In Search of: Cheap/common pressure switch

[NTL2009]

I'm certain these must be common in automotive applications, but I'm just not sure what search terms to use.

I'm looking for an on/off switch (to be electronically sensed, so low volt/current) that can be connected to a water pipe, and turn on/off at around 1~2 PSI.

The specific application is to monitor (with an Arduino-based NodeMCU) how often and how long my sump pump runs. Since it has to pump the water up ~ 10', the pressure in the discharge pipe (even w/o any added pressure from flow resistance) would be over 4 PSI. The discharge pipe is common 1 1/2" PVC, easy to drill/tap.

My first attempt was with one electrode in the pipe, just below the check valve, and another in the water at the bottom of the pit. But the check valve keeps the water trapped in that pipe, even with a 'weep hole' at the bottom (like putting your thumb on a straw).

I bought some AC current sensing modules, but I have to break into the line, or the cord to use them. There's not enough room in the box, and I don't really want to cut up the cord to isolate one line (void the warranty), so that gets a little messy (but do-able if needed).

I have three pumps in the pit, (1) a battery backed up one, (2) the (formerly) 'main' 1/2 HP pump, and I added a 3rd, cheap 1/6th HP pump that should be able to handle normal conditions, and save wear/tear on the 1/2 HP pump (which runs for only 5 ~ 7 seconds per cycle, and the integrated float is not adjustable). So a simple, common solution for all three would be preferred.

In the mean time, since my 1/6th HP pump uses an external float/Switch, I can put a multi-tap on that with a USB power tap to feed the input of the Arduino, that's a partial solution.

TIA - NTL2009

 

[jrmichler]

I bought some AC current sensing modules

Two suggestions:
1) Current sensing modules are typically installed in the circuit breaker panel. You need to use proper wiring practices inside the panel and for the low voltage wires coming out of the panel. If in doubt, call an electrician.

2) There exist pressure switches that switch at the low pressures of your application. Here's one: https://www.omega.com/en-us/pressure-measurement/pressure-switches/p/PSW-681-Series. I have had good experience with that supplier. They have other pressure switches. The one in the link is the first low pressure one that I found, then I stopped looking.

 

[NTL2009]

Thanks, but at $155 (and I need 3), it doesn't exactly fit my neighborhood of "cheap"! :) I'm thinking ~ $10 each?

When I search, I get inundated with the well pump switches at 45~60PSI. I tried automotive, can't find anything below 10 PSI.

I'm not sure I can get the resolution to detect the 1/6th, and 1/2 HP pumps separately and/or in combination. So I think I need a separate current sense on each, which I'd need to do at the outlet, not the main panel. Or I might just make a short extension cord, so I can split out one lead, and keep that sensor external, though technically, I don't think extension cords like that are approved for permanent use?

Again, doable, just looking for the same setup on each. The battery back up would draw from battery not the line, so that would also be different (but it also has a relay contact closure terminal, so that's easy anyhow.

 

[Baluncore]

I'm looking for an on/off switch (to be electronically sensed, so low volt/current) that can be connected to a water pipe, and turn on/off at around 1~2 PSI.

Step back and consider a third way. You could fit a level-sensing float-switch in the top of the rising outlet. That would not require power, and costs only $5.00

Pick the point close to the top of the outlet where the water is rising or level, so it would read high while a pump was running. When all pumps are off, the outlet could drain out and away from the switch to signal pump off.

 

[NTL2009]

Step back and consider a third way. You could fit a level-sensing float-switch in the top of the rising outlet. That would not require power, and costs only $5.00

Pick the point close to the top of the outlet where the water is rising or level, so it would read high while a pump was running. When all pumps are off, the outlet could drain out and away from the switch to signal pump off.

OK, that could work, but the outlet is pretty close to the floor above it, not sure if I can easily get a float mechanism in there. But my electrode idea would probably work just fine there, as the outlet pipe does drain out when the pumping stops (we have freezing weather here, so that's important). I'm not too worried about corrosion on those electrodes from the sample current, they normally won't be submerged, and I do a quick sample (a couple milliseconds every one second) so average current is tiny.

But it wouldn't tell me which pump was running. I really do want to know if the 2nd pump needed to assist the first pump. It also would not tell me if the pump was running, but not pumping (I once had the rotor slip off the shaft of the pump).

Eventually, I plan to have this email me if an out of ordinary condition occurs, but that has other complications, a bridge I will cross later.

 

[Tom.G]

A couple brain-storming approachs:

How about a microphone feeding the Arduino? You may be able to detect which pumps are running by the sound.

Or a vibration sensor (maybe a contact microphone) mounted on each motor.

Those at least avoid working with the power line voltage.

Cheers,
Tom

 

[NTL2009]

A couple brain-storming approachs:

How about a microphone feeding the Arduino? You may be able to detect which pumps are running by the sound.

Or a vibration sensor (maybe a contact microphone) mounted on each motor.

Those at least avoid working with the power line voltage.

Cheers,
Tom

Thanks, that thought crossed my mind, I haven't experimented with it yet, but I think it is do-able. Each pump has a rubber hose connection for it's check valve, so there would be some isolation. This board (Node MCU) has only one analog input, which I'm already using for a pot as an input for settings, so I'd need to use an op-amp on each microphone/vibration sensor to turn it into a clean digital on/off (I'd probably need an op-amp for the microphone any how though).

Though I really like the idea of a simple pressure switch, that's low voltage and easy to interface to some 'dumb' alarms.

 

[Tom.G]

This board (Node MCU) has only one analog input, which I'm already using for a pot as an input for settings, so I'd need to use an op-amp on each microphone/vibration sensor

Naw... just use a three-input summing amplifier configuration, and you can set the gain individually for each input.

For an example see:
https://www.electronics-tutorials.ws/opamp/opamp_4.html

Cheers,
Tom

 

[NTL2009]

Naw... just use a three-input summing amplifier configuration, and you can set the gain individually for each input.

For an example see:
https://www.electronics-tutorials.ws/opamp/opamp_4.html

Cheers,
Tom

I know about summing op-amps, but I fail to see how that would work here. The single ADC input is already connected to a pot that adjusts from 0 V to +Vcc which I use an a selector 'switch' for various things (setting the clock, clearing logs, set which info to display, etc).

Were you thinking each of the 3 pumps would have a different gain, and I'd know which one(s) are running by their combined voltages? That sounds kinda tweaky (I would expect the vibration sensors to be pretty noisy/variable), and I don't see how it can work with the pot.

At that point, I'd rather add an analog multiplexer, I've even got some on hand.

At any rate, I think the way I'll sense pump # 2 (which has an attached, non-adjustable, hardwired float) like this:

https://www.ridgid.com/us/en/1-2-hp-stainless-steel-dual-suction-sump-pump

is by attaching a metal rod or stiff wire to that switch arm to a micro-switch. That arm has hysteresis to it, the float flips it up and it stays up/ON until the float slides down that rod and pulls it down and it 'snaps' OFF.

 

[Averagesupernova]

So what's the problem with measuring the AC current? Has it not occurred to anyone that while you may not want to cut into the cord and not want to get into the building wiring, you can wire a 4x4 box with a heavy duty cord and receptacle on it and incorporate your current sensor into it. It's relatively simple to do. Make sure it is not possible for it to fall into the water the same way you would avoid that with any power strip or extension cord.

 

[NTL2009]

So what's the problem with measuring the AC current? Has it not occurred to anyone that while you may not want to cut into the cord and not want to get into the building wiring, you can wire a 4x4 box with a heavy duty cord and receptacle on it and incorporate your current sensor into it. It's relatively simple to do. Make sure it is not possible for it to fall into the water the same way you would avoid that with any power strip or extension cord.

Not really a problem, I've thought about doing it as you suggested.

It's just that the idea in my OP, a simple pressure switch to sense the pressure of the 10' head when the pump is on would be the simplest in my mind, and probably the cheapest if I could find a common, generic part that does this. Every pump would have the same configuration, they all need enough pressure to raise the water that high, regardless of gpm output, current draw, etc, no tweaking.

 

[Baluncore]

It might be possible to sense the position of the non-return valve by fitting it with a magnet and an external reed switch, or a hall effect sensor.

A differential pressure sensor from a washing machine, across the NRV for that pump would identify the state of the pump.

Maybe this cheap differential air pressure switch could work;
https://www.ebay.com.au/itm/284612251527

 

[rbelli1]

Sense the water capacitively. Then you don't have to break into the pipes or the wires.

BoB

 

[NTL2009]

Sense the water capacitively. Then you don't have to break into the pipes or the wires.

BoB

How so? Do you mean the water level? Or capacitance change of the PVC pipe, with the dielectric change between water and air?

One problem I found, and discussed above, is that with the check valves (needed to isolate the pumps from back-feeding each other), the water stays in the pipe, even below the check valve - like holding your thumb over a straw.

And water level is a bit tough to get calibrated to the pump run time. The water can sit just below the turn on threshold for many hours. You could look for deltas, and assume that dropping fairly quickly is the pump running, but what if there is a problem, and the water level isn't dropping?

With multiple sensors, I'll know that if the second comes on, that there is a problem with the first pump, or a very hard rain event.

I'm leaning to a micro-switch for the 2nd pump (the battery back up pump has a relay output I can use). It just seems the simplest at this point. But I appreciate any and all suggestions, it's good to get the mind thinking.

And as a side note, over the past two days, I got email/text alerts working. Finding the information was harder than actually doing it (Google made some security updates that take effect this month). But now I get emails 4 x a day (just because!) with the past 48 hours history of the # of cycles and total run time per hour for the first pump. And (when I get the sensors connected), I'll get an immediate email/text if either of the other pumps comes on, and again when it shuts off.

 

[Baluncore]

Consider a clip on current transformer to monitor on/off pump current.
"Non-invasive AC current sensor Split Core Transformer".
https://www.ebay.com.au/itm/363428381109

 

[NTL2009]

Consider a clip on current transformer to monitor on/off pump current.
"Non-invasive AC current sensor Split Core Transformer".
https://www.ebay.com.au/itm/363428381109

That's exactly what I bought, and had planned to use. It certainly would work, but it does involve splitting the cord, or getting into the electrical box at the pump to get to one wire on each pump, and there isn't enough room for that clamp, so I'd have to add on another box, or have them hanging from split cords.. I'd kind of prefer not to have to undo much of anything if I want to remove this monitor (like if I sold the house). I'm not sure what a home inspector would think of that stuff.

I can more easily pull out a micro-switch, or remove the external wall-wart power supply from the external switched float on the 1st pump. I guess I could make a short extension cord for each, that would keep the current transformer external and easily un-done.

And I'd probably need a little circuitry to get the AC output from the current transformer to reliably be detected by the digital inputs of the micro-controller. Easily done, but just one more thing that I don't need with a switch (I used a simple voltage divider from the wall-wart power supply to match the digital input level).

 

[Averagesupernova]

That's exactly what I bought, and had planned to use. It certainly would work, but it does involve splitting the cord, or getting into the electrical box at the pump to get to one wire on each pump, and there isn't enough room for that clamp, so I'd have to add on another box, or have them hanging from split cords.. I'd kind of prefer not to have to undo much of anything if I want to remove this monitor (like if I sold the house).

I gave you an option in post #10. Did you not understand it?

 

[NTL2009]

I gave you an option in post #10. Did you not understand it?

Sorry, didn't see the reply - yes, I understand, an external box would do the trick. For other reasons, I'm going to try the micro-switch on the pump#2, and if that isn't simple, I'll take this approach.