Below are the 8 most recent journal entries recorded in jeansensem's LiveJournal:

Tuesday, November 8th, 2005
3:52 pm	another week I've gotten all the hardware that I've currently ordered in - the computer, the temperature probe, and the pH probe. I've been reading through manuals to find out what I'm supposed to do to get information from the probes - more of that tonight. I have all of the equipment and keys I need - now the time. (1 Comment |Comment on this)
Wednesday, November 2nd, 2005
1:47 pm	late update Keep on forgetting to update and then figuring I"ll do it later. I've been working on writing my paper; its structure has completely changed. I figure that's alright I've had some problems with latexing it up, but have emailed charlie, and I'm sure he'll be able to help me The sensors and computer are ordered; they should get here sometime this week. I am dreadfully excited for being able to start real work I have keys to various places in dennis - I feel dangerously powerful. Back to working on my paper before class, woo (Comment on this)
Tuesday, October 18th, 2005
9:26 pm	token journal update I've been looking at sensors and found what I think I'll use for my pH sensor. Presentation went. Charlie has given me manuals to look through I have a better idea of what the hardware needs to look like. as far as I know, Charlie knows everything that I know about my project right now. (Comment on this)
Wednesday, September 28th, 2005
5:51 pm	preliminary code outline my software comes in two parts: software on the data logger and software to interact with it. /On the logger:/ This I'm not sure about. I know that the data comes in, but not in what form, and I don't know what the best language to use would be, so I don't know what constructs I'll have available. Here's my rough outline, though I don't know how much of it will end up being in the end result. memory pointer <= endofprogram wait for interrupt. start main main: beam all data (from endofprogram in memory to current memory pointer). wait for interrupt beam settings (should I include error checking, maybe just a set bitpattern between each record, eg?) if(confirmation) get settings (right now, that's a boolean wrap and a time interval) send confirmation of getting settings. set timer to the interval specified by user start normal loop else wait for interrupt GOTO main normal loop: write date, time collect sensor type write sensor type collect data from the sensor write data increment memory pointer by record size if(end of memory) check settings to see whether to wrap or stop if(wrap) memory pointer <= endofprogram else WAIT for interrupt GOTO main if(interrupt from elsewhere) GOTO main else WAIT for timer to go off. GOTO normal loop. [confirmation will have to wait for a bit, since it'll take my host program a second or two to make sure its stuff is alright. So if(confirmation) takes a little longer than just a check. Maybe it should be if(confirmation())] /On your pc:/ This is the more complicated program, since it can basically take up as much space as it wants. I don't have its details worked out yet; if it's alright, I'd rather make a basic version that does the minimum required (as shown below) and then allow whoever's going to be using this device to write their own complicated version that does precisely what they'd like. In any case, here's the basic-est basic version. main: What do you want to do? a: communicate with logger b: look at information (Comment on this)
4:26 pm	alright. Here's my preliminary paper outline, ya'll. AN EFFECTIVE AND ECONOMIC DATA LOGGER by Jean. abstract. 1. Intro: A data logger is... A data logger is used for... I construct my own (a summary of why) Included are the plans - feel free to make your own. 2. What are data loggers used for? waste management, scientific water + air monitoring (make it for the water and the air comes pretty much free), water quality management, laboratory experiments, fish farms (industry, science and resource management) are three main areas, as far as I've found out so far. 3. What is out there already? Products that do the same sorts of measurements, obviously, but not usually switchably, rather one for each, and they're really freaking expensive. data loggers entire range from 1000 - 4000 on average. (various catalogs, websites) Gross. also, they're /mostly/ for laboratory use. The rugged ones are on the high end of the price range. Hobo comes the closest to what I would like to do. sensors are about 40-500. less gross, but still not exactly what I'm looking for. chips are available for ? (waiting on information from charlie's friendperson) standard features include: longlife battery user-set sampling interval sometimes an optical usb connection sometimes self-contained long time between service. hobo uses infrared communications. wrap-around /or/ stop reading at end of memory ... 4. Why is mine better? less expensive. (include actual figures when they're figured out.) build it yourself and learn about hardware while you're doing it. Solar power rather than battery. open source - fiddle with it all you want, make it work for more sorts of things. (I'm playing with the idea of making the interfaces [sensor to converter if needed, converter to computer] as close to a standard as I can find out what the standard is. That way, theoretically, if you get things made to the same standard, you could just plug them in.) 5. Construction basic explanation of the parts involved, probably with pretty pictures. Explanation of choices made for parts so that potential constructors can make informed decisions. 6. How does mine work? (general outline, rather than code) To set settings, a piece of software on another computer (a nice interface, maybe) takes your choices and converts them to the simplest possible form before setting them on the logger. Every time the set time interval is reached, the current data is written to logger memory with a given format. When the logger is contacted, it beams over all of its collected data and waits for either confirmation of its current settings or new settings and then begins to record again according to those. The interface program saves the data it's been given and outputs it in at least one easy-to-look-at format (spreadsheet, graph?). 8. Conclusion So there. Appendix A: Construction (including cheapest places I've found to get parts) Appendix B: Code for logger itself. Code for interfacing program. (Comment on this)
Tuesday, September 27th, 2005
5:44 pm	lookin' at stuff Oxyguard's site is giving me information that may or may not be useful about pH probes. They say that it matters whether they're in turbulent water or otherwise and that something to do with electrolyte makes it work better. If I'm assembling my own probe, these things will be important to know. pH, as wikipedia tells me, deals with how many hydrogen atoms (H+) are swimming around - I vaguely remember that from chemistry Redox, measured in volts, is closely related and has to do with a substance's 'affinity for electrons' as compared to hydrogen's. So pH has something to do with electricity, so it's measured with an electrode, which is a device for measuring conductivity. Maybe you measure the redox with a hydrogen electrode and then convert that into pH? Obviously, I'm not very clear on any of that, so it might be difficult for me to build my own sensor. I mean, I know that if you stick a pH paper in a solution, it turns a color, which indicates its relative acidity, but that doesn't help so much with repeated measurements. You can only stick the paper in the solution once. Charlie's contact hasn't gotten back to me about chips that are available for this sort of thing yet - I'd rather go that way than buy sensors, but it does depend on the complexity of building the sensor once I've got the chip. He (Charlie, not the person who hasn't emailed me back) indicates that the chips may be much cheaper. I've learned some Latex and started on the part of the project that's due Friday. Charlie explained what the paper is actually supposed to be about, which is good - basically, it's what am I doing? what's it good for? what's already out there? why is mine better? how did I make mine? I can do that. Hobo (here) is the most relevant data source I've found - I'm working on looking at some of their technical manuals for information about their products (which are rugged, self-contained, immersible data loggers) for inspiration and clues to how to do this. Unfortunately, some of their manuals (the ones I want to read) are sideways. The only quibble I have with their stuff is that they want you to buy their software and they don't list prices for lots of their stuff, which is frightening. Other places mostly have laboratory equipment. Hobo's stuff, you pop into the water on a string, leave it for up to a year or two, come back, pop it out, and put it next to a device that reads the data from it. Unlike what I'd like to do, theirs is battery powered. They do have an optional transmit data in the field thing but that's more $s. The best news is that I have found out the correct name for the sort of thing I'm making - it's not a water monitoring device anymore, it's a DATA LOGGER. Much more succinct. (Comment on this)
Tuesday, September 20th, 2005
5:28 pm	sept 20 - not much Everyone else showed up last week with nicely latexed abstracts - Not me. The next iteration is due tomorrow, and I still don't have that much change to make to it. A little more redefining, a little more specificity and formal language. Oh well. I've been looking at the companies and their products and what is available and not. My current list of things I'd like to measure is: dissolved oxygen, pH and temperature, with optional salinity. Those sensors are available fairly cheaply and rather widely. I have a couple more resources to look through. Haven't talked to the pesticide people yet. The next question becomes where the area of my project is going to be. Obviously, I don't want to make sensors. I could interface directly with the sensors, or I could interface with the box they interface to. Unfortunately, interfacing with the box demands that I use one company's box; interfacing with the sensors demands that I use one sensor. Sensors, however, appear to be more universal than the boxen, and are certainly less expensive. At the moment, direct sensor interface is what I'm leaning toward. More research and talking to those in charge is called for. A refined abstract, a code outline, and a paper outline are due at the last day of the month. I believe that for me, that'll mean a refined abstract, a preliminary hardware design that addresses nature-proofing and which parts I'll use from where, and the outlines of the howto. Still am not sure what paper will be written on; hoping that will become clearer. (6 Comments |Comment on this)
Tuesday, September 13th, 2005
3:57 pm	sep 13 - investigating probes An introduction to my project seems to be in order. Here's the same one I repeated to teachers in the geosciences, biology and chemistry departments today: I, Jean Davidson, a CS senior, am making, for my senior seminar final project, a device which will monitor various things about an open water source. With a probe that dangles into said water, it will collect information that will be beamed back to a central location from time to time. To them, I added: I am interested in finding out what probes are available for a reasonable price and what things would be interesting or useful to measure. To those unfamiliar with my project, I add: I will make the device basically self-maintaining, solar-powered, and out of spare parts. I envision a tree-mounted thing with a cord that dangles into the water and has the probe attached to the end. I plan, when I have accomplished all the lofty things I set out to do, to make the plans (which will hopefully detail the construction and software of a cheap and easy thingy) publicly available. An important aspect of this project is making the device probe independent, so that with a new 'driver', a new probe could be attached and it would work just the same. I came back from my rounds through the various departments with a pile of catalogs and various scraps of paper covered with names and websites. Collating the answers to my questions, I discover that the water qualities that Earlham science teachers think would be easy, interesting and useful to measure are: pH temperature dissolved oxygen salinity (and conductivity, since they're apparently related) The water qualities one or another mentioned that might be useful or interesting but not necessary easy are: opacity (on a related note, how much light per day comes in) quantities of various dissolved things The water quality dismissed as unimportant by one faculty member is: turbidity It has been suggested that I investigate the following companies for probes: Aquatic Ecosystems Hobo Onset It is unclear to me whether HOBO refers to a brand name from a specific company or a specific product or an actually company but whatever. I'm sure I'll find out. Part of the project is to write a paper. I don't know what I'm going to write my paper on. Charlie suggests pedagogical issues related to using this device once it's made. Eh. I feel like I could write a survey about the field of ... dunno, samplers, maybe, or embedded systems of a scientific nature. I welcome suggestions. (3 Comments |Comment on this)