Two options to monitor current of my 12v pump, would these work?

Thread Starter

zirconx

Joined Mar 10, 2010
171
I'd like the monitor the current consumed by my 12v backup sump pump. I'll use this to determine if the pump is working. I will already be monitoring the float switch, so I'll know if it should be on. Comparing this to the current sensor should tell me if it's working correctly (if pump relay is on, but current consumption is not near the expected amount, then something is wrong). FYI the pump consumes around 13A when pumping.

I see two ways to get this data into my ADS1115 and then into the raspberry pi.

One would be a ACS712 current sensor. Unfortunately this is a 5v device so I'd have to add some complexity (a voltage divider?) to get the voltage into the range of the ADS1115 (which is a max of 3.6v since it's running on 3.3v to communicate with the raspberry pi). FYI a 20A version of a ASC712 circuit board outputs a base of 2.5v + 100mv for each amp.

Another idea I had was to measure the voltage drop between the battery and somewhere else in the circuit. I just used my voltmeter now and was able to see a .45v difference when the pump is running. Unfortunately when the pump is off I read the full battery voltage. I think this is because the relay switches the positive wire to turn on the pump, rather than switching the ground wire. So I'd have to figure out how to not to send the full 12v from the battery into the ADS1115 when the pump is off, or rewire my system so the ground is switched instead of the positive.

Thoughts? Thanks.
 

Thread Starter

zirconx

Joined Mar 10, 2010
171
Thinking about this more... from what I can tell the current goes through the ASC712 chip? That worries me a little because I don't want to introduce failure points into the system.

What do you think about using an INA219 break out board, and replacing the shut resistor with two wire leads that go to my +12v battery and somewhere else in the system, near the pump? Then I am using the main current delivering wires as a shunt. Of course it won't be calibrated, but I should be able to tell if current is flowing or not?

In this case the Vin+ and Vin- leads would not be connected.
 

Attachments

Dodgydave

Joined Jun 22, 2012
11,284
3 Shotkley diodes in series with your pump and an led with 100R in parallel with the diodes will indicate when the pump is on, if you need to measure the precice current then a current pcb monitor will work
 
Last edited:

LesJones

Joined Jan 8, 2017
4,174
If you use an INA219 you will need to use a micro controller to read the data from it's I2C interface. I have used a few of these with PIC12F1840 chips and HC12 modules to monitor voltage and current remotely. The system works well. Dave's system will work but it will reduce the voltage to the pump by a small amount. You don't give us any idea how much current the pump draws. If it is quite high a few turns of one of the wires to the pump wrapped around a reed switch would work.

Les.
 

Thread Starter

zirconx

Joined Mar 10, 2010
171
...You don't give us any idea how much current the pump draws. If it is quite high a few turns of one of the wires to the pump wrapped around a reed switch would work.
Les.
It's in my first post but I can see how it's easy to miss. It's 13 amps. I like the turns around a reed switch idea. Do you think that's enough current? I probably want a larger (physically) one, right? So I can get enough turns around it? The wire is probably 14g, so I might only be able to get 4-6 turns around it. Maybe these? https://www.ebay.com/itm/113949820199

What about a parallel circuit that grounds through the pump, a transistor or optocoupler or a relay? I'd have to switch the circuit around so I'm switching the ground instead of the positive input.
 

Thread Starter

zirconx

Joined Mar 10, 2010
171
How about INA170 ??
Thanks for the suggestion. Looks like there is not a pre-made breakout board available for this ic. On my circuit building abilities - I'm comfortable wiring up a switching transistor, or an optocoupler, for example, but including this bare ic in my circuit might be a little beyond my ability, and perhaps more work than I want to take on.
 

LesJones

Joined Jan 8, 2017
4,174
The reed switches in your link from post #6 do not give the ampere turn rating for them. I tried one I had in stock and that required about 40 ampere turns to close. I looked at reed switches on Farnell's website and the first one I found was rated at 20 ampere turns. You would probably need two reed switches. One to check the motor was taking enough current and a second one to check it was not taking too much current. Have you considered using a flow switch on the output of the pump or an optical sensor to detect that the motor was rotating ?
I do not understand what you are suggesting in the last sentence of post #6

Les.
 

Thread Starter

zirconx

Joined Mar 10, 2010
171
The reed switches in your link from post #6 do not give the ampere turn rating for them. I tried one I had in stock and that required about 40 ampere turns to close. I looked at reed switches on Farnell's website and the first one I found was rated at 20 ampere turns. You would probably need two reed switches. One to check the motor was taking enough current and a second one to check it was not taking too much current. Have you considered using a flow switch on the output of the pump or an optical sensor to detect that the motor was rotating ?
I do not understand what you are suggesting in the last sentence of post #6

Les.
I ordered some reed switches on Ebay. I couldn't find any that listed the ampre turns. I'll just experiment. I think I can switch to 18g wire for the magnet portion of the wire and get as many turns as I need.

Regarding measuring too much current - I have a 15A fuse in place to handle an over current condition. The fuse would blow, then my relay & reed switch tests would tell me there is a problem.

As far as pump output - I plan to measure the amount of water in the pit using a submersible pressure transducer. So I should be able to tell if water is leaving the pit. (My post about that sensor is here)
 

dendad

Joined Feb 20, 2016
4,451
The reed switch idea was popular years ago for car lamp monitoring so you could tell when a stop light blew, for example.
It may be worth adding a second reed with different turns calibrated for motor overload current so if the pump gets something jammed in it you can detect that.
I do prefer the ACS71x method myself. Then you can read the actual current. But if the reed is doing what you require, go with it.
 
Top