Configuring sharky 473 flow meter to monitor flow and Cop of my Heatpump

Hello and thanks for your support,
I just got an used sharky 473 flowmeter rated at qp 6 mc/h dn 32 which i would like to mount on my return flow pipe to the heatpump to monitor flow,
The sensor comes with a computer (sharky 458) a heat meter that extrapolates instantaneous and cumulative data. The flow meter comunicates via impuls out to the heat meter, whoch does Not have an impulse out.
Now, I would like to duplicate the impulse out of the flow meter and connect it to the emontx v3 pulse in and do the math with my other temp sensor to calculate energy and cop.
I could indeed connect the sharky 473 directly to emontx but i would like not lose the heat meter stand alone,
Is it possible? Do i need some sort of “pulse duplicator”? Or should just send the pulse to emontx and forget about the heat meter?
Thanks for your support

It might be possible to connect your emonTx to both flow meter and “computer”.

  1. What other items is your emonTx connected to?
  2. I can find no information about “Sharky 458”. What voltage does the “computer” output to the flowmeter (when the flowmeter is not sending a pulse)?

Thanks for your reply, the heat meter is the same as the scylar int 8, they just changed name, it has a direct pulse input, the flow sensor send 1imp/10l water, knowing that, I thought I could send the output to the emontx v3 pulse input AND to the heat meter, I tried that but failed.
The emontx is now fed trough a ds18b20 breakout board with 5 sensor, and the pulse input is now occupied by another heat meter, by sontex, which will be replaced by the sharky. The heat meter is the superstatic 440 and has a pulse output which sends just thermal energy calculated from the computer. With the sharky i’m trying to send the flow and then calculate in emoncms heat energy and cop.
Is ther a way to duplicate the pulse output from the flow sensor?
Thanks in advance

I think it will be possible to connect your emonTx directly to the pulse connection between the flow meter and the heat meter. The information that I am looking at is on page 19 of the heat meter manual:

That tells me that the heat meter provides a 3 V output, the flow meter has an “open collector” output, so that pulls the voltage supplied by the heat meter down to approximately 0.6 V when there is a pulse.

You can connect the GND pin of the emonTx ( 3 on the screw terminals ) to the negative side of the pulse terminals on the heat meter, and the pulse input of the emonTx (IRQ1, 4 on the screw terminals ) to the + 3V terminal (I think “I1” or “I2”) on the heat meter.

I cannot guarantee that this will work, but I think it will.

Thanks for your time,
On page 19 are described the optional modules, the heat meter has no output as it is, so the question is if i can route a cable from the flow sensor to the meter and to the EmonTx using the same output on the flow meter, or do I need a sort of amplifier to double the pulse signal?
If I connect it directly to the emontx pulse input. Can i use the wh to power process to obtain flow l/m?

Please read again page 19 and what I wrote.
Page 19 is not very clear, but surely it describes the Input of the “computer” as well as the optional outputs?

Hello, opening the case of the meter, I don’t see such connection as described in the picture, I could try to connect a secondary cable at the output of the flow sensor.
In case it will work, could you hint my ti the right process to apply to transform pulse to power, I would set up a wh collector to kwh to power, am I right? Thank you Sir

You have a cable - two wires - connecting the flow meter to the “computer”. You can connect the emonTx at whichever end is convenient. The important part is you must get the polarity correct.

I am not an expert on emonCMS, it will think a pulse is energy (≡ litres), therefore if you want flow rate, then yes, I think you need kWh → power.

Hello, i connected the flow meter to the heat meter androoted a second cable from the heat meter clamps to the pulse input on the emontx. And I get something, it seems to record a pulse, but I’m not able to figure out how to transform it in flow rate. i assumed the kwh power processor would do the trick, but the values given don’t match the flow showed on the flow meter.
The sharky sends 1 pulse every 10 liter, i’m testing with what I know to be a low flow rate assuming 10l/min (buket test) so it should send 1 pulse every 60 second. Any help is appreciated

I think @TrystanLea is the person best placed to help you. Somewhere, you will need to set the quantity that one pulse represents. One pulse = 1 Wh is common for energy meters, so you might need a " ×10 " somewhere if 1 pulse ≡ 10 litres.

Thank you Sir,
I did the conversion x10 but, correct me if I’m right, this way I just get a cumulative output, I would need a time correlation to calculate flow rate, is there a way to achieve that? Maybe I’m missing something

I was hoping @TrystanLea would help you with this, because I don’t know emonCMS well enough to tell you what you need to do.

@mo679 and I have a PM on this as well. I think the ‘Total pulse count to pulse increment’ process may work, though it may well need modifying so that it can be configured to give the incrememnt over a longer period e.g litres/hour, rather than litres in the last 10s which will often be 0

Thanks a lot for you support,
I came out using the total pulse count to pulse increment, assuming my flow meter sends 1pulse every 50ml( this is a test pulse output, normally are 10lt/pulse) are 20 pulse liter, I setup a discrete time to sample, assuming 30sec, and then doing the math. The resolution is obviously approximative. When I would use the norma pulse 10Lt/p, I can divide by 10minute measurement 600sec.
I still have to test it, what do you think?
It would be awesome to have an interrupt like function relating elapsed time between pulse (P2-P1)/(T2-T1)

Looks good @mo679 did it work for you in the end? are your graphs very spikey?

Hello, I just tested the high resolution pulse setup in 30sec time, and the readings are quite true, but there is an obvious approximation, which, I think, might be ok if you look at the mean value of flow rate needed to calculate cop and scop.
I’ll report later on when my setup is complete and running,
Any news about a function to time interval two pulse value? That would be the solution.