Subject: 50 HP Copeland Discus died - experiences ?
Date: 4 Jun 1995 03:20:53 GMT
{ Special thanks to Jonathan Elson & Ken Robinson for the great contributions on this one }
I lost a 50 ton Copeland Discus compressor last week, in a McQuay chiller.
( I know, it's so big, I ought to be able to find it somewhere ! :~( )
Initial symptom - This part-start unit has two 150 amp breakers for the
compressor contactors. On call, lead contactor picks, then one second
time delay ( as per spec ), then lag contactor picks, pulls 300 amp for
a few seconds, and lag breaker trips. BTW, it's a pump-down unit.
The other compressor works fine, BTW. Unit ohms and megs good.
No history ( 7 year old unit ) of failure. Oil good, freon good.
Changed contactors ( points were a bit smoky, but not terrible,
and new contactors were actually cheaper than point sets ! ),
changed breakers ( had spares already on-site ). No help. Voltage good,
all terminals tight, voltage holds well even at LRA, no sign of over-
heated wires or such anywhere. Reversed rotation, no help. All legs
exhibit LRA.
Teardown results :
Clean. All the way. Pulled off everything that pulls - 4 heads, end bell,
oil pump (and took pump apart), electrical plate. No problem found.
I can spin it by hand. There's a slight squeak to it somewhere, but no
bearing play and not hard to turn. Valves, knee rods, etc. all OK.
Oil level acceptable ( just up to the bottom of the glass ),oil almost
immaculate ( excellent for 7 years old ). Windings perfect. Electrical
connections perfect. No head discoloration. Unloaders OK. Oil pump
internals and engagement OK. Dip tube cast into body OK. Valves and
piston heads spotless. All screens spotless. Based on the above
( and I probably looked at things I'm forgetting to mention here ),
I couldn't give a good reason not to put it back in, except that I just
took it out ( and the new one's already in ).
Outside of flipping it belly-up and taking the bottom plate off
for a full tear-out ( which I don't think would tell me anything ), what
have I missed ? Gremlins ?
Paul
From: afn03528@freenet.ufl.edu (Ken Robinson)
You didn't say. Does the compressor attempt to spin? My only thought is
that you have a winding short. In some situations only a single winding or
two will short causing a transformer effect. The shorted coils draw a
large amount of current from the rest of the winding causing on overload
condition. With only a one or two winding short the resistance of the
winding will be very close to spec.
Are any of the windings discolored?
Three phase?
afn03528@freenet.ufl.edu (Ken Robinson)
Be sure to check the wiring between the contactors to the compressor for
a short to ground - don't laugh, I found one like this after
"professionals" condemned a 2 thousand dollar semi-hermetic compressor
(which had not one ____ thing wrong with it). Usually a short like this
makes itself fairly obvious (smoke and so forth), but if the lead(s)
is(are) able to weld themselves and create a connection that will trip
the breaker every time with no further deterioriation of the leads, you
will be hard pressed to see it.
: No laughing matter at all. I did check the wiring, to the extent of
stripping off the BX to examine every inch of every wire. I've seen the
wrong diagnoses things you speak of. I was called on one unit in Yonkers
(Queens, NY ) to second-guess a contractor who wanted to change a
chattering contactor. The owner didn't want to pay for a new 100 amp
contactor. Turns out it was a 240V coil, in series with a safety in the
compressor. When I popped the compressor open ( 15 HP Carlyle ), it was
major toast city - I couldn't have made it any blacker in there with
spray paint ! One side of the contactor control circuit was grounded
inside the compressor, the other side was hot w/ 120V, thus the chatter.
I got to tell the owner 'good news, bad news'... contactor's OK,
compressor's gone ! :~)
I've also seen where they run large power wires and small control
wires in BX, and the large wires end up on top of the small ones, and
rub them against the BX to where they short and/or open.
Sometimes they can wear such that they touch and ground out, then the force
of that spark pushes them away such that they ohm good, then the vibration
of start-up and running eventually brings the bare spot back into ground
contact ( another trip ), etc.
~~>I also ran across one case where a
contactor was shorted to the backplate of the relay panel it was
installed on - the short was in the back bakelite of the contactor (to
the backplate) and was absolutely impossible to see. BTW, I caught both
these problems with a Simpson 260 (no megger required) and have learned
to make sure that wiring downstream of the contactor and upstream of the
compressor is not shorted phase to phase or phase to ground.
Hope this helps you.
Glenn
Absolutely right. Thanks. BTW, the new compressor is in, and spins like
a top. I've had some email and seen some postings here that discuss
certain motor failure modes very thoroughly ( I think more postings are
coming :~) ). At this point, that's my inclination. I changed hi & low
pressure safeties and oil pressure safety just on general principles, too.
Don't you just know it - the hi pressure safety was bad right out of
the box ! Added immeasurable enjoyment to my day when I turned it on and
nothing happened !
Paul
From: jmelson@artsci.wustl.edu (Jonathan M. Elson)
The symptoms would first indicate bearing failure, but you looked at that
and found nothing. So, the only other thing it could be is a shorted
stator winding. Not a short to ground, which is easy to detect, but just
a few shorted turns. This will absorb energy from the stator flux, and act
as a heavy drag on the motor, causing increased current draw. the winding
resistance may not be altered enough to read the difference. The failed
turns are nearly always at the inside of the winding, so the burned
insulation is rarely visible. This is because the innermost layer of turns
in a winding are farthest from the cooling effect (in this case of
refrigerant) of the motor's coolant, and also have been bent to the sharpest
radius, and also are subjected to the most intense alternating magnetic
fields (from all the windings above it), and so have the most eddy current
heating. The innermost windings often run 10-30 C hotter than the outermost.
The heat leads to insulation breakdown earlier in the innermost winding.
An inductance bridge will show the problem, but you probably don't have
one of those. The inductance of the bad winding won't differ much either,
but the shorted turns will cause the 'Q' (quality factor) of the bad winding
be approximately ZERO! That will be the indication of shorted turns.
An impedance bridge is similar to a wheatstone bridge, however it is used
at AC instead of DC, to measure the complex impedance of capacitors and
inductors. Not only does it measure capacitance or inductance, but it
measures the quality of the component, as well. A stator winding (either
in or out of the motor) will show small resistance and large inductance.
It should have high 'Q' when out of the motor.
Since the bearings have no visible damage, it is practically impossible
that they are the problem. A 50 HP motor would turn bearings into crumpled
bits of metal and still keep turning. A 'sticky' bearing absorbing 50 HP
would be melted into ruin in several seconds.
From: jmelson@artsci.wustl.edu (Jonathan M. Elson)
: Could you explain the 'absorb energy from stator flux'? Is this something
like creating an opposing field ?
No, exactly the opposite. the iron in the stator (and the magnetic field
in the rotor once it is spinning) oppose the field of the stator, thereby
preventing it from drawing current from the line. The shorted turns remove
the magnetic field, thereby causing the rest of the winding to draw MORE
current from the line.
: Very interesting ! Are internal sensors usually placed there ?
yes, when a motor is tested for insulation class, a thermocouple is placed
near the innermost turns of the winding, or some other location that the
motor designer determines will be the hottest winding location in the entire
motor. The temperature rise is then determined at full load. This procedure
is described in some detail in some versions of the NFPA's National
Electrical Code book.
: What is 'quality factor' ? It's a new term to me.
Quality Factor, or Q, is an electrical engineering term relating to the loss
in an inductor. Higher Q means less loss.
: Is there a way to use a capacitance meter to gain information ?
Or to home=brew an inductance bridge ( assuming one owns a good digital
multi-meter with capacitance function ( Fluke 12 ) ?
From: jmelson@artsci.wustl.edu (Jonathan M. Elson)
: What tests would work while the stator is still in ? Does the inductance
bridge still work ? DOes 'Q' still apply ( or has Picard banished him ? )
Well, the problem with the rotor inside the stator is that the copper
conductors in the rotor are SUPPOSED to be shorted. And, the stator windings
definitely see the short. The stator draws large current from the line,
transfers large flux across the gap between stator and rotor, and causes
current to flow in the rotor winding, causing opposing magnetic flux.
This is what starts the motor spinning. But with the stator not conducting
current, and the rotor not spinning, it looks like a short, and will make it
much harder to see the difference between a good and bad stator. It doesn't
matter whether the stator is in or out of the motor housing, but it will
matter whether the rotor is inside the stator.
To answer another question, I don't see an easy way to convert a capacitor
meter on a DVM to read inductance. And for stator checking, you really
don't want inductance, you want 'Q'. A Digibridge from RLC (formerly
General Radio) will do it, but it is an expensive instrument. There are
cheaper instruments for radio-TV service, but I don't know if they have
the range for a big motor. The light-bulb idea is good, it might take some
tinkering to find the right wattage light bulb to get the best effect, but
it should work. Just remember, you are hooking the stator up to the power
line, so watch where the wires are dangling.
Path: From: jmelson@artsci.wustl.edu (Jonathan M. Elson)
: Thanks again for more really great information ! One more question,
if I may... 'light bulb idea' ? I haven't seen any post about that on
my server.
A reply to the original post said to connect 110 V power in series with a
light bulb to two of the stator windings. Try all three combinations of
stator windings, and look for differences in lamp brightness. With the
stator removed, the bulb should be dim with all three possible connections.
If the bulb is bright for one pair of wires, you have shorted turns on that
winding.
From: afn03528@freenet.ufl.edu (Ken Robinson)
If you still have access to the compressor, try this...
Using a large incandescent bulb, 150 watt or so, connect the bulb in
series with one leg of an extension cord with insulated clips on the hot
end. Using this you can apply power, current limited by the bulb, accross
each phase. Using a clamp on ampmeter, digital preferred, you should be
able to see differences in current between windings. If you leave the
power on the winding for a while the shorted? winding will heat up and
identify itself in that way.
My previous followup mentioned the transformer effect of this problem.
Having a shorted turn or two like this is akin to the old automobile,
starter dragging, problem. It isn't the physical dragging that causes the
problem, it is the shunting of the magnetic field. A shorted turn or two
shorts the fields of all windings.
If this problem is suspected in the future, on other calls, using the
technique described above usually will help convince the owner that you
might know what the problem wight be.
As always be careful with the power.
afn03528@freenet.ufl.edu (Ken Robinson)
Sorry, Paul, I posted the following response to your question, but found I
inserted it in the middle of your original post. You likely missed it for
that reason.
Re: Checking for shorted turns using a high wattage light bulb.
I keep a "zip cord" handy, somewhere, with insulated alligator clips and a
standard 110 male plug. When I suspect shorted windings I connect the
suspect winding, one phase at a time, to a 110 source, with a large, 100
to 200 watt incandecent light bulb in series with the circuit to limit the
current. Then using an amprobe, preferably digital, I measure the current
in the individual windings, on multi-phase motors. By measuring the
current, limited by the resistance of the light bulb, and measuring the
voltage drop across the winding, you get the old E=I*R stuff, which can be
related to impedence. You will find several things happening when you do
this. As expected, the current will be markedly differently between phases
if there is truly a shorted winding. If the motor is multi-phase there
will be a marked difference between the field generated by current flow in
shorted versus non-shorted windings, and the 'transformer effect'
generated voltage on non-powered windings will be dramatically different.
afn03528@freenet.ufl.edu (Ken Robinson)
[ Editor's note... Thanks again to Jonathan and Ken for the education ! As of a month after this series of posts, the chiller has lost one of 8 condensor fans due to mechanical failure, but otherwise is running fine with the new compressor. ]
Back to Interesting Threads Index