Fieldlines.com || Idea: Simple, low-cost datalogger for siting?

Front Page - [Homebrewed Electricity-- (wind) (solar) (hydro) (steam) (controls) (storage) (mechanical)] - Classifieds - Site News
Everything - Newbies - [Remote Living-- (housing) (heat) (light) (water)] - Reviews - Diaries - Our Products

Idea: Simple, low-cost datalogger for siting?
By gibsonfvse, Section Homebrewed Electricity Posted on Sun May 23, 2004 at 08:54:05 PM MST
Does anyone know of a low-cost datalogger suitable for siting in remote areas? Anyone built one? Read on for my thoughts...

I'm wondering if any of you know of (or have constructed) a low-cost datalogger that's suitable for siting in remote areas. When I say "remote," I mean some place where you'd set up these siting mechanisms, leave for a few months, and then come back. My hypothetical situation is a rural or otherwise developing region where one is doing a feasibility study for wind power (either mechanical or electric). Here are my ideas on functionality the siting device should have, and how to implement it:
Functionality: Record statistics on wind speed and direction for a year, rain or shine. Seems simple enough.
Implementation: The real work.
Brains: There are a few low-cost, easy-to-code microcontrollers, like the BASIC Stamp and PICmicro chips, that would make reasonable brains for such a project, IMHO.
Sensors: Anemometer and wind vane. Homebuilt anemometers have been discussed on this board before. A wind vane connected to a variable potentiometer would deliver a varying voltage based on position, though I don't know OTOH if there are any such potentiometers that can rotate through 360 degrees unimpeded. I'm also not sure what digital or A/D interfaces the proposed microcontrollers have.
Power: Solar power with battery backup. The silicon would probably need around 5 volts, but the current draw is relatively low. Power needs of the sensors would up the current needs, but hopefully not by much.
Data storage: There are many options... BASIC Stamp boards with flash memory have been advertised. I think the details to interfaces for certain memory cards are freely available information.
Interface: Memory cards are one option, as mentioned right above. Also, serial interfaces are an option.
Normally, I wouldn't write so much stuff without scrounging for sources, but terrible Internet weather here has locked me out of most websites except for this one. Odd, huh. Anyway, what I've written above is based on my knowledge and memories of previous things I've read about. It could be all wrong; I'm not an electrical engineer. Also, I'm not sure what the cost of such a proposed device would be, but it hopefully could be built for less than $100. It would have to be cheap for wide distribution... but then again, I don't know a darn thing about economics. I'm interested in hearing everyone's thoughts.

Idea: Simple, low-cost datalogger for siting? | 23 comments (23 topical)

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#1)
by drdongle on Mon May 24, 2004 at 05:37:49 AM MST

Check the Archive for an artical titled: "Log that data".

Carpe Vigor
Dr.D
Carpe Vigor, Dr.D

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#3)
by gibsonfvse on Mon May 24, 2004 at 08:35:07 AM MST

Intriguing. Thanks!

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#2)
by RobD on Mon May 24, 2004 at 06:07:26 AM MST

Implementation: The real work.
Brains: There are a few low-cost, easy-to-code microcontrollers, like the BASIC Stamp and PICmicro chips, that would make reasonable brains for such a project, IMHO.

Try the Atmel ATTINY15 (8 bit RISC). It's cheap and has A/D built in.

Sensors: Anemometer and wind vane. Homebuilt anemometers have been discussed on this board before. A wind vane connected to a variable potentiometer would deliver a varying voltage based on position, though I don't know OTOH if there are any such potentiometers that can rotate through 360 degrees unimpeded. I'm also not sure what digital or A/D interfaces the proposed microcontrollers have.

There are pots that can rotate 360 or you can use digital encoders (better). I don't like pots especially in wet environments. I use linear Hall sensors that have no contact and the magnet passing by or around them establishes the resistance at certain points just like a pot but with no carbon or scrapers.
Go to Mouser.com and look up Bourns digital encoders.

Power: Solar power with battery backup. The silicon would probably need around 5 volts, but the current draw is relatively low. Power needs of the sensors would up the current needs, but hopefully not by much.

You will need some type of power storage, let's say you use solar power, you'll need to store it to drive the circuit so this means charge controllers and you will want to keep the drain current as low as possible which means low drive voltage (5 volts is to high). I'm using 3 volts but you can go down as low as 1.5. With the new Seiko chips I have current drain down to 80uA (.000080 amps) before it hits the microcontroller and burst transmitter.

Data storage: There are many options... BASIC Stamp boards with flash memory have been advertised. I think the details to interfaces for certain memory cards are freely available information.

If you don't want to spend the time learning C, C++ or assembly language you can use the 'Stamp' but it is very expensive, you should be able to get a lot of info on it though. I think Jameco sells stamps and data loggers.

I can't disclose the project I'm finishing up here because I'm entering it into a microcontroller contest but the design will be on my web site once the contest deadline is past.
Next project is an inexpensive gauss meter using linear hall devices.
RobD

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#4)
by gibsonfvse on Mon May 24, 2004 at 08:51:33 AM MST

Try the Atmel ATTINY15 (8 bit RISC). It's cheap and has A/D built in.

Cool... even GCC already able to generate code for it. Nice

There are pots that can rotate 360 or you can use digital encoders (better).

I was looking for cheap optical encoders the other day... in fact, that's exactly what I had in mind originally, but I didn't know where to go to get a cheap one. Thanks for turning me on to the Bourns ones. A Hall sensor (or multiple sensors) and a magnet/magnets should be a good option for the wind vane.

You will need some type of power storage, [ . . . ] (5 volts is to high). I'm using 3 volts but you can go down as low as 1.5.

I should have used a better choice of words; I wasn't necessarily thinking of the solar cell as the main driver, but your point is still valid. I'm assuming the Atmel is ok at 1.5 volts; I haven't checked yet.

I'm already familiar with C and C++, and assembly language for these simple microcontrollers should be relatively straightforward. Probably the best thing to do is to keep the circuit "always on" and use a serial interface, so I don't need an interface to flash memory. Storing wind speed and direction as bytes every half hour for a year requires less than 64 kilobytes of memory. Then again, you've probably dealt with data storage already, so you would know better on storage and interface options than I would. Thanks for the great help!

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#5)
by Chagrin on Mon May 24, 2004 at 09:19:46 AM MST

usdigital.com has encoders with online pricing. For example, model "E4" is $29.

Probably be a lot easier and cheaper to buy a "one wire" weather station ($79 at www.aagelectronica.com) and attach a data logger that interfaces with its serial port.

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#6)
by gibsonfvse on Mon May 24, 2004 at 12:36:13 PM MST

I misspoke again; actually, the Bourns encoders are not optical encoders, but that shouldn't be a problem. Optical encoders are way too expensive....

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#7)
by RatOmeter on Mon May 24, 2004 at 03:11:59 PM MST

The cheapest optical encoders are about $9 at Wal*Mart or free if you've already got'm. I'm talking about a PC mouse. Obviously you can't use them without a little hacking, but you've got 2 choices, depending on the balance of your electronics vs programming skills.

If you're more up on electronics and mechanical things you can desolder the IR emitter detector pair, add a transistor buffer circuit to get you a TTL compatible output. The optical interrupter wheel would need a mount of some sort and a physical connection to your anemometer shaft.

If you're more whizzy at programming (and mechanical things, hmmm) and your mouse is of the serial (RS-232) flavor, you can just connect it to the serial port of your microcontroller and interprete the data stream. It's been a while since I've looked at the output of a serial mouse, so I don't recall the exact format, but you could figure it out empirically or by internet search.

Either way, you've got the mechanical issue of connecting the wheel to your anemometer and weatherproofing it, but it's a lot cheaper than the $30 or so dollars for Clarostat optical encoders from DigiKey.

Kinda depends on how much money you've got to burn and how much fun you get out of little hacking projects.

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#8)
by RobD on Mon May 24, 2004 at 05:46:35 PM MST

You'll have to go to the Tiny13 if you want lower voltage. The 13 goes to 1.8 and the 15 only goes to 2.7 volts.
Have fun!

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#9)
by mercurous on Mon May 24, 2004 at 10:42:53 PM MST

Many microcontrollers (PIC series is what I use the most) have a low current "Sleep" function that will only wake up on interupts, etc. If you take advantage of this, you could keep the uC sleeping most of the time, only waking up to take and log data. There are probably many people on this board who would be interested in the final product so let us know what you turn out!

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#10)
by gibsonfvse on Tue May 25, 2004 at 12:55:59 AM MST

Yes; at least the Atmel chips have a sleep function, which I plan to use. The PIC chip is very popular (and, I believe, forms the brain of the BASIC Stamp), but currently, I haven't looked into it very closely. I spent a lot of time looking at the Atmel chips and found a large amount of good things, including AVR codegen capabilities now rolled into GCC, and even some simulators (this one with a fairly functional demo; this one I haven't browsed yet). I'm sure similar tools and features are also available for the PIC, though there doesn't seem to be codegen abilities for it rolled into GCC. I'm not an electrical engineer, so I don't know how it's going to go, but I'm curious. We'll see if my curiosity lasts long enough for me to start and finish such a project while not working the usual workweek.

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#11)
by XJames on Tue May 25, 2004 at 02:43:53 AM MST

Go to www.avrfreaks.com for AVR projects

search on Google for CoadeVisionAVR (AVR C compiler)

search on Google for BascomAVR (AVR BASIC compiler)

Look on www.microchip.com for application note AN582 which has information on a low power clock circuit that is the basic concept for a battery powered data logger which can run for a year or two on 4 D cells.

Have fun. The project will be a lot of work.

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#12)
by RobD on Tue May 25, 2004 at 05:15:49 AM MST

I'm about half way through my project here and I'm into some specialized chips on this that only come in SMT.
Yes, the AVR's do have sleep and idle functions that cut the current down to a few uAmps but you have to coordinate the sleep with wind speed and sensing for times when the wind doesn't blow at all. It makes the project trickier and using optical or hall sensors adds several ma. to the design. I think you'll like some of the solutions I came up for this.
The Bourns rotary encoders I spoke of only run about $3.00. I try to stay away from or design my own sensors when the prices get high. If you have a small lathe you can make some neat sensors very inexpensively.
One of the reasons I use the AVR over the Pic is speed. The AVR runs one instruction per cycle and the Pic needs four. That means a 1 meg. AVR can work as fast as a 4 meg Pic. Once you get up to the higher freqs. the dividends start to pay off. Also I think the AVRs are cheaper overall but I may be wrong. I program in assembler and I like the AVR better than the Pic assembler. I find the instruction sets are more powerful. Pic Basic is very slow and expensive but it's a great start. My preference, you can do great stuff with either chip but it's very hard to beat the ATTINY, the 11s cost me $1.32 a hundred even the 2313 that I use in the Tach jr. is cheaper than the Pic I would need for the same application. Aren't the F84s up to about 7 or 8 bucks now?
RobD

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#13)
by gibsonfvse on Tue May 25, 2004 at 10:37:30 AM MST

My thoughts were to have the CPU wake up every half hour and read the wind speed and direction, and then go back to sleep (regardless of current environmental conditions). Of course, this would mean I need a realtime clock chip. Apparently some of the PICs have such a clock built-in, but I'm sure such clocks are sold stand-alone as well. It's probably not necessary to store the actual date in memory so long as we know when the device was powered up, but it might be convenient. Well, all of this stuff is worthy of investigation. Thanks for the ideas!

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#14)
by gibsonfvse on Tue May 25, 2004 at 10:39:29 AM MST

Before I forget: do you have any recommendations for info on how to do 8-bit math? I don't think I'll be doing anything sophisticated, but I'd like to find out just in case (the complexity of the math may dictate what features I choose to implement).

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#16)
by RobD on Tue May 25, 2004 at 05:40:10 PM MST

Perhaps waking up the cpu every 1/2 hour is to long a period you can lose a lot of data and have false , my take is every 3 minutes tops. Most chips have one or two timers and a WDT (watch dog timer). You can drop the operating freq. to extend your clock times. (using a 32kc watch crystal will give you very long times but you may lose resolution so you'll have to do the balancing dance between resolution and power savings.
Most of the chips do the basic add and subtract with boolean operations. You can use shifts also. Things start to get hairy when you go to floating point math but it's possible. (The Tach jr's. 2313 chip has floating math routines for accuracy). I don't think it is required in my anemometer though but I'm just getting to the code now.
Cheers

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#15)
by XJames on Tue May 25, 2004 at 11:44:25 AM MST

I have a schematic and some assembly language code for a battery powered PIC based data logger. If anyone would like a copy I can email the files. Or if you can wait a day or two check out www.safetypinmotor.com/logger where I will post the schematic and code. In the meantime, check out plain old www.safetypinmotor.com where you can learn how to build a super-simple electric motor.

Buy one of the previously mentioned $79 weather stations and build the logger and you will have a $100 solution.

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#17)
by gibsonfvse on Wed May 26, 2004 at 01:54:08 AM MST

Cool... looking forward to seeing it. Thanks for the help.

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#18)
by geoffd on Wed May 26, 2004 at 04:37:26 AM MST

Check out the following URL, it seems to implement a fair amount of what you are trying to do.

http://home.arcor.de/d_meissner/d_logg2_engl.htm

Cheers,
Geoff

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#20)
by gibsonfvse on Wed May 26, 2004 at 02:15:05 PM MST

Very thorough... thank you.

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#19)
by Dan M on Wed May 26, 2004 at 09:08:36 AM MST

It looks like you have plenty of suggestions for the logging. Here's one for the sensors.

So called "instantaneous reading" or "zero inertia" (actually they're just very low inertia) anemometers are constructed from a piece of wire and some strain gauges.

You need a piece of wire light enough and long enough that when it's placed vertically the wind will bend it but will NOT cause permanent deformation. A car radio antenna might work for this but you might want to shorten it some, or at least cut off the lump at the top to help reduce the possibility of harmonic vibration.

Glue the strain gauges vertically at the base. Strain gauges are simply resistors. The resistance changes as they are "stretched". A small amount of lateral deflection in the vertical wire will induce a longitudinal stress in the wire and will reqister as a change in resistance. If you glue 2 strain gauges at the base 90 degrees to each other you would be able to use one to get the east-west component and one to get the north-south component of the wind. The vector sum is the velocity and the arcTan[(ew)/(ns)] will give you the angle with respect to north.

The advantage of this method is that it is cheap and all solid state. There is also very little inertia associated with the wire (as compared to a rotary anemometer) so it will more accurately measure gusts (maybe measuring gusts is a disadvantage).

Disadvantages are that you will have to calibrate it, and you might have to smooth your data because of the low inertia.

Have Fun,

-Dan M

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#21)
by gibsonfvse on Wed May 26, 2004 at 02:21:08 PM MST

This is quite an interesting idea. I don't know how it would stand up in very high winds, and I also don't know how much strain gages go for. I believe they are relatively cheap, but I'll have to do some searching. Problem is, I believe the relationship between wire deformation and strain gage reading is nonlinear for reasonable windspeeds since the wire is flexible... but I'm curious about this option, and I may be able to think of some configurations that wouldn't have this issue. I could be wrong, as I'm suspecting you've constructed this type of anemometer, or are at least familiar with it. I'll do some musing... thanks!

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#22)
by RatOmeter on Thu May 27, 2004 at 05:34:56 AM MST

Look at Omega

http://www.omega.com/toc_asp/subsectionSC.asp?subsection=E02&book=Pressure

You'll be interested in their "general purpose" strain gauges at that link, they cost $50 plus for a package of several. You'll also need their epoxy for mounting them and it; that stuff ain't cheap and I've heard you could use other (cheaper) epoxies but there may be reliability tradeoffs.

Be sure to read up on strain gauges in their technical reference section.

Don't expect an amateur strain gauge device to have linear and temperature stable output, but you should be able to compensate for drift and non-linearity to a large degree in software.

Also read up on the signal conditioning electronics required.

[ Parent ]

Re: Idea: Simple, low-cost datalogger for siting? (none / 0) (#23)
by gibsonfvse on Thu May 27, 2004 at 11:31:48 PM MST

I think this thread's now dead in the water, but:

My latest thought was to have two "hanging paddles"... Imagine a paddle (of sorts) that hangs below a bearing that allows it to swing. The paddle is raised as the wind speed increases, and the raising changes the position on a potentiometer which is rigged to give a varying voltage with position. Two of these paddles can be set at 90 degrees to each other to give all the information needed to resolve wind speed and direction. Potentiometers come very cheap (cheaper than strain gages, encoders, and Hall effect sensors), and microcontrollers like the ATtiny15L also do A/D for relatively cheap. Just a thought of mine... feel free to comment if you desire. Thank you, everyone, for the great ideas.

Idea: Simple, low-cost datalogger for siting? | 23 comments (23 topical)

Menu

� create account
� How to use the board
� FAQs
� search the board
� Google search the board

Login

Make a new account

Username:

Password:

Total Views

��102 Scoop users have viewed this posting.

Related Links

� BASIC Stamp
� PICmicro
� Also by gibsonfvse

You must be a registered user to post here. It's easy and free, and the link is on the upper right side of your page.
All trademarks and copyrights on this page are owned by their respective companies. Postings are owned by the poster, but may be deleted or moved at the ADMIN's sole discretion. The Rest � 2009 Forcefield.
You can Email the board ADMIN here. PLEASE include the username you signed up with!