Sinking 12V through a Reed switch into a 9425

[dest2ko]

We have successfully had a system fielded for about a year now, and one of its several functions is to measure product flow through its channels. Given the application, chemical compatibility, cost, size and overall flexibility, I chose a Reed switch actuated flow meter and am sinking it into a 9425. I am using period-based calculations on the FPGA to derive a tick count and go from there, utilizing a known K-factor to calculate rate. The system is pretty accurate and has measured within manufacturer specified tolerance when compared to a Weights & Measures calibrated positive displacement flow meter that is far better.

Over time, especially recently, we have seen the flow meters get returned to us with tags describing two particular issues:

1.) Massive GPM rate (well beyond the range of the meter)

2.) No function

I recently tested a flow meter tagged "Broken, massive GPM" and replaced the side of the shell containing the Reed switch with a new one from the manufacturer. It worked correctly thereafter. My question is... am I protecting my Reed switch well enough?

I don't drive inductive loads. I don't heat lamps. The length of cable (containing 22AWG 2 conductor) is, at most, 20 feet each way to the meter. This obviously isn't an A/C circuit. The 9425 has an input impedance of 30kOhms +/- 5% and I additionally supply a 1kOhm in series for good measure. This is a 12V power supply with a common ground. The flow meter manufacturer is very poor on communication, but has more or less stated this is fine. I question their tech support.

I assume the 10V (ON state requirement)  / 30kOhm input impedance = how the minimum input current is derived? (330uA)

See attached image (apologies for the quality.)

Thanks,

R

[dest2ko]

Also curious about debouncing circuits and whether or not that is a concern until the reed contacts become worn or damaged. 

[mterz]

Do you think there are any issues on the 9425 input side? I think the protection concerns would be more on the other side of the circuit (where you have the resistor). Was the 1 kOhm based on specs for the Reed switch?

[Minions]

• Is your cable coiled?  Coiling of the wire can actually create an inductive path.
• Is your cable shielded?  Unshielded cable can easily couple up with AC power.
• Is your facility near a Radio/TV antennae tower?  We had a facility that was 20 miles from a tower and were able to read their broadcast carrier frequency with a scope and a small length of wire. An additional overbraid may be required for even a shielded cable.

[dest2ko]

No, not coiled.

Yes, shielded (foiled and braided).

As far as RF goes, there are usually no radio towers nearby. 

I will have to look into the RF a little further. These particular configurations have about 2 feet of the (shielded) cable exposed, max. The rest is contained in an aluminum with a custom weld job for the conduit that extends the length of the machine, so I'd figure that's at least some sort of Faraday cage in a way? I'm assuming a lot on this one. 

[dest2ko]

The 1K was recommended by the flow meter manufacturer, though they did say keeping it to less than 1mA would prolong the life.

I just didn't realize 12mA would damage it and render them useless after 3 months; I thought I was keeping it pretty conservative with a 1k.

If this is the case, to get it to 1mA would require a 12kOhm minimum resistor. Perhaps this is where my interpretation of input impedance regarding the NI 9425 may be off. I was under the impression that a 30kOhm input impedance means it's grounded through a 30k resistor and so shorting a signal from some VDC source would still be tolerable since it at least has that current protection built in. I also notice on the data sheet that 330uA is the minimum excitation current to get it to register an ON state. I also notice 10V is the minimum voltage. Is it coincidence that 10V / 30k = ~330uA? Can I add more resistance to the circuit? I think the 12V for the flow meter is what's killing me here.

Do I assume the current going through the Reed switch is:

1.)  12V / (30k + 1k) = 387uA, or

2.)  12V / (1k) = 12mA?

[Minions]

Is this for the activation of the relay or for passing through the common to the NO/NC contacts?  Even though a relay may take in a 12Vdc source for activation, it is usually good practice to place a simple 1N4148 diode across the relay coil with the cathode attached to the +Voltage input.

[dest2ko]

I don't have any relays in this particular circuit; it is just a 12V source going through a resistor, then into an isolated Reed switch that is actuated open and closed by a turbine (magnetic) within the flow meter to complete the circuit, sinking it to the 9425.  (tick counter) 

[dest2ko]

I will add that the units are staked with grounding rods upon deployment to a location.