Power Requirements for Arduino, various parts
From BenningtonWiki
Telemetry
Hannah Wolfe and I were tasked with coming up with rough estimates of how much power is going to be needed to run these things for a respectable amount of time. Here are the results as of yet, all secondary sources; I have yet to run any of my own tests, but I'll try just seeing some rough estimates with a AA battery.
Contents |
[edit] Research
[[1]] This link from Rob Faludi provides us with a whole bunch of real world tests from the first new generation iteration of the Arduino unit; it would be surprising if our power requirements for each of these tests cost us any more or less power than his tests, as many of the parts are exactly the same, or practically the same in terms of power draws.
[[2]] According to this site here, the GPS module that we will be using draws about 70 milliAmps when being used. So, if we wanted to use it for one second, 0.0194444444 mAh would be used, I think. I'm not good with electrics at all, but this seems right. Either way, we're looking at using 70mA with the gps.
I used a Razr as a good example cell phone, as it's these days pretty middle of the bottom of the road. I was able to figure out that with a 680 mAh battery, we are given about 6.5 hours of talk time, which translates into roughly. 120 mA used when talking.
Arduino unit: site says that it uses 40mA max per port, but it really depends how we write everything; this could turn out to be the largest drain on power. on Faludi's blog, I noticed one test that used 4AA NiMH 2500 mAh as the power source; with that, his program ran 178 hours, 13 minutes. The program was a smallish size, but was still pretty power-draining: "ATMEGA1688 microcontroller transmitting one serial byte every 500 ms at 9600 baud." So, lets just round off initially until we have tests at about 100mA liberally for usage.
So, totalling all of these things, we can start to get a rough idea of our total power draw for the project:
- Arduino: 100mA
- Cell phone 120mA talking
- GPS: 70mA for location ping.
added up, we're looking at 300mA of current needed per message, on a somewhat rough scale. Hannah, can you check through this since i'm a noob?
I think the arduino power is going to be the most variable depending on what we are writing, how many ports we use etc. For the bookmarks we used 2 750 mAh/3.7 volt batteries which would power it for 3 days straight, a week with the arduinos turning off at night. I think we were only using one port, though the program on the arduino was pretty extensive. (I am going to need to brush back up on power usage, it's been a year.)Hwolfe 18:59, 4 March 2009 (EST)
One problem that I was looking at is while arduinos are relatively hardy when it comes to temperature, the chemical reaction though in batteries is slowed at colder temperatures, which leads to less of a current. We are going to need to take this into account, or have really good insulation on the battery. If a car battery in the winter has trouble operating at low temperatures in -50 degrees C temperature we are going to have major problems.Hwolfe 18:59, 4 March 2009 (EST)
[edit] List of things to do for next week
- Test arduinos with batteries
- explore/order batteries to test
- if we are going to be powering some sort of heater to keep it warm that is more power consumption
[edit] Power Consumption Tests
We intend to, in the next week, run three side by side power consumption tests, starting them nearly simultaneously, then letting them run until the batteries die. The programs are all slight variations off of the telemetric motion sensor test we all ran as a class. The Difference is that one will not be sensing anything, just writing the number of seconds to EEPROM, another will measure the sensor and write it once, and another will take the results from the sensor and write it a second time, as well as have another port on high, basically mimicking the idea of having two binary sensors running simultaneously and writing. The next step after that is to plot the three times on a chart, and try to draw a general line as we make one more EEPROM write and open up one more port for use. We can get a general sense of how much time an arduino will run for a given number of ports being used from this preliminary research. Although its kind of trivial, the code for the three programs are as follows:
[edit] Batteries, Round one
After one week, we had a 6 volt battery that started out of the packaging, and ended up with 5.23 volts after a week of running code that simply counted time and assigned it to EEPROM. This is very good news; from this, we can make this equation to find out how much power an arduino uses: 6 - 5.23 = 0.77 volts used. At 5 amps, we can come up with how many mAh the arduino consumes at a minimum rate of just running the chip without any of the ports at about 9.16666 mAh. Then, for every port, we're going to be adding 40mAh, assuming we would be using them for a full hour constantly. For one second of use, it would drain .01 mAh
A 9 volt lasted approximately 60 hours running the motion sensing experiment, so it's probably not the best choice. Hwolfe 15:33, 20 March 2009 (EDT)
ACCORDING to ladyada, These be the stats for the GPS, the SD Card, and the arduino:
- The Arduino chip is always using about 10mA of current, maybe a little more if there are LEDs. You can reduce this a lot by making it go to sleep
- The USB chip uses about 10mA as well. If you dont have a USB cable plugged in, this will save you a bit of power
- The GPS module uses between 20mA and 60mA of current, depending on whether it is trying to acquire a lock or whether it is merely tracking. You can reduce this by using the GPSPWR pin to turn the GPS off between reports.
- The SD card uses 20mA but only when writing data. You can reduce this by sleeping between reports and only logging fix data from the GPS which translates to less data written.
