Monitoring Water Level in a Tank

Well there are all kinds of choices to be made here.  I have my engineering preferences.  Another poster in another thread said that all engineers have their unique histories that lead to their unique preferences.  It would be helpful to know more about you, you seem like an educated man.  Do you know any programming languages?  If I have a crude map of your expertise, I can maybe communicate more effectively.   

I will make some undefended assertions, 50/50 chance someone will jump on and disagree.

If this were my project I would probably strive to get the individual functions to act autonomously.  I would seek to rely on the microprocessor more for data logging than control.

People come in here with their apps not realizing how important the use context is and think we are blowing them off.  We are often middle-aged men with time on our hands and some of us even have our bills paid.  If you ask specific questions that are answerable, your case will be attended by someone here.  If you are real lucky you can huck a coupla of us into a (redacted) contest, then you will nearly drown in data and alternatives. 

Let me give an example about use-case:  I don't know how important it is that your system never overflow.  If it is in your backyard that may be a non-starter.  If it is in your living room, and your wife is waiting to accrue one more excuse to boot you to the curb, overflow could be a very important consideration.  I don't know and this and that is causing my mind to vibrate.

I don't know that your rig is going to be in the sun or not.  Or how remote from its operators.  If we are working with water and line power we better reckon stuff like isolation transformers and GFIC.  If we can use a solarcell and battery to run this party, we don't have to worry about the ol' zapperoni so much.  This is causing cognitive frisson in me.  I can't answer in a singular way an incomplete question.

There can be all of these environmental existentials.  Maybe you have deer.  Maybe you need to put up an electric fence to ward away the deer, who knows?

Another unstated parameter is the precision you seek in the data you require.  Without a notion of this, the project is simply uncostable.

Don, I am an engineer.  I work in a lab.  we do this test often. It saves me time to have the test run and collect the data at the end.  I have no electronic or programing knowledge.  Accuracy of the test is to 1/8".  No environmental issue such as weather, animals, etc.  The test must be battery operated, solar cell sounds awesome.  Not worried about overflow, both reservoirs are over a sink.  There will be someone around the test, so making a mess is not an issue.

Thanks, that simplifies things greatly.  In a delicate environment more money has to be spent on covering anomalies than the process itself when working with volumes of liquid.

The way to go with these things is to continue with a divide-and-rule method.  We repeat our attack until the problem granularizes to the point where comprehensibility ensues.  We can work backwards from the end-effectors for a while.  Formal documentation is your friend.

You should have two sources of power.  One can run the electronics, it needs to be clean but not capacious.  The other runs the electromagnetics, it must be at least momentarily pussiant, but not necessarily terribly clean. 

Talking about the solenoid, I have called echotech on another occasion and they informed that they want a 'hard landing' when de-enerqized.  I've already developed the driver for this working on another project.  But, it is out at the house and I am at my apartment.  I am not going to the suburbs today, probably, and it will take me a day or two to dig it out.  I know, experientially, that it works nicely.

If this were my project I would use an MSP430 for the controller.  It has good hardware, and it programs in C++, which is good because I am pretty familiar with C++.  My kid even has a coupla letters from Bjarne Stroustrup!

http://www.element14.com/community/docs/DOC-69317?ICID=TILaunchPads-BoosterPacks-space

Maybe this one.  Perhaps we could use the on-board wi-fi to communicate with a host PC.  This would be more effort in software, but less in hardware as we wouldn't have to mess with an SD card.

I don't have any direct experience with either of these two level sensors.  I always solved the problem by massing the container.  A liter of water really does weigh a kilogram.  You could wrap the 'get_level' functionality

in software so that you could vet the both of them while running the same high-level code.

Aside from the continuous detection of water level, you may want to implement a discrete sensor as a 'fail-safe.'

This is for if your continuous detector goes on the Fritz.  It can interrupt the processor and tell it to quit pouring, something's wrong.

The link you gave shows WiFi C3200 Launch pad.

Should I purchase it or the MSP 430.

They have 4 different units of the MSP 430.

The link you gave shows WiFi C3200 Launch pad.

Should I purchase it or the MSP 430.

They have 4 different units of the MSP 430.

There is a lot of variety in the MSP430 line, sometimes it seems to a fault.  The version you seek depends upon the functionality you seek.  If you are planning to use an analog level sensor, you almost definitely want one with an on-board A/D converter.  If you want direct communication with a host computer, get a wi/fi or Ethernet variant.  If you just want to datalog onto an SD card, get a cheaper variant.  I think TI can sell you an SD-card adapter, if not SparkFun surely can.  Playing around on their website is both rewarding and fun.  The SD card interface is SPI, they may have a software library for that functionality pre-fabbed.  What are you going to do with the data?  Does it go into a spreadsheet or database?  IIRC, there is a version with an SD interface on-board.

If I were you I would read over the material for a coupla days or evenings.  Then I would call TI and discuss my application.  They are very eager and helpful, at least whenever I have interacted with them.

Let me get back with you after I receive all of the items

If you need the WIFI then go for the CC3200, it can do everything the 430 can do and way way more and has the benefit of the WIFI stack with little impact on flash space, the board is also much faster, not that you need that for your app but you will get the experience right away and this can become your goto board for many future projects

The 430 based boards are great too but for more local IO use, like all these level boards (Arduino’s included) they can rapidly run out of program space the minute you add IP networking and WIFI, regular radios etc as they typically use less program space . you will not have this issue with the CC3200

Peter

Thank you Peter.  I am willing to try any controller that can get this project done.  Like you said I have more projects to do if this ends up a success.

Thanks again