Looks like the best and safest method is going to be to sense the position of the lever somehow. You mentioned having an external handle that attaches to that yellow tab and sticks out through the front of the panel. You could possibly attach a microswitch to the outside of the front to sense the position of the external handle.
Switch circuit connection
- Thread starter Ira
- Start date
The external handle for the lever is only supposed to be in the hole in the lever when it is being used to reposition it. There's a small bracket on the ATS face plate that holds it when it's not in use. It's just a "slip fit", so unless something was done to secure it in place, it would fly off when the ATS switched automatically. Plus, you would have to cut a slot in the front cover to allow for the motion of the handle.
The ATS is kinda like a breaker panel in that there is a "face plate" that covers everything but the two disconnects (and has a slot for the lever handle). There's a front cover that goes on after the face plate that covers everything and helps make the ATS waterproof.
Ira
BraveSirRobbin
Moderator
Get a FLEX SENSOR which changes resistance as it "flexes" and have it mounted so one end is anchored, and one end is somehow attached to a part of the mechanism that moves when transfered. When the switch moves, this resistance will change and you can monitor this via the M1. You may have to provide some excitation voltage and have the Elk in analog mode, but I believe this is all doable. Since the exictation voltage will come from the Elk you will know the position of the switch, even during a power outage. As a matter of fact, you may be able to fine tune this enough to know if the switch is in an 'in-between' state.
The solenoids are connected by a stamped steel sleeve (aluminum colored). I don't think the lever floats if neither coil has power. There are some pretty big springs (can't see them in the picture) above and below the micro-switches that I think probably hold it in place and augment the motion when it changes. The picture shows it in "normal" mode (utility power present). If I lose utility power, the lever will stay in the same position until the utility power has been off for 15 seconds and the genset has started and run for 15 seconds (after the initial 15 seconds of no power). In other words, when I lose power, I don't get it from the genset for about 30 seconds. During that thirty seconds, the lever stays up, but I don't know what force or how much force is holding it up. What I do know is that when it switches in either direction, it slams hard. My guess is the springs are responsible for that, or at least helping a lot.
Ira
tmbrown97
Senior Member
Hey Ira,
I actually understood exactly what you were saying from your very first post - especially the part about wanting detection even if neither circuit has power. Unfortunately, I just didn't explain my option very well. I wish I had time to diagram it out...
Here's the short of it though - I know that the normal relay trigger doesn't accomplish your goal because relays don't work without power. That's why my suggestion was to take the existing two switches and totally repurpose them. Meaning, instead of them connecting 120V across a circuit, instead have them connect something like a ground. A ground can be detected across an M1 zone and can also be used to trigger a relay at the same time. This way regardless if either source has power, you'd have an accurate indication of which position the switch is in.
Now of course you're lacking the operation that the switch originally provided. That's the part that really only matters if there's some power moving through the device. That's where I'd use the relays. Depending on which position the switch is in, use that ground signal that's already acting as an indicator to the elk also trip the relay that connects that 120V source. Use the same thing on the other end of the switch.
Now this leaves open the question of how to get a ground signal that can trip a relay and not hurt the elk. This is where a better electrical engineer could probably answer some specifics better; I know that two 12V transformers powered off the respective 120V legs could do it then you could use two 12V relays; but I'm sure there's a simpler way than even that.
In the end, it works within the constrained space of the existing switch; retains all existing functionality, and still provides detection even if neither power source is on. Doesn't that about cover it? It looked like to do have room in that box for some other stuff; just can't tamper with the switch much. Of course, this is dealing with a lot of 120V, so you better know what you're doing and protect everything inside, etc; and if it shorts something out, warranty isn't going to do anything for you.
If I'm still not making sense, let me know and I'll see what I can diagram out.
I actually understood exactly what you were saying from your very first post - especially the part about wanting detection even if neither circuit has power. Unfortunately, I just didn't explain my option very well. I wish I had time to diagram it out...
Here's the short of it though - I know that the normal relay trigger doesn't accomplish your goal because relays don't work without power. That's why my suggestion was to take the existing two switches and totally repurpose them. Meaning, instead of them connecting 120V across a circuit, instead have them connect something like a ground. A ground can be detected across an M1 zone and can also be used to trigger a relay at the same time. This way regardless if either source has power, you'd have an accurate indication of which position the switch is in.
Now of course you're lacking the operation that the switch originally provided. That's the part that really only matters if there's some power moving through the device. That's where I'd use the relays. Depending on which position the switch is in, use that ground signal that's already acting as an indicator to the elk also trip the relay that connects that 120V source. Use the same thing on the other end of the switch.
Now this leaves open the question of how to get a ground signal that can trip a relay and not hurt the elk. This is where a better electrical engineer could probably answer some specifics better; I know that two 12V transformers powered off the respective 120V legs could do it then you could use two 12V relays; but I'm sure there's a simpler way than even that.
In the end, it works within the constrained space of the existing switch; retains all existing functionality, and still provides detection even if neither power source is on. Doesn't that about cover it? It looked like to do have room in that box for some other stuff; just can't tamper with the switch much. Of course, this is dealing with a lot of 120V, so you better know what you're doing and protect everything inside, etc; and if it shorts something out, warranty isn't going to do anything for you.
If I'm still not making sense, let me know and I'll see what I can diagram out.
Todd,
I think I understand what you are saying now. Have the two switches be used solely as NC dry contact zones for the M1G, i.e., remove the existing 120vac wires from them completely. Low voltage outputs from the M1G could then be used to power a couple of relays that would emulate the state of each switch. The 120vac wires that were originally connected to the switches would now be connected to the new relays. Since the M1G is backed up by battery, it's impervious to the presence of 120vac.
So the first scenario, in the old setup where 120vac is present at the COM terminal and the switch is closed, therefore 120vac is present at the NC terminal...in the new setup, the zone loop thru the switch would be closed (but no 120vac power thru it), the M1G having a rule that says turn output nnn on when zone nnn is closed would have output nnn on which would close the contacts on an external relay. Since the original 120vac circuit thru the switch is now going thru the relay contacts, the original circuit is complete the same way it was before. I have the desired benefit of knowing that the lever is pressed against the switch because of the closed zone.
Second scenario, in the old setup where 120vac is not present at the COM terminal and the switch is closed due to lever position, the M1G still knows the position of the switch from the closed zone, so it turns on output nnnn which closes the external relay completing the original 120vac circuit. But since 120vac is not present at the original COM wire, it isn't present at the original NC wire (both of which are now connected to the external relay contacts).
Third scenario, in the old setup where 120vac is present at the COM terminal but the switch is open due to lever position (so 120vac is not at the NC terminal), the M1G knows the lever position because the zone is open, so it turns off output nnn which opens the external relay which keeps 120vac from flowing from the COM wire to the NC wire.
Fourth and final scenario (?), where 120vac is not present at the COM terminal but the switch is open due to lever position, the M1G knows the lever position because the zone is open, so it turns off output nnn which opens the external relay, which doesn't really matter because there's no 120vac present.
Unless I wanted to be able to detect actual mechanical failure, e.g., lever or switch breaking/malfunctioning, I could do everything by just re-purposing one switch. Since the whole exercise is solely for monitoring and not taking any action other than notification, re-purposing one switch would probably do fine.
Assuming that delays caused by the M1G and external relays don't cause a timing problem (because the "action" would be slower), I think it would be a really nice solution. I'm not sure I would want to do it on something that is still under warranty (manufacturer may also use ATS modifications as a reason to void genset warranty). Then again, I may not do anything due "warranty concerns". The other concern would be the added complexity in what is now a pretty simple design. What would happen if somehow things got out of sync, e.g., for some reason the M1G changed the state of the output when it shouldn't have, maybe if it had to be restarted. Maybe additional controls can be built in to have a failsafe mode. If everything was out of warranty, I would be wiring it up instead of writing this long, drawn out post.
I wish manufacturers would give more consideration to simple things like adding dry contacts to devices for monitoring purposes. I would have paid the extra $5 for it.
Regards,
Ira
tmbrown97
Senior Member
I'm actually sitting in a meeting right now so I can't get into too much detail... but t seems like what you've understood is very close to what I was describing... the only exception is that I actually wouldn't use the Elk to control the original functions that the switches handled; I would repurpose the switches as you understood, but instead of using the switches to signal the elk exclusively then letting the elk exclusively control the generator... I'd find a way to mechanically let the ground signal that triggers the elk ALSO trigger the relay; that way the elk is only involved to monitor, not control; control operation would continue completely independantly of the Elk, just with a slightly different circuit design of the switches/relays.
I think I understand. Another posiibility (and this may be what you have in mind) that may serve to best isolate the M1G...maybe the switch should be re-purposed to simply provide low voltage. The switch's COM terminal is connected to the negative side of a low voltage source. The switch's NC terminal goes to an external relay coil. The other side of the relay coil gets the positive side of the low voltage source. The relay is a DP?T, so it has two sets of contacts. One set of contacts is connected to the M1G zone. The other set of contacts is connected to the original wires that were on the switch's COM and NC terminals. Alternatively, two SPST relays could be used with the low voltage source connected to both of them in parallel.
I could use a small UPS to power the AC/DC adapter to provide the low voltage source so that the relay(s) would always mimic the actual switch state regardless of whether or not utility/genset power is present, at least until the UPS ran out of power, which could be days. I guess if I use a relay that has a 120vac coil, I don't even need the AC/DC adapter.
Thanks,
Ira
I could use a small UPS to power the AC/DC adapter to provide the low voltage source so that the relay(s) would always mimic the actual switch state regardless of whether or not utility/genset power is present, at least until the UPS ran out of power, which could be days. I guess if I use a relay that has a 120vac coil, I don't even need the AC/DC adapter.
Thanks,
Ira
pipeman
Active Member
Ira, I've been following this, but still may have missed something.
Have you checked the voltage on the NC side of the switches? It think your first post says the common terminal goes to 120VAC when that switch is depressed. That indicates to me that the common terminals of each switch are driving some other component, and then the "other" side of those components would be connected to ground. IF (notice the big IF), that is the case then a signal might be acquired safely from the NC contact.
Switch is closed = open on NC
Switch is open = possible ground on NC (through the downstream component)
If the downstream component is a coil or capacitor, you could have problems with an inductive spike (?) or stored charge?
In any case, it seems if you're interested in the switch position only and if both switches are mechanically interlocked through the common shaft, wouldn't you need to monitor one switch only?
Have you checked the voltage on the NC side of the switches? It think your first post says the common terminal goes to 120VAC when that switch is depressed. That indicates to me that the common terminals of each switch are driving some other component, and then the "other" side of those components would be connected to ground. IF (notice the big IF), that is the case then a signal might be acquired safely from the NC contact.
Switch is closed = open on NC
Switch is open = possible ground on NC (through the downstream component)
If the downstream component is a coil or capacitor, you could have problems with an inductive spike (?) or stored charge?
In any case, it seems if you're interested in the switch position only and if both switches are mechanically interlocked through the common shaft, wouldn't you need to monitor one switch only?
If the solenoids have the same common could you wire in an external latching relay? Use the solenoid common to be the common to the latching relay. Use the trigger wire for each solenoid to control the status of the latching relay. The relay has a magnet in it to maintain the last status regardless of power. This won't give you the actual xfer switch position but it would follow the status of the direction of the switch electrically. This would also give you dry contacts to send the signals to the M1. What do you think?
