There are tens—if not hundreds—of ways to instrument a generator so you can tell it is running, and producing a voltage. As has been pointed out the tough thing is determining if it can do that into a representative load.

I think the TS‘ original idea was a sort of path-of-least-resistance approach. Since the generator is going to produce power, the relay seems a low-effort, low-bother method. But, I am not sure the reason for the signal was ever articulated.

What is the problem that the signal will solve? That will go a long way to deciding if the relay, or any analogue of that approach, really is a solution.

 

If one wants to know if the Gen is powering a load, and at what load, just wrap #24 wire around one of the Gen's power-out lines, to make a pickup coil. If the Gen starts then we should safely assume a transfer switch put load onto the generator, so there should be at least some load on the generator if it's running. A pickup coil can sense the amp range 0-Gen(max), etc. However, that all requires way more stuff than say just a micro switch.

The "is it running" question can be solved many ways. Even the floating FET method can detect when that electrically-noisy gen is running.
http://amasci.com/emotor/chargdet.html

 

Since the generator is going to produce power, the relay seems a low-effort, low-bother method.

I've got $5 that says it won't output a single volt at half speed.

But, I am not sure the reason for the signal was ever articulated.

Agreed. Is this really a "test?" If so, what is criteria for pass/fail? That the engine runs? That it outputs voltage unloaded? With a load? I personally think it's not a "test" so there is no pass/fail criteria; it's just a weekly exercise of the engine (not the generator). So what does the signal signify?

We are saying a whole lot of words (a couple of pages full) that probably aren't worth reading much less writing, and based on various assumptions that stand on some very thin ice. We should probably wait for TS to take some measurements, provide some more insight, before we say more.

 

If gen is connected to mains,

The generator does not connect to mains. It is mains when utility is absent and the transfer relay connects the generator to the load.

 

The reality was discussed in posts #17 , #18, #38, and #41. The reason for checking is to verify that the generator is ready, willing, and able to supply power when the mains fail. Many of the comments presume that if the engine is running that all is well. I have seen where that is not valid, because a bug gut stuck to a slip-ring, in a much older generator. Engine spinning and no power at all.
So the only way to verify it can, short of having it take over and replace the mains, is to verify the voltage at the connection to the change over contacts.
AND, by the way, the relay in that very early photo was not a change over relay because it had no normally closed contacts. A REAL CHANGE-OVER device would have the mains contacts normally connected, and transfer to the generator contacts when it was able to deliver power.

 

The reality was discussed in posts #17 , #18, #38, and #41. The reason for checking is to verify that the generator is ready, willing, and able to supply power when the mains fail. Many of the comments presume that if the engine is running that all is well. I have seen where that is not valid, because a bug gut stuck to a slip-ring, in a much older generator. Engine spinning and no power at all.
So the only way to verify it can, short of having it take over and replace the mains, is to verify the voltage at the connection to the change over contacts.
AND, by the way, the relay in that very early photo was not a change over relay because it had no normally closed contacts. A REAL CHANGE-OVER device would have the mains contacts normally connected, and transfer to the generator contacts when it was able to deliver power.

It is also possible for voltage to read correct on no-load, but then bad low voltage when some load is applied. The best test is a test while device is supplying some % of rated power output.

Can I say it? Sure I can. Next time buy a Kohler home gen with remote monitoring.

 

The reality was discussed in posts #17 , #18, #38, and #41. The reason for checking is to verify that the generator is ready, willing, and able to supply power when the mains fail.

There goes that same assumption again, and thanks for cross-referencing it back to previous iterations.

Consider that the purpose of the test is to verify that the generator IS ABLE TO SUPPLY POWER, not just able to start. So every bit of the system needs to be verified as functional.
In a good quality well engineered system, the check would verify that the generator was delivering the specified voltage, not simply spinning. Thus verifying the output of the generator would be a requirement.

That is the assumption I was referring to here:

It keeps being referred to as a "test" so imaginations are filling in gaps about what is being tested, and drawing all kinds of conclusions based on those assumptions. This does not sound to me like a "test" and there is no pass/fail; I think it is just automated routine maintenance. The same thing you do when you have a spare car that doesn't get ran often; you need to start it up every few weeks and let it get up to operating temp for a while

What makes you think this is any kind of test at all when it doesn't even get past half-speed? Why would you expect it to be ABLE TO SUPPLY POWER at half speed? Why would you expect any output voltage at all? If this were any kind of test then there would already be some kind of output for pass/fail; a light, an alarm, a relay contact, something, and TS would not be looking for a way to determine if the engine is running.

 

What makes you think this is any kind of test at all when it doesn't even get past half-speed? Why would you expect it to be ABLE TO SUPPLY POWER at half speed? Why would you expect any output voltage at all? If this were any kind of test then there would already be some kind of output for pass/fail; a light, an alarm, a relay contact, something, and TS would not be looking for a way to determine if the engine is running.

Yeah, sounded like a "is it running or not" sensor question. If the gen has a cooling fan, a 99cent micro switch will work.

Low-tech low-cost. You can use two micro switches (with thin paddle) to sense a low airspeed and a higher airspeed, simply by making paddles the right size, smaller paddle for the high airspeed detection. I assume low rpm's has lower cooling vs std rpm's run condition.

 

The generator does not connect to mains. It is mains when utility is absent and the transfer relay connects the generator to the load.

So the relay is automatic ?
Wonder if there is a free switch contact . The contactors I use have these signal contacts

 

Consider, if you would, that this IS A GENERAC system, not a CUMMINS system. ( No additional comment so as to avoid censure.)

 

So the relay is automatic ?
Wonder if there is a free switch contact ...

Yes, it is automatic. Nobody runs to switch anything manually to swap mains sources when utility fails nor returns. When the circuitry senses utility absence, starts the generator engine on. If after a preset number of seconds is still absent, the relay changes the load to generator instead of utility.
Spare contact ? As the black microswitch attached at the back in picture at post 7. It senses the relay has changed state. Same as sensing if the coil is energized, as in post 7.
There can be dozens of ways and dozens of devices attachable and useable for the original poster intentions.

 

Yeah, sounded like a "is it running or not" sensor question. If the gen has a cooling fan, a 99cent micro switch will work.

Low-tech low-cost. You can use two micro switches (with thin paddle) to sense a low airspeed and a higher airspeed, simply by making paddles the right size, smaller paddle for the high airspeed detection. I assume low rpm's has lower cooling vs std rpm's run condition.

Yeah probably the best option so far.

 

Once again, there is a big difference between a generator running and that generator being able to deliver the specified power. Sort of like a car: Just because the engine can start does not mean it is driveble. It might have flat tires or a broken transmission, or some other serious fault. Yes, the engine starts and runs very well, BUT......

 

Once again, there is a big difference between a generator running and that generator being able to deliver the specified power. Sort of like a car: Just because the engine can start does not mean it is driveble. It might have flat tires or a broken transmission, or some other serious fault. Yes, the engine starts and runs very well, BUT......

Load test seems to be outside scope of the ask.
A true load test inherits the "running or not" question.
I not even sure if that gen senses load side. No load (xfer switch did not do cutover) then the gen probably just idles at lower rpm's (maybe, depends on gen type). The gen might also go full rpm's if it knows util lost power but the xfer switch failed to cutover.

I guess the periodic test is their normal "can we start" test. The gen side assumes the xfer switch gear is good. There's several fail modes to be had, like the gen starts because of lost util power, the xfer switch does it's thing, but gen just stays idling at low rpm's.

I just just all that familiar with the Generac gen systems.

Sensing running or not, possibly even at low speed vs high speed, is easy.

 

A simpler solution would be to use a resistor divider network to reduce the ac generator voltage, and rectify and smooth to dc (with a reservoir capacitor) – then select a suitable resistance value for this dc voltage to supply the current to the diode of an opto isolator (with the generator running at both full and half speed). The transistor output of the opto isolator can then be used to provide the required signal (or operate a relay coil).

I thought about this. What size capacitor would I need? I could feed this directly into a relay that has the opto isolators - I have some on hand.

 

A still simpler way will be to put a 120 volt relay across the 240 volt output. At half voltage that should be about 120 volts. The extra benefit is that it would be verifying both halves of the output. Then, to simplify protection for that relay when it puts out full voltage, have a PTC device in series set to reduce the relay current if it rises beyond normal. Just like regular power, the two sides are outof phase.

I didn't think about that. I only found one relay that had the wide AC control voltage. But they cost $85 each so my solution is cheaper.

 

Sensing the transfer switch coil instead of generation tells it is not a weekly exercising event but a real happening. No complications, no circuitry, no equipment. However your transfer relay looks like, as this or other:

View attachment 304489

I just want to know if the generator is running, either in exercise or due to an outage. I have a UPS/Raspberry plus power monitors on some 3 phase motors that will notify me of any real outage.

 

I think I would use a current transformer to galvanically isolate the sensing part, then use a simple solid state solution for the signalling.

Thought of this, but there's no current when they are exercising. It doesn't transfer the load, just runs with an open circuit.

 

The problem can be solved much more easy. Use the ETI relay type CRM-91H/UNI. Operation voltage is 12 to 240 VAC/DC. Set delay-on function , it's marked as "a" and a minimum time.

The cheapest I could find was $38 - my solution is cheaper. Yours is simpler.

 

I thought about this. What size capacitor would I need? I could feed this directly into a relay that has the opto isolators - I have some on hand.

I didn't see Hymie's post before I said the same thing in #27. Again, if the resulting signal is fully isolated and talking to a rPi, Arduino, etc, then you don't need a filter cap; the software can filter it. In fact, since a pulsing signal is way more noise-immune, no filter cap is better. The uC can detect signal presence and *frequency* before making a decision to produce an indication.

ak