Voltage regulators - you can set a clock by. 1

[itsmoked]

We have two 50kW, 240V, 3ph, 4 pole, generators. They both ran approximately 10 years before their voltage regulators failed. New ones were installed that ran several more years before failing. New ones were installed that ran about a year before failing. The latest set ran about 4 months....

Theories include:
1) Something changing in the generators that require more,(excessive), field current.

I'm having a little trouble with this since the two generators live different lives. They have lots of fuses in the field circuit but I'm not sure they would open on just a minor long term overload.

2) The regulator quality is getting worse.

It was noted the original VRs had one adjustment screw - voltage. The next set had two adjustment screws - voltage and (?). The latest has three - voltage, (?), and droop.

These generators are both using retrofitted electronic governors that are isosynchronous.

Ideas, suggestions, or further theories welcome.






Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

The second screw will probably be stability. With no load, turn the screw until the voltage starts to fluctuate. Turn the screw back until the voltage levels out. Stability increases with loading, so if it's good at no-load, it's good.
How about dirt buildup and possible tracking/flash-overs on an older, dirty machine.

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

 

[rasevskii]

I suppose that these are brushless exciters, and the fuses that you refer to are on the rotating rectifier assembly. If a fuse has failed, that would in fact require more exciter field current to get the same ac voltage. Usually there is a protection that would signal this (too high fld current over a certain time) but on such small units likely this is not provided.

just an idea

rasevskii

 

[itsmoked]

Thanks Bill on the mystery screw. We'll check that out.

These #&$^# regulators are fully potted so no tracking and no easy repairs.

rasevskii; The fuses have been checked and are intact.


Extensive web searching has turned up 'obsolete' documents that show that the original VRs had a slightly lower maximum field excitation voltage compared to the newer ones. It had a higher maximum current limit. And, the older ones were rated at 750W of field power and the new ones are rated at 625W of field power.. I think I'm seeing the writing on the wall.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[catserveng]

Do know the rated field volts and amps? How about the field resistance? Basler makes a number of good retrofit AVR's, their AVC series I use quite a bit.

I'd say a lot of AVR's certainly aren't as robust as they used to be, guess that's the value engineering thing. But after 10 years there may be something else, even on different units with different applications. Have you done a basic megger test on the stator and rotor? How do the diodes and surge suppresser (if fitted) check out? What is the no load rated voltage field volts and amp readings? Are they stable? As you add load how does the readings look, are you running out of field output before you get to full load?

I had a 60 kW unit with a Marathon tail end about 6 months ago, 12 years old, on it's 5th AVR, similar the what you describe. Last AVR lasted 2 weeks. Everything checked ok, but I felt the diodes maybe flakey, replaced them and so far so good.

Hope that helps, Mike L.

 

[electricpete]

I'm not a generator guy and you'll probably get better comments from the others, but I'll wade into unknown waters...

What does it mean when a voltage regulator "fails". Was there any check of what subcomponent failed?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[waross]

The magic smoke in a regulator is sometimes sort of greenish. As Kieth said, many of them are potted. Some are not but invariably have unidentified proprietary integrated circuits.
Catserveng and Keith, Do you suppose that the rotating surge arrester may be due for replacement? Could a failing surge arrester be letting load induced transients reflect back to the regulator?

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

 

[catserveng]

Bill,

If I come up on a generator with "odd" problems, including repetative AVR failures, I usually replace the diodes and surge suppressor even if they check good. I also won't replace just diose of a surge suppressor, the whole set get's done if it's my say.

When CAT had serious diode and surge suppressor issues a few years back we learned a lot about the components of a rotating rectifier assembly. I think the diodes and surge suppressor do age, and become capable of causing intermittant problems. It's also hard in the field to really test a diode properly, you check it with an ohm meter or a DMM in diode check mode in both directions, and if you get reasonable readings you call it good. But what happens under load, especially if you have things like lightening strikes, load harmonics, poor paralleling control, etc.

My own rule of thumb is 5 years on a prime unit, 10 years on a standby, then change them. Funny how many tail ends I see come back from a rewinder after a clean, dip and bake with the old rectifier components. What's not funny how many times I end up replacing them because the newly repaired tail end didn't work as expected.

Pete, most common failure modes I see in AVR's on the units I usually deal with (100kW to 10 MW)is field output failures and failures in the voltage sensing circuits. As Bill said, new units are potted, but seem to still have similar failure modes. One of the most common failures I see is after an AVR retrofit where the old analog regulator had a field breaker, and the new regulator is instaled with the breaker still in the field output circuit. First time field breaekr opens, field output of AVR usually dies. Most AVR suppliers I'm familiar with clearly state NOT to open the field circuit under load, if you want to disable the AVR, open the PM or power input. But if everyone followed directions I wouldn't be so busy!

 

[ScottyUK]

Aside for Mike -

On bigger grid-connected sets we tended to have more problems with the rotating diodes, the immediate cause usually being a lightning strike on a line close in to the station or a major grid disturbance. I am fairly certain that the underlying cause of some our problems was our installation practices, although given the access we worked through to replace diodes it is unsurprising that we didn't always get the compression right on the hockey puck diodes. A fault on the static part of our AVRs was quite unusual and usually temperature-related when it did happen. Brush ran some of their components - particularly resistors and zeners - at a scorching temperature which accelerated their demise.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[itsmoked]

electricpete; "Fails" means either no obvious visible problem, or the bottom of the housing is charred into a cratered trench where some now-vaporized component once existed.

Thanks for all the tips. Cat if I get there I will run the tests you suggest.

After a tideous web search much has been gleaned. The original Marathon regulator was the size of a lunch pail, non-potted, and did not need any forced air across it. It was rated:
63Vdc 4A continuous
105Vdc 7A forcing (735W)

The new ones available are the size of a large deck of cards:
73Vdc 3.5A continuous
105Vdc 5A forcing (525W)

There are also issues with cooling. These generators are in boxes with little air exchange so the ambient is pretty horrendous. These new regulators are supposed to be mounted in the cooling air flow of the fan forced air being sucked across the rotor. They haven't been and it's hot air anyway..

I am wondering about the air inlet for the prime mover. That has to be cooler air and a lot of it. Perhaps mounting these diminutive regulators in that air stream would work.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[catserveng]

Keith,

You may want to consider a Basler AVC63-4, here is link,

http://www.basler.com/downloads/svd3bull.pdf

or

http://www.basler.com/downloads/svc4bull.pdf

I have used these quite a bit and find them very good for smaller units. They have two versions, one with the adjustment pots on the "back" or nonheat sink side (AVC63-4D) and the other with the pots on hte heat sink side, a quote from the bulletin,

"Both the AVC63-4 and the AVC63-4D are intended to be
mounted inside the generator's terminal box. For applications that require the voltage regulator adjustments
to be made from outside of the terminal box, the
AVC63-4D should be considered."

Hope that helps,

Mike L.

 

[itsmoked]

We are just now debating what to buy, Thanks!

Keith Cress
kcress -

http://www.flaminsystems.com

 

[JIMGEN]

Or the mov could be shorted. But kind of interesting that both units are doing the same thing??