• If you like antique engines, vintage tractors or old iron machinery, register and join us. When you register on Smokstak, please give complete answers and fill in all blanks. IF YOU ARE ON WIRELESS OR SATELLITE, GIVE YOUR CITY AND STATE! NO ZIPCODES! All registrations are manually approved.

Setting up an automatic transfer switch for a 120v generator

wh1630

Registered
Good afternoon, everybody

Now that my 1961 Kohler 2.5R21 is up and running, I have changed my focus on setting up an automatic transfer switch to start it whenever the power goes out. However, I've reached a little bit of a conundrum.

This is how I have the switch set up:

The actual switching of the load is accomplished through a 40 amp double-throw relay. One 'throw' is for normal power and the other is for emergency power. This relay is electrically held in the normal position and drops into the emergency position when the normal power is cut. This ensures that the generator and utility power can never meet.

When I designed this system, I wanted to have a time delay before the generator starts and stops. On the utility (normal) side, there is an agastat 7012 on-delay relay that switches the power to an agastat 7022 off-delay relay. The 7012 is to provide a return-to-normal time delay by supplying power to the 7022 after ten seconds. The start wires to the generator are wired to the 7022's delay-on-dropout terminals, so when the normal power is removed, it will close after ten seconds. Once the normal power returns, there will be that initial delay from the 7012 before the coil of the 7022 receives power, thus opening the start terminals and stopping the generator.

I have tested the switch and it works as I intended it to.

Now for the tricky part:

The building I'm going to install the generator in is fed by a 30 amp 120/240 circuit. There is a subpanel with only two 20A circuits on it. I only want the generator to power one, as it can only produce 120 volts at 21.5 amps (I do have a 20 amp fuse on its output and the internal control wiring of the transfer switch is fused at 5 amps). I have realized that if I put the transfer switch behind the 20A breaker and the breaker were to trip for some reason, It would cut power causing the generator to kick in. Obviously not ideal. The thing is though, I'm not quite sure how to approach this...

Thank you.
 

LWB250

Registered
Based on what you describe, the only way it could work fail-safe is if you put the transfer switch between the main (30A) breaker and the 20A breaker you want it to feed.

Is the potential load from the second 20A breaker such that it would overload the generator with the intended 20A breaker in the circuit as well? If so, I don't see any other way to do it.
 

Wayne 440

Registered
Is this building a new occupancy for you? If not, how many times has the subject 20A breaker tripped "for some reason"??? IF the answer is not a significant number, I wouldn't worry about the 20A normal supply breaker tripping. IF the reason for the trip still exists, the generator fuse will blow as well; if it doesn't, the generator will be doing what it is supposed to, supplying the load(s), since the utility supply is interrupted.

What I would consider is adding a circuit to shut the set down if it does not have output to the load terminals. No point wasting fuel if the load is not getting power.
 

wh1630

Registered
Based on what you describe, the only way it could work fail-safe is if you put the transfer switch between the main (30A) breaker and the 20A breaker you want it to feed.

Is the potential load from the second 20A breaker such that it would overload the generator with the intended 20A breaker in the circuit as well? If so, I don't see any other way to do it.
That was what I was thinking, but the problem is that the generator has only two wires (l1 and neutral) while the subpanel has three (l1, l2, and neutral), so I think putting the transfer switch before the panel is out of the question.

Is this building a new occupancy for you? If not, how many times has the subject 20A breaker tripped "for some reason"??? IF the answer is not a significant number, I wouldn't worry about the 20A normal supply breaker tripping. IF the reason for the trip still exists, the generator fuse will blow as well; if it doesn't, the generator will be doing what it is supposed to, supplying the load(s), since the utility supply is interrupted.

What I would consider is adding a circuit to shut the set down if it does not have output to the load terminals. No point wasting fuel if the load is not getting power.
The building isn't a new occupancy for me. The subpanel was installed about three years ago and none of the circuits have ever tripped.

I'm curious as to whether or not I can tap one phase and neutral off of the incoming 30A feed (using appropriately sized wire), run that through the transfer switch, and then have a dedicated, separate breaker for the circuit I wish to provide backup to. Gonna see what the NEC has to say about that, although if anyone else has an idea, feel free to chime in.

Thanks.
 

LWB250

Registered
If it's as you describe then the only choice is to break the 20A circuit out separately from the panel and feed the ATS from it. You didn't say what was on the other 20A circuit and why you couldn't include it. There may be a way to load shed it.
 

Zephyr7

Registered
Note that the NEC states that an ATS must be “electrically operated and mechanically held”, which means it can’t just be a relay with the coil energized all the time. Your “ATS” technically does not meet code.

The way I would approach your problem is like this:

Feed the subpanel through the ATS with a 40A single pole breaker. This gives you full capacity to both 20A circuits if you ever need it.

Bus the two legs of the subpanel together so that all spaces feed from that single incoming supply wire from the 40A breaker through the ATS. Label the subpanel “120V ONLY”

Leave the two 20A breakers alone. This precludes any possibility for the subpanel to support 240v loads, but since your genset is 120v only, you don’t really have any 240v options anyway.

Bill
 

countryboy07

Registered
If you are going to wire it that way I'd recommend using a starter of the appropriate size over a relay. It would be a nema #2 I believe. The coil is alot beefier in a starter if your going to hold it energized 24-7 how ever I'd consider looking into wiring it n/o and energising it when the generator starts. You'd probably have to use a reversing starter to interlock it so only one side can be pulled in at once.
 

Zephyr7

Registered
I would NOT use a starter solenoid! Those are only rated for starter voltages, maybe up to 24v or so. Do NOT try to use those on a 120+ volt power system!

A 40 amp contactor would be fine energized all the time as they are intended to operate that way. As I mentioned before though, an ATS really shouldn’t be built with relays or contractors with “energize to stay on” coils.

Bill
 

Vanman

Subscriber
Last Subscription Date
07/10/2019
I believe he means a motor starter. I hope!

Nice thing about a starter is that you can put the correct heaters in it to very nicely protect the generator from overload.

And using a reversing starter is very clever! Mechanically and electrically interlocked. Sweet!

Keith
 

countryboy07

Registered
Yep I ment motor starter. A contactor is just a heavy duty relay, a starter has overloads. The main thing to remember about a xfer switch is it needs to isolate the load from the line first and foremost, so a reversing starter is mechanically interlocked so if you were to have contacts weld or something like that you can't inadvertently backfeed the line
 

wh1630

Registered
Note that the NEC states that an ATS must be “electrically operated and mechanically held”, which means it can’t just be a relay with the coil energized all the time. Your “ATS” technically does not meet code.

The way I would approach your problem is like this:

Feed the subpanel through the ATS with a 40A single pole breaker. This gives you full capacity to both 20A circuits if you ever need it.

Bus the two legs of the subpanel together so that all spaces feed from that single incoming supply wire from the 40A breaker through the ATS. Label the subpanel “120V ONLY”

Leave the two 20A breakers alone. This precludes any possibility for the subpanel to support 240v loads, but since your genset is 120v only, you don’t really have any 240v options anyway.

Bill

I think that's the most logical approach. There isn't anything in that building that runs on 240 volts anyways. I do have an Allen-Bradley 50A reversing contactor on hand too, so I'll implement that instead. I don't think it has any thermal overloads on it, though. I suppose it's just a matter of figuring out how to connect the thing to meet the conditions of being "electrically operated and mechanically held" and which of the two coils on the contactor gets connected where.

Thank you for the information, too.
 

Zephyr7

Registered
Be sure to wire in any limit switches your unit may have. Some of those mechanisms don’t want the coil to stay energized after going over, so there will be a limit switch to cut power to the coil once the unit has completed the changeover cycle. You’ll want to make sure the system can only energize the unit in the correct order too, which is easy to implement with some relay logic in addition to the limit switches.

Nothing a little experimentation can’t figure out :)

Bill
 

wh1630

Registered
Be sure to wire in any limit switches your unit may have. Some of those mechanisms don’t want the coil to stay energized after going over, so there will be a limit switch to cut power to the coil once the unit has completed the changeover cycle. You’ll want to make sure the system can only energize the unit in the correct order too, which is easy to implement with some relay logic in addition to the limit switches.

Nothing a little experimentation can’t figure out :)

Bill
So it looks like the unit I have is electrically interlocked with microswitches to prevent both coils from being energized at the same time, in addition to being mechanically interlocked. I would consider myself fairly knowledgable when it comes to creating circuits with relay logic, too. Given the configuration of the contactor I have, I would imagine that the normal power would have engage one side of the contactor, thus isolating the generator side. When the normal power drops out and the generator kicks in, it would engage its side of the contactor, thus isolating the normal side.
Using relay(s) I would make sure that the power on the normal side takes priority over the generator side, allowing it to automatically re-transfer when the outage ends. Does this sound about right or make sense?
 

Zephyr7

Registered
That sounds like it will work. You’ll probably want to put a time delay relay in so that “return to normal” doesn’t happen immediately when utility power is restored. A frequency sensitive relay would be best for “transfer to emergency” to make sure the generator is up to speed before it gets hit with load. If you can’t find a frequency selective relay, just use a time delay relay that gives the unit enough time to get up to speed before the ATS goes over.

Usually ATSes are configured so that the coil to transfer to emergency is powered from the emergency source and the coil to transfer to normal is powered from the normal source. This ensures the ATS can’t go over to a source that isn’t energized regardless of what the logic tries to do. All the relay logic does is open or close the power to the coils based on whatever events you want to incorporate in your system.

Bill
 

wh1630

Registered
That sounds like it will work. You’ll probably want to put a time delay relay in so that “return to normal” doesn’t happen immediately when utility power is restored. A frequency sensitive relay would be best for “transfer to emergency” to make sure the generator is up to speed before it gets hit with load. If you can’t find a frequency selective relay, just use a time delay relay that gives the unit enough time to get up to speed before the ATS goes over.

Usually ATSes are configured so that the coil to transfer to emergency is powered from the emergency source and the coil to transfer to normal is powered from the normal source. This ensures the ATS can’t go over to a source that isn’t energized regardless of what the logic tries to do. All the relay logic does is open or close the power to the coils based on whatever events you want to incorporate in your system.

Bill
If it's an interlocking reversing contactor set, adding an ICM-102 to EACH coil, and setting it for say... 5 seconds... would be about right:

https://www.icmcontrols.com/productdetails/ICM102
Thank you, guys.

@Bill, Funny you would mention the frequency or time delay, because when I constructed the first prototype of the transfer switch, I noticed that as the generator would start up and come up to speed, it would cause the old relay to 'flutter'. Because I did not want this to happen, I installed an amperite thermal time delay relay tube as a 'buffer' to give the generator five seconds to come up to speed before transferring.

@dkamp, the time delay for transfer and re-transfer is accomplished by the two agastat relays. I had considered using a solid-state timer like the one you suggested, but I was able to get the two agastats for pretty cheap. Because they are electro-pneumatic, they can close a set of contacts after the coil power has been removed. Also, I just happen to find electro-mechanical things so much more fascinating. But thanks anyways!
 

dkamp

eMail NOT Working
WH... the Agastat, being mechanical, has a clear advantage with respect to electrical transient damage, and if you already had, or acquired them for ELP (Extremely Low Price) or MFP (My Favorite Price), then by all means yes!

The ICM102 is an inexpensive, dandy, and handy gadget, I use them for making 'normal' relays and contactors into delayed-action. I particularly like that they work at both low, and high voltage ranges, but they ARE solid-state, so a good solar flare might zortch them.

In a similar, but not the same (?) circumstance, I have a contactor panel on my generator shed wall consisting of three contactors... appropriately sized, but they're 'electrically interlocked'... the auxiliary relays of two, are wired in series with the coils of the third... in effect, there's three 'knockout' loops... making them exclusive... only one coil can be energized. I have them wired to each of my generators, as an automatic selector.

When one of the generators powers up, it's respective contactor pulls in, and the aux contacts disable the other two.

Unfortunately, there's no product (of if there is, it's not within my ELP/MFP grasp) that will MECHANICALLY interlock the contactor shorting bars, and I don't have any reversing contactors (I'd need two sets for three machines) to make it fly, I just built it with what I had, and oversized the contactors so much that they'd 'probably' never weld. If they do, fuses will blow, lights will go out, and we'll sigh and take a walk with the flashlight, right?

But I see same flutter on contactors when the output is dropping out. Usually not coming up... and there may be some wisdom in this thought:

The difference between a "RELAY" and a "CONTACTOR" is often disputed, but I will identify a CONTACTOR as being designed such that it consumes less power to 'hold' than it does 'pull in'.

Most frequently, this characteristic is found in an AC device, because the shorting-bar's armature has a lower inductance when open, but rises substantially when closed. When AC voltage is applied to the contactor coil, the armature pulls in, and once there, closes (and 'contacts') the electromagnet base, completing a magnetic circuit that holds with great strength, and at that point, inductance of the coil rises substantially, limiting incoming AC current to the coil.

Which doesn't mean much to this thread, except for... that once pulled in, the generator's output needs to fall LOTS before that contactor will drop out... and once pulled in, the contactor will WANT to hold (because it takes much less to hang on, then it does to grab, right?)

If your relay flutters on the way IN... then a CONTACTOR (even a tiny one) is what you really want... it has that 'holding hysteresis' that you need... without adding a delay relay.

But I personally would KEEP the delay relay anyway, because I'm a Belt-And-Suspenders guy when it comes to building things that break.

As a totally aside, unrelated, and irrelevant note... In a DC application, inductance has no effect, BUT, the result of a closing armature under DC is the same as AC- the magnetic circuit is completed by contact, thus, the amount of current required to HOLD is much higher. In this case, the coil consists of two separate windings- one is a high current (shorter, thicker) 'pull in', while there's a secondary 'holding' coil, both on the same polepiece, and in the same polarity. They're wired in parallel, however, an auxiliary contact of the shorting bar opens to disengage the 'pull in', leaving just the 'hold' contact active.

I've seen a variety of clever uses of these characters, and the ingeniousity of our predecessors never ceases to amaze me... one of my favorites, is when the dual-winding DC polepiece contactor's coils are used as logic... like... the Pull IN is used to pull in the bar, the HOLD coil keeps it pulled in, and then the PULL IN is later energized with polarity REVERSED to counteract the HOLD's magnetism, to spit the bar OUT...

And I realize, when I see really clever stuff like that, just how much of a moron I am. ;-D
 

Zephyr7

Registered
Contactors also generally have two sets of contacts in series per pole instead of just one like a typical relay. Contactors are expected to be able to interrupt power to inductive loads like motors.

Dkamp: check out “letching relays” sometime they come in two flavors: two coils, one set and one release (pulse the “set” coil to latch the relay in one position, pulse the “release” coil to latch it in the other position), and single coil where you pulse it with one polarity to latch in one position and you pulse the opposite polarity to latch it in the other position. Magnetic wonders at work.

I totally understand your appreciation of the cleverness that went into a lot of old devices. Nowadays everything tends to be brute force with a faster microprocessor. Every now and then though, you’ll see something that combines a fast microprocessor AND cleverness and then you can do some really amazing stuff!

Bill
 
Top