Since my Motorola repeaters are connected to the power system I'm discussing, technically this is radio-related (barely).
The FM transmitter site where my repeaters are located is fed three phase power from the street to their own outside delta-wye transformer, which feeds the building with 120/208 power. A 400A generator transfer switch follows the main breaker and feeds everything in the building.
A week ago one of the three wires coming from the street broke its insulator and pulled loose. This meant only two phases of the delta circuit were being supplied with power. The standby generator started, but the transfer of power never occured. LEDs on the transfer switch indicated good generator power but bad commercial power. The station engineers had to manually (mechanically) select the generator to get things working again.
The transfer switch is supposed to switch to the generator if the incoming power is 20% low or more than +/- 2 Hz in error. Obviously it thought enough was bad to start the generator - it just never switched to it. Also, it was running when the engineers arrived about an hour after the transmitter went off the air - normally it would have shut off if not in use.
Everything works properly if the street power fails completely, i.e. all three phases go to zero. But when just one phase dies, apparently strange things happen.
The building lights, remote control sytem, and electric gate all were operating on the "bad" power. The transmitter had shut off - it has a very sophisticated phase monitor in it. The air conditioning compressor eventually blew a circuit breaker and smelled bad, but other than that, everything recovered on generator (and commercial power, when that was restored). Neither of the FM transmitters would power up on the street power.
Measurements on the incoming street power were as follows: 120VAC from each phase to neutral, 120, 135, and 195VAC phase-to-phase. I suppose the incoming power from the street was just two phases at 120 degrees apart. Normally, the building power is 208VAC phase-to-phase.
When one leg of the delta transformer goes away, the two windings that connect to that point would look like a center-tapped transformer winding, operating against the one remaining powered winding. The fact that there was still good phase-to-neutral power probably kept a lot of things running. My repeater, which is on an APC SmartUPS 1000, never even burped !
It would be interesting to have the power company pull one phase and let us go measure the resulting building power to see what we're dealing with. Unfortunately, this would mean putting the station off the air for a while, plus the power company would probably want to get paid to run the experiment.
Question: what would the building power look like, phase-wise? Any electric power engineers out there, or suggested web sites where I could ask such a question?
Thanks.
Bob M.
3-phase power failure
Moderator: Queue Moderator
Street power feeds the delta side of the transformer. 120/208V 3 phase with neutral (wye) comes out of the transformer and feeds the main breaker. The power then goes into the three phase transfer switch and that feeds power to all the other distribution panels. The air conditioners and FM transmitters are the only stuff that uses 3 phase power - everything else just picks up one of the three phases and is normal 120V power.
The generator is an 80kw, 6-cylinder propane unit. It's output is 3 phase, 120/208V, that also feeds, and is controlled by, the 400A transfer switch. The building can't tell the difference between commercial and generator power, except for the extra noise created outside.
Bob M.
The generator is an 80kw, 6-cylinder propane unit. It's output is 3 phase, 120/208V, that also feeds, and is controlled by, the 400A transfer switch. The building can't tell the difference between commercial and generator power, except for the extra noise created outside.
Bob M.
Hello,
I would check the transfer switch..... Almost all of the switches I have worked with monitor each phase, thus any loss of a phase will start the generator and transfer the load. Every so often you find an older one that does not and what we have done is install a 3 phase monitor (they make a small plug in unit that is about the size of a timing relay and not only monitors a loss but low voltage)
http://www.macromatic.com/products/phase-monitor-relay
They work great and you can have it do a force start of the generator.
Gary
I would check the transfer switch..... Almost all of the switches I have worked with monitor each phase, thus any loss of a phase will start the generator and transfer the load. Every so often you find an older one that does not and what we have done is install a 3 phase monitor (they make a small plug in unit that is about the size of a timing relay and not only monitors a loss but low voltage)
http://www.macromatic.com/products/phase-monitor-relay
They work great and you can have it do a force start of the generator.
Gary
I'm sure the transfer switch is the culprit. The problem is: we lost one leg of the DELTA which feeds the step-down transformer. The output is three phase WYE configuration, and all of those legs still had power: 120 to neutral and anything from 120 to 190 phase-to-phase. So it didn't sense any loss of power on a single phase because there really wasn't one. Sure, the phase relationship probably got all screwed up, as did the voltage, but the transformer did its best to keep the WYE side running.
I agree that a more sophisticated phase monitor might have caught it. The difficulty will be integrating it into the existing transfer switch control circuitry, which probably is not designed for any external sensor. The generator WAS started, and was still running an hour later; the switch just never transferred to it even though it indicated NO commercial power.
Bob M.
I agree that a more sophisticated phase monitor might have caught it. The difficulty will be integrating it into the existing transfer switch control circuitry, which probably is not designed for any external sensor. The generator WAS started, and was still running an hour later; the switch just never transferred to it even though it indicated NO commercial power.
Bob M.
3 phase power sense.
Hello.
Most 3 phase sensors have something that actually follows the phase loop.
I have never found a unit that could not take an upgrade, just a matter of looking at the schematics.
ALL of the sites we run that use 3 phase power are set up with power supplies that can deal with only 2 phase power.
This may mean a 50% reduction in power, but everything will still work.
We can run single phase at 30% power.
And, the experiment is not hard to do, just split the breaker handle and trip only one breaker.
Most 3 phase sensors have something that actually follows the phase loop.
I have never found a unit that could not take an upgrade, just a matter of looking at the schematics.
ALL of the sites we run that use 3 phase power are set up with power supplies that can deal with only 2 phase power.
This may mean a 50% reduction in power, but everything will still work.
We can run single phase at 30% power.
And, the experiment is not hard to do, just split the breaker handle and trip only one breaker.
Re: 3 phase power sense.
Let me ask a loaded question here. Does the generator system have an exerciser built into it? The second question is does the load transfer to the generator when it runs on the exerciser?
I use to get into heated debates with a cellular system engineer that just refused to allow the emergency generators to pick up the load when they were automaticaly exercised each week. He thought it would cause a glitch and cause the cellular site or the switch office to go down each time it ran and transfered.
The problems that not running the generator under load cause are sometimes hidden. The more obvious is that if it is a diesel engine, you will load up or wet stack the engine over time and it won't be able to pull the required load when needed in some cases.
The big hidden problem is the transfer switch tends to freeze in the normal position. The actuator coil can and will rust up. The actuator may jam or rust in the normal position. Even the contacts could get welded and not release.
If you have a generator, run it and put the load on it everytime. If you have critical equipment on the load, put it on a good UPS to carry it over the outage until the generator picks up the load.
One other point to make here on the voltages you measured. Many if not most of the 120 / 208 systems that people lable 3 phase have a high voltage on one leg to neutral. This is called the stinger leg. Generaly you will get someplace around 190 volts from the hot side to neutral. If you look closely, none of the normal 120 volt circuits are connected to this leg. Only the 3 phase loads use it. If you look in the electrical pannel, you will probably find orange tape wrapped around this phase leg wire.
Last comment on what happened at this site. If the generator started, the transfer switch control circuits did their thing. They sensed the phase failure. You just need to look and see why the switch didn't transfer. I would be willing to bet the generator system is set so it doesn't pick up the load when it exercises.
Jim
I use to get into heated debates with a cellular system engineer that just refused to allow the emergency generators to pick up the load when they were automaticaly exercised each week. He thought it would cause a glitch and cause the cellular site or the switch office to go down each time it ran and transfered.
The problems that not running the generator under load cause are sometimes hidden. The more obvious is that if it is a diesel engine, you will load up or wet stack the engine over time and it won't be able to pull the required load when needed in some cases.
The big hidden problem is the transfer switch tends to freeze in the normal position. The actuator coil can and will rust up. The actuator may jam or rust in the normal position. Even the contacts could get welded and not release.
If you have a generator, run it and put the load on it everytime. If you have critical equipment on the load, put it on a good UPS to carry it over the outage until the generator picks up the load.
One other point to make here on the voltages you measured. Many if not most of the 120 / 208 systems that people lable 3 phase have a high voltage on one leg to neutral. This is called the stinger leg. Generaly you will get someplace around 190 volts from the hot side to neutral. If you look closely, none of the normal 120 volt circuits are connected to this leg. Only the 3 phase loads use it. If you look in the electrical pannel, you will probably find orange tape wrapped around this phase leg wire.
Last comment on what happened at this site. If the generator started, the transfer switch control circuits did their thing. They sensed the phase failure. You just need to look and see why the switch didn't transfer. I would be willing to bet the generator system is set so it doesn't pick up the load when it exercises.
Jim
My two bits worth:
So, what does your generator tech have to say about it?
They DO have some one service it don't they??
I have noticed that the generator panel seems to be the first thing that gets zapped in a lightning strike.
It may be that it won't transfer during exercise on purpose, but then loosing one phase of the power line is usuially the failure mode.
I would then presume that the logic in the transfer panel would then allow transfer during a partial failure.
Perhaps there is a failure in the transfer switch control circuits, or they were never properly set up in the first place.
Running a 3 phase motor such as in an air con on single phase will eventually burn out the stator.
Such motors should have low voltage/lost phase protection. Such protectors are much cheaper than a compressor change.
Most critical systems have back up batteries, or a UPS.
There should be no service interruption when the generator transfers.
If there is, then the cause of the "glitch" should be found and remedied.
In other words Jim202 has it right.
But then, some of us do not make our living behind a monitor in the office, we go out in the field and fix these things.
I would suppose that a broad cast station might have a problem during transfer, isn't it better to know that it will actually work on that dark and stormy night when you would be risking life and limb to go to the site even if it were possible?
So, what does your generator tech have to say about it?
They DO have some one service it don't they??
I have noticed that the generator panel seems to be the first thing that gets zapped in a lightning strike.
It may be that it won't transfer during exercise on purpose, but then loosing one phase of the power line is usuially the failure mode.
I would then presume that the logic in the transfer panel would then allow transfer during a partial failure.
Perhaps there is a failure in the transfer switch control circuits, or they were never properly set up in the first place.
Running a 3 phase motor such as in an air con on single phase will eventually burn out the stator.
Such motors should have low voltage/lost phase protection. Such protectors are much cheaper than a compressor change.
Most critical systems have back up batteries, or a UPS.
There should be no service interruption when the generator transfers.
If there is, then the cause of the "glitch" should be found and remedied.
In other words Jim202 has it right.
But then, some of us do not make our living behind a monitor in the office, we go out in the field and fix these things.
I would suppose that a broad cast station might have a problem during transfer, isn't it better to know that it will actually work on that dark and stormy night when you would be risking life and limb to go to the site even if it were possible?
Aloha, Bernie
Great comments and suggestions. Thanks.
The generator does exercise itself once a week, and I believe it also picks up the load - the equipment sees a small glitch but either rides through it or turns off and back on automatically. The critical 120V stuff is on a UPS. The 20kw FM transmitters are not.
The generator and transfer switch was tested manually just three days before this situation. Now by "manually" I mean someone went up there and pressed the "test" button and observed everything happen. So under the condition simulated by the test button, the switch did transfer. This is NOT the same as pulling the main breaker, however doing that would cause everything to dump and the station would take 45 seconds to come back on after power was restored. The CE and I might meet up at the site late some evening and he's going to pull the main breaker and make sure things really work the way they should.
I don't believe that the weird high leg is necessary with three-phase WYE service in the building, as each leg is giving us 120V to neutral and the phase-to-phase is normally 208. I have seen DELTA feeds - one of the AM stations I work on has it - where one of the secondaries is center-tapped and provides a neutral and two 120V circuits, which would mean some weird voltage (190) from the opposite delta tap to this neutral, while still providing 208V phase-to-phase. But the FM site is pure three-phase WYE power with each leg identical, plus a neutral wire.
By the way, this same AM site has a very old Onan 4kw propane generator, 120/240V two-phase, that is wired to only switch the 120/240V circuits, which are fed from their own panel. The only unit that uses three-phase power is the A/C compressor, and if the power fails, that unit will just be off until power is restored. I suppose they could have a similar failure of one leg of the delta feed too.
Right now a schematic would be nearly impossible to obtain. I understand that a new logic board for the transfer switch would cost something like $1200. I have to wonder why any company would design their own phase and voltage monitoring circuits inside the unit when so many really good ones already exist and just plug-in.
It certainly is possible that one of the transfer solenoids has failed or locked up. The engineers that visited the site had to turn the mechanism by inserting a screwdriver into the hole provided for that purpose. I don't know if they were present when power was restored, but the switch apparently reverted back to commercial when this happened.
My GE transfer switch at my house uses one solenoid to toggle the transfer switch back and forth, and relies on microswitches and available voltages to decide whether it should activate the solenoid or not. So if mine fails, it won't switch either way.
Bob M.
The generator does exercise itself once a week, and I believe it also picks up the load - the equipment sees a small glitch but either rides through it or turns off and back on automatically. The critical 120V stuff is on a UPS. The 20kw FM transmitters are not.
The generator and transfer switch was tested manually just three days before this situation. Now by "manually" I mean someone went up there and pressed the "test" button and observed everything happen. So under the condition simulated by the test button, the switch did transfer. This is NOT the same as pulling the main breaker, however doing that would cause everything to dump and the station would take 45 seconds to come back on after power was restored. The CE and I might meet up at the site late some evening and he's going to pull the main breaker and make sure things really work the way they should.
I don't believe that the weird high leg is necessary with three-phase WYE service in the building, as each leg is giving us 120V to neutral and the phase-to-phase is normally 208. I have seen DELTA feeds - one of the AM stations I work on has it - where one of the secondaries is center-tapped and provides a neutral and two 120V circuits, which would mean some weird voltage (190) from the opposite delta tap to this neutral, while still providing 208V phase-to-phase. But the FM site is pure three-phase WYE power with each leg identical, plus a neutral wire.
By the way, this same AM site has a very old Onan 4kw propane generator, 120/240V two-phase, that is wired to only switch the 120/240V circuits, which are fed from their own panel. The only unit that uses three-phase power is the A/C compressor, and if the power fails, that unit will just be off until power is restored. I suppose they could have a similar failure of one leg of the delta feed too.
Right now a schematic would be nearly impossible to obtain. I understand that a new logic board for the transfer switch would cost something like $1200. I have to wonder why any company would design their own phase and voltage monitoring circuits inside the unit when so many really good ones already exist and just plug-in.
It certainly is possible that one of the transfer solenoids has failed or locked up. The engineers that visited the site had to turn the mechanism by inserting a screwdriver into the hole provided for that purpose. I don't know if they were present when power was restored, but the switch apparently reverted back to commercial when this happened.
My GE transfer switch at my house uses one solenoid to toggle the transfer switch back and forth, and relies on microswitches and available voltages to decide whether it should activate the solenoid or not. So if mine fails, it won't switch either way.
Bob M.
Many if not most of the 120 / 208 systems that people lable 3 phase have a high voltage on one leg to neutral.
A 120/208 WYE service is 120 v to ground from all three phases. There is no "wild leg" like that seen in a typical 120/240 delta service where there is 208v from one phase to ground and 120v on the two others.
The transfer switch probably doesn't reference the voltage across all three phases (a-b, b-c, c-a), but instead looks at each phase to neutral(a- N, b-N, c-N). Some older switches use two separate sensing circuits, one for gen-set start/run and one for transfer to allow the engine to run a moment or two before load comes on.
I had the same thing happen at one of our facilities, the set started but never transfered. The generator tech told us that the "start/run" circuit was wired to monitor phase to phase, but the "normal power" was referenced to a single phase to ground.
The manufacturer of the switch claims they are monitoring phase-to-phase voltage, but who knows without a schematic. It's a lot harder to do that without some form of isolation to the sensing circuits.
This switch was installed new 4 years ago, but I don't know the brand nor how long it's been made. I'll find out more in a couple of days.
The idea of two sensing circuits sounds reasonable, but what good is it to start the generator and leave it running, without ever switching power to it, especially if the commercial power is NOT OK. Seems to me that there is a design flaw in the system.
MASTR: Glad we're not alone. Did your transfer switch get repaired or replaced so that wouldn't happen again, or is there very little you can do about it short of redesigning it to monitor all possible voltages and phases?
Bob M.
This switch was installed new 4 years ago, but I don't know the brand nor how long it's been made. I'll find out more in a couple of days.
The idea of two sensing circuits sounds reasonable, but what good is it to start the generator and leave it running, without ever switching power to it, especially if the commercial power is NOT OK. Seems to me that there is a design flaw in the system.
MASTR: Glad we're not alone. Did your transfer switch get repaired or replaced so that wouldn't happen again, or is there very little you can do about it short of redesigning it to monitor all possible voltages and phases?
Bob M.
-
psapengineer
- Posts: 175
- Joined: Thu Oct 07, 2004 10:00 am
Wild Leg Open Delta
I'm thinking that a "wild leg open delta" on the primary side of the utility transformer, for a lighty loaded secondary (the primary FM Station kicked off, right?), would indeed transform power to all three phases on the secondary Y side. Lightly loaded is the key.
Sounds like the transfer switch logic did sense the failure as it did start the engine. It either then hung on the "Transfer to Emergency" delay or the transfer switch stuck.
If this is an Asco design type switch it uses 1 coil, a bridge rectifier and 2 micro switches to toggle the switch between normal and emergency. We had 1 that would get stuck returning from emergency, turned out to be a bad limit switch
It's important to exercize this type of switch every week as any gumming of the lube or anything else that slows down the movement of the coil plunger will cause a failure to complete the transfer.
It's hard to trouble shoot what happened here as you need to know what the position of the switch was before it was moved manualy. If it moved at all I would say the logic controls did their job and the problem is in the switch.
The logic energizes the bridge rectifier thru 1 of the limit switches(depending on the direction of movement) and the DC output of the bridge powers the coil. The coil pulls the switch from 1 position to the other. When the switch reaches the end of travel it contacts the limit switch and removes the power from the coil so it does not burn up.
The limit switch contacts have been known to oxidize and not pass the power, the limit plungers sometimes stick also. A weak rectifier will sometimes leave the cycle mid way and usuly results in a smoked coil and rectifier as it is still powered because the limit switch never opens.
You need to cycle the switch a few times and see if you can get it to hang up again and trouble shoot from there.
If this is an Asco design type switch it uses 1 coil, a bridge rectifier and 2 micro switches to toggle the switch between normal and emergency. We had 1 that would get stuck returning from emergency, turned out to be a bad limit switch
It's important to exercize this type of switch every week as any gumming of the lube or anything else that slows down the movement of the coil plunger will cause a failure to complete the transfer.
It's hard to trouble shoot what happened here as you need to know what the position of the switch was before it was moved manualy. If it moved at all I would say the logic controls did their job and the problem is in the switch.
The logic energizes the bridge rectifier thru 1 of the limit switches(depending on the direction of movement) and the DC output of the bridge powers the coil. The coil pulls the switch from 1 position to the other. When the switch reaches the end of travel it contacts the limit switch and removes the power from the coil so it does not burn up.
The limit switch contacts have been known to oxidize and not pass the power, the limit plungers sometimes stick also. A weak rectifier will sometimes leave the cycle mid way and usuly results in a smoked coil and rectifier as it is still powered because the limit switch never opens.
You need to cycle the switch a few times and see if you can get it to hang up again and trouble shoot from there.
Cause Motorola said so that's why
"I'm thinking that a "wild leg open delta" on the primary side of the utility transformer, for a lighty loaded secondary (the primary FM Station kicked off, right?), would indeed transform power to all three phases on the secondary Y side. Lightly loaded is the key."
Yes, the xmtr dumped, probably almost immediately, as it DOES have a sensor in it for phase/voltage. The only "big" load was the 3-phase A/C compressor which eventually popped a breaker after overheating. The remaining load was most likely under 10 amps per phase. As the outage happened during the day, the tower lights weren't even on.
This is exactly the kind of problem that'll be hard to reproduce since it involves dropping one leg of the delta feed (primary) of the transformer. There are no switches or fuses we can pull at the transformer, and I don't think we'd even want to. I don't know what voltage the lines carry, but I would guess it's more than 480. I'll have to check the ratings on it when I'm up there.
I wish there was an easy way to set up some transformers (i.e. audio), and generate three signals 120 degrees apart to feed a delta-wired primary, then measure the wye-connected secondary to see what really comes out (phase and voltage). Then we could go back to the switch manufacturer and suggest they come up with a better design.
The transfer switch apparently uses TWO solenoids to swing it each way. My GE unit at home has just one, and each time it actuates, the switch toggles. It only does so if necessary - microswitches sense the position.
The easiest way to exercise the switch would be to turn all the loads off, then push the appropriate buttons to flip it each way several times with the generator running. This way, it won't shock anything too badly by momentarily interrupting power. But I don't think the radio station will go along with that kind of outage, even for testing. The best we can do is pull the main breaker late at night and see what happens.
I have electric door locks in my truck. If I press the lock button quickly, the key-in-ignition chime will sound, but the solenoids won't lock the doors. I have to hold the lock button down for about 1/4 second to have it actuate the solenoids. Why couldn't they just do everything with one switch contact (they actually DO use a DPDT switch) or feed one signal to the computer and let IT lock the doors.
Bob M.
Yes, the xmtr dumped, probably almost immediately, as it DOES have a sensor in it for phase/voltage. The only "big" load was the 3-phase A/C compressor which eventually popped a breaker after overheating. The remaining load was most likely under 10 amps per phase. As the outage happened during the day, the tower lights weren't even on.
This is exactly the kind of problem that'll be hard to reproduce since it involves dropping one leg of the delta feed (primary) of the transformer. There are no switches or fuses we can pull at the transformer, and I don't think we'd even want to. I don't know what voltage the lines carry, but I would guess it's more than 480. I'll have to check the ratings on it when I'm up there.
I wish there was an easy way to set up some transformers (i.e. audio), and generate three signals 120 degrees apart to feed a delta-wired primary, then measure the wye-connected secondary to see what really comes out (phase and voltage). Then we could go back to the switch manufacturer and suggest they come up with a better design.
The transfer switch apparently uses TWO solenoids to swing it each way. My GE unit at home has just one, and each time it actuates, the switch toggles. It only does so if necessary - microswitches sense the position.
The easiest way to exercise the switch would be to turn all the loads off, then push the appropriate buttons to flip it each way several times with the generator running. This way, it won't shock anything too badly by momentarily interrupting power. But I don't think the radio station will go along with that kind of outage, even for testing. The best we can do is pull the main breaker late at night and see what happens.
I have electric door locks in my truck. If I press the lock button quickly, the key-in-ignition chime will sound, but the solenoids won't lock the doors. I have to hold the lock button down for about 1/4 second to have it actuate the solenoids. Why couldn't they just do everything with one switch contact (they actually DO use a DPDT switch) or feed one signal to the computer and let IT lock the doors.
Bob M.
It's going to be a tough one, It's hard to explain that you need to cause a power failure to trouble shoot the thing thats only purpose in life is to protect you from a power failure !!
If you are woried that the logic is not protecting you properly from a failure where the failed phase does not drop to 0 then you should look into the phase monitor relays that someone else sugjested previously. They are even avaiable from Grainger and they not only monitor voltage but also frequency and phase rotation which I assume that your phase angle problem would triger.
When you get the transfer switch schematic look for an option called "Area Protection" It's intended purpose is to put the load on generator when a failure is expected or to do a utility load shed turing peak utility load periods. It most likely will have a jumper installed on it's terminals, removing the jumper will cause a generator start then transfer. It is a perfect place to hook up an external phase monitor relay. I think the Area Protection is sometimes called Accessory or Option 27 by some manfacturers.
If you are woried that the logic is not protecting you properly from a failure where the failed phase does not drop to 0 then you should look into the phase monitor relays that someone else sugjested previously. They are even avaiable from Grainger and they not only monitor voltage but also frequency and phase rotation which I assume that your phase angle problem would triger.
When you get the transfer switch schematic look for an option called "Area Protection" It's intended purpose is to put the load on generator when a failure is expected or to do a utility load shed turing peak utility load periods. It most likely will have a jumper installed on it's terminals, removing the jumper will cause a generator start then transfer. It is a perfect place to hook up an external phase monitor relay. I think the Area Protection is sometimes called Accessory or Option 27 by some manfacturers.
Cause Motorola said so that's why
MASTR: Glad we're not alone. Did your transfer switch get repaired or replaced so that wouldn't happen again, or is there very little you can do about it short of redesigning it to monitor all possible voltages and phases?
The switch in our case was an Onan branded unit, I don't know who the OEM really is. I was told that the repair (really a re-fit, as the switch was apparently not built with sensing on all 3 phases), was done in an hour or so. I've no idea about the cost, as it wasn't an issue as far as I was concerned, other administrative people deal with that aspect
Remember that the DELTA side of the power company transformer is what lost one phase. As far as I can tell, all three phases on the WYE side (the side that feeds the transfer switch and the building) were still alive and had 120VAC to neutral. It was only the phase-to-phase voltage that suffered, along with, probably, some phase angle.
If the unit has provisions for an external device to trigger the switchover, that would be something to consider. If the switch manufacturer had done a proper job, this should have never happened. One sensing circuit is all that it should need to start the sequence of events rolling, but it should sense everything possible, not just the voltage on any or all phases.
I'll report back with anything more specific, especially if that test happens later tonight.
Bob M.
If the unit has provisions for an external device to trigger the switchover, that would be something to consider. If the switch manufacturer had done a proper job, this should have never happened. One sensing circuit is all that it should need to start the sequence of events rolling, but it should sense everything possible, not just the voltage on any or all phases.
I'll report back with anything more specific, especially if that test happens later tonight.
Bob M.
The testing never happened Saturday, but should happen tonight (Monday). I was up at the site today and looked at the power company transformer. The only info I found on it said 120/208V, and some white chalk markings that showed Y-Y, which would imply that the primary is 3 phase WYE service, which I think means there should be a neutral in the cable. I didn't check the incoming lines on the pole to see how many there are (3 would imply DELTA and 4 would imply WYE). The transformer is very securely locked shut, not that I'd want to open it while live anyway. I'm assuming that it's connected directly to the three highest wires on the utility poles, any one of which would feed a standard gray pole-mounted transformer to provide 120/240V single-phase residential service. I've always thought that the primary voltage is around 2200V, but I don't know for sure. I'm supposed to hook up with a power company EE over the next few days, and maybe he can enlighten me (no pun intended).
Bob M.
Bob M.
Update. I met the CE up at the site tonight. The transfer switch IS an ASCO brand. The logic board inside has a microprocessor and three small transformers on it, along with several DIP switches. There's a separate exerciser timer, an RS-232 interface module, and of course the front panel switch/LED assembly. Nothing inside to adjust except perhaps the wait times for generator start/available and turnoff after commercial power resumes.
I pulled the main 400A breaker on the output of the transformer. Everything got nice and dark and quiet. 6 seconds later the generator started cranking. It probably did so for another 5 seconds then finally started. This seemed rather long for a propane unit, but what do I know. The transfer switch changed over to the generator very non-ceremoniously, just like it should. We ran it this way for probably 10 minutes, after which I turned the main breaker back on. Several minutes later the transfer switch flipped back to commercial power, and we manually shut off the generator, rather than wait the 15-30 minutes it probably would have taken. In other words, it worked perfectly.
I can only assume that the three small transformers on the logic board would be used to measure phase-to-phase voltage, as they could easily measure phase-to-neutral without a transformer. If, however, they are using the transformers to measure phase-to-neutral, then they would have no way to measure phase-to-phase. A schematic would be quite helpful.
If they're just monitoring phase-to-neutral, then a lightly loaded secondary would likely have equal voltages even if the primary was fed with a single phase. I'll know more after talking to a power company engineer.
Bob M.
I pulled the main 400A breaker on the output of the transformer. Everything got nice and dark and quiet. 6 seconds later the generator started cranking. It probably did so for another 5 seconds then finally started. This seemed rather long for a propane unit, but what do I know. The transfer switch changed over to the generator very non-ceremoniously, just like it should. We ran it this way for probably 10 minutes, after which I turned the main breaker back on. Several minutes later the transfer switch flipped back to commercial power, and we manually shut off the generator, rather than wait the 15-30 minutes it probably would have taken. In other words, it worked perfectly.
I can only assume that the three small transformers on the logic board would be used to measure phase-to-phase voltage, as they could easily measure phase-to-neutral without a transformer. If, however, they are using the transformers to measure phase-to-neutral, then they would have no way to measure phase-to-phase. A schematic would be quite helpful.
If they're just monitoring phase-to-neutral, then a lightly loaded secondary would likely have equal voltages even if the primary was fed with a single phase. I'll know more after talking to a power company engineer.
Bob M.
Hello,
Most switches I've seen with a rs232 port are programmable and have a bunch of settings that can be adjusted. You may want to check withsome who can read and check the settings. I know that when the switches are delivered they are not set up in most cases this allows the electrician to install the unit and not have it operational till the system is complete. As a matter of fact we are in the process of installing a 400 amp transfer swich and new generator in my firehouse and the switch is in and wired but can not transfer and will only supply utility power till the outside work is done and they come and program it.
Gary
Most switches I've seen with a rs232 port are programmable and have a bunch of settings that can be adjusted. You may want to check withsome who can read and check the settings. I know that when the switches are delivered they are not set up in most cases this allows the electrician to install the unit and not have it operational till the system is complete. As a matter of fact we are in the process of installing a 400 amp transfer swich and new generator in my firehouse and the switch is in and wired but can not transfer and will only supply utility power till the outside work is done and they come and program it.
Gary
We had a similar problem a while back. Transformer feeding our building failed but we lost only one leg of the 3 phase feed. The generator didn't start. Had to manually drop the supply feeding the generator for it to start. We have an almost new (4 years) generator and transfer switch. I too would like stress exercising the generator under load. Ours does, each week, but the old one at our courthouse did not. Only ran to exercise the engine. After a couple of years, the first time the power went out the generator caught fire.
The weekly test at this facility DOES switch the load, so everything really is exercised to the fullest. This particular failure probably should have caused some ganged breaker to pop, or maybe another phase fuse should have given up, but no such luck.
I think the only people who would know how to talk to the unit with a computer would be factory personnel or their service technicians. Too much could go wrong if the settings weren't properly adjusted. The liability issue alone would take years to sort out.
Bob M.
I think the only people who would know how to talk to the unit with a computer would be factory personnel or their service technicians. Too much could go wrong if the settings weren't properly adjusted. The liability issue alone would take years to sort out.
Bob M.
I do know that Asco System 7 does monitor Phase to Phase as there is no neutral connection to the logic controls that I can find. The system 7 does have dip switches to adjust pick up and drop out voltage for both normal and emergency source.
There is an option for an rs-232 monitor option but all setup is done via the dip swirches. The switch I am refering to was manfacturered in 1993 and I think this line continued till around 2000, how old is yours.
The manuals I have only show the installation and setup of the switch and accessory connections it shows nothing about what happens inside the logic controls.
What model switch and control is this?
I am still not convinced that the logic did not sense the power failure as it did start the engine. Kind of puts it back to square 1, sorry
There is an option for an rs-232 monitor option but all setup is done via the dip swirches. The switch I am refering to was manfacturered in 1993 and I think this line continued till around 2000, how old is yours.
The manuals I have only show the installation and setup of the switch and accessory connections it shows nothing about what happens inside the logic controls.
What model switch and control is this?
I am still not convinced that the logic did not sense the power failure as it did start the engine. Kind of puts it back to square 1, sorry
Cause Motorola said so that's why
The facility is about 4 years old, so the switch was made in 2000-2001.
Naturally I failed to remember to write down the model number late last night. I usually go to bed at 2200 and I was at the site until after midnight. I will be back up and will write down the model number.
I know there's a neutral bus inside the cabinet, but I don't recall if any internal wiring went to it. I did see some microswitches on the transfer mechanism. All of the electronics is on the inside of the hinged front panel.
I'm not at all surprised that no one has a schematic of the logic board. These things are usually proprietary and also sample relatively dangerous voltages, so they really don't want anyone fooling around with it. There are two board-edge connectors, one labeled "high voltage" which probably picks up the sensing voltages,and the other labeled "low voltage" which probably deals with the microswitches and actuator(s).
I would think that ASCO has the ability to replicate our failure and could either modify the firmware or add circuitry to the sampling network to detect the strange phase angles and voltages we probably experienced.
Bob M.
Naturally I failed to remember to write down the model number late last night. I usually go to bed at 2200 and I was at the site until after midnight. I will be back up and will write down the model number.
I know there's a neutral bus inside the cabinet, but I don't recall if any internal wiring went to it. I did see some microswitches on the transfer mechanism. All of the electronics is on the inside of the hinged front panel.
I'm not at all surprised that no one has a schematic of the logic board. These things are usually proprietary and also sample relatively dangerous voltages, so they really don't want anyone fooling around with it. There are two board-edge connectors, one labeled "high voltage" which probably picks up the sensing voltages,and the other labeled "low voltage" which probably deals with the microswitches and actuator(s).
I would think that ASCO has the ability to replicate our failure and could either modify the firmware or add circuitry to the sampling network to detect the strange phase angles and voltages we probably experienced.
Bob M.
I used to work with an older engineer who did power engineering when he got out of college; it's possible to use a delta/delta transformer with just two transformers. The catch is that they have to carry 33% more load, so they must be appropriately sized. The missing transformer will cause harmonic effects in the power system, which must be balanced by two more of these funny deltas in use elsewhere to achieve equal loading on all phases.
I have seen a weird delta connection where only two windings were used on the three wires, but I don't think that's the situation here. I'm sure there was all sorts of distortion and strangeness going on.
I'd like to find out what will come out of the WYE side (120/208) when one or two of the input phases goes away. If I had a source of relatively low voltage three phase power, or even an audio signal, I'd run it into three separate 1:1 transformers and look at the output. I know how to make square waves with a microprocessor, but making three of them 120 degrees apart at the same time would seem to be quite a bit more difficult. I guess I could turn one output high, wait a bit, turn a second one high, wait a bit, turn the first one off, wait a bit, turn the third one on, wait a bit, etc., and probably get something that looks like a three phase signal.
While there's a 400A breaker between the transformer and the building, I'm sure the current being drawn is much less than that, as the 21.5kw transmitter is only running at 11kw out (around 15kw DC in), and it only has a 150A breaker feeding it. I'd be surprised if the total building power was more than 25kw right now. The transformer seems to be big enough to handle the entire 400A load - I would guess it's 5-6 ft high and 4x6 around. A utility asset number is on it, but no ratings info at all. I'll have to go up and take a closer look, plus get a model number off the ASCO switch.
Bob M.
I'd like to find out what will come out of the WYE side (120/208) when one or two of the input phases goes away. If I had a source of relatively low voltage three phase power, or even an audio signal, I'd run it into three separate 1:1 transformers and look at the output. I know how to make square waves with a microprocessor, but making three of them 120 degrees apart at the same time would seem to be quite a bit more difficult. I guess I could turn one output high, wait a bit, turn a second one high, wait a bit, turn the first one off, wait a bit, turn the third one on, wait a bit, etc., and probably get something that looks like a three phase signal.
While there's a 400A breaker between the transformer and the building, I'm sure the current being drawn is much less than that, as the 21.5kw transmitter is only running at 11kw out (around 15kw DC in), and it only has a 150A breaker feeding it. I'd be surprised if the total building power was more than 25kw right now. The transformer seems to be big enough to handle the entire 400A load - I would guess it's 5-6 ft high and 4x6 around. A utility asset number is on it, but no ratings info at all. I'll have to go up and take a closer look, plus get a model number off the ASCO switch.
Bob M.
Voltage should be 120V on two sides and 0V on the third. cross terminal voltage would be 208-?-?... I think I see what you're saying... without the other side driven it could load the availible power out down... seems like the system is engineered to take it anyway since it would be doing it one way or the other...
The incoming power IS delta, as there are only three lines run from the street to the transformer. The output is 120/208 wye. The transformer has 75kw written all over it. The building load is probably 1/4rd of that at most, since the FM transmitter is only making 11kw.
My analysis of the delta-wye situation is getting me to try an experiment. But first, some thoughts. The three separate windings on the transformer are all on one core, so there will be some magnetic effects to the entire system if one leg of the delta feed goes away. You'll then end up with one winding being fed voltage where the lines are 120 degrees apart, instead of the usual 180 degrees, and the other pair of windings are in series and being fed with the same voltage. I have no idea what the two phase delta voltage really is. All I know is that there will be some voltage on each delta winding, so there should also be some voltage on each wye winding.
Assuming the voltage is evenly split across the two windings with the missing leg, then I would expect one phase of the wye side to be normal voltage, and the other two should be 1/2 voltage. But since everybody is on one iron core, this may not be the case. I'm assuming a very light load here, if anything at all, as in only a few kW. If a big load was maintained, then all bets are off.
I'd like to find a site where they have 480v delta feeding a delta-wye transformer with no load, where I could remove one of the 480 lines and observe the output voltage with a true RMS meter, and put a scope on the wye side of the transformer to check the phase relationship. My Fluke 199C scope can give me the phase difference and RMS voltages rather nicely.
I found a manual on the ASCO 300-series transfer switch. One section says they monitor the voltage on all three phases of line power. Another section says they monitor the line-to-line voltage. Elsewhere they say they only use or look at two phases of the emergency power, assuming that the generator is working properly if those two phases are energized, and the voltage and frequency is within range. They wouldn't make a schematic available, but CT has this very nice law on the books that says manufacturers MUST provide schematics of electronic equipment. I think this was meant to pertain to home electronics like stereos and TVs, but NEC extended it to cover a manual for an old VGA monitor once, so this might be a way to get something from ASCO.
Bob M.
My analysis of the delta-wye situation is getting me to try an experiment. But first, some thoughts. The three separate windings on the transformer are all on one core, so there will be some magnetic effects to the entire system if one leg of the delta feed goes away. You'll then end up with one winding being fed voltage where the lines are 120 degrees apart, instead of the usual 180 degrees, and the other pair of windings are in series and being fed with the same voltage. I have no idea what the two phase delta voltage really is. All I know is that there will be some voltage on each delta winding, so there should also be some voltage on each wye winding.
Assuming the voltage is evenly split across the two windings with the missing leg, then I would expect one phase of the wye side to be normal voltage, and the other two should be 1/2 voltage. But since everybody is on one iron core, this may not be the case. I'm assuming a very light load here, if anything at all, as in only a few kW. If a big load was maintained, then all bets are off.
I'd like to find a site where they have 480v delta feeding a delta-wye transformer with no load, where I could remove one of the 480 lines and observe the output voltage with a true RMS meter, and put a scope on the wye side of the transformer to check the phase relationship. My Fluke 199C scope can give me the phase difference and RMS voltages rather nicely.
I found a manual on the ASCO 300-series transfer switch. One section says they monitor the voltage on all three phases of line power. Another section says they monitor the line-to-line voltage. Elsewhere they say they only use or look at two phases of the emergency power, assuming that the generator is working properly if those two phases are energized, and the voltage and frequency is within range. They wouldn't make a schematic available, but CT has this very nice law on the books that says manufacturers MUST provide schematics of electronic equipment. I think this was meant to pertain to home electronics like stereos and TVs, but NEC extended it to cover a manual for an old VGA monitor once, so this might be a way to get something from ASCO.
Bob M.
Asco transfer switches
Bob,
Good luck on trying to get a diagram out of Asco. When I was working for Verizon Wireless, we used both the 165 and the 300 series switches. Never could get them to provide the internal electronics prints.
About the best you can do is get the transfer switch diagram and manual off the Asco web site.
We had run into some interesting problems with some time delay issues with some of the older Coleman 10 and 15 KW generators and the Asco 165 switches. Managed to replace most of them over time, but played hell in the process of doing it. Hope I never see those generators again. Briggs and Straton even had a short 90 degree breather hose that cracked with age after about 2 or 3 years. They wanted $45 for this 3 inch length of thin hose.
Jim
Good luck on trying to get a diagram out of Asco. When I was working for Verizon Wireless, we used both the 165 and the 300 series switches. Never could get them to provide the internal electronics prints.
About the best you can do is get the transfer switch diagram and manual off the Asco web site.
We had run into some interesting problems with some time delay issues with some of the older Coleman 10 and 15 KW generators and the Asco 165 switches. Managed to replace most of them over time, but played hell in the process of doing it. Hope I never see those generators again. Briggs and Straton even had a short 90 degree breather hose that cracked with age after about 2 or 3 years. They wanted $45 for this 3 inch length of thin hose.
Jim
The CE of the station will be trying to get the schematic. I wasn't able to find anything of use on ASCO's web site except sales brochures.
B&S parts are usually quite reasonably priced. I'm surprised. Maybe that was the dealer's price to get the part and install it? I mean you need real special tools to work on thse engines (hammer and screwdriver at least) !
I'm still working on trying to measure the wye voltage and phase by finding another delta-wye transformer where we can disconnect one feed line. I know a few electricians and they might have some sites that are partially wired.
Bob M.
B&S parts are usually quite reasonably priced. I'm surprised. Maybe that was the dealer's price to get the part and install it? I mean you need real special tools to work on thse engines (hammer and screwdriver at least) !
I'm still working on trying to measure the wye voltage and phase by finding another delta-wye transformer where we can disconnect one feed line. I know a few electricians and they might have some sites that are partially wired.
Bob M.