Voltage regulator questions II 1

[itsmoked]

Please recall the thread: thread237-399922

A Newage Stamford tailend.

Fried regulator.
Dual Turbo'd Cummins
NO PMG.

Well glory be! The generator, living in a converted baggage car, actually came to my town and only about 8 blocks away.

I meggered the exciter stator: 8MΩ
I meggered the main stator: 15MΩ

Exciter Resistance is: 23.5Ω
With a 12.8V battery hooked to it I get about a 1/2 amp and a mondo inductive spark (as expected).

Fired up the Cummins - after having to replace the two 8D batteries that were left to rot (were only two years old).

And did I get 480V between all phases with less than a 1% variation as suggested by the test manual?

NO.

I got 390V, 226, 228V.

I'm starting to see why MikeL said other things often break the regulator. I can see it straining a gut trying to reach a level of excitation that could get the output up to 480V.

I'm guessing this is going to be a faulty main stator winding isn't it? Not likely the diodes or all the phases would, perhaps, be low but close to the same. I did not get a chance tonight to run it long (10pm) as it's next to a trailer park and I figured I'd cut them some slack. I wasn't able to check for subtle engine loading with the field battery engaged. I'm scratching my head as to how one can tell if a winding is heating up.





Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

Check for phase to wye point voltages.
I have had a couple of generators that were started up with wet windings that failed quickly. (After the hurricane, not from sitting), but your windings may have been saturated with condensation.
The neglected batteries are a good suggestion that there was no anti-condensation heater in service.
Proceed carefully. If the windings start arcing to ground, it may cause irreparable damage to the core, quickly.
You may try a low voltage AC across each phase in turn. (Phase to wye point.)
Measure the current. A winding with shorted turns may be expected to have a lower impedance than a good winding.
If you don't get an indication with low voltage, try a Variac and a fuse and increase the voltage. You may get a transition from a good reading to a bad reading as the voltage rises.
If the wye point is grounded, you may consider lifting the ground connection. If a turn to turn short starts arcing, it may quickly develop into a turn to ground fault with rapid core damage.
You are aware of the dangers of an arcing fault and possible high voltages. Proceed with caution if you decide to lift the ground connection.
A good visual inspection of the end turns is in order, possibly with a bore scope.
I am not concerned with the low voltages that you saw.
I am very concerned with the wide variation in the voltages.
The uneven voltages are strongly suggestive of stator winding damage.
I am somewhat pessimistic about this one.
As always, Mike will have some good suggestions.
Good luck.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[catserveng]

When you meggered the stator I'm assuming you went phase to ground with the neutrals still tied together, right?

Can you seperate the neutrals (if its a 6 lead machine) and read phase to phase? Usually reading like yours indicate a phase to phase problem, two low phases and one higher but not quite to rated. I also wouldn't have expected a "mondo" spark to excite an unloaded machine to rated voltage, at least not in my experience.

MikeL.

 

[ScottyUK]

Winding is definitely connected correctly isn't it? There's a very-near-to [√]3 ratio there.

 

[itsmoked]

Check for phase to wye point voltages.

Wye points? There are none I know of. This is a delta only.

I have had a couple of generators that were started up with wet windings that failed quickly. (After the hurricane, not from sitting), but your windings may have been saturated with condensation.
The neglected batteries are a good suggestion that there was no anti-condensation heater in service.

Heaters?? This 'place' has no power normally.

Proceed carefully. If the windings start arcing to ground, it may cause irreparable damage to the core, quickly.
You may try a low voltage AC across each phase in turn. (Phase to wye point.)

Can I do this line to line?

A good visual inspection of the end turns is in order, possibly with a bore scope.

OK

I am not concerned with the low voltages that you saw.
I am very concerned with the wide variation in the voltages.
The uneven voltages are strongly suggestive of stator winding damage.
I am somewhat pessimistic about this one.

Indeed.

Thanks Bill


Keith Cress
kcress -

http://www.flaminsystems.com

 

[itsmoked]

When you meggered the stator I'm assuming you went phase to ground with the neutrals still tied together, right?

Yes..
You also think there's a neutral somewhere in this steel cylinder..?

The "Box" mounted on the exciter end top of the generator has exactly three ~0AWG cables coming up to three large short busbars. The three buss bars each have two parallel ~500mcm cables leaving them to the breaker/GFI box elsewhere. Also in this Box a single 14AWG wire leaves each busbar to a small terminal block and continues to the breaker box where the voltage regulator also lives. The exciter/field wires leave the exciter stator and also land on the little terminal block where they continue to the voltage regulator. That's it. I see no other neutral connections and no place for them to congregate short of being jammed somewhere in with the stator windings.

Can you separate the neutrals (if its a 6 lead machine) and read phase to phase? Usually reading like yours indicate a phase to phase problem, two low phases and one higher but not quite to rated.

I will look further for the 'other' ends of these stator phases.

I also wouldn't have expected a "mondo" spark to excite an unloaded machine to rated voltage, at least not in my experience.

Maybe I've mislead you on this. I'm just referring to the spark you get when you hook up the battery to the exciter stator as being obviously more than what you'd see to a resistive load. (Without the machine running even.)



Keith Cress
kcress -

http://www.flaminsystems.com

 

[itsmoked]

Winding is definitely connected correctly isn't it? There's a very-near-to √3 ratio there.

I got that feeling too. Yeah, no one has messed with the 'setup'. I will do the test over with the leads lifted to make sure there's no unexpected load across two phases sending me down the wrong path. I've tried to trace things out but the panel is packed.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ScottyUK]

If you can separate all the winding ends and identify them, arrange them so all windings are grounded at one end and open at the other. That will reduce the likelihood of destructively large current flow, but be aware that you are still vulnerable to inter-turn faults in this configuration and you are vulnerable to inter-turn faults any time excitation is raised. With the set running measure each winding voltage line-ground - all should match, and windings of the same phase should pretty much be equal. If they aren't then you're probably looking at a stator rewind.

 

[catserveng]

Keith,

Sorry, I thought this was a wye connected machine, what I mostly see at that rated voltage. Ideally you should isolate each stator winding and make sure that there is no phase to phase short.

You also want to test the main rotor to see if it is ok as well

MikeL.

 

[waross]

I missed the root 3 ratio.
As a possibility, the windings may be wye connected internally and one winding is completely shorted. That may give those readings.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[ScottyUK]

Bill - I only spotted it because the voltages are very close to the old European standard distribution voltages of 220/380.

 

[LionelHutz]

If you have decent a motor shop nearby, paying them for the time to come in and do a surge comparison test on the winding would be worthwhile. Then, you will definitely know if there is a problem, and probably what the problem is too.

 

[waross]

If one winding is shorted, the voltage regulator may have died trying to develop full voltage.
I would resist the temptation to run the engine until the windings have been inspected.
The first step in a generator failure is often winding damage. This is often repairable.
The second step in a generator failure is often core damage. This is often NOT repairable.
Yes, I know that you know this Keith. The comment is meant to re-emphasize this point for the benefit of those following this thread for general knowledge.
We don't want to see anyone having a repairable machine become an unrepairable machine as a result of inappropriate troubleshooting methods.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[itsmoked]

Follow up:

Well.
Scotty nailed it.

I went back to figure out how to isolate the stator windings taking a 6 digit Kelvin ohmmeter with me. I took the top off the 'generator box' to look at all the busbars where the stator cables land. There are two cross bars at different levels with 4 inch busbars. While pondering why I was seeing two stator cables landed on each bar, like a face slap with a mackerel, I realized there were FOUR busbars, each with two landed cables. Further inspection made me realize that the three 14AWG wires across the room in the control panel that I was reporting the badly different voltages across were not all going to the three Phase busbars. One was going to the 'fourth' busbar. In my defense the insides of the generator box is completely homogeneously coated in a hard gray/brown dust that does not blow-off but must be scraped off with a sharp tool. You can't see the ORANGE, YELLOW, RED tape indicating the phases on the big cables. Ney you cannot even see that there is marking tape.

Once the discrepancy was noted I scraped the ends of the three and only three 14AWG wires learning that they were W, U, and.... N.
Hence the 390/224/226 readings.

With that knowledge it became clearer that a transformer was not needed to get the 480 down to 277V the Vreg could swallow as I had 277V available.

I redid the 12V/field battery test getting 390.2/390.1/390.2V.

I proceeded to hook-up the already purchased and installed by others VR902 regulator. After spending a bunch of time finding a shunt trip breaker and installing it. I discovered whoever had initially mounted the regulator had just tossed a bag in the bottom of the enclosure that contained the bizarre shunt-trip (18VDC!) and a required voltage adjust 3,500Ω pot.

So I yanked my 100V~415V shunt-trip made a bracket to hold the factory shunt-trip and pot, soldered leads on the pot and wired it all up.

Checked the water level. A bit low. Checked the oil level about in the middle of the allowed range and surprisingly immaculate.

Walked back thru the power car to the F type locomotive past the 1,500HP prime mover and stuck my head out the door to tell the security guy to standby for some noise. Man. This is parked 100 feet from about hundred apartments and it's 8pm. Hopefully Monday Night Football will keep them distracted.

I back tracked and hooked up the brand new pair of batteries, battery leads to the Cummins starter. Turned on the fuel pump breaker. Set the engine switch to run. Noticed the main 700A breaker was off and left it that way. Pressed the starter switch and waited while the engine got it together. When it finally caught the sound was astounding. With headphones on it sounded like I was in a gunfight with only short barrel 44 magnums blazing away - point blank. I have never heard such a racket. Smoke started pouring out of the sides of the twin turbos. Which appeared to be covered with black oil.

After about 6 seconds of this sonic maelstrom the noise stopped replaced by a smooth but slow running engine. After about ten more seconds the engine smoothly accelerated to synchronous speed. The smoking turbos tapered off to no smoke and the black oil coating utterly vaporized to 'classic' rusted iron.

Measuring the voltage at the breaker was found to be 360V. Turning the voltage adjust pot up all the way resulted in 407V. The built-in voltage range pot on the reg was adjusted until the voltage was 500V then the voltage was adjusted down to 480V with the voltage adj pot. Well actually since it was pretty course 479.6V. The balance was 479.6V/479.6V/479.6V. I don't think I've ever seen that balanced a set of readings.

I tried measuring the frequency but my meter completely spazzed-out showing 810Hz. I switched to a new clamp-on Klein meter. I clamped it around one of the big phase cables going to the breaker just to hold it. I turned the meter to Hz and it read 60.1Hz without even hooking the leads up?? When I hooked the leads across phases it too read garbage 1100Hz, 800Hz, etc. Once I got garbage it stayed garbage. I'd have to turn the meter back off then on then set it to Hz and it would show 60.1Hz again until I hooked the leads up. I don't get it.

By this time the engine was up to 120F. Still standing in a cloud of oil smoke I started wondering why the six foot diameter 20HP radiator fan was not running. I guessed maybe the main breaker had to be on. I decided to close the panel before trying that particular festivity. On closing it low and behold there was a large panel frequency meter reading 60Hz.

Both-handing the breaker lever closed there was a sudden hurricane blowing past me. The fan! Checked it and it felt smooth.

Temp got up to a steady 160F.

The room is surrounded by 200W light bulbs that didn't light up. Why? Had to dig out the breaker panel. No power found in it. There is a contactor that is twice the size of the main breaker in the generator control panel. It hooks the generator to the Head End Power bus when the ON-LINE knob is turned. Unfortunately a few days ago the yahoos incorrectly connected the HEP cable up and it came loose to be dragged by the train down the ties. That pulled half the wires loose and the pins out of the connector. Since that hadn't been repaired there was no way I was going to close on HEP. We'd done the whole job on drop lights from my truck via an inverter and a Milwaukee battery work light.

I shut it down and asked the security guard if he'd seen anyone with torches and pitch forks. He said AFTER the start it wasn't very loud really. "He'd actually forgotten about it after a few minutes." Guess the two 8 foot long 2 foot diameter mufflers do the trick.

The only thing left to do is load test it. I have no idea what the field current needs to be so my fuses may be wrong. I have no manual for the tail-end and not even a plate for the generator. My plan is to run the full train, about 10 cars, with heat and lights while measuring the field current. Add a fudge, factor and put those fuses in.

Thanks everyone for your support and guidance. It all went together to flesh-out a successful result.



Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

Hi Keith;
No load speed is typically 61.8 Hz --- Unless the governor is set to isochronous mode.
If the frequency dips slightly and returns to normal when you load it up it is probably in isochronous mode.
If the frequency drops aHertz or so and stays down under load, the then the governor is in droop mode.
The normal no load setting in droop mode is 61.8 Hz at no load.
Thanks for the update.
Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[ScottyUK]

I always enjoy your write-ups Keith, this one 'generated' a few early-morning smiles.

 

[itsmoked]

Bill; These generators NEVER run paralleled and the gov'ner is isosynchronous.
But, I did not know those values - noted.

Scotty; I try to paint The Picture so everyone can 'be there'.

Now why did NONE of my freq meters work? Does a 1/2MW generator put out horrific harmonics when completely unloaded and the meters didn't know what to make of them? Seems far fetched.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

Hi Keith. Governors are proportional controllers.
With 3% proportional band a set point of 60 Hz will give a full-load frequency of 57 Hz.
3% offset is added so that the no load frequency is 60 Hz + 3% or 61.8 Hz.
This drops to 60 Hz under full load.
For 50 Hz systems no-load frequency is generally 51.5 Hz.
Rather than call this "3% proportional band plus 3% offset" it is called 3% droop. Clocks generally run a little fast and some UPSs complain if the frequency is not exactly 60 Hz. Other than that users are never aware that they don't have exactly 60 Hz.
Isochronous or isosynchronous adds integral or reset action to the proportional controller. The first response to a load change is proportional followed by the integral correction.
Mike; Are any of the new load control panels and governors using load monitoring for a faster response to load changes?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[catserveng]

Keith,

Have you taken a DMM right to the back of the freq meter? What is the voltage and frequency on your meter? Seems you had some unexpected wiring on this unit, so maybe this is another one of those situations? Of all the panel meters I replace, I end up doing VAR and frequency meters the most, especially in a mobile application like a railcar mounted unit. You might want to make sure you're feeding the meter the correct voltage, not common but I do see 240 VAC frequency meters sometimes in the US, mostly on units that are 50/60 Hz rated and where the panel meters are fed from the same voltage as the AVR sensing.

Bill,

It is very rare these days that I work on a unit with governor droop, even in paralleled applications. The some of the Woodward digital governors can be programmed to use load response for helping improve governing dynamics, along with some of the load controllers like the DSLC2 and the EasyGens. I'm sure many others like the Basler, ComAp and Dief controllers pretty much do the same thing, but the bulk of my experience is with the Woodward products. I did a job earlier this year with some larger medium speed diesels (5MW) using Woodward 828 governors where we used a kW signal input to help respond better to large load changes. Not only for load pickup, but to help reduce overshoot as well. Took awhile to get it all dialed in but the results seemed to make all involved happy with the overall performance.

For anyone interested this Woodward manual does a nice job illustrating Bill's explanation of droop governing (along with a bunch of other load control info),

http://www.otherpower.com/images/scimages/3409/Governing_Fundamentals_and_Power_Management.pdf

Or you can look up Woodward Application Note 01302

Hope that helps, MikeL.

 

[waross]

Thanks Mike. Nice manual.

Bill
--------------------
"Why not the best?"
Jimmy Carter