Aiber Water flow sensor

[bushtech]

This water flow sensor stopped working .It's part of the control mechanism in a gas (LPG)geyser.

The 3 legged magnet sits in the pipe to the left of it. It senses water flow by by being forced up the pipe to in proximity to the 2 green wire bit which is a switch that gets closed when the magnet gets near it. Water flow through device is from the 3 leg magnet to the outlet near pcb. Normal orientation of device in geyser is with inlet facing down.

Here is other side of pcb:

I have checked the 3 leg magnet switch. It does close and allow current to flow when water flows through device.
The 3 leg transistor thingy sits in proximity to the magnet on top of the rotor wheel which spins when water flows.It does not sense the rotor magnet spinning so suspicion focused there. Very difficult to read anything on the transistor thingy but it could say M42.

I wouldnt expect there to be too many variants of that 3 legged flow sensor. Can somebody point me to some. Not too sure what to google for
That is if I have diagnosed this correctly

 

[Harald Kapp]

That "3 leg transistor thingy" is possibly a Hall effect sensor. I could not find one that matches the supposed "M42" label, but for example SM451R.

 

[bushtech]

Thank you so much Harald. I hope you are spot on. I think all I can do now is the"suck it and see" approach. I have found these locally. Although rather expensive better than having to buy 3000
https://www.mantech.co.za/ProductInfo.aspx?Item=14M5636

 

[73's de Edd]

Green wires associated with reed switch and its round linear magnet.
Then for the hall effect :
Blk . . .pwr gnd
Red . . . pwr to hall . . .probably +5vdc
Yel . . .expect pulses out, as the rotary magnet on the end of the impeller rotates.

 

[bushtech]

Thanks St Edd. In what format would these pulses be on the yellow line

 

[73's de Edd]

Just one pulse after the other, as the round magnet rotates.
Use a mini iron /steel screwdriver shaft to run around the periphery circumference to count the poles.

 

[bushtech]

Thanks St Edd. The only way I'm going to know that my replacement hall sensor is working correctly is by taking readings on the new one and comparing them with readings on the fixed one. But then I need to know what sort of signal to look for. Voltage spike, current spike, some 5v on off business?

 

[73's de Edd]

I would expect a "some 5v on off business? '" situation in accordance to the rotational position of the units round magnet . . . . .unless it incorporates a schmitt trigger . . .sort of unexpected, though.

 

[bushtech]

Would a scope be able to show it?

 

[73's de Edd]

AB-SO-ROOT-ELY ! . . . . .
( Particularly, if being auto triggered and used in its appropriately ranged . . . DC coupling mode . )

Aside:
I jes now dun gots mesefs some mores smarts !
I now dun noes whuts a " quantity surveyor " iz be in.

 

[bushtech]

Ok then
Newsflash!
I am now the proud owner of this:

But I am still an absoloot noobie when it comes to using this baby so a bit of help on where to connect the 2 bits of the probe will really come in handy.
Before I have to start a new thread on how to fix a scope

 

[Harald Kapp]

You connect a probe with BNC connector to one input, e.g. "INPUT CH1".
With the controls for "VERTICAL (V/DIV) you set the sensitivity and the 0 V reference position of the trace.
With controls for "HORIZONTAL (TIME/DIV)" you set the resolution with respect to time. Depending on the angular speed of the magnet probably something around "10 ms /div" is a good starting point. If the pulses are too far apart, reduce resolution in time to a higher number (e.g. 50 ms/div). If the pulses are very close to each other, increase the resolution to e.g. 2ms /div or lower.

This description may get you started better.

 

[bushtech]

Thanks Harald
Trying to get it to work. Does the earth wire of the probe go onto the common earth. Black wire?.
@73's de Edd There is also a green wire. It's a 4 pin plug.

 

[Harald Kapp]

The tip of the probe you place onto the signal you want to measure.
The ring is ground and usually goes to ground of the system being measured.

Note two things:

1. The grounds of the probes are connected to each other and the internal ground of the oscilloscope. Therefore do not connect them to different potentials. This would create a short circuit.
2. Ground of most scopes is connected to protective earth of the mains plug. Therefore when measuring in mains connected circuits take great care where you attach ground of the probes to avoid a short circuit. If you encounter strange signals with a mains frequency component, stray currents through the scope's earth connection may be the cause. In that case an isolating mains transformer may solve the issue. Or use of a battery powered scope.

 

[73's de Edd]

I would try the simplest possibility first.
By hooking up AC metering mode between the Black ground and Yellow pulse out wiring and manually slowly rotate the impeller to see if there is outputting.
Or actually, just try both of the loose rod magnets ends to come up to and move away from the hall sensor and see if it outputs on the Yellow connection.

BTW . . . .. here is the manual and schematic for that beast.
I last met up with one circa 68-70' ish.

http://bama.edebris.com/download/nonlinearsystems/ms215/ms215.pdf

 

[bushtech]

Thanks Harald
The scope is battery powered which I suppose makes it a bit less risky. Connecting the probe between the green and yellow wires I am starting to see some blips on the scope. But I need constant flow through device , blowing through it is not exactly a constant flow. Now busy McGyvering some constant air flow through device without involving wife's hair dryer. Without that I cant get repeatable readings that I can compare with the replacement hall sensor.

St Edd
Thanks for that. Fortunately I did get the manuals with the scope and its accompanying Touch Test 21 multimeter.

 

[73's de Edd]

Connecting the probe between the green and yellow wires I am starting to see some blips on the scope.


My gut is seeing the GREEN as being an isolated SPST switch circuitry

Whereas, the Hall related circuitry is separate . . . put scope grounding clip to BLACK and expect pulse outputting on probed YELLOW . . .if the unit is powered up with Red as +5VDC.
And you are magnetically activating the hall with a magnetic field presence going into and then out of its sensing range.

PEE ESS . . . .
Doth mine tired eyes . . . see a 250 Mhx scope probe that cost more than the current monetary evaluation of that whole little "Un Linear " Olde-Silly Scope "

https://www.tiepie.com/en/usb-oscilloscope/accessories/oscilloscope-probe-250mhz-x1-x10



.

 

[73's de Edd]

22.08 South African Rand equals 1.49 United States Dollar

That sounds reasonable to me . . . but . . . . BUT . . .BUT ! that unit that your tech da-da sheet is referring to, is being a LINEAR unit.

In your flow sensor device application , one would expect that unit to be using a digital type with a definite snap on - snap off action. . . . . .so it would have an internal schmitt trigger function design included within its outputted signal flow processing.

To test on your original unit . . . . just use DC meter functioning in ~ 5VDC range and meter between black ground and Yellow output.

Then introduce a magnets field from afar on up to kissing the hall sensor face, then try with the other pole pieces magnetic polarity.
If having a linear unit, there will be a graduate and proportional changing of reading in accordance to the magnets proximity.

If having a digital / switch type of unit there will be a definite snap on - snap off type of voltage reading at their detected two magnetic proximities.

And thats just the way it is . . . today . . . . here on FEB 8 2021 . . . .

ADDENDA . . . . . last post
PEE ESS . . . .
Doth mine tired eyes . . . see a 250 Mhx scope probe that cost more than the current monetary evaluation of that whole little "Un Linear " Olde-Silly Scope

I initially made out the units HP9??????? prefix . . .but then saw the downgrade to a 60 MHX hi end spec . . but even that . . . exceeds the needs of a mere 200khz bandwith / (audio scope+ )


73's de Edd . . . . .



.

 

[bushtech]

@73's de Edd

Notes: Geyser uses 2 x D size batteries =3V How do we get 5V
Scope probe HP9060 Current price in SA about $17
So how do you spec a scope? Vertical bandwidth = 15MHz
Horizontal bandwidth = 200kHz

I am still battling to get constant airflow through device.
Current iteration using a small battery powered vacuum cleaner

 

[bertus]

Hello,

It can well be that the flow sensor will not work for air.
The wheel looks to small for air.
The sensor is designed for liquids, as you said in the begining of the thread.

Bertus