water sampler

[timothyp]

Would like to build a device which can collect (via pump)a 100mL sample on to filter paper and then rotate 10 degrees for next sample. Sample interval is twenty minutes thus filter paper would be changed every 12 hrs. Filter paper diameter from 20 to 30 cm. Any suggestions on pieces and parts? Thaks in advance

[Will]

Are uou sure that you mean a `100 mL ' sample ? Unless I am completely misunderstanding then by the time it has collected 36 samples you will have dumped close to one US Gallon of water (Approx 3.75 L = 37.5 * 100 mL) on the filter paper

[Chris Foley]

Could you please provide more information?<p>* How many axes of motion do you want to have? Please describe.
* If you want a circular motion going from sample to sample, how long should the rotating arm be?
* Again, if you have a circular motion, what degree of accuracy do you need in placement?
* Does the system specification include weighing the filter paper? How about data collection?
* Does it include automatically changing filters? If so, what should be done with the "used" ones?
* 100 ml with what degree of accuracy?
* How slowly should the sample be put into the filter paper to ensure proper drainage without splashing?
* Should there be some kind of liquid storage while the sample is dripped into the paper?
* Do you require protection from cross-contamination of samples?
* What is the liquid? Is it safe? Do you have to think about OSHA, or UNSCOM?
* What's your budget? Is this a home brew, a "government job", or is the company paying for it?

[skrallman]

I don't quite follow you about the 20min sample interval and changing the filter every 12hr---but I'm guessing that you mean a sequential sampler, i.e., take a sample then rotate to the next sample filter/bottle.
The commercial samplers I've seen have either a rotating arm with a slip joint that rotates to each bottle, or the tray that holds the sample bottles rotates. I think a rotating tray would be the simplest to homebrew.
For a first cut at it, I'd use a lazy-suzan to hold the bottles and use a stepper motor, or geared motor with a rubber wheel to drive it from the outer rim. Then you could put the motor and pump on timers.
Scott

[timothyp]

Thanks for the replies. Let me clarify the project application. We would like to collect the fish fecal material from the bottom of conical culture tank in order to evaluate diet energetics. The "lazy susan" saple tray would hold the filter paper above the tank allowing water to fall back into the tank. The volume of water pumped onto the tray would be ~100-200mL at about 1 Lpm. The volume can be adjusted via the time the pump is on. The key is rotating the tray to minimize discharging over a previous sample. We are not trying to obtain discrete samples but want to avoid overlapping as much as possible.
We are a small gov. lab and do not want to invest in a commercial sampler as we may use apparatus once and shelve it until further use. We are trying to avoid the expense of manual sampling. Filters will be changed every 12 hrs.
ScottK idea is kind of what we had in mind. Ideas for locating stepper motors or geared motors wold be helpful.
Thanks again for the insight and help.

[Chris Foley]

Hmmn. Filtering floating fish feces. Fascinating. (Sorry -- I couldn't pass it up!)<p>I'm assuming you would require a certain amount of accuracy in the pump volume(say, +/- 1 or 2%) If you have a budget, I'd recommend a small peristaltic pump. They're available from Cole-Parmer (general-purpose lab supply, rather pricey but they usually have just what the doctor ordered). If you get one that operates on 115VAC, the motor speed will be dependent on frequency, so you just have to set on/off time to determine total flow. Get their free catalog, if you do science for a living, or if you just want a source of good lab equipment.
http://www.coleparmer.com<p>
The tubing coming from the pump can be extended out on a balanced arm which is connected to, and rotated by a gearmotor. You want external 1:36 gearing, so one revolution of the gear motor will correspond to 10 degrees. You can do this fairly accurately with a couple of belts and pulleys. Given a sufficiently slow movement of the motor, you shouldn't have to worry excessively about belt slippage. They're a bit easier to set up than gears. You might use an opto-interrupter to indicate a complete turn of the motor. Just put a slotted disk on the shaft, connected to one of the pulleys. A good source of DC reversible gearmotors is Herbach and Rademan. They have a 4RMP 24VDC motor here with high enough torque to get the job done, and can do the 10 degree turn in 15 seconds...
Herbach & Rademan 24VDC Reversible Gearmotor<p>The rotating arm with the filter attachment underneath should be counterbalanced with a weight on the other side. You should set it up with bearings so the weight of the arm and counterbalance aren't resting on the motor shaft -- that tends to burn the motor bushings.<p>Peristaltic pumps are self-priming, so you don't have to worry about keeping fluid in the tubing. You have to work out a setup to 1) get the end of the hose into the tank, 2) turn on the pump for a set period of time, corresponding to 100ml., 3) turn off the pump, and withdraw the hose from the tank, and 4) turn the pump back on to allow all the water/fishpoo to run through the hose/pump/hose/filter. If you have shop air available, I'd recommend using an air cylinder to do the dipping of the hose inlet into the tank. Many sources for these -- I'm feeling patriotic today, look at Clippard Instruments
http://www.clippard.com/<p>You might want a slower pump than 1 l/min (100 ml = 6 sec). Try for a speed that will stretch out the time enough so you don't have to worry about extreme accuracy on the timing.<p>You can either use a PC, a PIC, or a PLC to control this job. I hope some of this has been of help. If you have any other questions, feel free to e-mail.<p>[ April 11, 2003: Message edited by: Chris Foley ]</p>

[skrallman]

Another motor that comes to mind is a RC servo motor. They can be easily modified to rotate continuously, they're geared low, have quite a bit of tourqe and they're cheap. I think you can get them for as little as $9 at servocity.com. People use them alot for robots so there is also alot of info about controlling them with microcontrollers.

[bodgy]

<blockquote><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><hr>Originally posted by Chris Foley:
Hmmn. Filtering floating fish feces. Fascinating. (Sorry -- I couldn't pass it up!)<p><hr></blockquote><p>I take it there would be 1,048,575 different ways of tackling this project. <p>Colin

[Chris Foley]

Yes, multiple permutations, lowest installed cost, meeting project specs. I was thinking, with a 36:1 ratio, it might be better to use gearing instead of multiple belts/pulleys. Most projects are that way. The longer you look at them, the more ideas and permutations come to mind. Eventually you have to just choose something, and go with it. For the FFFF, I think most of the cost will come from the choice of the degree of precision required in the 100mL sample, unless there's something out there in surplus in a metering pump. The good ones are hard to come by, especially ones that are still in production (for a piece of equipment like this, you want to be able to fix it if it breaks).<p>One other issue here. One of the prime directives in something like this is to keep it simple. The designer may be working on a contract basis, or may have moved on to something else, or happens to be out sick on the day the FFFF breaks down. On that day, they'll call in somebody from the maintenance department to fix it. You want to at least try to make something which is straightforward enough so that maintenance tech (who may only have limited electronics experience, and may distrust computers entirely) has at least a chance to fix it. Otherwise, the owner may just scrap it if they don't understand it, and can't fix it. That doesn't help the owner (who feels cheated), the maintenance tech (who feels like he's been showed up, and will spend his lunch break convincing everyone he knows that the designer did a bad job) and the designer (who has a dissatisfied customer who won't recommend him for the next cool job). Good luck, tim. Happy hunting.<p>[ April 10, 2003: Message edited by: Chris Foley ]</p>

[Dimbulb]

The tube at the bottom of the tank is positioned at the radius of the rotary base that is supported on a mechanical timer and a vacuum pump made from a bass ackward fish air pump.<p>The base is fitted with ten vacuum filter disks and below has floating stops. The vacuum pump turns on by the timer but turns off after fifteen seconds as well as a solenoid that releases the pinch on the tube.
A time delayed relay connects timer to pump solenoid junction.<p>First forget foolish filanderings from formidable former foolish friendly foodchain fanatics, Frankly finding free facility functioning for fresh feeces fashioned from flat fibrous felt frustrates freeflow for famous fixed frame fish filters.<p>A water sensor insures device shutoff in case it malfunctions which prevents tank water loss and getting in deep do do<p>[ April 10, 2003: Message edited by: 1206DX ]</p>

[Colinr]

just a thought if it is every 12 hours that uou will change the filter paper just get a mechanical clock mount it horisontally with the hands uppermost. remove the minute hand and fit the filter paper onto the hour hand this will then rotate once every 12 hours. you then need to arange for the output pipe of the pump to sit just above the edge of the filter paper. then use any timer to switch the pump on and off wvery 20 minutes.