Can somebody please help me understand this?

[GrizzlyAK]

I'm trying to connect a ZWave sensor to my Septic's High Water alarm to control a powered water main valve. I have it working, basically, but am having trouble understanding some of the things which are beyond my current level of circuit knowledge.

What I started with
I have an Orenco Septic system that has an indoor alarm panel that is connected to the main PLC outdoors by a two conductor 16g wire. Here's what I know:

1. The voltage across the wires going to the PLC outside (when disconnected from the alarm) is 0V, and there is no continuity. Open circuit.
2. The voltage across the terminal on the inside alarm to which the wires are connected is 15.2V (There is an AC adapter powering it).
3. When I short the wires (or the terminals on the alarm box), the alarm sounds, and the voltage across those terminals drops to 0V.
4. Orenco tech has confirmed that this is normal behavior.

I assume this circuit is just a test for circuit closure by the outdoor PLC when the high water float goes high. What I don't understand is how this circuit works. Is this a 'typical' type of circuit for measuring dry contact, i.e., apply a voltage across the wires? Are there schematic examples you can provide? How does it not require a load to short?

What I want to do
I have a Fortrezz MIMOlite Wireless Interface/Bridge Module which has a single input sensor (digital and analog functions) and a single (relay controlled) output switch, which communicates over ZWave. It is powered by an AC adapter to about 13.5V nominal (according to specs). You can find tech specs here. I want to configure this unit so that the input sensor reads circuit closure on the alarm wires coming from the PLC to the Alarm box, so I can then signal my powered water main valve to shut off via ZWave (to prevent flooding). Here's what I know about it:

1. The voltage across the input sensor terminal is 2.6V when powered with nothing connected
2. When I short the terminal (or wires attached to it), the voltage drops to 0V. So this circuit appears to operate like the one above; a 'circuit closure sensing' circuit.
3. When I do #2 above, the MIMO sends a signal via ZWave to my automation controller that the sensor has been triggered.

OK, so far so good. Initially, I thought I'd just connect the MIMO in parallel to the alarm, i.e., connect the wires from the MIMO input sensor to the terminal where the wires from the PLC are attached to the alarm. As soon as I do this, the alarm sounds, which means the MIMO shorts the circuit. Then I tried connecting the MIMO in series, and that seems to work. It doesn't short the alarm (see real-world test below).

Aside from above, another thing I don't understand are the voltage readings across the circuit when I connect the MIMO up in series. If I treat each unit as normal voltage sources (like a battery) in series, then a simple circuit diagram would indicate either 12.6V (15.2V - 2.6V) or 17.8V (15.2V + 2.6V) end-to-end depending on how I connect the MIMO into the series circuit. Although the trend in voltage change is right, the absolute numbers do not match. I get 13.4V and 18.4V respectively. This discprepancy is probably explained by the type of circuits used in each device, which I don't understand. I'm hoping someone can explain it to me.

Finally, I did some real-world tests with everything connected in series by actually raising the high water floats to see what the system did. It DOES work, although I don't quite understand what I'm seeing. As the float was raised, I measured the voltage between the alarm terminals (which is normally 15.2V) which didn't drop to 0V, but to either 8.6V or 12.2V depending on which way I wired the MIMO into the circuit. This same voltage was also measured across the MIMO input terminals, which makes NO SENSE TO ME. (NOTE: The MIMO allows me to set upper and lower thresholds so I was able to bound the 2.6V nominal closely, so that even though the voltage across the alarm terminal didn't drop to 0V as it does when shorting, it was enough to trigger the sensor.

What I need
Besides answers to the above questions about the circuits, I'm hoping someone can help me understand what is going on here in my real-world test. I want to make sure I'm not causing any damage to my system with this particular setup.

Please let me know if you have questions, and I'm hoping someone out there can help.

Cheers,
Shane

 

[rjenkinsgb]

Hi,

A typical input circuit to work with a "dry contact" system - such as a PLC, or an alarm system that senses switch closures / openings, has something like an LED in an optocoupler or a transistor base connected to the input terminal via a combination of series and parallel resistors, that allow an appropriate current through the sensing device when there is a voltage above the defined input threshold, and also provide some loading below that to prevent or minimise false inputs due to electrical noise or static.

They also often have other components to filter noise spikes and possibly protect from excess voltage.

Depending on the system, that circuit is either connected to 0V / "ground" and needs a positive voltage to switch it, or connected to the unit positive supply and uses a connection to ground to switch it.


In principle you can think of it just as a lamp or LED and resistor.
In the alarm, one side is connected to the nominally 12V feed, and connecting the input to ground / 0V turns it on.

In a PLC it's often common to ground and an external supply (eg. 12 to 24V) turns the input on.


From a quick look at the mimolite data:

I'd connect 12V power from the alarm to the device.

Connect the wires from the water alarm to the input terminals on that; it should read supply voltage with the water alarm not triggered and 0V or nearly so when triggered.

Then use the Z-Wave settings, Parameter 3, to slave the relay to the input - and connect the relay contacts back to the original alarm panel input.


Edit - a Fibaro "binary sensor" may be a better option if you have problems with the mimolite unit.

That's designed for converting switch inputs to Z-Wave and the user manual is a lot clearer. They also have two separate channels plus the facility for a few temperature sensors.
The principle of connecting the switch / dry contact to the unit then using a relay contact to pass that on is still the same.

Data:

https://manuals.fibaro.com/content/manuals/en/FGBS-321/FGBS-001-EN-A-v1.1.pdf

I believe the Fibaro "Smart implant" is pretty similar, except in a box rather than being a bare PCB module.


That should pass the alarm signal and give a Z-Wave trigger.

 

[GrizzlyAK]

That should pass the alarm signal and give a Z-Wave trigger.

rjenkinsgb Wow! Thanks for the detailed answer. Sorry for the delay in responding back. It all makes sense, even though I couldn't draw out the particular design of each of the circuits. I suppose there are myriad ways one could configure such circuits, and I knew my 'view' of the voltages across the devices as batteries was incorrect. Your descriptions of the process was helpful.

I followed your recommendation and now have it configured as you suggested and it works like a charm! I initially considered configuring the MIMO this way, but rejected the idea purely because I wanted to connect it as a completely passive observer to the existing system (i.e., I was afraid I'd mess things up). I also didn't know that putting 2.6V on those wires running to the PLC instead of the 15V the Alarm supplied wouldn't be an issue. But now I realize that (within a range), it really doesn't matter since it just closes the circuit. Plus, all those various weird voltages I was seeing, which I didn't really understand, ended up feeling anything but passive!

But, now, having it configured this way, it works as expected, and is indeed the best approach. When I short the wires from the PLC, the terminals on both the MIMO and the Alarm drop to 0V, as they should! I have the input set to trigger the relay automatically through Parameter 3, and also get reports on my hub via ZWave that I will use to trigger the Water Valve to close.

Thanks again for taking the time to help a guy out.
Cheers,
Shane

 

[rjenkinsgb]

Glad to hear it's working!

Just a thought; if the PLC uses relay output, 2.6V may be too low to reliably "wet" the relay contacts if they are unused for a long time.

Adding a resistor from one of the MIMO input terminals to either power or ground, whatever combination puts a higher voltage across the PLC output, may improve reliability.

I'd try a 1K half watt resistor temporarily in each position, one wire combination at a time, while monitoring the voltage across the PLC output, to see which one works - then connect it like that permanently.

If the PLC outputs are actually solid state, it's not needed.

Unfortunately the diagrams I can find for that device are not clear on exactly how the input sensing works & the internal configuration of the input.

 

[GrizzlyAK]

Hi, I'm not sure I understand what you mean by

if the PLC uses relay output, 2.6V may be too low to reliably "wet" the relay contacts if they are unused for a long time.

How does the time unused come into play in this circuit? Whenever I've measured the voltage across the wires coming from the PLC in any condition, the voltage is always 0V, and no continuity when untriggered, continuity when triggered. It appears to me that the Alarm unit is just a trigger/untrigger-and-forget device to the PLC. When asked about how the Alarm (Sentinel) works, the Orenco tech told me the following:

...as far as the connection point 24 and 25 that's what we call a dry contact and when in Alarm condition it closes. The sentinel panel connected to it is a low voltage panel due to the plug/transformer converting 120vac to the range of 6 to 12volts. So simply put when there is no condition 24 and 25 are open and when in alarm it's closed and a low voltage signal is sent.

24, 25 are the switch block terminals on the main PLC control box outside to which the two wires going to the Alarm are attached.

Based upon the state history of the MIMO analog sensor, voltage across the inputs both when disconnected and when connected, but untriggered, stays pretty much right at 2.6V. When triggered, it goes to 0V. I don't understand how a resistor in that part of the circuit would have any effect, other than lowering the voltage seen by the PLC when connected (before dropping to zero). Let me know if you think I still need to do something with a resistor.

Thanks!
Shane

 

[rjenkinsgb]

Hi,
relay contacts (and most types of simple switch contact) can slowly tarnish when unused for some period, and some level of voltage, the "wetting voltage" is then needed to guarantee the oxide layer breaks down and the contacts conduct when they close.
The longer the relay is unused, the more the voltage increases from zero, up to whatever maximum wetting voltage it needs; that may be 1V, 2V, 5V or higher.
More info:

Wetting current - Wikipedia

en.wikipedia.org

One of the input terminals on the MIMO unit is probably connected to either ground or power, with the other connected through a fairly high value resistance to the opposite (power/ground) to pull the 2.6V across the actual input measuring circuit.

Adding another resistor from one of the inputs to power or ground should increase the 2.6V to a rather higher level, so improving the wetting voltage.

Just measure the voltages on both MIMO input terminals, relative to the 0V / ground terminal and I can tell you which combination of connections to try adding a resistor to.

 

[GrizzlyAK]

Oh! I get that!! Is that what you'd call a 'pull up' resistor?

So, thinking through this, if the MIMO power brick supplies a 'nominal' 13.5 V and there's 2.6V across the terminals, does that mean you could calculate the existing impedance? And that 2.6V is the 'wetting voltage' supplied by the Alarm to the PLC relay which allows it to close the circuit and drop the voltage to zero, dumping the current in the existing resistor? And there may be an instance in the future where 2.6V won't be sufficient to do that, compared to the 15V provided by the Alarm originally?

I also understand what you're saying about the contacts tarnishing, increased voltage, etc. What that indicates then, is that on the PLC end, that relay requires some level of voltage to 'close' the circuit, once the relay closes the contacts? I guess I never considered that.

Very cool (assuming I'm understanding it). I have a degree in physics, but my basic electronic course was back in 1983'ish, when 6502's were all the rage!

Also, I've uploaded a couple of PDFs from Orenco's technical library to my Dropbox here, if you are interested. One shows the connection of the Alarm and the other is pretty much my control box (mines a little older). Not a lot of detail.

 

[rjenkinsgb]

You got it, exactly.

I just need the voltage measurements for figure out the internal connections.