so from now, when everything has to be ready.
I wonder if there are other folks out there somewhere on the internet who have experience with this sensor, and can provide some help. This might be a good time to do a little Google searching, and maybe some communicating.
How many MG811 sensors do you have now? Please plan to test each of them for a few minutes at or before your Saturday meeting. I'd like a few more data points for ambient mV levels.
Tracy,
We'll have $eve$ $ensor$, and four modules when I open the package that just arrived from Parallax.
TRACY: I agree with Sylvie, why would you want to record the potentiometer voltage?
I suppose the only reason was, "The data is there; why not record it for posterity in case we would want to refer to it later?" That was the only reason. Data for the sake of data.
TRACY: "... we should also record the raw mV in a separate column. We need to work that out before your Saturday meeting. There are still sooooooooo many uncertainties in the sensor performance.
Yes, there seem to be a few uncertainties. We'll begin by running trials on all the sensors and then recording the data for each. Does that sound like a good starting point? I agree with Sylvie's suggestion that we continue to record the raw mV data from the CO2 sensor in its own column. Maybe we could also add a partial pressure column once we get the constant figured out, or the sensor that we're going to use characterized...? What are your thoughts?
Tomatoe sandwiches for dinner, with tomatoes from the garden! Trials after tomatoes.
sylvie369 said: Given that last point, I agree that you should plan to record the mV output of the sensor as part of your data stream. If you don't do that, you'll forever lose the raw data... It's an easy safeguard.
Sylvie,
Agreed! It makes good sense to keep then original raw data intact. As I said in my post to Tracy, maybe we could also add a column for CO2 partial pressure, or not. What are your thoughts? I defer to you and Tracy on this decision.
Bummer that you can't join us at Black Rock for the launch... It'd be great to connect with you and meet you. Another place, another time maybe... (no icon for crocodile tears.)
We're run trials on the different sensors after dinner and we'll post the data and results. Tracy did note that there's a "right" and "wrong" way to install the sensor in the module: "Red dot goes toward TP4." I hope that I don't have it in backwards... not that we've ever crossed wires or anything! (*I've already checked it and it's installed correctly. Not to worry.)
Attached is a graph comparing six different CO2 sensors. It's easy to see there's variation between the sensors. :rolleyes: All six (sensors #1-5 and #9) had been previously tested and seemed to get up to "flat line" stabilization millivoltage within the 5-minute test period. Sensor modules #7-9 just arrived today and the millivoltage seemed to climb more slowly toward stabilization; they are not included. Sensor #7 is 21 minutes into its very first trial and the millivoltage seems to be stable around 3318.
NOTES:
1) The red dot on both sensors #3 and #4 was exactly halfway between both sets of three pins...! What a conundrum when it came time to plug them into the module.
2) Sensor #5 has one pin that's only 2/3 the length of the other five pins.
And the moral of the story is... there's variation between the sensors. Now, how do we use that information most effectively?
Tracy and Sylvie (I'm still crying crocodile tears),
Here's the graph for one of the three new CO2 sensors, sensor #6. I was premature in saying the millivoltage had stabilized at 3318 mV after 21 minutes. It dropped and then stabilized in the mid-2500 mV range (see graph.) This seems to support the conclusion that there's significant (?) variation between sensors. Maybe they all should have been tested for 60+ minutes...?
To add one more data column in the program, make the following changes:
CO2mV VAR Word ' one more variable
' edit the following line in the V_write_record subroutine.
SEROUT TX\CTS, vBaud, [ $8,$20,0,0,0,41,CR, HEX2 hour, ":" , HEX2 minute, ":", HEX2 second, "," , degC.BIT15*13+32,DEC3 ABS (degC/10),".", DEC1 ABS degC,",",
DEC3 (RH/10),".",DEC1 RH,",", DEC5 feet,",",DEC5 CO2pp, ",",DEC5 CO2mV,CR, LF]
' note that this adds one comma and one 5 digit number,
' and bumps the total number of bytes per record up from 35 to 41.
' In the MCP_get subroutine, add the following statement
' just after the ADC acquires the mV reading from the potentiometer
CO2mV=CO2pp
Later on we can use CO2mV to store the raw mV and CO2pp to store the estimated CO2 partial pressure.
The most recent version of the program that I see is the one called "ASP-2 PROGRAM - August 4, 2010.txt", that Mark posted here on 08-05-2010, 02:08 PM. Is that the most recent version?
Thanks for contributing that graph with multiple sensors. That is great to have. I haven't seen any such comparison data on line, and as a matter of fact, not much statement of if and how people are using this sensor.
Attached is a graph comparing six different CO2 sensors. It's easy to see there's variation between the sensors.
Wow. I think I see why Parallax built these sensors into simple "alarm" devices rather than introducing devices intended to give you a true reading of CO2 level. There's a 2:1 ratio between one sensor and another, in the most extreme case.
Maybe you should choose the sensor that responds most like the one in the datasheet, and make sure it reads consistently, and work with that? I'm at a loss otherwise.
Comparison graph for seven CO2 sensors, with our new sensor #6 added. The average stabiization level seems to be somewhere around 2500 mV. That's where the graph libes for most of the sensors are clustered.
============================
Tracy, thanks for the program code! As Andrew says, "We'll 'mash it in'" tomorrow and see how it works.
Sylvie, yes, now I see why the CO2 sensor has an alarm, too..!
I think the one chosen for launch should be one that stabilizes to a nice, flat line with minimal noise and drift. For example, #9 seems to be kind of noisy and is still drifting downward after 4 minutes. It might stabilize after a while longer. #6 and #4 look good.
"ASP-2 PROGRAM - August 4, 2010.txt",
posted on 08-05-2010, 02:08 PM
Is that up to date?
If you get a chance, please try the vacuum chamber experiment again using sensor #3 or sensor #5.
I agree with Dr. Allen's thoughts. I think that those drops could be from events that we are not noticing. Like cracking a window open, or just having a pet enter the room, or just looking over the sensor. These all could increase or decrease the mV count, since all can change the environment. This, like Dr. Allen said, would be best a situation for a controlled environment.
_________________________
I have not been posting as much as I have in the past because of a recent camp I have enrolled in. It is the Launch Robotics Engineering program at WPI in MA. We are now dealing with VEX and easyCpro. It turns out that I have the most in depth robotics experience in the class, due to ARLISS. I really appreciate all of your help because everyone here has made that possible for me!
What a wonderful opportunity to be able to participate in a technology and robotics summer camp at WPI. It's validating to hear that your experience with the ARLISS project helped you get even more out of your techno-camp experience. That's why Dr. Allen, Sylvie and I do projects like this, to help guide bright young students like you in the right direction. I think you're beginning to see the connection...! Are you thinking about applying to WPI, Cal-Tech, Carnegie-Mellon, and MIT? You should. They would be lucky to have you as a student. Keep up the good work! You've really impressed me with your participation this year.
See you on Saturday; a newspaper reporter is coming to interview you and your teammates.
I've attached the most recent version of the ASP-2 program ("August 8") , the version that was used to run the ASP and generate all the previous graphs and data.
Here is a homework assignment I would like you complete, the sooner the better:
1) Add the new program code that Dr. Allen just posted for us into the attached program.
2) After each line of new code briefly explain what you did, when you changed the program, and who changed it (your name or initials.)
3) Comment out-- *don't delete*-- any line(s) of program code that are not used or needed.
4) Upload the updated program to the forum so we can compare your program to everyone else's, and so we can test it on the ASP-2.
5) Name the new program "ASP-2 test program (your name) - August 11, 2010"
The first Rocketeer to upload a fully operational program, one that does what the new changes are supposed to make it do, gets extra homemade cookies at Saturday's practice (and Mrs. Kibler makes BIG cookies!)
6) Explain briefly what the new program code will make the ASP do differently.
To add one more data column in the program, make the following changes:
CO2mV VAR Word ' one more variable
' edit the following line in the V_write_record subroutine.
SEROUT TX\CTS, vBaud, [ $8,$20,0,0,0,41,CR, HEX2 hour, ":" , HEX2 minute, ":", HEX2 second, "," , degC.BIT15*13+32,DEC3 ABS (degC/10),".", DEC1 ABS degC,",",
DEC3 (RH/10),".",DEC1 RH,",", DEC5 feet,",",DEC5 CO2pp, ",",DEC5 CO2mV,CR, LF]
' note that this adds one comma and one 5 digit number,
' and bumps the total number of bytes per record up from 35 to 41.
' In the MCP_get subroutine, add the following statement
' just after the ADC acquires the mV reading from the potentiometer
CO2mV=CO2pp
Later on we can use CO2mV to store the raw mV and CO2pp to store the estimated CO2 partial pressure.
The most recent version of the program that I see is the one called "ASP-2 PROGRAM - August 4, 2010.txt", that Mark posted here on 08-05-2010, 02:08 PM. Is that the most recent version?
Thanks for contributing that graph with multiple sensors. That is great to have. I haven't seen any such comparison data on line, and as a matter of fact, not much statement of if and how people are using this sensor.
Just tested the next-to-last new sensor we got and the CO2 millivoltage stabilizes around 2500 mV like four of the other CO2 sensors. It is responsive to CO2 (attached graph #1.)
We also made a make-believe "CO2 spike" graph by using the inverse of the data points (attached graph #2)
Tracy, when the trial started the "other" millivoltage showed slight interference. When we changed the BS2 stamp the millivoltage re-stabilzed with minimal (3 units variation.) Why would that be?
We also added the program code you wrote but the stamp/BOE/datalogger didn't like it. The datalogger wouldn't initialize until we first ran the "August 8" version of the program, turned the BOE off, then ran the program with the new code. The millivoltage did appear on the datalogger. We'll come back at it again tomorrow and start from the beginning with your instructions. We were back and forth at the computer today, "distracted" by yard mowing, a trip to the grocery, etc. We'll stay focused tomorrow...!
Good evening,
Mark and Chris, and Bandit the Rocket Dog, and Poochie the Cat
If you get a chance, please try the vacuum chamber experiment again using sensor #3 or sensor #5.
Dr. Allen,
Andrew here. I just tested both sensor #3 and sensor #5 in the vacuum chamber as you have asked. Attached are the Excel files/graphs for both experiments. However, I'm rather curious of why you have asked us to specifically test sensors #3 and #5. We'll mash the new program together today and report back to you.
Interesting--There was no response from the CO2 sensor, but a very noticeable response from the RH sensor. From the standpoint of theory, I thought the sensor would respond directly to the partial pressure of CO2. If you read the data sheet, that is how the theory is presented, but then the graph is in terms of parts per million. The two units of measure can be directly related when the measurement is made at standard atmospheric pressure at the earth's surface. Humidity is a similar kind of measurement. There is a complicated relationship (psychometric) between what the sensor measures and units of pressure, water content, and relative humidity. The humidity reading changes as the pressure goes down, and I was expecting a similar change in the CO2 reading. So I am puzzled. But scientific insight depends on puzzles, right?!
============
Thanks for the invitation to participate in the launch. It would indeed be great to meet you all in person and to see the fruits of your (our) labors rocket skyward. Looking at my calendar, I'm sorry to say that won't work out for me either. There is too much business I will have to attend to that week, and as the proprietor, the buck stops with me.
The reason I asked for those two is that they are outliers on the low side, so the reading has lots of room to increase. For the actual flight, you can choose one of the sensors like #6 or #4 that is near the center of the distribution around 2.5 Volts resting.
Attached is my version of the program with my corrections made. I put *** every time I made a change so it was a little easier to find. Sorry for such a late reply, I got my braces on today, and I was at the doctor's office for 3 hours.
I'm having trouble downloading your attachment. I click on it, and it asks me if I want to save it. I say yes, and it asks me where to save it to. I do that, but then it says that it doesn't know what program to use it under. It has 2 options: 1: use a Web service to find it, or 2: pick a program from a list of installed programs. I tried both, and when I tried the first one, it came up having no results. I tried the second one, and it didn't have the program I needed. I'm really confused, since i thought it was as easy as downloading it and plugging it in, but I guess it's not that simple.
I'm having trouble downloading your attachment. I click on it, and it asks me if I want to save it. I say yes, and it asks me where to save it to. I do that, but then it says that it doesn't know what program to use it under.
Justin
Justin,
This means you haven't downloaded the free Parallax Stamp Editor software from the Parallax website yet (www.parallax.com). You need the software to program and operate the Parallax BOE/breadboard kit I got for you and each of your teammates.
Download the Stamp Editor software, then download my attachment and let me know what happens. I'll look for your reply.
Comments
"From the raw data, can you tell us if it is 3000 mV or closer to 3300 mV? It is hard to read from the graph."
Tracy,
The maximum millivoltage for the trial for the attached graph #1 was 2960 at 56 seconds, 2962 at 58 and 61 seconds, and 2964 at 63 seconds. At the point it dropped, maybe in response to turn on the vacuum pump (?)
It's important to note:
1) The trial for the data above was only 3:24 (minutes and seconds) and it was conducted inside a vacuum chamber.
2) In previous trials where we got 'clean' data, trials NOT done in the vacuum chamber, the average maximum millivoltage was typically between 3000 and 3500 mV after the heater had warmed up and the millivoltage had stabilized (see attached graph #2.)
Can you help me understand why this is significant?
Also, I have a nagging question, probably with a simple answer. When we solder resistors in-line with some of the wires, why do we use a 1 kilo-ohm resistor? Why not a 1/2 ohm, or 2 ohm resistor? What's the "magic" in using a 1 K-ohm resistor?
=============================
The Rocketeers are doing awesome work, aren't they? And we sure were glad, at 1:30 AM last night, that the glitch with the MAWD was simply two reversed wires. That's what happens when you work late, and when you're tired. It's not the best time to do quality work. I hope the Rocketeers will remember this lesson if they're thinking about staying up all night to study for an important exam the next day... not a wise decision!
Best regards from the patio,
Mark
"Question #1: What do you suppose the sensor would read if you seal it in the air space inside a bottle half full of soda water? Would that reading change with temperature?"
The sensor would read CO2 because like Justin and Andrew said, CO2 would come from the soda water and go into the air space and thats where the sensor is so it will read CO2.
TO SEAN: Sean, can you be more specific? Would the CO2 concentration increase, stay the same, or decrease?
Question #2: do you expect the mV to go up, or down, as the ASP-2 ascends to 10000 feet? (you looked at this question before, but it is time to think about it again!)
Like Mr.Kibler and Justin said,CO2 decreases as millivolts increases and CO2 increases as altitude increases. So as I understand it, mV would decrease as the ASP-2 ascends to 10,000 ft.
TO SEAN: In saying this, are you saying that CO2 increase with altitude? Is that your hypothesis? Also, does pressure affect CO2 concentration? If so, how? (*That's why Dr. Allen gave you a "CO2 sensor in a soda bottle" homework question. I think.)
Thanks for getting your homework questions answered promptly!
Mr. Kibler
For Dr. Allen's question #1, " What do you suppose the sensor would read if you seal it in the air space inside a bottle half full of soda water? Would that reading change with temperature? "
**(I did not think of the pressure of the container being a factor, but more of evaporation being the main culprit)
I believe this has something to do with the common chemicals in the soda. Since CO2 is trapped in the soda by the carbonation, and the temperature is increased, causing evaporation, the CO2 trapped by the carbonation in the water, would evaporate causing the sensor in the air portion of the bottle to sense the CO2, causing mV to decrease...
I might be wrong on this one, I just researched Soda briefly to learn how it works, in both sweet, (coke) and water, (sparkling water) based solutions.
And for Dr. Allen's second question, "Do you expect the mV to go up, or down, as the ASP-2 ascends to 10000 feet? (you looked at this question before, but it is time to think about it again!)"
Well, looking back on our conversations, mV decrease as CO2 increase for the sensor's inner workings. And as I remember from our past conversations, CO2 concentrations decreased as altitude increases. So then in retrospect mV would increase as altitude increases.
Also note that if you divide 2964 by 7.6667 (the gain of the amplifier on the #27929 module), the result is 386 mV. That is a bit higher than the value on the typical MG-811 graph (the one in Andrew's most recent post). The typical value would be around 320 mV at an ambient CO2 concentration of 390 ppm.
I have two MG-811 on running side by side, on on one BOE as I write this, and one of them has equilibrated to around 1100 mV and the other to 1000 mV after 1 hour of warmup, although they are still increasing gradually. I am puzzled, because that is quite low on the curve. Very far below the typical value. They do respond well to CO2.
By the way, it is possible to install the sensor backwards in the socket. The documentation does not really mention this, but there is a red dot on one side of the sensor, at least on the ones I have here. The red dot indicates the positive output side, and the red dot has to be on the side facing TP4 as you install the sensor in the socket. Remember that the sensor element is a little battery, and like any battery, it has a positive and a negative pole.
Mr.Kibler,
For Question 1: Pressure does affect CO2 concentrations but I can't seem to figure out how. I think that CO2 concentration would increase.
For Question 2: I was thinking that CO2 increases with altitude earlier, but now, I checked it , and CO2 decreases as altitude increases.
I won't be on the forum for the rest of the night, my mom is having a cookout for my birthday today
Sean
Mr. Kibler
"... the output from TP1 to your MCP3202 analog to digital converter cannot exceed 3.3 Volts... Also note that if you divide 2964 by 7.6667... the result is 386 mV... By the way, it is possible to install the sensor backwards in the socket."
Thanks for the helpful information Tracy. It will help us dial in the 'constant' we'll use in the final version of the program (if it really is a constant.) The data we got from this afternoon's trial showed the millivoltage stabilized at at 3315 (see attached graph... and EXCEL data!)
For some reason I knew someone was going to say that the sensor had a right and wrong pin orientation: "When in doubt, read the documentation and directions...!" I recalled that you asked about that before and then I remembered no one replied. Now we know. Thanks!
How do we 'calibrate' the sensor before flight? What do we use as a starting point, as a baseline? How do we use the data that the Rocketeers calculated for your previous homework assignment (below)? Where do we use the constant they calcualted previously? Is it valid since we're seeing variation between sensors? Could you explain how we change the program to make the potientiometer mV appear on the flsh drive too?
ppm mV
200
300
400 324
500 318
600
700
800
900
1000
1500
2000
3000
4000 281
5000 277
6000
7000
8000
9000
100000
========================
We've finished the hardware portion of the ASP for the most part. We still have a few minor things to do at Saturday's practice. I would like to spend most of it walking the Rocketeers through how to add program code for the (elusive!) constant, making a pre-flight checklist, and a post-flight data sheet and power-down sequence so we don't lose any data. This will be one of our last "build" sessions as a team.
The following weekend the Rocketeers are helping the Henniker Rotray Club serve at their annual "Chili Fest." They're one of our sponsors and the Chilli Fest is great fun! Thouands attend. There'e a chilli contest for amateurs and restaurants, a classic car show, a big cookout, lots of displays and crafts, etc. It's a small way to say 'thank you' to the rotary for their support, and an excellent way for the Rocketeers to be involved in a community service project. Rocketeers from (many) years gone by come each year to help. It's held at ski area... without snow! If it's like previous tears it'll be hot, especially parking car or behind the grill.
Mark
Hmm. In the past your sensor was taking many minutes to settle in, and settled in at lower levels. Tracy's sensor acted like that as well, if I remember correctly.
But these data show it settling in at 3319 mV (that is, 3.319 V) in about 30 seconds (attached - a graph made from your Excel file). I suspect that it reached the full voltage available (which could certainly be a few 10s of mV over the nominal 3.3V level) and then stopped there, which means that you don't have that "overhead" that Tracy mentioned before.
There's no sign of any meaningful variation in these data either. Once it settles in, the values only vary from about 3309 mV to about 3317 mV, which sounds well within the margin of error of the system, and not a reflection of real CO2 level changes.
Does the sensor as set up still respond to CO2? Can you verify that?
QU: "In the past your sensor was taking many minutes to settle in, and settled in at lower levels. Tracy's sensor acted like that as well, if I remember correctly."
ANS: Sylvie, I didn't mention that the sensor had been running in another room-- warmed up-- before we put it in the vacuum chamber for the trial. It didn't really stabilize in 30 seconds; it was already warm.
QU: There's no sign of any meaningful variation in these data either. Once it settles in, the values only vary from about 3309 mV to about 3317 mV, which sounds well within the margin of error of the system, and not a reflection of real CO2 level changes.
ANS: Yes, there is still very little variation in the data... Hooray!
QU: Does the sensor as set up still respond to CO2? Can you verify that?
ANS: Yes, the sensor still responds very nicely to CO2 and the ASP responds well to temperature, humidity, and pressure altitude (see the attached graph.) WE didn't test for that while it was in the vacuum chamber though; we should do that. Susanne and I had the ASP running on the table on our patio yesterday when we ate breakfast outside. It was interesting to see the CO2 data fluctuate every time the wind blew even a slight breeze. That's a good sign, yes?!
MY QUESTION: How do we get the potientiometer voltage to read out on the flash drive? What do we add as far as program code? The data is there on the screen so we may as well record it as flight data, right?
QUESTION #2: So what about calibrating the sensor before flight, or adding the 'constant' into the program? Or should we simply record CO2 millivoltage and worry about that after the flights?
QUESTION #3: Are you and Tracy joining us in Nevada for the launch? We'll put you up in a hotel (you get to room with me!), feed you, and even chauffeur you around (we'll have a 15-passenger van.)
We might even speak to the students at Gerlach High School again this year. It would be really cool to have the whole project team there so everone has a full understand and appreciation of the scope of the project. Two of the Rocketeers' parents are coming on the trip, too...!
Mark
Ok, good. Phew. You had me worried. I do still wonder about the level of those readings, though. 3319 mV is right up there at the top of the scale. Tracy?
I'm not sure why. The voltage on that point is simply a function of how you have the potentiometer set, and not of anything about the environment in which the rocket is flying. I suppose it might tell you something about how the sensor is working, but you have no reason to expect it to change in flight, right?
Well, that's the $64,000 question. I think that Tracy is still working on that part. The issue is the fact that the three sensors we're looking at give such different output, and that they don't correspond neatly with what's in the datasheet. I suspect that means that each sensor has to be "characterized" separately, and calibrated to some known values, as you suggested in that long list of mV and ppm figures you posted a little further up. I could say more, but it'd probably just muddy the waters - this is really Tracy's specialty again.
Much as I'd love to, and much as I'd seriously considered it, I just don't think I can spare the time or the money right now (yes, even with how little it would cost). I've been away quite a bit lately, and I'm going to have to be a bit more responsible with respect to indulging my whims, at least for the fall.
Note the difference between these two sensors, and the relatively low resting output level, around 1 Volt. I have confirmed the readings with a multimeter. Compare with the typical value which would be around 2.5 Volts or with the close to 3.3 Volts that Mark found in the recent ASP-2 experiment .
How many MG811 sensors do you have now? Please plan to test each of them for a few minutes at or before your Saturday meeting. I'd like a few more data points for ambient mV levels. It might be a good idea to run them all in the same circuit with the same software.
Rocketeers, there is a distinction between parts per million (ppm) and partial pressure (pp). Do you understand that distinction? I think the mV output of the sensor is primarily a function of partial pressure, and secondarily a function of parts per million. Do you understand what I mean by that hypothesis and how it could apply to the ASP-2?
I agree with Sylvie, why would you want to record the potentiometer voltage? It has nothing at all to do with the reading from TP1. Is the mV reading currently recording okay on the ASP-2 flash drive?
My feeling is that we can make a stab at translating the mV reading to CO2 partial pressure, but if we do that in the data stream, we should also record the raw mV in a separate column. We need to work that out before your Saturday meeting. There are still sooooooooo many uncertainties in the sensor performance.
Given that last point, I agree that you should plan to record the mV output of the sensor as part of your data stream. If you don't do that, you'll forever lose the raw data that might give you something to work with if the rest of what you do doesn't work out as planned. It's an easy safeguard. It also makes it far more likely that you'll have a unit ready to fly a month or so from now, when everything has to be ready.
I wonder if there are other folks out there somewhere on the internet who have experience with this sensor, and can provide some help. This might be a good time to do a little Google searching, and maybe some communicating.
Sylvie,
Agreed! It makes good sense to keep then original raw data intact. As I said in my post to Tracy, maybe we could also add a column for CO2 partial pressure, or not. What are your thoughts? I defer to you and Tracy on this decision.
Bummer that you can't join us at Black Rock for the launch...
We're run trials on the different sensors after dinner and we'll post the data and results. Tracy did note that there's a "right" and "wrong" way to install the sensor in the module: "Red dot goes toward TP4." I hope that I don't have it in backwards... not that we've ever crossed wires or anything! (*I've already checked it and it's installed correctly. Not to worry.)
Mark
Attached is a graph comparing six different CO2 sensors. It's easy to see there's variation between the sensors. :rolleyes: All six (sensors #1-5 and #9) had been previously tested and seemed to get up to "flat line" stabilization millivoltage within the 5-minute test period. Sensor modules #7-9 just arrived today and the millivoltage seemed to climb more slowly toward stabilization; they are not included. Sensor #7 is 21 minutes into its very first trial and the millivoltage seems to be stable around 3318.
NOTES:
1) The red dot on both sensors #3 and #4 was exactly halfway between both sets of three pins...! What a conundrum when it came time to plug them into the module.
2) Sensor #5 has one pin that's only 2/3 the length of the other five pins.
And the moral of the story is... there's variation between the sensors. Now, how do we use that information most effectively?
Mark
Here's the graph for one of the three new CO2 sensors, sensor #6. I was premature in saying the millivoltage had stabilized at 3318 mV after 21 minutes. It dropped and then stabilized in the mid-2500 mV range (see graph.) This seems to support the conclusion that there's significant (?) variation between sensors. Maybe they all should have been tested for 60+ minutes...?
Mark
Later on we can use CO2mV to store the raw mV and CO2pp to store the estimated CO2 partial pressure.
The most recent version of the program that I see is the one called "ASP-2 PROGRAM - August 4, 2010.txt", that Mark posted here on 08-05-2010, 02:08 PM. Is that the most recent version?
Thanks for contributing that graph with multiple sensors. That is great to have. I haven't seen any such comparison data on line, and as a matter of fact, not much statement of if and how people are using this sensor.
Wow. I think I see why Parallax built these sensors into simple "alarm" devices rather than introducing devices intended to give you a true reading of CO2 level. There's a 2:1 ratio between one sensor and another, in the most extreme case.
Maybe you should choose the sensor that responds most like the one in the datasheet, and make sure it reads consistently, and work with that? I'm at a loss otherwise.
============================
Tracy, thanks for the program code! As Andrew says, "We'll 'mash it in'" tomorrow and see how it works.
Sylvie, yes, now I see why the CO2 sensor has an alarm, too..!
Yawn!
Mark
"ASP-2 PROGRAM - August 4, 2010.txt",
posted on 08-05-2010, 02:08 PM
Is that up to date?
If you get a chance, please try the vacuum chamber experiment again using sensor #3 or sensor #5.
_________________________
I have not been posting as much as I have in the past because of a recent camp I have enrolled in. It is the Launch Robotics Engineering program at WPI in MA. We are now dealing with VEX and easyCpro. It turns out that I have the most in depth robotics experience in the class, due to ARLISS. I really appreciate all of your help because everyone here has made that possible for me!
What a wonderful opportunity to be able to participate in a technology and robotics summer camp at WPI. It's validating to hear that your experience with the ARLISS project helped you get even more out of your techno-camp experience. That's why Dr. Allen, Sylvie and I do projects like this, to help guide bright young students like you in the right direction. I think you're beginning to see the connection...! Are you thinking about applying to WPI, Cal-Tech, Carnegie-Mellon, and MIT? You should. They would be lucky to have you as a student. Keep up the good work! You've really impressed me with your participation this year.
See you on Saturday; a newspaper reporter is coming to interview you and your teammates.
Aim high,
Mr. Kibler
I've attached the most recent version of the ASP-2 program ("August 8") , the version that was used to run the ASP and generate all the previous graphs and data.
Here is a homework assignment I would like you complete, the sooner the better:
1) Add the new program code that Dr. Allen just posted for us into the attached program.
2) After each line of new code briefly explain what you did, when you changed the program, and who changed it (your name or initials.)
3) Comment out-- *don't delete*-- any line(s) of program code that are not used or needed.
4) Upload the updated program to the forum so we can compare your program to everyone else's, and so we can test it on the ASP-2.
5) Name the new program "ASP-2 test program (your name) - August 11, 2010"
The first Rocketeer to upload a fully operational program, one that does what the new changes are supposed to make it do, gets extra homemade cookies at Saturday's practice (and Mrs. Kibler makes BIG
6) Explain briefly what the new program code will make the ASP do differently.
Have fun, now go get 'em!
Mr. Kibler
We also made a make-believe "CO2 spike" graph by using the inverse of the data points (attached graph #2)
Tracy, when the trial started the "other" millivoltage showed slight interference. When we changed the BS2 stamp the millivoltage re-stabilzed with minimal (3 units variation.) Why would that be?
We also added the program code you wrote but the stamp/BOE/datalogger didn't like it. The datalogger wouldn't initialize until we first ran the "August 8" version of the program, turned the BOE off, then ran the program with the new code. The millivoltage did appear on the datalogger. We'll come back at it again tomorrow and start from the beginning with your instructions. We were back and forth at the computer today, "distracted" by yard mowing, a trip to the grocery, etc. We'll stay focused tomorrow...!
Good evening,
Mark and Chris, and Bandit the Rocket Dog, and Poochie the Cat
Dr. Allen,
Andrew here. I just tested both sensor #3 and sensor #5 in the vacuum chamber as you have asked. Attached are the Excel files/graphs for both experiments. However, I'm rather curious of why you have asked us to specifically test sensors #3 and #5. We'll mash the new program together today and report back to you.
Thanks again for your assistance,
Andrew
============
Thanks for the invitation to participate in the launch. It would indeed be great to meet you all in person and to see the fruits of your (our) labors rocket skyward. Looking at my calendar, I'm sorry to say that won't work out for me either. There is too much business I will have to attend to that week, and as the proprietor, the buck stops with me.
The reason I asked for those two is that they are outliers on the low side, so the reading has lots of room to increase. For the actual flight, you can choose one of the sensors like #6 or #4 that is near the center of the distribution around 2.5 Volts resting.
Attached is my version of the program with my corrections made. I put *** every time I made a change so it was a little easier to find. Sorry for such a late reply, I got my braces on today, and I was at the doctor's office for 3 hours.
Sean
I'm having trouble downloading your attachment. I click on it, and it asks me if I want to save it. I say yes, and it asks me where to save it to. I do that, but then it says that it doesn't know what program to use it under. It has 2 options: 1: use a Web service to find it, or 2: pick a program from a list of installed programs. I tried both, and when I tried the first one, it came up having no results. I tried the second one, and it didn't have the program I needed. I'm really confused, since i thought it was as easy as downloading it and plugging it in, but I guess it's not that simple.
Justin
Justin,
This means you haven't downloaded the free Parallax Stamp Editor software from the Parallax website yet (www.parallax.com). You need the software to program and operate the Parallax BOE/breadboard kit I got for you and each of your teammates.
Download the Stamp Editor software, then download my attachment and let me know what happens. I'll look for your reply.
Thanks,
Mr. Kibler