smart implant as voltmeter

[Łukasz997 117]

I need to measure presence of voltage 1.8V DC, minimal current (~10-20 mA). I would like to know not even precise measurement, but if the voltage is present or not (other words - if LED is on or not).
Can I use smart implant to do this? Based on the documentation I hope that I can - implant acting as a 0-10V sensors driver.

I don't want to supply power to the "sensor" (as at the wiring in manual), just take voltage read-out.

 

Can anyone with practical experience advise me what can I pay attention? Especially, how to wire it for two independent "sensors".

In advance - thanks!

Edited January 24, 2025 by Łukasz997

[m.roszak 763]

You can measure the voltage by inputs no problem, just make sure the implant has common GND with the LED. 

[Łukasz997 117]

Thank you. I've checked, it works. What I noticed - is delay of 30 secs between change and read-out on the HC3. Is it possible to shorten this time? The best I would find it working as binary sensor, immediately... Or at least every 10s.

There are some parameters as time gap between reports and minimal reported change. But gap can be set as low as 60s. - so from where those 30 secs come.

I thought that minimal reported change is about the change which is immediately reported - bu no...

Is it something I can do to increase the frequency of reports?

[Łukasz997 117]

I wat to share one more info: I've said that I've checked. Well, I connect a 9 V battery to implant (two first wires, red and blue). And put the controlled voltage from power unit to yellow and second blue wire. Voltage readout from power supply and on HC3 was the same. Seem to be working.

But when I put voltmeter to sensor cables - it was 7 V there. When powering the implant with 24V, it was 18V there.

This is unacceptable when I have to insert the cable to measure to fragile point on living electronics.

It's clear that I did something wrong, can anyone tell me what exactly?

What colours should have the cables connected to measured point?

Edited January 24, 2025 by Łukasz997

[Brors94 115]

22 minutes ago, Łukasz997 said:

I wat to share one more info: I've said that I've checked. Well, I connect a 9 V battery to implant (two first wires, red and blue). And put the controlled voltage from power unit to yellow and second blue wire. Voltage readout from power supply and on HC3 was the same. Seem to be working.

But when I put voltmeter to sensor cables - it was 7 V there. When powering the implant with 24V, it was 18V there.

This is unacceptable when I have to insert the cable to measure to fragile point on living electronics.

It's clear that I did something wrong, can anyone tell me what exactly?

What colours should have the cables connected to measured point?

 

I guess you checked the manual?  

I think it will be the 2-wire you need? 

 

 

[Łukasz997 117]

Yes, two wired. This is my diagram.

Please login or register to see this attachment.

Edited January 24, 2025 by Łukasz997

[Martin_N 154]

@Łukasz997

 

 

Quote

I wat to share one more info: I've said that I've checked. Well, I connect a 9 V battery to implant (two first wires, red and blue). And put the controlled voltage from power unit to yellow and second blue wire. Voltage readout from power supply and on HC3 was the same. Seem to be working.

 

Ok, the power for the Smart Implant is on the first two wires, Red and Blue.
However, the SI is powered in the range of 9VDC-30VDC. So I don't understand why you are using a 9V battery for power when a small discharge will take you outside the limits.

 

 

Quote

But when I put voltmeter to sensor cables - it was 7 V there. When powering the implant with 24V, it was 18V there.

 

 

 

I don't understand this a bit. When you use a 24VDC SI power supply, where does 18V appear? So you have a too soft source that drops by 6V when connecting the SI? Really?
Another thing that is described incomprehensibly, are you measuring the 18V on the input wires?

 

Try to draw or describe it in more detail. After all, you have that diode connected using a red and blue wire.

 

 

[Martin_N 154]

5 hours ago, Łukasz997 said:

Thank you. I've checked, it works. What I noticed - is delay of 30 secs between change and read-out on the HC3. Is it possible to shorten this time? The best I would find it working as binary sensor, immediately... Or at least every 10s.

There are some parameters as time gap between reports and minimal reported change. But gap can be set as low as 60s. - so from where those 30 secs come.

I thought that minimal reported change is about the change which is immediately reported - bu no...

Is it something I can do to increase the frequency of reports?

No, it is not possible to set a smaller timeout. This is a protocol limitation to prevent network congestion. It is a pooled value.
Unlike a binary sensor that actively sends a state change "immediately".

[RH_Dreambox 189]

Please login or register to see this attachment.

[Łukasz997 117]

12 godzin temu, Martin_N napisał:

Ok, the power for the Smart Implant is on the first two wires, Red and Blue.
However, the SI is powered in the range of 9VDC-30VDC. So I don't understand why you are using a 9V battery for power when a small discharge will take you outside the limits.

Oh, this is only for testing purpose, "laboratory run": I have only one programmable power supply and use it to connect 1.8V to sensor wire and see readout on the HC. That's why I use the battery to supply SI - only temporary.

When really working the SI power supply will be from the mainboard where 24V DC is present.

[Łukasz997 117]

12 godzin temu, Martin_N napisał:

I don't understand this a bit. When you use a 24VDC SI power supply, where does 18V appear? So you have a too soft source that drops by 6V when connecting the SI? Really?
Another thing that is described incomprehensibly, are you measuring the 18V on the input wires?

 

Try to draw or describe it in more detail. After all, you have that diode connected using a red and blue wire.

I know that's hard to understand. I will make a photo from this experiment and it will be completely clear.

Now say what can I achieve.
I have an Immergas Vitrix heater which do not share any data of it's functioning (old model). All I can do is to engage and disengage the heater by smart implant. And I actually do it.

 

But I want more data about how heater works - o.e. how much energy is spent for water heating for house purpose and how much to heat the water for radiators.

There are several mini-LED diodes connected to the mainboard showing various heater state - among others those I'm interested on. I measure by a voltmeter the voltage of such led when lit on - 1.8V (powering the diode is managed by heater mainboard, of course).

So I think that I establish constant measurement of diode voltage by SI and will know when the function is activated or not.

I've tested and it works - 1.8V put to the yellow and blue wire gives readout on th multi-level sensor on HC3. Assuming all wil be ok I mounted it to the heater.

Unfortunately - after connecting (power to SI taken from the mainboard: 24V DC) the mini-LED start to lit constantly, despite it shouldn't, because function on heater was not active. It comes clear that on yellow and blue wires there's a voltage. In other words - smart implant is not working as ordinary voltmeter which is only waiting for voltage to measure, but it's also emitting a voltage on sensor wires.

Here's how it was connected. It should work and it didn't.

Please login or register to see this attachment.

 

By the way - thank's for all help. I decided not to give up ☠️!

[RH_Dreambox 189]

Have you configured your Smart Implant to a voltmeter?
The example below shows configuring the input 0-10V which is converted to 0-400 cm.
In your case you configure Volt to Volt.

 

Please login or register to see this link.

[Łukasz997 117]

Godzinę temu, RH_Dreambox napisał:

Have you configured your Smart Implant to a voltmeter?

Well, this is the issue - I've configured the test smart implant, but... forgot to do the same with other implant in heater. The voltage on sensor appears when inputs are not configured...

I apologize to all of you for your involvement and, frankly, for wasting your time. Much ado about nothing...

I will be more careful in the future.

 

And thank you all!

Edited January 25, 2025 by Łukasz997

[Martin_N 154]

So I'm back now and I see you've solved it. It looks like an interesting application. Although I must point out that the LEDs are usually output directly from the processor, so any failure on the SI and switching back to the binary sensor can damage the processor over a longer period of time (you've already tried that the SI provides its own measuring voltage in BI mode). Replacing the board is not a cheap thing.

I would rather add a relay for galvanic isolation, and monitor the status of the power outputs using the BI sensor on the SI inputs. If of course this coincides with the status of the LED you are now monitoring.

 

Good luck and thank you for sharing such an interesting solution.

 

[Łukasz997 117]

52 minuty temu, Martin_N napisał:

So I'm back now and I see you've solved it. It looks like an interesting application. Although I must point out that the LEDs are usually output directly from the processor, so any failure on the SI and switching back to the binary sensor can damage the processor over a longer period of time (you've already tried that the SI provides its own measuring voltage in BI mode). Replacing the board is not a cheap thing.

I would rather add a relay for galvanic isolation, and monitor the status of the power outputs using the BI sensor on the SI inputs. If of course this coincides with the status of the LED you are now monitoring.

 

Good luck and thank you for sharing such an interesting solution.

 

You are definitely right.

Anyway, realising this idea in practice was very risky.

 

And - it can't be realised so easy.

First of all, at startup smart implant (also properly configured as a voltmeter) passes its all power supply voltage to the sensor cable (yellow, green) and gnd (blue) - even 30V... After very short time it drops to zero, but possibly can damage fragile electronic. That's the information that does not make this discussion a waste of time - for someone who can use SI this way.

Second, the controller of this sub-board with leds (it's actually not the mainboard) sense that something additional is connected (even if no voltage peak at the begin) and informs about board fail. After disconnecting - it turns to normal.

I will try to find an optical micro-sensor who can catch the led emitting light or dimmed. This is safe & universal and can be possible.

Here's the QA's interface I wanted to enhance with new data - to be honest, completely without the need, just to prove that I can.

 

Please login or register to see this attachment.

 

Edited January 25, 2025 by Łukasz997

[Martin_N 154]

You can continue to monitor the diode status only by connecting an optocoupler. Pay attention to the Uf voltage and possibly connect a suitable resistor. Connect the SI input in binary mode to the optocoupler output.

It is still a connection to the motherboard. You must not exceed the maximum output current from the processor (or from where the LED is connected). In practice, it is amplified using some transistors (so you can consider a value of up to 50mA), but it must be verified.

 

The option with an optical probe (phototransistor) is also possible, even safer, because you will not affect the motherboard. However, there may be a problem with the excitation of the phototransistor to open (the LED may have low brightness / may be affected by other LEDs). Be careful then, it depends on the polarity of the phototransistor connection to the SI input!

 

I keep my fingers crossed, I would be happy if you add some more information here and enrich the community here.

[Łukasz997 117]

1 godzinę temu, Martin_N napisał:

You can continue to monitor the diode status only by connecting an optocoupler. Pay attention to the Uf voltage and possibly connect a suitable resistor. Connect the SI input in binary mode to the optocoupler output.

It is still a connection to the motherboard. You must not exceed the maximum output current from the processor (or from where the LED is connected). In practice, it is amplified using some transistors (so you can consider a value of up to 50mA), but it must be verified.

 

The option with an optical probe (phototransistor) is also possible, even safer, because you will not affect the motherboard. However, there may be a problem with the excitation of the phototransistor to open (the LED may have low brightness / may be affected by other LEDs). Be careful then, it depends on the polarity of the phototransistor connection to the SI input!

 

I keep my fingers crossed, I would be happy if you add some more information here and enrich the community here.

Both solutions are OK. But I would rather vote for second one. As I found out the controller is not so straight and stupid. I'm wondering what reacts with it generating message faulty board. Probably SI has inner resistance or capacitance... So the approach what need to connect something to board is still risky and not universal.

The photosensor seems to me a good solution. When it's micro-size and we add some "blackbox" to it, the LED light should be enough to activate. And it can be used to cheap integrate many "impossible" devices and some very interesting an unusual cases.

This is more experiments, searching, money spending and sooner or later a success.

Of course I will share with you any valuable info I found. But for now, I'm little bit tired by all above, and creating gas consumption statistics QA (not exactly easy task, 1400 lines).

So I will take my time with this!

Once again - thanks a lot!

Edited January 25, 2025 by Łukasz997

[Tim__ 126]

I have connected a Smart Implant to a kitchen range hood, using the status LEDs to determine the state of the fan and lights. What you need to ensure is that your ground wire (blue) is connected to the ground of your appliance. This is not necessarily on the output side of the LED as you have shown. You may have introduced a short circuit between LED output and ground, and might be why the mainboard is reporting a fault. In my case, the sensor wire reports about 4V when the LED is on, and 2V when off. I use LUA to read the voltage and convert that into a state (and use it to toggle the range hood's buttons as appropriate).

 

As to your concern about the voltage on the sensor cables when not connected to anything, i don't believe you need to worry. These are high impedance floating inputs and so that voltage should not appear once the the wire is connected into a circuit. You can check this by connecting the sensor wire to a dummy circuit of two resistors (one to + and one to -) and use a voltage meter to see if the voltage is stable when the SI starts up.

[m.roszak 763]

On 1/24/2025 at 11:41 PM, Martin_N said:

I don't understand this a bit. When you use a 24VDC SI power supply, where does 18V appear? So you have a too soft source that drops by 6V when connecting the SI? Really?

So, 2-Wire is a parasite configuration  - it powered up by the implant. It is stated in the manual to use 12V there as power supply. It works only with specific sensors.

If you want to read the voltage directly you need to treat it as a 3-wire sensor and disable pull-up in Implant input. 

 

On 1/24/2025 at 6:03 PM, Łukasz997 said:

But gap can be set as low as 60s. - so from where those 30 secs come.

Z-Wave specification does not allow to send this more frequent, sensor values has specific limitation for that. 
 

[Łukasz997 117]

15 godzin temu, Tim__ napisał:

I have connected a Smart Implant to a kitchen range hood, using the status LEDs to determine the state of the fan and lights. What you need to ensure is that your ground wire (blue) is connected to the ground of your appliance. This is not necessarily on the output side of the LED as you have shown. You may have introduced a short circuit between LED output and ground, and might be why the mainboard is reporting a fault. In my case, the sensor wire reports about 4V when the LED is on, and 2V when off. I use LUA to read the voltage and convert that into a state (and use it to toggle the range hood's buttons as appropriate).

 

As to your concern about the voltage on the sensor cables when not connected to anything, i don't believe you need to worry. These are high impedance floating inputs and so that voltage should not appear once the the wire is connected into a circuit. You can check this by connecting the sensor wire to a dummy circuit of two resistors (one to + and one to -) and use a voltage meter to see if the voltage is stable when the SI starts up.

It's very likely what you're saying - the shortcut to gnd. Only basic knowledge about electronic circuits does not help me.

It's possible that when the implant would be powered from different (isolated) source that board, the problem will not consist, but it can't be powered other way.

 

So, what do you mean ground of appliance - the heater chassis? The electronics of the heater consist of two board: main and connected to it this one with led. Connecting is made with micro-micro-tiny ribbon, of course not described...
I'm not pretty sure that it helps, by I'll try.

50 minut temu, m.roszak napisał:

So, 2-Wire is a parasite configuration  - it powered up by the implant. It is stated in the manual to use 12V there as power supply. It works only with specific sensors.

If you want to read the voltage directly you need to treat it as a 3-wire sensor and disable pull-up in Implant input.
 

Thanks, but could you write on my schematic how exactly wires should be connected? Pull-up is disabled.

[Tim__ 126]

8 hours ago, Łukasz997 said:

what do you mean ground of appliance - the heater chassis?

Ground is the "negative" or 0V side of the power supply. And you will already have a blue ground wire connected there, so there is no need to connect the other blue wire (they are simply connected to each other inside the smart implant). So you'll just use 1 blue, 1 red and 1 yellow wire to connect to your heater.