>Given your parameters, I would fool around with using a SINGLE wire >hooked up to the input of a CMOS gate (inverter). You might have to bias >the input with a VERY high value resistor, but once you figure this out, >the wire hitting the liquid should be able to consistantly change the >output. You'll have to protect the input also, maybe with some zener >diodes. > >If you want to mess with analog, it would be better to use an op=amp or >a comparator, but I used a similar arrangement for a non-contact prox >sensor (another application where there was no room for the price of an >off-the-shelf sensor) and it worked fine. That way, you have six input >channels for about 15 cents. You may even be able to run the CMOS off >15V and run the PLC input directly. Now were talking. I'll have to look into this idea. >Don't you have to clean the goop off the wire in between jugs? Or is this >a static appplication, where the wires just sit in the jugs? Yes it is static. Once the jugs are placed into position, they are not moved until empty. We know how much is being dispensed with each application ~ 10-20ml. We can then set our limits to know how much is left BEFORE we run out. Trying to be proactive for a change.

 

Here are some possible candidates: Canongate Technology Limited makes a non-invasive ultrasonic point level switch called a SpotCheck. You just put it on the side of the tank where you want to detect the level. They are represented in the US by Instrument TAGS in Pasadena, TX 713-477-9780. Their web site is: http://www.canongatetechnology.co.uk A matchbox size non-invasive level device can be found at http://www.etacbe.com/uk/PR/613.htm Another similar meter can be found at http://web1.pipemedia.net/katronic/meassono.html along with a continuous type at http://web1.pipemedia.net/katronic/prod_04.html A capacitive proximity switch might also be used to measure the level. Bill Mostia ======================================================= William(Bill) L. Mostia, Jr. PE Independent I &E Consultant WLM Engineering Co. P.O. Box 1129 Kemah, TX 77565 [email protected] 281-334-3169

 

We have a laser contrast photo sensor that we have used for this at about 8-10" away. It doesn't seem to matter what colour the liquid is. Email me if you would like a data sheet. Rick Flokstra Baumer electric, Canada

 

>As always the question "What will do the job?", fails to be answered with >"What is the job?". People try to answer an unspecific question, not >having nearly enough information to give a conclusive answer. Everyone >having their own idea of how much is too much money to spend. Eight people >give a reply without knowing what the measurement will be used for. Having >no idea of the required accuracy, no concept of the process or how it is >controlled, we don't know if the process fluid is explosive, corrosive, >toxic and yet ..... we know an instrument that will do the job [Kirk S. Hegwood] OK. The job is to monitor the level of 5L jugs, 8 total. The chemicals are "siphoned" from the jugs with a pump and are non-(corrosive, explosive, toxic) chemicals. Control voltage is 12VDC. We do have in place a plc to control the operation. This is a machine which is planned for mass production and the target industry does have a price ceiling. Therefore, we have to keep the cost low. Ingenuity, which I'm seemingly short on, or some other technology is what I'm looking for. >This guy could be a kid trying to distil "wood alcohol" and one of you told >him to put a bare wire into the jug [Kirk S. Hegwood] I'm the one that did that . For the record, snow mobiles and other such machines use this for oil level indication.

 

You can consider to measure the weight of your container. Depends on too much things (dirty place ? , resonating machine ? different specific weight of the material ?) but it is an alternative, too. Regards, Csaba AtySoft Ltd., * TEL: (+36 1) 316 3251 * FAX: (+36 1) 212 0250

 

Improve away. You could make it proportional through a valve attached to the mechanical arm (ala toilet tank). Oh, and maybe even make an integrator out of a second mechanical arm controlling dripping water to a cup with a slow drain (integrated error) which further offsets the first arm. ;-> Anthony Kerstens P.Eng.

 

> You're going to make something that's cheaper (installed) than a capacitive prox? What exactly is your target price per jug then (including engineering, fabrication, installation, etc.)? Even if you were going to make a contraption out of springs and micro-switches (to weigh the jug), you still have to design, build, and tweak it. > I'm trying for $15 to $20 per point. Don't know if that is possible but I can try.

 

Hi Kirk... In one of your replies, to Willy Smith, you mentioned that the rate of application was 10 to 20 ml. Given the fact that there is 5 liters in a full jug, could you not just count applications per jug in a plc register then estimate remaining volume from that. 5L = 5000 ml = 250 applications @ 20 ml before we may be close to empty. This could warn your operators and would use no parts. Best Regards... Rick Kelly Chief Electrical Technician Natural Cuts Cheese Operations Kraft Canada (613) 537-8069 V (613) 537-8057 F [email protected] http://trondata.on.ca

 

Now that you have mentioned this, I recall that we have some Videojet ink jet printers that have two methods of measuring liquid level. One method uses a magnetic float switch. A ring shaped magnet slides up and down a tube which has several magnetic reed switches inside (as Mr. Jansen mentioned seeing). This is used for the ink pump and mixing system reservoir. The other method is used for the ink and make-up fluid bottles (which are changed regularly by the set-up personnel). There are two tubes which go through the lid of the bottles. One tube is used to suck out the fluid, the other is used to detect when the fluid level is getting low. The detection tube is cut slightly shorter than the suction tube. A weak stream of air is blown into it, and some sort of pressure or flow switch is used to detect the amount of air which passes through. When the level in the bottle gets low enough, the air can flow out the tube and this flow is detected and used to indicate the level is low. Both tubes go through a hole in a bottle lid which is a permanent part of the machine. When you put a new bottle in, you throw away the lid which came with it, and screw on the lid which is part of the machine. The two tubes are held together with a band near the bottom to ensure that one doesn't curl up away from the other. Since the detector tube is slightly shorter than the feed tube, it will indicate a low fluid level before the end of the feed tube is exposed. No air appears to bubble out until the level gets quite low (if even then). The first method sounds like the easiest of the two to implement, provided you use a well designed system. If you can get a good off the shelf unit it would probably be preferred, as there would be nothing worse than having the float jamming half way up the tube all the time. Since you mentioned that you expect to buy a lot of these, you should be able to get a good volume discount on them. The air bubbler method sounds simple, but I imagine that getting it to work just right may be more difficult since you are probably working with fairly small differential pressures. Also, the liquid must be able to tolerate having some air bubbles in it. You would also probably need to have a fairly clean air supply (or good filtering) so that you don't contaminate the fluid. The machine I mentioned does use it, but I don't know how much trouble the designers had getting it to work. I also don't know how reliable it really is, since we check the bottles every shift and generally change them before they get close to empty anyway. ********************** Michael Griffin London, Ont. Canada [email protected] **********************

 

It may be easier to measure the volume before it is placed in the jug. I'm assuming the jug has a wide variance in interior shape/volume of the jug.

 

Yes. We will actually being doing this for inventory control. Whenever the door to the machine is opened, the operator screen will prompt the "user" to ask if they changed any of the products. At this point we are making an assumption that the operator inputs a correct response - Yes, No, etc. If we had a hardware low level indication, then we help omit some human error. Thanks for the input, Kirk S. Hegwood President Signing for Hegwood Electric Service, Inc. [email protected] Phone: 770-447-8853 Fax: 770-447-5310

 

>It may be easier to measure the volume before it is placed in the jug. >I'm assuming the jug has a wide variance in interior shape/volume of >the jug. Yes, but generally they are all 5L.

 

Your 3/12 response looking for a small float actuated reed switch sounds like an easy, low-cost solution. Dwyer Instruments makes several different variations of these for both vertical and horizontal installation. You can get more detailed information at www.dwyer-inst.com, it is the series F7. Tony Kohl Dwyer Instruments Inc.

 

Ok, this is a late replay, but anyway. Are there anyting else wrong with the bare ended wires idea than that the liquid get stuck to it and thereby stil conducts? If no then I wonder do you have one two-part wire or two separate wires? Two separate wires that are separate all the way up and with enough distance from each other could never conduct between them regardless of how much liquid stays on each cable if they are far enough from each other. I might have missunderstod the problem however. /Johan Bengtsson ---------------------------------------- P&L, Innovation in training Box 252, S-281 23 H{ssleholm SWEDEN Tel: +46 451 49 460, Fax: +46 451 89 833 E-mail: [email protected] Internet: http://www.pol.se/ ----------------------------------------

 

>Johan Bengtsson wrote: >Ok, this is a late replay, but anyway. >Are there anyting else wrong with the bare ended wires idea than that the liquid get stuck to it and thereby stil conducts? [Kirk S. Hegwood] No not really. What we actually did is connect one lead to a metal tube which is the siphon and the other lead (with only 1/2 inch bare) strapped to the tube and placed at correct length. >If no then I wonder do you have one two-part wire or two separate wires? Two separate wires that are separate all the way up and with enough distance from each other could never conduct between them regardless of how much liquid stays on each cable if they are far enough from each other. [Kirk S. Hegwood] Your correct this would work electrically, but all of this would have to mechanically fit through a roughly 1 1/2 inch opening at the top. Also, this mechanism is removed occasionally from the jug for product replenishment.

 

Johan, Conductive sensors have long suffered from corrosion eventually making them ineffectual in liquids. The most reliable method, if there is room in the jug, is a buoyancy type sensor. The electrical sensor remains out of the liquid and a float rests on the surface of the fluid. A connected arm actuates a potentiometer type sensor. Many off-the-shelf float designs exist. Charlie Griswold Automation Engineer Sequenom, Inc. Ok, this is a late replay, but anyway. > > Are there anyting else wrong with the bare ended wires idea than that the liquid get stuck to it and thereby stil conducts? > > If no then I wonder do you have one two-part wire or two separate wires? > Two separate wires that are separate all the way up and with enough distance from each other could never conduct between them regardless of how much liquid stays on each cable if they are far enough from each other. > > I might have missunderstod the problem however. > > > /Johan Bengtsson > > ---------------------------------------- > P&L, Innovation in training > Box 252, S-281 23 H{ssleholm SWEDEN > Tel: +46 451 49 460, Fax: +46 451 89 833 > E-mail: [email protected] > Internet: http://www.pol.se/ > ---------------------------------------- >

 

Dependent on what you want to spend, an RF Capacitance or Sonic gap switch might do the trick. Prices range from $250US on up depending on probe cofiguration and length. A local rep could help you out if we knew where you were. L.Kolbert [email protected]

 

Charlie Griswold <[email protected]> wrote: >Johan, > >Conductive sensors have long suffered from corrosion eventually making >them ineffectual in liquids. [Kirk S. Hegwood] True. Even in our limited testing, corrosion was beginning to take place in some of the chemicals. >The most reliable method, if there is room in the jug, is a buoyancy type >sensor. [Kirk S. Hegwood] Still playing with some ideas others have suggested.