ESI Project: Water Usage Monitoring

[abecedarian 330]

Yeah, umm... okay. I'm not liking the requirement for 5v for the MAX9926.

 

It seems there should be an easier way to count the transitions the water thing would make, and I'm open to suggestions.

 

Using one leg of the generator, through a diode and transistor / mosfet, to signal the positive phase has occurred may suffice?

[greeeg 460]

Easiest. Diode rectification. Either full wave or half wave.

 

Since the flow sensor generates an AC signal, is that signals amplitude going to increase with high flow? If that's the case you'll need to add some signal conditioning to ensure signal never goes over MSP's Vdd. A zener could work. But would clip the top of your waveform.

[abecedarian 330]

Since I don't have one of these to verify operation, I would assume +/-0.15v to +/-9v, but it could be +/-0.3 to +/-18.0

I guess I need to get one of these here so I can better quantify how it works.

 

In a similar vein, the sample rate should be faster than one full cycle at maximum flow. Essentially saying that consecutive samplings with the same value do not indicate the unit was moved / any flow has occurred.

 

 

It's no wonder I don't work for Edison or SoCal Gas.

[greeeg 460]

To make use of background processing of the ESI, you could use an envelop detector fed into a variable gain amp. In fact there is likely an off the shelf company for this. Like the MAX9926 but that could run off 3.3v

 

Then you could use the ESI processing state machine to track a full revolution.

[abecedarian 330]

Maxim offers the MAX9924/5/6/7 which operate from 5v, though their outputs are active low so only require a weak pull-up to whatever voltage source fits.

 

TI offers the LM1815, which operates from 2 to 12v, but research I've done suggests it has problems with differential sources; i.e. works better with single ended VR sensors. This would conflict a bit with being able to generate voltage from the water flow device's AC voltage output, I think.

 

ON semi has NCV1124 and a few other variants that similarly to the LM1815, work better with single-ended sensors, but are 5v components.

[abecedarian 330]

Simple hall effect flow sensor inbound.
Thought being get this working, then worry about scavenging power from one of those water flow / hydroelectric A/C generators.

Oops... forgot to tell the wife about this.

[Lgbeno 189]

Interesting project, I'm actually doing something similar with cc110l and msp430G2553. I call it "droplet". It runs off of a cr2450 battery, sleeps most of the time then wakes up sends measurements to a Hub that is wall powered and the hub ships the data to data.sparkfun.com. Here's the droplet:

 

 

And here is a hub prototype (just ordered a real pcb for it)

 

 

You can see the electric imp that receives data via uart the pushes it to sparkfun.

 

My web app project imp.guru fetches the data from sparkfun to view, here's the data imp.guru/f1

 

 

I'm willing to share any/all of this work for your project if interested. My recommendation would be to skip the whole energy harvesting thing and use batteries. With the right design you should last years on a pair of AA's.

 

I've seen projects on hack a day where people use Hall effect sensors to pick up the spinning magnet in their utility water meter. I think this is it:

 

http://hackaday.io/project/1460-Remote-Water-Consumption-Display

[Lgbeno 189]

Btw I've been away from the forum for awhile, what is ESI?

[abecedarian 330]

Btw I've been away from the forum for awhile, what is ESI?

Extended Scan Interface.

Basically, a peripheral that can control and sense various things like inductive and other sensors, and wake up the MCU periodically to allow for transmission of accumulated data. So, whereas your 'droplet' has to wake the processor to do the measuring, the ESI peripheral does it on its own. Given the MSP430 involved uses FRAM, even more power savings are earned.

 

The power harvesting thing was / is a side-thought, that I think would make sense for something under the sink or in the wall. It seems too bulky for something attached to the shower head. And as you mention, a pair of AA's should last a while at least.

 

I'm looking at it more as an educational thing for my nephew, trying to get him to be more responsible for things.

 

I am interested, however, in your droplet device.

[Lgbeno 189]

Interesting, definitely like the educational opportunities for the kids, better to learn this stuff then spend all of their time playing mine craft!

 

Here's the details on droplet, the schematic ain't pretty but I didn't feel like spending time bringing it to a professional caliber.  I built up 6 of the original revision, found a few bugs taht were all fixed in this revision, it should be good to go.

 

https://upverter.com/imp-guru/0609e08f25aa7f3d/Humidity--Temp-Guru/

 

I just finished up the hub board, have not tested it yet

 

https://upverter.com/imp-guru/84ba884b5851ecba/Imp-Guru-Sensor-Hub/

 

The hub has the cc1190 frontend chip as well so I'm hoping for long range, we'll see what I can get.

 

for either design you can click on "Download associated files" and grab a zip of the gerbers which are perfectly formatted to submit to oshpark

[bluehash 1,581]

Interesting project, I'm actually doing something similar with cc110l and msp430G2553. I call it "droplet". It runs off of a cr2450 battery, sleeps most of the time then wakes up sends measurements to a Hub that is wall powered and the hub ships the data to data.sparkfun.com. Here's the droplet: ImageUploadedByTapatalk1412796802.981967.jpg

 

ImageUploadedByTapatalk1412796834.415790.jpg

 

And here is a hub prototype (just ordered a real pcb for it)

 

ImageUploadedByTapatalk1412797145.122317.jpg

 

You can see the electric imp that receives data via uart the pushes it to sparkfun.

 

My web app project imp.guru fetches the data from sparkfun to view, here's the data imp.guru/f1

 

 

I'm willing to share any/all of this work for your project if interested. My recommendation would be to skip the whole energy harvesting thing and use batteries. With the right design you should last years on a pair of AA's.

 

I've seen projects on hack a day where people use Hall effect sensors to pick up the spinning magnet in their utility water meter. I think this is it:

 

http://hackaday.io/project/1460-Remote-Water-Consumption-Display

- HiJack.

@@Lgbeno You should start a project thread. Nicely done!

[abecedarian 330]

@@bluehash - I don't mind the diversion.

 

Well anyhow.

 

Still waiting for the flow meter to arrive. I've been looking online at it and its brethren and realized I have a few AS5140 magnetic rotary position sensors here. They have an incremental mode that provides quadrature output signals so this may be an option to get more accurate flow readings over the basic hall effect sensor. Can't wait to see if I can get one of those working.

[abecedarian 330]

Flow meter arrived and it's, unfortunately, not going to lend itself over to using the position sensor. The sensor uses a hall effect sensor that is recessed into the housing, meaning the magnet on the shaft is too far distanced from the circuit board to be detected by the sensor I mentioned previously.

 

I have another sensor / generator en route and am going to be evaluating that.

[Lgbeno 189]

I've heard that MAG-3110 can sense these mag fields quite well

[abecedarian 330]

@@Lgbeno - thanks for the head's up on that.

 

This sensor can be modified to support two hall-effect sensors- the board actually has pin-holes for another sensor, so I am looking at that possibility. The idea of floating a magnetic position sensor over the end of the shaft is pretty much moot though, at this point. Suffice it to say that the o-ring and such separating the fluid housing from the sensor housing is enough to tell me I shouldn't muck with it.

 

 

I haven't said much about this for a while, but it is slowly morphing into an aquarium bio-filter / water temperature moderation thing. This seems much more practical a use than where I was going, initially.