Switch circuit connection

[Ira]

I want to detect the postion of this switch, preferably using a M1G zone, but other ideas are welcome.

The switch is inside a genset automatic transfer switch, so I don't have the option of changing it out. Only the common and NC terminals are being used. The NO terminal is empty. There are actually two switches that are mechanically opened/closed by the position of a lever in the ATS. When the lever is in the "up" position, one of the switches is open and the other switch is closed. The state of the switches reverses when the lever is in the down position. It is a "break before make" setup, so for a split second during transition, both switches can be open.

When utility company power is present, the common terminal on one switch has 120vac on it. When genset power is present, the common terminal on the other switch has 120vac on it. Both common terminals could have 120vac at the same time (utility power present and genset running). Also, both may not have 120vac present at the same time (utility power not present and the genset hasn't started, or failed to start).

Bottom line is I can't depend on the presence of 120vac to tell me the position of the switches. What I think I need to do is treat the switches as dry contacts, even though there may be 120vac present. For example, tap on to the switch's common terminal and the NO terminal and wire them to an M1G zone. My guess is, the 120vac on that zone would not make the M1G happy. So is there a way to effectively remove the voltage from the wires going to the M1G zone without affecting the NC side of the switches? It's like I need to remove the voltage from the NO "zone loop" without introducing resistance.

Keep in mind that the primary goal in this is to be able to detect the switch(es) position even when there is no power at the switch's common terminal. The only way an error should occur is if there is a mechanical/physical breakdown somewhere.

Thanks,
Ira

 

[Ira]

Adding email notification.

 

[BraveSirRobbin]

I'm not sure if I'm following this correctly, but first read of this I would think about adding two 120 volt coiled relays to monitor the AC voltage present between the NC and C of each switch. The contacts of these new relays would then be monitored by Elk input zones. Then just setup a logic to determine overall state (both off, both on, relay one on, relay two on).

 

[jwilson56]

I'm not sure if I'm following this correctly, but first read of this I would think about adding two 120 volt coiled relays to monitor the AC voltage present between the NC and C of each switch. The contacts of these new relays would then be monitored by Elk input zones. Then just setup a logic to determine overall state (both off, both on, relay one on, relay two on).

I would agree with BSR. Adding a relay across each switch and use the NO contacts of those relays into Elk input zones would be the cleanest and safest way to monitor the switches. I would make sure the 120AC relays you get are continuous duty if either is going to be energized for long periods of time.

 

[tmbrown97]

since you can't swap out the switches, i'd consider some minor rewiring - probably bringing some relays into the mix. leave the switches as they are, but remove the power connections that are there now and repurpose those switches to activate a relay or two.

Of course, you're screwing with 120V here, so don't do anything in there if you're not sure what you're doing.

 

[Ira]

I'm not sure if I'm following this correctly, but first read of this I would think about adding two 120 volt coiled relays to monitor the AC voltage present between the NC and C of each switch. The contacts of these new relays would then be monitored by Elk input zones. Then just setup a logic to determine overall state (both off, both on, relay one on, relay two on).

Reading back over my post, I probably over-complicated what I want to do, so let me try again (and sorry if I was confusing).

What I really want to know is the position of the lever in the ATS (not really the positions of the switches as I previously said, but that is how I would derive the lever position). The position of the lever determines the state of both switches, regardless of whether or not there is power at the switch(es). The ATS lever physically contacts (or "un-contacts") the hinge lever on the switches to make or break them. That gives me an indication of how the ATS is attempting to provide power to the house (utility company or genset).

If I rely on the voltage at the switches as the means to determine lever position, there is at least one scenario that can't be accurately determined -- the lever position when power is not available at either switch. Also, the lever can be manually repositioned which brings up other uncertainties. There are also some transition phases that would make determining lever position more difficult and less accurate. For example, when I lose utility company power, it takes 15 seconds for the genset to attempt to start (so it won't start on a momentary power loss). That means, there is a fifteen second period where neither switch has power. The reverse is true when utility power is restored -- there is a fifteen second delay where the lever stays in the genset position even though utility power is present, followed by a 30 second second cooldown period when the lever is in utility power position but power is present at both switches (similar to the exercise cycle).

This is probably a good example of the 80-20 (or 90-10) rule, especially using multiple M1G rules to attempt to more accurately derive the ATS lever's position from voltage at the switches, but if there is a simple way to treat the existing switches as M1G zones (dry contacts) without affecting the existing performance of the switches, that would be ideal to tell me the actual lever position in all scenarios that didn't have a mechanical failure.

Thanks,
Ira

 

[fcwilt]

Do you own this equipment?

Can you access switch or lever in any way such that you might detect the position optically?

 

[Ira]

Do you own this equipment?

Can you access switch or lever in any way such that you might detect the position optically?

It's my equipment. The switches and lever are in a pretty tight spot with a lot of "unprotected" 120vac stuff in very close proximity. I don't think you could get an optical sensor in there.

 

[jwilson56]

Going back to what BSR and I said.... have those two switches feed a pair of relays. The relay contacts can be monitored by the M1. I presume you want to have some alert when neither switch is being made after a period of time. This can be done with a rule. Unless I am missing something.

Switch one turns on relay one that feeds zone 9

Switch two turns on relay two that feeds zone 10

Have a rule for zone 9 and 10 that after a certain period of time output x turns on (or whatever)

If that is not what you are trying to accomplish then maybe take a step back and explain what it is your trying to do.

 

[fcwilt]

Do you have enough room to run a couple of light pipes from the outside into a position where light transmitted by one pipe could reflect off something shiny attached to the moving part and the reflection be "seen/not seen" by the other pipe?

This would put the light source, light sensor, etc on outside of the box.

For the light pipes you could heat and bend some clear plastic rod or use some optical fibers like http://scientificsonline.com/product.asp_Q_pn_E_3053301

You could perhaps pick up a light source and sensor from some place like edmund scientific or radio shack or the like.

With a few other parts you could end up with a contact closure to connect to your M1G.

 

[gatchel]

I am only getting part of the story because I am heading out now, but for what it's worth I have had success stacking micro switches using 4-40 screws or equivalent to add more dry contacts to a device to be monitored. Maybe this could be done in your application.

 

[Ira]

Let me see if I can explain why relays powered off the two existing switches won't cover all the bases. Remember that I want to be able to determine the position of the lever, i.e., is the lever up or down. How can that be done if utility power is gone and the genset won't start, meaning there's no power at all (but the M1G is on battery)? I guess I could write a rule that says if neither switch has power, the lever is in its last known position (before all power was lost), but that assumes the lever position wasn't changed manually. I know I'm splitting hairs, but I would rather not rely on the presence of voltage at one or both switches to determine the lever position since it can get a little complex.

I thought about adding another switch beside each of the two existing ones earlier today, but I didn't make it out to where the ATS is located to see if it's physically possible. Even if it does have room, I'm a little concerned about stability. When one of the ATS coils kicks in to move the lever, it's a pretty violent movement.

Here is a link to a picture of what I'm working with. The switches are between the red "shield" and the two coils that are in the center of the box. The bright yellow piece right below the top coil is actually part of the lever. You can stick a little "handle" provided with the ATS in the hole in the yellow piece when the cover is on the ATS to manually switch between utility and genset power if necessary. The blue wires with the "pink-esh" terminal lugs on them between the two coils and the red shield are connected to the terminals on the switches.

Ira

 

[BraveSirRobbin]

Could you mount/glue a small round magnet on the part of the switch/lever that physically moves up and down? How much distance does the 'throw' use (distance between xfer and normal positions)?

Also, you could monitor the status of the two coils, but I'm guessing this wouldn't be of any use for the same reason as the switches (i.e. no power at all)?

 

[Ira]

Could you mount/glue a small round magnet on the part of the switch/lever that physically moves up and down? How much distance does the 'throw' use (distance between xfer and normal positions)?

The throw is maybe two inches. I wouldn't trust glue because of how violent the action is. The position change is fast enough that you don't even notice it in the house. If I could find a way to mount a magnet, it would still be difficult to find a place to mount the sensor. In the picture, the fuse to the left of the coils is a regular 1.25" fuse. That gives a perspective on how little room there is to work with.

Also, you could monitor the status of the two coils, but I'm guessing this wouldn't be of any use for the same reason as the switches (i.e. no power at all)?

That's correct.

I'm not all that great on electronics, but can't something like a diode be put in series to eliminate the voltage, or would the diode also make it look like an open circuit? For example, for each switch, tap on to the common wire to one side of an M1G zone and put the appropriate diode in series in that line. Then run a wire from the NO terminal on the switch to the other side of the M1G zone. So when the switch is in a state where there should be continuity between common and NO, would the diode eliminate the voltage but not affect the continuity? Would the diode affect anything "upstream"?

Just for clarification, when the ATS lever is up against the switch lever, there is continuity between the common terminal and the NC terminal. So when the lever is in the up position, the upper switch has continuity between common and NC. When the lever is in the down position, the upper switch has continuity between common and NO. I'm pretty sure these switches are break before make, but I haven't tested them.

Thanks,
Ira

 

[Ira]

Took a look at the ATS again. Mounting additional switches next to the existing switches is a non-starter. The bracket that holds the coils in place fits up against the existing switches so there's no room to "stack" any. Also, the ATS lever that hits the switch lever has a small plastic tab that actually makes contact with the switch lever. The tab is the same width as the switch, so even if the coil bracket wasn't in the way, the tab wouldn't contact the added switch lever the same way. There's really no room in there to add any additional components.

Ira