Posts made by zboblamont

@Tmaster Never assume the original PCB was just a circuit of components, the PCB can incorporate RF shields and antenna radio ground planes...

@zboblamont OK, so experiments concluded that the return data from the latest version of Domoticz is a String whether it be V_VARx or V_TEXT.

I tried the dual declaration using the same CHILD_ID as intended and as VAR_X (using an RF24 - not sure of the relevance) and the actual value sent but failed to make any sense of abstracting the Ulong conversion or the string.
Only by presenting a fresh CHILD_ID and duplicating the send of data (water meter) as a V_VAR1 could I recover the last sent reading as a V_VAR1 from that fresh CHILD_ID, thence eventually arriving at 'strtoul' as a solution to restoring the last known value.
For clarity my Gateway and Controller run on a UPS, and I hope shortly to add the spare router to that UPS so I'm not left in the dark on what is happening in the dark on a battery powered laptop...
Hope this helps others trying to keep track of meters between relentless power cuts and reboots...

@Tmaster It would certainly imply the PSU or mains are causing noise to appear, but it could be on anywhere in the combination, but it's not necessarily at the rfm itself.
I'd surmise your problem is related to proximity of the psu or mains power.
eg - I have 3 un-shielded nodes with rfm69 variations running on 5v wall wart PSUs and battery backup and never had a single problem - The difference to your scenario is that the closest PSU and mains is 250mm away.
One has been running on an old Nokia "charger" and USB breakout now for 6 years straight between power cuts (and christ almighty we do get them) yet it is still going strong.

Perhaps reconsider your chosen method of supplying shore-power...

@Tmaster What I meant by my previous comment was to tap on the Node name to see what sensors were connected - Had you clicked on TANQUE you would have seen a box pop up below giving you all the children to that particular Node with what value was last updated and when, so you could easily have found out all about your mysterious 13 children.

At least you have it sorted now, but it pays to do a bit of housekeeping in the sketch setup and Domoticz - eg which of the 6 "Unknown" Nodes is the one you want to next have a look at ?

All my Node IDs are set at the Node, and the sketch name labels the Node ID also, the Nodes are named in Domoticz for easy identification.

In the sketch my sensors are numbered sequentially and commented on what they are, so naming them in Domoticz becomes very much easier.

@Tmaster Ah, didn't spot that...
My next thought was it was not enabled under Devices in Domoticz but you said nothing shows up, yet you presumably see that specific child updating on the Domoticz log at intervals ?

Have a look at Hardware, click Setup for the Gateway line, then the Node, and verify what is there is what is expected for child 99.

Something's not making sense here.

@Tmaster The problem cannot be your gateway if some messages are being passed but not others, it is likely an issue at the Node or has not been enabled at the Controller.

Most Controllers have a live log of what is coming in, and a devices table from which you can see what was last updated and when.

I suggest editing the Node sketch to print locally over serial to verify the results you are getting then investigate from there.

Long story short, what is returned by Domoticz when a V_VAR1 is requested back to the Node ? What was sent as a Long originally, or does it return a string ?

@Tmaster Oh well, told you it was a wild guess I'm using default settings of D2 here so never encountered the need to find out.
You need complete separation of the interrupts, I'm not sure the gyroscope will function without it.

Looking again at https://www.mysensors.org/download/sensor_api_20 for perhaps a better understanding, I note the explanation of both"#define" lines you originally posted which now makes sense. Pin D3 is indeed INT 1, so you were redefining both the pin and the interrupt.
So long as these are set before including the MySensors library it should indeed override the defaults.

Not sure if it's relevant, but could it be something to do with the rfm69 driver ? Many of the guides relate to the original driver.
I recently upgraded to the NEW rfm69 driver thinking it was more efficient, but with the Gateway and 2 Nodes reprogrammed, ended up with repeat messages and the system slowing to a crawl.
Reversing the procedure to the original driver restored system reliability, speed, and sanity.

@Tmaster Sorry, I should have been clearer, but I have little knowledge of the details

• I you do a search on Google (I haven't checked if the site search function has been restored) for that entire define statement you will land here https://www.mysensors.org/apidocs/MyConfig_8h.html
  Scroll down and you will find "#define MY_RFM69_IRQ_PIN DEFAULT_RFM69_IRQ_PIN" as one single line statement which I presume you append it with "3" to redirect the pin and solve your current problem - Perhaps try that first.

I've no idea what the "#define DEFAULT_RFM69_IRQ_NUM 1" does at all as it sounds like it's altering an internal RFM69 register.

@Tmaster I'm making a wild guess the define is wrong
See #define MY_RFM69_IRQ_PIN DEFAULT_RFM69_IRQ_PIN
Define this to override the default RFM69 IRQ pin assignment.

@Jalina-He Having re-read all 6 of your contributions since you first joined this forum 11 days ago to make sure, it appears you have no interest in nor experience of this hobby.

On your latest odd and final question I am prepared to answer - Plastic can be recycled when it is no longer of use - Batteries - The PCB is as good as the day it was installed - PCB's don't become (sic) overdued.

@Jalina-He Why would I wish to design a product when commercial manufacturers already manufacture them for other functions which can be re-purposed ?

I have no vested interest in Gewiss, but I do know from experience that these weatherproof light switch boxes work exceedingly well, which is why I first commented.

eg - My Gas meter node has been in such a box for 6 years now with no signs of any damage to any of the unprotected parts within it, or degradation of the plastics over 6 winters and summers - The only interruption to continuous service has been changing batteries or reprogramming the Node.

@Jalina-He I'm not sure I understand your question over position of the ultrasonic board - It is mounted forward to enable quick access/replacement via the hinged front access, there is no advantage to pushing it further back.

The hinged access door and the two parts of the box have waterproof seals - The clear plastic window allows quick inspection and to see what leds are blinking - The mounting points on the rear box are on the outside of the sealed compartment.
If Gewiss is available in your market I can totally recommend these waterproof lightswitch boxes.

@Jalina-He If that question was directed at me - No, it is not new, but 6 years old but as yet unused - You can see the original 4 module tangs still uncut through the clear plastic window.

• I bought 2, 3 and 4 module cases at the same time and tried slightly cheaper copies which failed over time (leaked, transparent plastic degraded), the Gewiss originals remain almost as new despite years of exposure to snow, rain and UV - Higher quality.
  eg - This is the smallest one used, a 2 module unit serving as a wiring junction box between the pump station and the Node, with an ultrasonic sensor control board mounted in the front section for quick swap should it become faulty.

@รอเร-อ I've used Gewiss modular light switch boxes for years and no problems experienced - Winters go down to -20 and summer up to 45, and no appearance of condensation.
The front section has a clear flexible plastic window which hinges at the top, the front and main box are separated by a gasket, by undoing 4 screws.
I separate the front from the back then cut off the modular catch tangs with a dremel to give more space.
Inserting the antenna down through one of the grommets makes it totally weatherproof.
If opening it up in damp or humid weather, I chuck in a dessicant bag before closing it to mop up any residual moisture in the box.

That's good news.

Moisture would not necessarily have damaged it permanently, it's more a case of humidity can play havoc with tiny electronics and wiring when it condenses.

The advantage of a weatherproof box in my environment is obvious, but it has advantage elsewhere as it's a controlled and sealed environment inside - If closed up dry it stays dry.

@Oumuamua Sorry but I've no idea - I didn't update the MySensors side of the system once everything on the Gateway and Controller were working perfectly 5 years ago - Only Domoticz updates on the Pi Controller have interrupted it, the arduino gateway and disc drive have run 24/7 through multiple power cuts on the UPS.

It does seems unlikely the driver is at fault given so many others updated and none have reported such a bug.

IIRC the new driver has to be installed on ALL devices using the rfm, so I'll update the Gateway in summer when swapping out the UPS battery and drive.

@Oumuamua I strongly suggest your problem is moisture related, been there done that.

My external nodes are all in weatherproof boxes (Gewiss surface mounted modular light switch box with the transparent hinged front cover - I dremel off the switch tangs) with a roasted clay desiccant pack thrown in to cope with any moisture which may creep in on the cables over winter, then the front is latched shut.

No oxidisation is possible so long as the box is weatherproof (sealed), as any internal moisture is absorbed by the desiccant (a reactivated silica gel pack will work) when you close it up.

My gas meter node was like that, running for 4 winters down to -20 and never skipped a beat - It was like new when I changed the batteries, and ultimately took it apart, zero corrosion.

It's an awfully convoluted way of doing things but suspect your error lies in misunderstanding calculations mixing ints and floats and something screwy with your resistor bridge - It is always worth checking the bridge with a multimeter and raw voltage to verify the expectation.
If you have 3v at the top of your voltage divider, you should be getting 0.9593v applied on the ADC pin which will be read as 892 against the 1.1v internal reference.

The easiest way I found was to define a multiplier needed to derive the raw voltage, in this case (((R1+R2)x1.1)/R2) - For your resistor arrangement the max you can read is 3.4404v before exceeding the 1.1v internal reference.
To reverse the ADC reading of 892 to raw voltage is (892x3.4404)/1023 = 2.9984 v.

Perhaps this is a simpler way to do it

float MULTIPLIER= (((1000+470)*1.1)/470);//Resistor bridge values in k and Vref
int sensorValue = analogRead(BATTERY_SENSE_PIN);
float batteryV = (sensorValue * MULTIPLIER)/1023.0; //Note the trailing decimal point on the 1023 for calculations involving floats

@Doubletop Thanks for that.
I haven't range limited the readings on this version (ie ignore bad results) but do check for two consecutive readings, thus far no bogus readings recorded.

In my own case, the 3v node had no spare pins, so had to expand anyway to avoid using to a second dedicated node.

Having read on the 5v returning stable results, went with a second pro-mini at 5v controlling the v3.0 (also enabling further expansion) running off a PSU, talking I2C to the Node.

My principal problem was winter condensation on the transducer head due to the pipe used, aside that the v2.0 ran two years before finally dying.
This time it's on plastic pipe and the v3.0 is socketed for easy swap should it go faulty, fingers crossed for a trouble free winter.

@Doubletop Interesting, as this made me check my own program for the V3.0, which finds a typo - I'd set the trigger at 100 instead of 10, but it never failed to get two consecutive readings.
Can you clarify your experiment on this was at 3v or 5v ?

I've gone back to try ultrasonics again in the underground tank - The transducer is now installed in a plastic pipe which should solve the condensation issues from the previous metal pipe below -10 blinding the transducer.
The 5v requirement of the V3.0 is met by a pro-mini talking over I2C to the 3v node unless it's running on batteries.

One facet of the v3.0 I fell over in setting this up was it's inaccuracy at short range - Measured level 298mm, read level 275mm, from 1m down the error vanished - both devices bought responded identically.
The solution here was to scale the pulseIn for top and bottom for the known volume, and thus far, working reliably.

Having decided to have another go at ultrasonics for level in the underground bulk water tank (this time in plastic tube not steel - Condensation issues below -5c), I ordered up a pair of sensors, which on arrival turned out to be version 3.0.

Looking up the differences, they state 5v so no longer specced as 3v, trigger exceeding 10 microseconds, and the blind spot had increased from 20cm to 23 over V2.0.
Curious whether 5v would restore readings on the original V2.0, was surprised to see it responding accurately again.

Once installed in plastic in the tank, an error was evident but only for near objects.
eg - At 1300mm physical depth to surface the reading near enough correlated, but with the tank full and surface 295mm below the transducer it reads 273, as the level falls to 330 it reads 309 etc., etc..
I bench tested the second set and got the exact same error on short range.

A correction factor will solve my own use case, or perhaps even refit a V2.0, but for any looking to deploy these new V3.0 types it is worth checking short range accuracy.

@Jens-Jensen I've successfully used a 5psi pressure sensor to get raw water tank readings. down to about 1.3m.

@JonAndersen If you find a line of sight either direct or via a repeater site, Yagi or Quad might do the trick....

@Nca78 Agreed.

@NeverDie I may be wrong in my understanding of the argument -
Omicron is crowding out the more dangerous version Delta and will soon be the only variant in circulation.
The cross immunity being documented suggests any subsequent variant would be rendered ineffective since the immunity imparted is for the whole virus not specific proteins.

John Campbell has had a regular update for some time, is very analytical in his approach and is excellent in explaining in layman's terms.
He's still fighting the UK authorities over not making aspiration mandatory for intramuscular vaccinations, and there is mounting evidence and professional opinion he is correct in doing so.

@NeverDie The latest update from Prof John Campbell you may find of particular interest since it incorporates the latest data from the US.

Whereas the Omicron variant is indeed causing a massive increase in cases presenting, the hospitalisation, ITU and death rates are markedly different to Delta which is being rapidly supplanted.

On vaccines and immunity some interesting observations also https://youtu.be/TrVGymR-jFU

@NeverDie If you are experiencing the same as Europe, just as the Delta variant was more infectious, Omicron is worse with a doubling time of 2 days, hence a rapid escalation in cases.
Increased hospitalisations are a side effect of such rapid escalation in numbers, the only saving grace appears to be Omicron is less serious a variant on effects, but it is not benign by any means.
If you look up Prof John Campbell on Youtube you'll find plenty of information on what is happening, although he concentrates more on England, where hospitalisations have gone through the roof (not that you'd know it watching the media - Much as Brexit, Omerta)...
His opinion is none will escape infection, but precautionary measures such as vaccinations and masks will minimise effects and help slow spread to more prevent strain on medical facilities.

Thought I remembered a teardown and examination, and finally found it... Enjoy...
https://youtu.be/9Pw7U0XFgUM
At least I hope it's the correct one...

@TheoL Old fashioned perhaps but self-reliant.
I avoid cloud services as much as possible, hence the early choice of MySensors, Domoticz and a dedicated router, still trundling away on a UPS 24/7 in the background through the many power cuts.
I may not have internet, and may well be sitting in the dark, but can still see what's happening in and around the house.

I'd thought recently to get replacement "smart" light switches for this place and the radio equipped version might solve a potential rewire, but on discovery the "smart" solution depended on a server thousands of miles away, it was back to the drawing board.

@Dizzwold No need to apologise, it's more a case of waiting until somebody comes along who might know the status of Candle, such as @alowhum who has just been tagged.

@TheoL It comes in waves...
A lot of the energy over the "crazy ideas" peaked interest in the commercial world to capitalise on the IoT craze, a financial example of IFTTT.

I'm nowhere close to smart on development as the majority here, having bought (not developed and built) solutions to my own specific needs, experimenting only on the margins, but with the MySensors and Domoticz environment core as it was central to what I required for other things, data acquisition and on-site containment.

However, I strongly suspect a second wave of interest is about to land as the chicken of the brick walls erected for commercial interests (a la Apple) come come to roost.
Proprietary protocols and incompatibilities are a nonsense in this digital age, as is the "Cloud" being pivotal to whether Light Switch A turns off Light Switch B, and whether device Z is incorporated.

I'm certain the "crazy ideas" will return, it's only the when which is at issue

@skywatch My apologies for derailing the topic, yep, got two of the boiler feed and return also, and tapped into the trigger wire to notify boiler ON/OFF so Domoticz tracks the run time per day.
12 degrees is recommended across the radiator but it's not quite so simple, it also matters WHEN you get that differential -
That's where the portable node came in. Plastic pipe clips held sensors on the inlet and outlet pipes of the rad, the downstairs 6 all had a 11-12c differential at around 16/17 minutes, but no way would the upstairs 3 get there before the boiler cycle ended (room thermostat), which I put down to longer pipe runs.

I increase the rate on the thermo-valve (pressure agnostic) to get the same heat on the return at ca 18 minutes. Rinse/repeat the remaining 2, eh voila. Balanced, a nice little peak on the return flow temperature at the boiler, reduced gas bills and every room +/- 0.5 of it's set point.

Actual differential of these 3 upstairs radiators will be WAY under 12 for a constant feed, but for a heat cycle of 19-25 minutes depending on outside temperature it's perfectly balanced before the cycle ends.

@topgun78 I had a notion to zone control as I viewed it much the same as you have outlined, why heat the house when you're out for the day, but was surprised when I experimented to verify it.

Gas and temperatures were monitored on Domoticz, but we do get pretty harsh winters here, ymmv. All the rooms had their own thermovalve settings.
Most expensive to cheapest energy users were:
Turn off rads in unused rooms.
Turn down heating overnight.
Turn down heating when out.
Heating all the house 24/7.

It sounds counter-intuitive until you consider every house has an over-rated boiler, and a combi boiler's efficiency tails off as the return temperature increases.
There is a minimum heat you must dump before the boiler protects itself by shutting down. ie- how few rads must be available to sink the heat.

I made a portable node with two clip-on DS18B20s to ensure all my rads return high temperature almost simultaneously to the boiler (ie balanced) over numerous heat cycles, so the majority of the heating is at max efficiency.

I fully understand what you are trying to do and why, all I'm suggesting is decent rad valves and a fully balanced system will save you much more.

Good luck anyways..

@topgun78 Controlling a motor is easy enough, but what you are considering making will almost certainly bring you a world of pain and probably exceed the cost of a commercial version off the shelf.
Plenty have hacked into the control side of it with active forums in Holland and Germany if that interests.
If the real challenge is to improve heating, perhaps you might reconsider approach for a more physical one ?

From my own experience, upgrading the insulation yourself brings the single greatest and most cost-effective benefit, the other is fitting decent thermostatic valves and heads.
Here in 2017 the house got additional insulation, original heads and valves changed over to Heimeier Eclipse, one radiator and towel-rail uprated and it all paid for itself from the first year's gas savings which halved my gas bill.

From +/-1.5 in 2016 to +/-0.5c in 2018 in every room at -20 outside, running 24/7 and not a battery in sight (except for the monitoring system UPS).
Once you get a heating system balanced and set up, it simply works.

@alec44 Induction ? Pressure switch with a roller ?

@siddarth-sid Perhaps irrelevant since form-factor featured in your post, but...
I put in a UPS which featured on Instructables a while back based on a MeanWell AD-55A PSU, which is still working since 2018. It's neither not neat or compact but has sailed through multiple power cuts suffered here with aplomb. Simple enough to wire up, no processing required, the SMPS does it all.
Never did get round to powering the router from it as originally intended, but ain't dead yet...
I think this links https://forum.mysensors.org/post/100533

@KevinT Hopefully it helps, the Sleepy Dog implementation was mentioned here https://forum.mysensors.org/topic/6908/adafruit-feather-m0-rfm69hcw-with-mysensors/6?lang=en-GB
I'm still at the screwing around stage, even managed to get a puff of magic smoke when I removed the battery while still on USB power. Oops... Presumably I've blown the battery charging circuit, but it's still functional for fiddling with and testing otherwise.

For mains plus battery backup nodes I'm less concerned over sleep than continuity through power cuts, the extra pins over the 328 are a godsend. If frugal consumption is possible so much the better.

The recommended Adafruit library and FeatherM0 board definition are fine for Blink etc., but anything MySensors related fails to compile, so will try the Sensebender definition next and hopefully make some progress.

@KevinT Suggest a keyword search - I stumbled across a post which mentioned "Sleepy Dog" working for SAMD0, whilst trying to figure out how to similarly deploy a Feather RFM69.
I'm still a novice with this device, but I do recall the author saying Adafruit's "Sleepy Dog" worked, and IIRC there was some discussion over the power consumption not being spectacular, but best find and read with fresh eyes.
The stumbling block for MySensors appears to be that the only SAMD0 implementation developed was for the Sensebender Gateway, which logically was never intended to go to sleep.

@rejoe2 Anti-freeze.

@skywatch Just to add to your points on 254 max children, definitions, presentation, etc., which may help @mrmuszynski at some stage.
I've an array in a sketch with 14 DS18B20 addresses, all paged in turn using the one wire protocol on a single pin, so it is entirely possible to hit the 254 children limit before looking at I2C devices.

The beauty of the DS18B20 array is individual chips are "defined" sequentially, hence making presentation and fetching values available to a FOR loop, which makes for a highly compact sketch.

@CrankyCoder A third the volume I'm looking at, but in your case no pump issues.
Perhaps you might have a look at some of the less industrial sensors for what you want - a little differential sensor from Digikey on the end of a some poly pipe to a bib connex on the water line might give you all the info you need, ready compensated against atmosphere...

@CrankyCoder Sorry, didn't see this post appear - Yes it could work IF you can get a sensor of sufficiently low value.
The 5psi sensor was the smallest stud type I could find cheap - ca 6mm steps equates to ca 15 litres here, a 2psi would be even more accurate at a range would be 1.4m.

Interconnected barrels will all balance hydraulically, so with a single sensor on the common discharge you could report a per barrel value and do the multiplication for the barrels in service.
Do remember this is a static hydraulic value, velocity at the point of sampling has an effect, hence why readings are inhibited when the pump is running, since the sensor's fitted to the suction line.

This is the plot I'm getting - the 40 litre variation is presumably drift from the sensor, atmospheric pressure changes, or something I've missed, but not critical at less that 2% of tank capacity.
When the borehole pump kicks in to fill the tank the graph is a straight inclined line on the graph, so overall happy with this.

Would any users of these devices be able to advise whether the power transition between USB and Lipo is per a UPS ?

I'd looked at these initially but decided on 328 devices which at that time I knew worked with MySensors.
Only 4 remaining pins available on the 328 was one limit, the other was power transition - From USB to battery no problem, but restoration of USB power caused a reboot - In reality only a problem for one Node on dual power sending in water meter readings.
Having now damaged two of the 328s (don't ask), now looking for replacements with a little more capability (multiple interrupts + extra pins), so looking again at the Feather M0.

Thanks in advance

@rvendrame You're welcome.
I suggest mounting the sensor on a 1/2" ball valve, that way you can isolate for cleaning/replacement etc., and if it means draining the tank to make the fitting it's a one time event as the valve isolates.

@4994james Mine is a non standard Arduino with booster and LDO.
Possibly something wrong with your divider wiring ?
If you've got A0 as input and do an ADC read, the value should change with varying voltage.
I use a 10uF electrolytic as have loads of them, stable as it gets.

@rvendrame No problem, this is the link to the last one I bought https://www.ebay.com/itm/283251717007

The picture shows a female thread so I ordered an adapter for it to 1/2" but it arrived as a male. Luckily being a standard 1/4" thread meant a local sourced adapter was available.
Be careful of the thread type, NTF requires an adapter (also available via Ebay), but don't presume it's female from the photo, verify it with the seller.

If you know the head of water to the location of the sensor you select the closest value.
In my case I'm using 820mm of a ca 3500mm sensor, so only using 23% of it's range.
You lose 10% of the voltage top and bottom of supply, so you abstract 102.3 from the ADC reading in your sketch then multiply by the constant. They are never spot on, so worth reading the ADC with a temporary sketch or whatever to derive the constant for a known height.
Good luck

@4994james The internal reference is the more stable, but do remember that the ADC sample needs a stable voltage and low noise.
I've used high M value (prone to noise) resistive dividers with a simple electrolytic capacitor from the ADC pin to ground, no problems, perhaps worth a try.

I can only speak from my own experience, MySensors filled the gap at a time when commercial offerings still hadn't taken off, and as has been mentioned already, proprietary protocols and internet reliance put me off then and still does.

The combination of Domoticz and Mysensors suited my own needs of data acquisition rather than control, and although I've considered modifying and incorporating commercial devices since, they ultimately hit the brick wall of what CAN be done rather than what I need.

Hobbyist interest will come and go, but the forum inactivity is probably more testament to MySensors reliability for the vast majority who once flooded the place, and any new users find most of the answers they need here, so never pose a question
If it ain't broke don't fix it sort of idea - eg I had put off updating the IDE and MySensors until around a month back, a bit of a headache when I did, but it was soon sorted and back to silently doing what I need.

Is there competition with the huge variety of plug and go devices in elegant cases now available ? I suggest not.
For those wanting off the shelf solutions with limits there is a product, but why would I want 5 individual devices in a single space rather than the one currently, or the mobile apps for each, or the app to save me walking 3m to switch the light on/off ?
Domoticz and MySensors are not flawless, neither the Pi3 nor the Arduino are the latest MCUs, but the combo is plenty capable of doing what I require, and for an old fart like me, plenty.

A quick update on my long running battle to measure clean water tank volumes with ultrasonics, plagued by unreliability and periodic condensation problems on the ultrasonic head.

As the pump well is alongside the underground bulk water tank, and the pump intake is below the pump cut-off level, it was ideally suited to trial a pressure sensor.
This solution would equally suit overhead tanks.

In this case the level range is 820mm, the smallest value screw-in sensor found was 5psi so ca 3.5m of water column.
With the intended range, this gives ca 6mm steps on the ADC, so not so different to the ultrasonic in terms of performance. Though advertised as 5v units, they work just fine on 3.3v.

After some plumbing mods the sensor was fitted and hooked up to the Node over 3m away, and has now been working for a month flawlessly (once I inhibited readings when the pump was running ).

In all this time the ultrasonic has failed to read despite no condensation issues and several attempts to coax it into life, but at least I have a reliable alternative now...

@nagelc Heat detection ?
Most modern stoves have electrical connection for ignition with flame detectors built in to ensure the gas is cut off if it hasn't ignited.
Although they look like thermocouples poking through just below but alongside the burner head, they must be sealed pressure tubes acting mechanically as they don't shut down if there's an outage and you ignite them externally.
Perhaps that could be tapped into or the same technique used with pressure switches ?

@Olaf-Jacobs It's a straight analogue read on a single ADC pin relative to Vcc and ground of the Arduino Node.
In my case it's a pro-mini, the ADC pin in this example earlier defined as "Pressure", the result radioed in only if the ADC reading has changed. Three other binary inputs for contactors are polled every second on separate digital pins (one of which is pin 3), which uses all 4 pins available on this pro-mini due to the radio etc.

if ((!hidro&&!Hidro&&(counter2>=132))||initialise){
  int incoming=analogRead(Pressure);
  if (prevolume!=incoming){
    prevolume=incoming;
  sleep(50);
  float tankvolume=(((incoming-102.3)*6.704199635)+0.5);//5psi sensor
  send(msg8.set(tankvolume,0));//This is actual volume
   sleep(100);
}}

I'm not clear on the rest of what you're trying to incorporate or what pin 3 (digital interrupt) has to do with it, the pressure sensor signal goes to an analogue pin to enable the voltage to be read.

@Olaf-Jacobs If it's a typical screw in pressure sensor from China I've found they work on 3.3v fine. You lose 10% voltage off the bottom and top, so from your ADC reading of 1023 full scale to Vcc you lose 204.6 from the ADC reading. So your pressure range is 818.4 increments of the sensor pressure value... Voila...
When you do a float calculation on the pressure value, you simply send it in to however many decimal points you want (rounded up 0.5, 0.05, 0.005 etc)
I have 5 bar and psi units down to 5psi for a tank all working perfectly.
Hope this helps

@rvendrame Timely - I haven't put it into operation yet as only just finished the plumbing mods in the pump well for a 5psi water pressure sensor. Because I've saddled the sensor to the pump suction line I'll have to inhibit readings when the pump runs, hence delayed operation.

5psi is equivalent to 3.516m head, so you will measure over 80% of the ADC range giving a resolution of 4.25mm..

My 3,000 litre underground water tank suffers the same issue over condensation when winters drop below -5.

Hope this helps.

@Julie-Adam Have you considered disconnecting the air-line from the controller and reconnecting it to the air-head ?
The cranium woodframe is probably interrupting signals which possibly
explains the coupon fascination...

@tssk I hope I've not misunderstood what you intend, but possibly a further consideration ?
I have 14 sensors defined in an array with their digital addresses.

During presentation a similar "for" loop to the above ties each sensor digital address with a dedicated "i" to the Controller as in your sketch. That way the sensor location is fixed - if one goes faulty it only requires that specific address edited in the array for the replacement device and reprogram the Node.

In reality the last 2 devices in the array (Boiler) are on a separate data line to the first 12 (Spaces), but it is relatively simple to call each defined data line and address them by altering the "i" and "MAX_ATTACHED..." for each instance.

@nekitoss No problems with alkaline power, 6 - 18 months now on 2xAA down to ca -15c, battery voltage reported ca 2 or 3 times per day.
I vaguely recall reading an article on lithium and alkaline battery behaviour in extreme cold weather, the alkalines were the better performers.

Note that the pro-minis here use onboard booster and LDO which provide stable 3.3v all the way down to ca 1.8v battery death, so battery voltage is unrelated to system voltage unlike your setup. They also use rfm69s (433MHz) which are better able to penetrate structural obstacles.
Bear in mind that if you have comms problems this can substantially reduce battery life as it does not quickly waken, perform the task and go back to deep sleep

@jvdk I agree with the point made by @evb , why would you want to wire connect the ultrasonic node rather than the usual radio connection ? You are adding levels of complexity and power demands for what reason ?

My ultrasonic water tank node (pro-mini+rfm69) is now 2 years on the same 2xAA batteries sending in levels every hour (RTC), the only problem encountered - condensation forming a drip on the face of the ultrasonic head during very low temperatures.

@skywatch Rounding up to the nearest int ?

@tssk My misconception.

@tssk Most of these boosters have a grounded backplane, perhaps trying to re-orientate the inductor so the backplane acts a shield to direct RF may be an option to try ?
If the noise in on the power line I've heard of some managing to block it with a VR circuit if the voltage drop is acceptable, but never tried it.

@skywatch Perhaps of interest, this possibility (lots of power cuts here in the early days) led me not only interface serial but power for the Gateway from the Pi header rather than separately, the current draw is tiny.
The power cuts I later solved with a UPS but even before that it rebooted and went to work without issues, since the two reboot in sequence under the control of the Pi.
As @Yveaux however has highlighted all my Nodes are fixed IDs.

@Olaf-Jacobs Good.
I did read short spurs CAN work and the waterproof versions are on pre-made cables, but not sure of the limit.
If you hit occasional problems it may be worth trying cross wiring the Cat5e to maintain a series connection and see if that fixes it, that's what I did to stitch in an intermediate sensor and it worked fine.
Good luck.

@Olaf-Jacobs DS18B20s are usually reliable, but seconded on the comment from @Grubstake regarding cable type and topography for reliability, although generally if working at the start it should continue uninterrupted.
To illustrate the point, all of mine are connected to a Node rather than a Gateway on a single line of Cat5e for the house, and a short length of phone wire for two on the boiler. A random weird reading on all the house sensors (not simultaneously) turned up when I'd extended the last sensor on the Cat5E with a length of phone wire until I got some more Cat5E. Uniform cable characteristics seem important.

You might also check the Arduino version and libraries being used, have seen strange things happen with mismatched libraries etc..

@mrdisco Ok interesting, may have a play with that, a couple of spare boards doing nothing in the spares.
Situation here is a below ground 3,000 litre tank, the head mounted in a 3/4" reducer on the end of 1/2" gun barrel, no false echoes, surprising given the moulded tank has internal shoulders.
Though the gun-barrel proved robust, what I neglected to consider was condensation forming on the pipe when temperatures drop, occasionally forming a large drip on the sensor face, notably when it drops below -10c. Water is usually ca 5c.
Fitted a little formed foam this year held by a tie around the head to see if it cures or limits the issue, no fun wiping a drip off in winter.

I let the Arduino Node work out litres available and % used which it sends in every hour to Domoticz, now on year two of 2xAA so pretty happy with it.

@skywatch Perhaps I'm not understanding the problem of using a serial gateway for radio other an additional component such as a pro-mini?
Here the Pi3 Controller and Arduino 3v3 Gateway/radio are adjacent and direct wired on serial and power/ground to the pins.
From vague memory the default configuration on the Pi had to be swapped for wifi, found relatively simply instructed on the internet.

@mrdisco Would you mind expanding on this as I'm curious - A stilling well, typically 100mm plastic pipe is frequently used in industry with the US firing from the top and the pipe submerged below the surface..
When you refer to using 32mm pipe, how high above the fluid surface is the sensor within the 32mm pipe and what in the sensor the range you're measuring ?

@Thomas-Weeks Interesting comment on enclosed spaces, I had no such experience - two v2s in separate fluid tanks one of which failed after a year having otherwise worked flawlessly. I looked for 3 consecutive matches on a reading before radioing in.
One tank is concrete so smooth sides, perhaps 1.5 by 1.2m in plan, depth 1.7m.
The other is a circular section plastic moulding depth to the bottom ca 1.5m but with internal shoulders I expected to give reflection problems, yet no issues and it's still banging away 2 to 3 years later on 3v3.

@NeverDie This may well be a different scenario, but most agricultural electric fences generate the HV from a 12v battery, and have an inbuilt indicator if the fence is live.... Viz you could tap into the indicator...
Some I've seen running 24/7 for years in remote locations (deer fences) kept charged up only by DC wind generator..

@frydrik For a population of 500,000, say 200,000 customers, you are into commercial scale data gathering such as EON etc might deploy, a principal driver for the power companies fitting smart meters.
I would have to guess in a city of the size you mention, smart meters are already in place...

@NeverDie Short blasts while keeping an eye on the bag, the steam vents off quite vigorously stretching the porous bag but having only one perhaps I'm overly cautious with 15 second bursts...
If I ever get spares may test one to 'destruction' in the micro, but it is way faster re-generation than traditional slow oven, 10 mins v 2 hours..
Clay grains survive the nuking intact, silica beads don't ...

@tssk So long as your box is near enough sealed, dried gel or clay packs are effective in drying the trapped air even if the box is occasionally opened for say reprogramming.
If the packs are permanently exposed to external air they will quickly saturate and cease to be effective.
I prefer clay packs as they can be quickly regenerated in a microwave if you're very careful.

As a postscript to my own question should someone be looking at similar later, in practice the sensors were found to work perfectly well at 3v3.
Described as 5v working, 0.5-4.5v signal range, the effect is pro-rata at 3v3 in terms of voltage but the ADC output is identical for VCC scaling - viz, ZERO pressure is 10% of 1024 the range is 80%

@tssk A basic pro-mini clone needs modified (VR and LED), but with a few mods can run direct on batteries to fairly low voltages.
I bought these modules in as it would have cost more to make them even I was skilled enough (I still ain't), but they are essentially 3.3v Pro-mini with an onboard booster, regulator, rfm69 and various other refinements which gave it the edge over a Moteino.
Sundberg84's board might be an alternative worth considering, plenty of knowledge on it and help always at hand.

As to sensors, have a DHT22 somewhere but never deployed it when I found DS18B20s did all I required on a single run of Cat5e throughout the house and the outside, a second line covers the boiler.
I went for the 433MHz version of the RFM from the start, it easily passes through multiple masonry and concrete with a quarter wave antenna, so your floors may not be the problem you think later. The higher the frequency, the more attenuation effects you encounter, if .433 is permitted in your country I'd recommend it.

@tssk said in Which radio / wireless module to choose? Please recommend:

I prefer to use same module for everything but from some forum post I got impression that the GW might better use different module?

Your Gateway has to use the same radio as your nodes to communicate. Generally speaking the easiest and simplest Gateway is identical to your nodes, connecting by serial to your Controller.
It is the Controller which has to be a more powerful device such as a Pi, it and the Gateway will live in the same box and will run 24/7 on mains power, usually communicating via a router.

Your Controller+Gateway should be accessible, so would not recommend putting it into a loft with no easy access.
If your reasons for the location are radio communications (to the hives?), a 433MHz RFM will penetrate walls with ease...

Further to the point made by @mfalkvidd on batteries, a 9v wastes 60% of it's capacity on a 3v3 Node, 2xAAs will actually have a higher mAh capacity.
All my Nodes have 2xAA and are still going after 2 years..

@ady_e_n Presumably a low current DPDT switch.
Since I have them anyway I'd be using a latching DPDT signal relay fired by two logic level mosfets controlled by two pins, hence emulating the switch itself with no running current. The relays I have are NEC EE2-3TNU around 7mm thick and 9mm tall, 5v version should be available if that is your logic voltage level.
If space permits you could hot glue the relay to the board on it's side with the mosfets/resistors/3 pin control mounted onto it's base, hard wired to the original switch pads.

@wiredfrank Not sure what you meant by 433MHz not being allowed, so far as I recall that band is permitted worldwide but has transmit duration restrictions as it's a shared band.
I started off with the 433MHz RFMs and never had the slightest issue, penetrating walls and floors with plenty of range outside, even at one stage buried in snow.
The range and penetration of 868 v 433 will be slightly reduced, but performance should still be better than 2.4GHz.
The point about the recommended wire length being close enough remains valid

@wiredfrank It's a bit of overkill to be honest for hobby purposes, a quarter wave wire works fine when cut to the recommended dimension, or if concerned over the groundplane, a dipole as outlined by @scalz above.

A VNA is typically used to determine the exact resonant frequency of an antenna, if it's out by a fraction it won't make much difference in performance.

@Yveaux Yes, but was hoping someone had already gone through similar experiments, and could advise what did/didn't work
Your post prompted a reminder there is a 1/2" valved drain cock for the water meter/PRV in the bathroom... This would simplify testing without having to make plumbing alterations for the final locations

Didn't find exactly what I was looking for in a forum search...

Bought two of these (Chinese?) pressure transducers and 1/2" adapters with the intent of monitoring pressure (initially at 2 points) in the pumped water supply which has various headlosses in the line.
These are listed as 5v supply, 0.5-4.5 output for 0-0.5MPa, 20mA consumption, beyond that only claims the response is "perfectly" linear.

The Node is 3.3v, 5v is available on-board from an external PSU, but the voltage output will have to be scaled down to suit the ADC pin, even more so if aiming for Vref 1.1.
Normally a high value resistance ladder would suffice, but concerned over effect on response (presumably a variable resistance device), checked output resistance to ground, at atmospheric ca 6.7 Mohm, purportedly corresponding to 0.5v.

It occurred to me that if it IS a variable resistance it should also function the same at 3v, so any here able to offer any insights, experience, or tips ?

@amkochaki You imply the sensor sends two different voltages, one for standby and one for active, perhaps you might establish what these voltages are for each condition.
All your ADC is doing is measure the voltage, your sketch takes that value and converts that to a logical 1 or 0, standby/active.
If the voltage swing is large, say 5v for stable and 1v for triggered, perhaps conversion via resistor divider can provide a logical 1 or 0 input simplifying the process ?

@Encrypt It sounds as if you are making progress, moving off breadboard will certainly make for greater reliability.

My reasoning for checking HIGH and LOW is that these are fixed states rather than in transition. FET type sensors are very fast, but the Elster Water Meter Node would send an intermittent false increment (No idea why/what), which is why the additional check on pin state was introduced.
That pin check was spurred by experience with the Gas Node, which gave rapidly escalating readings, found to be caused by the reed closing for ca 6 seconds. Checking in a Sleep loop that the pin had changed back from LOW to HIGH before going deep sleep resolved it, 3 years on it's still in complete sync with the register.

@Encrypt Sorry, meant to respond to this when you posted it but got distracted again...
If I recall correctly the calls to Gateway are limited to 7 or 8 attempts or thereabouts, but if your Gateway falls over every increment it will do the same. This can hammer Node batteries as I discovered after a power failure (regular feature) killed the Pi3, now solved with a DC/DC UPS.
Never had comms issues with the RFM69 on 433MHz but they do use proper whips.

Have a similar setup with an Elster sensor, but by the time the water meter was done I'd learned from similar "issues" from my first ever Node on the gas meter which uses a Reed...

Interrupt on FALLING was causing rapid cycles of the increment the trick was to ensure the increment only happened once on every wake from Sleep.
Set the interrupt to LOW, verify that the pin does indeed read LOW after wakening (in case of false triggers), increment the count, send the updated value, then verify in a SLEEP loop that the pin has indeed gone HIGH again before going to deep sleep.

Hope this helps

@mhkid If the suggestion by @mfalkvidd doesn't resolve the problem you might look at moving both Gateway and Node marginally to see if it it is some effect.
Clearly the Gateway can "hear" the Node but not the reverse, and as most transceivers are matched between power out and receive sensitivity, polarisation or influence from nearby objects may be in play. There can be some strange effects at 2.4 GHz, particularly where "shrunk" antennae are used.

It's kind of a hassle wasting 5 minutes, 2 - 3 times per day waiting on the gate to open, and the LONG wait for it to close just to make sure nobody sneaked in...
Looking at fitting a cantilevered sliding gate here so it traverses the -20 ice and snow (no bottom rail), but will be opening and closing it manually as it's way faster..
Plenty of manufacturers of gates depending on your locale, most of the kit here seems to be rebranded turkish, italian or chinese origin, but as sliding gates are usually driven by a fixed motor on a castellated rail they are slower than the hinged manual gates I already have.
And no remotes to fumble for...

@NeverDie Actually no. Because this is an EPA list denoting the effectiveness of listed products, it's much too complicated to find out concentrations of the active ingredient, and I would guess that the kill times have a Factor of Safety built in.
Saw nothing on what surfaces are referred as the Virus is known to survive on different surfaces for different durations, again I suspect a high FoS.
Sorted on "Follow the directions..." and looked for Human Coronavirus, ethanol and isopropol both pop up at 30 seconds contact (normal terminology for disinfection) time.

Hypochlorite is a few minutes but varies according to brand name, so not really helpful if you don't know concentrations.

My point remains that alcohol is readily available, and as an additional precaution post washing surfaces probably more than sufficient in a domestic environment.
If 70% is adequate for handwash on skin where it evaporates in a few seconds, no reason why on a cold non-porous surface it would be no less effective

@NeverDie An alcohol mist is pretty fast and effective as a surface disinfectant and almost universally available.
HOCl is a weak acid from vague memory and not so far removed from hypochlorite (same family?).
You have to remember that if a 20 second wash is good enough to clean your hands, the same should be true of surfaces, and raw produce rarely needs much other than a wash...
Sterilisation is a blanket attack so yes kills the bad stuff but also the good bacteria. eg - Chlorination provides you safe drinking water but kills some useful stomach bacteria on ingestion, which is why foreign waters give tourists problems the locals don't suffer....

@GLAB No worries, was in a similar situation.. My perspective was on what the Node could be made to provide, not what could be done within Domoticz to address the requirement - eg Had heard of Dummy counters but had no experience nor understood their purpose, nor had ever used scripts.
I had been sending a logical ON/OFF as well as a cumulative total from the Node and hit the same hair-pulling scenario as yourself.
Now only the ON/OFF is sent and the short script makes the transposition within Domoticz to a dummy meter (I think kWhr) using the internal clock in Domoticz.
Have fun

@GLAB Indeed my tank controls the pumps (borehole in and supply out) on standard level controllers to the contactors within a full/empty range, the Node is only used to verify tank levels with the ultrasonic. However...

As I understood it, the time default of 5 minutes in Domoticz is your principle issue, which is why you have the parallel pair taking the same data to operate "electromechanical timers" which does satisfy your requirements, but only locally.
My point was that Domoticz can do the "electromechanical timers" part internally, all it needs is a stop/start for pump OFF/ON to control it, you can count it in seconds if you want but minutes is probably sufficient for what you are after.
The second pic showed my boiler running times as a "Switch" in Domoticz. The third pic shows this transformed to a counter/meter which is a Dummy or Virtual Meter counting the minutes run as set up here.

The 8 line script under "Events" is what updates the meter from the ON/OFF every minute of every day, a second script resets it every midnight, but it could be let roll over if that's what you wanted.

Trust me, you may be 9 years older than me, but it's not that difficult once you look at it from another perspective

@GLAB May have misread this but perhaps a different perspective? Currently using mysensors with domoticz, all on rfm69 connections, my water supply is from an underground bulk tank monitored by ultrasonics and reported every hour, so the pump rate is not so much an issue here, I only need know that it is been topped up by the borehole at silly o'clock and how many days supply I have left should the borehole pump fail.
However, when I wanted to monitor the CH boiler run times I hit the same 5 minute default in Domoticz .
The Node wakens on an Interrupt from the boiler contactor changing state On/Off, it also reads feed and return temps at 1 and 5 minute intervals but the Interrupt is deadly accurate.
This is notified by the Node to a logical switch in Domoticz, but the problem was it is complicated rather than simple to read.

What I used within Domoticz was a short script to update a dummy meter. The horizontal scale on the graph is the default accuracy of 5 minutes, BUT the vertical value is to one minute of accuracy, reset every midnight via a second script. You could presumably do this for any time value on the vertical scale or even to estimate a cumulative volume used I guess, but again only really useful to compare daily run times.

The Utility page from which this graph is obtained holds the overview of the daily cumulative (left side 3rd down - 41 minutes).

Is that the sort of thing you are trying to achieve ?

@skywatch I understand your confusion on the wealth/jobs aspect, "the financial markets to go back to what they were doing before " is precisely what we do NOT need if we're not going to avoid boom bust austerity cycles... Mark Blyth does some good lectures on it which may help reset perspectives, his software/hardware computer analogy is particularly helpful.
Those same vested interests have been provoking public frustration of late in England, but I sincerely hope folks stick with it, idiots are not the only victims if they don't.

Plenty to keep me occupied here and beer is cheap enough (ca 50p ), currently fitting IP65 spots in the bathroom, so laying off the beer and looking forward to a shower....

@skywatch Agreed, but I'm in a shielded group due to age, and since retired my "normal" ain't everyone else's.

The immediate problem of lockdown (beyond getting pissed off) is political not health. With so many now on the government payroll (that quaint US term furlough) essentially MMT is in place and working, and it's scaring the sh1t out of those who make a nice living out of MMT denial. If the penny drops as lockdown extends, the big con is over, hence the heavy campaigning to exit lockdown soon, and politicians bricking it. I gather you're a Brit so this may prove interesting if you're not already acquainted with Murphy... https://www.taxresearch.org.uk/Blog/2020/05/05/people-and-jobs-or-wealth-the-government-has-to-decide-which-to-prioritise-and-there-is-only-one-right-answer/
Our lives will be fundamentally changed for probably 18 months to 2 years with trials and setbacks to keep things under control, what form the steady state takes is currently anybody's guess, but we should know better by 6 months in, around October.

@skywatch Sorry, read it in two separate press articles but didn't pay attention which ones... Essentially virologists and epidemiologists reckon this is the most watched and researched bug worldwide, yet ONE lab published a non peer-reviewed paper ?

@NeverDie There is no one-size-fits-all approach to exit lockdown, the trouble is there are multiple demands to do so none of which are based on science, which is why most of Europe is feeling it's way slowly and with an abundance of caution, the alternative is a lot of dead and overwhelmed health systems.
Romania has 48/M dead, UK 460, US 236, even if eradicated now there will be 2nd, 3rd etc waves unless you have another strategy to prevent spread. It's all about breaking chains of transmission, and that means dramatically changing "normal" human behaviours.
Social distancing and sanitation appears to be key, Test/Track/Isolate is also being examined, but no jumping in a plane to Paris for a weekend, it's now 14 days in quarantine.

@NeverDie @skywatch Be careful, this has already been outed as fake news...

@NeverDie Most all normal water meters use magnetic drives to isolate the register from the main flow (even the wet registers).
Common water meters are single or multi-jet Class A or B to actuate the magnetic disc but meter orientation alters accuracy. The positive displacement/piston variety simply rotates volumetrically so orientation does not matter. There are also ultrasonic types so again no contact, very accurate but crazy expensive.

The critical point is the K factor of the factory built meter, K=1 means 1 litre output. This is not something you can alter, eg My outdoors Zenith is a multijet with K=500, the sensor is about 50 euro, but a 5c reed fits just fine. Useless at 500 litre/pulse but it illustrates the point it's how the meter is built.

@OliverDog the V200 I have has a mosfet type sensor (triggered by an inductor?) rather than a reed, normally very expensive, but I got lucky with one supplier, from memory the meter and sensor were less than 100 quid, the battery for the sensor lasts maybe 10 years. If it dies I'll install a V100.

Modern versions such as the V200H are usually RF or cable to a specialist receiver but it starts getting complicated with proprietary protocols as they're really designed for the water service industry, although some wizzes have hacked various models. May be worth a google as I'm out of touch these days..

It may be an old Class D design, but the V100 I nearly bought instead has a plugged socket in which the reed is inserted into the meter body. The Elster official reed is expensive but you can slip a normal reed in with a bit of preparation, it works fine as did one for a friend. Pretty sure the V100 is now made in Malaysia by the ex-director or similar of Elster, demand in Europe etc for the drive-by reading capability and Class D has tilted the market. No idea whether the Chinese market has clones or they're made properly.

@NeverDie "A patent pending Flume sensor"... Flumes are restrictors for open channel flow measurement been around for over a century or more, a pipe version is called a Dall Tube, see Bernoulli's equation.
Sounds like marketing hype, but it's just mag measurement sent to an app.
A lot of folks have tried sensing magnetic fields or reading part of the register, think some on here had some success with photo sensors, so plenty of options...

@NeverDie I'm not conversant with the US marketplace, my own is an Elster V200 from memory.. They make all sorts of variants eg https://www.johnsonvalves.co.uk/tag/iso4064-class-d/ but these are modern versions of their original V100 which I believe is still manufactured in Malaysia (?), and I see from a quick Google for "volumetric water meters" Chinese manufacturers are copying it also.
Other european manufacturers are Itron, Sensus etc, but they are heavily committed to service companies hence built in radio for drive-by meter reading etc...
One which uses a reed activated sensor is by far the simplest and most reliable to incorporate for domestic monitoring, but YMMV.
The only the thing to watch for is maximum flow rate, if your max demand can be satisfied by a 15mm instead of 20mm the headloss should have no impact while detecting lower flows (or leakage).

Shutting down supply is failsafe but impractical as you've highlighted. You're smart enough to figure out how to monitor flow rates and say shut down supply (electronic valve) if a "burst" is picked up by the node, hence normal usage will not be a problem.

@Puneit-Thukral Saving power by floating the pin may return spurious triggers, pull-up resistors give a solid open/close.
The gas meter here has a simple reed, now in it's second year on 2xAA at ca 30,000 actions/year, simple high value resistor bridge with a cap across the lower to ground to minimise noise and deal with "bounce" yet minimise current used.

@NeverDie There's a danger of trying to find a overall solution for distinctly different scenarios even if they all concern water escape.

Leakage or loss within a domestic supply can be readily picked up by a Class D water meter with K-1. My domestic meter is Class D with a 1 Litre sensor linked to Domoticz so any leakage shows up overnight. Your a drip per minute sounds miniscule, but if you sleep the average 8 hours, that's 480 drips overnight, chances are it will trigger the meter.
The beauty of modern positive displacement meters is they are mounting agnostic, have remarkably low pressure drop and are built for 20+ years reliable life. Usually the sensor costs more than the meter, but even $150 is cheaper than the insurance loss if something goes wrong.

For leakage in a say a bathroom, wet socks are the surest indicator there is a problem to investigate. Almost built one to alarm for a water softener drain overflowing, but ended up altering the plumbing with a backup overflow. Problem solved but a sensor node remains on my to-do list.

A burst will cause damage within minutes and chances are it will occur when you are not there or within earshot. Only had one instance where a valve literally burst (faulty casting?) but heard it and managed to isolate quickly. Had I been elsewhere and received an alarm because a sensor detected it, by the time I responded the damage would have been horrendous (3m3 tank and pumped supply ), but a public water supply will continue to feed regardless.

For the vast majority of normal sensors they rely on detecting water on the floor to alert you to a partial but unusual leak from a stuck ballcock, stuck/split air valve or a loosened drain connection, the only question is the detection area.

However, returning to that Class D, it is equally possible that the system alarms for unusual triggers when nobody is in the house (say intruder alarm active) or at night times when asleep, one litre of water is a lot easier to deal with. It is equally possible to trigger an electronic shut-off valve if the flow exceeds normal use as with a burst.
Just my 2c...

@evb Having had various problems with two US devices using the ping library, found the simpler timing method gave consistent results after only an initial false.
Setting the upper and lower thresholds in the sketch (if that is appropriate - viz tanks) and checking for two consecutive readings within that over up to 10 attempts then validates the value, only then reporting the result. Converting the distance to mm is perhaps overkill, but for me it works 99.99% of the time.

@biffhero List of sketches on the folder eg Node8BoilerT3vWNrtc3 tells me it's Node 8, BoilerT tells me it's the Central Heating Boiler temperature monitor, 3vWN is the voltage and Node device, rtc means it has one, and finally the version number 3.
By embedding the sketch name and version number in the sketch, Domoticz always holds the current sketch and version it's programmed with, as at one point I was switching programs around and lost track which one was on it.
So, some old sketches I retain as they feature a different technique or method which may come in useful later, but usually old versions are deleted if satisfied with the final.
It's easy enough to find the sketch in the folder as they follow in sequence.
Just my methodology but hope this helps...