# How to wire-up a flow sensor with Hall effect to operate a 12V relay

#### bowsprit

Joined Oct 2, 2018
51
I'm still assuming that the output is open-collector. I found a photo of the internal workings, and you can see the pull-up resistor, which goes to the positive supply.
So with a 5V supply, it will be TTL compatible.
With a 12V supply, it will have a 12V output, so my circuit will work.
I built your circuit as per diagram:
By just running the sensor under the tap, at what i guess will be the water flow in reality(when the engine in running) the alarm switches on /off very fast.
If I speed up the sensor, by blowing into it, the alarm stops ( a much faster rotation)
On the rotor slowing down( run out of breath) The alarm starts to switch on/off rapidly until a contentious sound is heard.
So the circuit does work.
What do you suggest I change , to set the threshold lower? thanks

#### Ian0

Joined Aug 7, 2020
7,445
Either the 100k or the 100n between the two 555s.
Larger = alarm at a lower flow rate.

#### JACKET002

Joined Sep 3, 2020
3
The sensor works by measuring the cooling effect of the gas passing over a thin, heated wire. The higher the cooling effect, the higher will be the flow of gas. In order to be accurate, the sensor needs to know the density of the gas, which is provided to the measurement system by the gas analysis data. The Hall effect is the production of a voltage difference (the Hall voltage) across an electrical conductor, transverse to an electric current in the conductor and to an applied magnetic field perpendicular to the current.

#### Ian0

Joined Aug 7, 2020
7,445
The sensor works by measuring the cooling effect of the gas passing over a thin, heated wire. The higher the cooling effect, the higher will be the flow of gas. In order to be accurate, the sensor needs to know the density of the gas, which is provided to the measurement system by the gas analysis data. The Hall effect is the production of a voltage difference (the Hall voltage) across an electrical conductor, transverse to an electric current in the conductor and to an applied magnetic field perpendicular to the current.
That you for that factoid. I’m sure it’s relevant in some way.
What we are dealing with here (in case you haven’t read the earlier posts) is a WATER floW sensor, which uses a paddle wheel, which has a magnet attached, the movement of which is detected by a Hall effect sensor, giving a pulse output for every rotation of the paddle wheel.
No gases, nor heated wires are involved.

Last edited:

#### bowsprit

Joined Oct 2, 2018
51
The sensor works by measuring the cooling effect of the gas passing over a thin, heated wire. The higher the cooling effect, the higher will be the flow of gas. In order to be accurate, the sensor needs to know the density of the gas, which is provided to the measurement system by the gas analysis data. The Hall effect is the production of a voltage difference (the Hall voltage) across an electrical conductor, transverse to an electric current in the conductor and to an applied magnetic field perpendicular to the current.

Now that's very interesting . I am sure you are correct for YOUR sensor. Thanks for the info..

#### Ian0

Joined Aug 7, 2020
7,445
Now we know that, shall we have a go at measuring the air intake to the engine next?

#### bowsprit

Joined Oct 2, 2018
51
Now we know that, shall we have a go at measuring the air intake to the engine next?
Great idea, If you can you email me the transverse hot-wire than I'll solder it to the flux capacitor.

#### bowsprit

Joined Oct 2, 2018
51
Ok........Results: --> I doubled R3 to 200K,--> It does not 'seem' to make much difference in stopping the alarm going of earlier than before.
Than I put in a 1M as R3 -->nothing worked anymore, --> replaced both U1 & 2 U, still nothing is working.
Replaced R3 (1M )back to 100K again nothing working. -->replaced U1 & U2 again, now it's working.

I tested the circuit for a few minutes, all seems OK. -->5 min later it's no-go again.

By now I am fed up replacing those smd 555's . (Every time I replaced a component, I checked for shorts etc).
I think that there is a vulnerability in the circuit,... does not seem to last., Maybe I'm wrong?, Maybe it's not my day, either way I need an other solution, preferably one without those tiny pesky 555 smd's..
Oh.. The last lot I put in are 555 DIP8, (Bonus with these, at least I don't need a magnifying glass)

#### Ian0

Joined Aug 7, 2020
7,445
Have you looked at it on a scope?
where does the supply come from and how stable/noisy it is?
how much supply decoupling do you have? Old bipolar 555 LIKE decoupling! (CMOS ones not so fussy)
I’m assuming, though I never did check, that the buzzer Current IS less than the 555’s maximum. If it’s a bit close, joining pin 3 to pin 7 will give you a bit extra.

#### bowsprit

Joined Oct 2, 2018
51
Have you looked at it on a scope?
where does the supply come from and how stable/noisy it is?
how much supply decoupling do you have? Old bipolar 555 LIKE decoupling! (CMOS ones not so fussy)
I’m assuming, though I never did check, that the buzzer Current IS less than the 555’s maximum. If it’s a bit close, joining pin 3 to pin 7 will give you a bit extra.
I'll try to answer your questions lin by line:
1)I don't have a scope, only multi meters
2)Supply come from 12v lead-acid batteries, fed by solar panels & wind gen. Nothing else(running) is connected to the batts.
3)Decoupling is zero, I built the circuit as designed by you, Only added a led & 330R resistor
4)Buzzer current 20mA(max) 3-24V SMF-27-1( https://www.aliexpress.com/item/32829087510.html) max current of 555 is 200mA .

All resistors are 1/4W the 100nF are 'green caps'. 4n7 is ceramic, Diode is 1N4007

#### Ian0

Joined Aug 7, 2020
7,445
Add 10uF electrolytic and 100nF ceramic across the supply and see if that sorts it out.

#### sghioto

Joined Dec 31, 2017
4,203
Getr you
either way I need an other solution,
Get the 8 pin LM2907 it will work using a coupling capacitor on pin 1. Excellent results with the schematic below.
SG
EDIT: Added R5, 47K resistor.

Last edited:

#### bowsprit

Joined Oct 2, 2018
51
Getr you

Get the 8 pin LM2907 it will work using a coupling capacitor on pin 1. Excellent results with the schematic below.
SG
Ok, LM2907 are on order. -->I'll try this solution also . Thank you

#### Ian0

Joined Aug 7, 2020
7,445
Could I suggest a pulldown resistor from input to ground? Just to establish a known DC level on the input. About 10 times the value of the pull up resistor.
(It will also need some supply decoupling - its of the same age as the original NE555)

#### bowsprit

Joined Oct 2, 2018
51
Add 10uF electrolytic and 100nF ceramic across the supply and see if that sorts it out.
I made a circuit board for 555 testing, as yesterday i was guessing which one was faulty.
That sorted, I got the 555 circuit running again....

I have added the 2 x 100n green caps to both 555 ic's and a 10uF to the 12v+ in & ground, and increased the 100K to 1 M.
The circuit has been running for about 1 hour and is working fine (switching on & off) .
The 1M resistor shuts the buzzer of, at 3 liters/min.
So I'll put a 1M trimpot in. And when I get my engine up and running, I can adjust to suit the amount of water flowing through.

Than when the LM2907's arrive (coming from China), I'll build that circuit also with all recent suggestions added from sghioto and Ian0 .

I will report back here, when it is all been up and running for a while and have used both circuits..
In the mean time many, many, thanks for all your patience, help and time spent gentlemen.

In the future I use

#### Ian0

Joined Aug 7, 2020
7,445
A pleasure to be of assistance.
In the future I use
What do you use in the future?
May I recommend the CMOS versions of the 555?

I wonder how many people have been put off electronics, because their first circuit was a simple 555 oscillator and it failed due to lack of decoupling.

If you look at the internal circuits of the old NE555 (you need an old datasheet - they have stopped publishing the internal circuits for new devices) you’ll see it has a lot in common with LM393, LM358 and LM2907. They are all of the same age, and They all like a lot of decoupling! The 741 is unrelated.

I learned about the LM2917 in the early 80s when everyone my age had an old Mini and they all wanted a rev counter.

Thinking about Jacket’s post - if you get an NTC thermistor and put enough current through it that it heats up (probably doesn't need much), the flow of water would cool it down and the resistance Would increase. If it works it would just be a simple comparator circuit. . . .

#### sghioto

Joined Dec 31, 2017
4,203
Than when the LM2907's arrive (coming from China), I'll build that circuit also with all recent suggestions added from @sghioto and @ Ian0 .
Just a heads up, I added a resistor R5, 47K ( post #72 ) to buffer the relay driver circuit from the LM2907.
SG

#### bowsprit

Joined Oct 2, 2018
51
I added the recommended components (By @sghioto and @ Ian0 . ) to the attached (LM2907) diagram and a Cermet pot-meter (R5), I hope I have done all this correctly?

Also, I like to get rid of the relay and use a 2N2222 transistor instead to switch the buzzer off.
Can anyone please advise on how to change this. Thank you.

#### Ian0

Joined Aug 7, 2020
7,445
We seem to have lost C1 from the input and a different capacitor has now become C1 , but apart from that it Looks correct to me.
The buzzer just goes where the relay coil was - simple as that.
I didn't know that they still made 2N2222 - so I checked in Farnell, and they are still there after all these years, but over £1 a piece. There are 10p transistors that will do the job!
Alternatively, a small mosfet (2N7000) would mean that R7 was not required.