xfmr stiff enough... Elevator soft starts - lights dim

[wyreman]

I assumed the five pulses are the result of
going over to the garage elevator door and pushing the buttons for all 5 floors.
the elevator goes up, unloaded, stops doors open and close, and then it moves to the next floor,

but now that you mention it, i think these are one second intervals... it must take at least twenty seconds - prob more- for the elevator to do its business at each floor.

one more thing to check!

the whole series of pulses looks like it is about 2 minutes long with about a minute gap.
I'll have to check if the elevator makes all 5 stops in two minutes and then it takes one minute to get back down.
https://1drv.ms/i/s!AkuU8mYuf3eEeOpa6tgvi7cApXM

 

[Electric-Light]

The amp imbalance looks rather high.

But as I mentioned earlier, it is helpful to just set up three L-N connected light bulbs. The exact measured values are less important as long as you're able to see visible difference in the way three bulbs flicker from each other and my guess is that you will see the difference with the readings I am seeing. It is a whole lot easier than comparing a tiny difference in a graph each time.

You can rotate the input to elevator equipment. Google rotating phase to balance current. Also read the controls manual. Keep watching what the light bulbs are doing. The flicker location changes with the load if it's a load issue. Think as if you're moving wheels around. Is it a bent wheel/rim or something on the vehicle side? Move tires around. Does drifting or vibration change or stay the same?

 

[Ingenieur]

What type of driving machine? Traction? Etc
#not checked
under what conditions does it happen?
#ascending --- every time


has fla under worse case been measured and compared against nameplate?
#yes --- as expected

have rate of accel and decel been measured/adjusted?
#unknown

Ascending every time
that is interesting
without knowing the machine type that is the lowest load
the counterweight is ~140% x empty car wt
meaning gravity pulls the car up and the motor is braking
if the car is 1/2 loaded then neutral, motor does little work
and if fully loaded is a its max

you need to see what the load is
record while
running empy
then with a full load per cab nameplate

 

[gar]

170225-0927 EST

wyreman:

Referring to your plots in post numbered #60 there is some interesting information.

Now that we know that each 100 A pulse corresponds to travel from one floor to the next it appears the motor is on for 10 seconds for this travel, and then the elevator stops for 10 seconds at each floor.

Disregarding the transient current pulses the current for the elevator rising looks like about 100 A at start, rises by about 20% to a peak over 5 seconds during travel and acceleration, settles back down to about 100 A while finding its floor level (possibly some of this last time period is non-acceleration, just linear up motion). This is a consistent pattern. Similar for all three phases.

The large transient current pulses, about 3 times the needed current to run the elevator are random, but occur at turn on of the motor. These are some sort of turn on transient. They simultaneously occur on all three phases and are about the same magnittude for each phase. Their apparent magnitude differences and non-existrence on some elevator cycles could result from the data logger, but I doubt it. This could be determined with a Fluke 27 or other similar min-max meter. A 20 second time period for the elevator cycle is plenty of time to read and do the min-max reset.

As I roughly calculated before the transformers have about an 800 A full load capability, and one might expect a 3% voltage drop at full load. 3% of 120 V is 3.6 V. 3.6*300/800 = 1.4 V out of 120 V is the voltage change that might be expected at the transformer for a 300 A transient load. This is to address some comments by others.

There is certainly a greater voltage drop on one phase vs the other two.

Some have mentioned the load is a delta, but from what I conclude from this thread the source is a wye and the sourcre being a wye is the important factor in looking at the problem.

Power input to the elevator is about 3*120*100 = 36 kW. Power factor won't be 1, but probably not 0.6 . About 10*36/3600 = 1/10 of a kWh is used to raise the elevator about 10 ft. Use half of that as an estimate, or 1/20 kWh. 40,000 ft-pounds of work equals about 0.015 kWh, So this systrem is likely quite inefficient.

.

 

[Ingenieur]

It is a hydraulic machine
I would get the elev guy in the loop
looks like running 60 A
peak 170 (probably higher)
looks like 250-300%
can be set at 200% or 120 A

it looks like the controller is rated at 80 (guess based on observed fla)
Controller setting is based on rating 116-425% x rating (80 in this example)
so if set at 200% that is 160 A or 160/60 = 270% of motor fla
to set at 200% x running A = 120 A
120/80 = 150% = controller setting

I would start here
set the starting amps as low as possible
200% is the recommended min
But you could go lower as long as it starts smoothly with full load

imho you do not have a power quality/capacity issue

 

[wyreman]

Somebody asked what am I going to do next site visit ?
My clear statement of the problem is

Has pge new source got a bad splice ?
Or is the elevator too wobbly to keep the lights steady ?

Set up
3 dataloggers
3 phase to neutral lightbulb's at main

Clear log of events and record measurements
Synchronize time

Look for p-n dimming at main
All 3 simultaneous lightbulb's - visual

Is there one bad leg? Rotate conductors!
Did the bad leg follow pge or elevator?

Mini 30a 120v load bank at main
Multiply instantaneous min max vd by factor of 10
Proves main has excessive vd even w/out elevator
Or not!

 

[wyreman]

Motor shop question

Motor shop question

According to my friend

Your graph shows Ph 3 bit low, both E and I. Curious.
If E source decreases I would decrease.
If load current decreases E would go up???
Only if there is restive element up stream
This is at meter, PGE.

Did you meter at Motor? Results were?
How would we measure IR drop for all 3 phases under load?
Using elevator
Has to be almost instantaneous read due to very short inrush duration

Does your graph show higher current on the lower voltage leg.
Are run currents equal?
Would I vary in motor differently at greater load ?
Dont think so. Motor shop question. Suspect linear.

Load bank pain to connect. Major down time

 

[wyreman]

It is a hydraulic machine
I would get the elev guy in the loop
looks like running 60 A
peak 170 (probably higher)
looks like 250-300%
can be set at 200% or 120 A

it looks like the controller is rated at 80 (guess based on observed fla)
Controller setting is based on rating 116-425% x rating (80 in this example)
so if set at 200% that is 160 A or 160/60 = 270% of motor fla
to set at 200% x running A = 120 A
120/80 = 150% = controller setting

I would start here
set the starting amps as low as possible
200% is the recommended min
But you could go lower as long as it starts smoothly with full load

imho you do not have a power quality/capacity issue

It's rated at 156a and elevator guy wouldn't set it lower than 200% which is where we are now

 

[wyreman]

170225-0927 EST

wyreman:

Referring to your plots in post numbered #60 there is some interesting information.

Now that we know that each 100 A pulse corresponds to travel from one floor to the next it appears the motor is on for 10 seconds for this travel, and then the elevator stops for 10 seconds at each floor.

Disregarding the transient current pulses the current for the elevator rising looks like about 100 A at start, rises by about 20% to a peak over 5 seconds during travel and acceleration, settles back down to about 100 A while finding its floor level (possibly some of this last time period is non-acceleration, just linear up motion). This is a consistent pattern. Similar for all three phases.

The large transient current pulses, about 3 times the needed current to run the elevator are random, but occur at turn on of the motor. These are some sort of turn on transient. They simultaneously occur on all three phases and are about the same magnittude for each phase. Their apparent magnitude differences and non-existrence on some elevator cycles could result from the data logger, but I doubt it. This could be determined with a Fluke 27 or other similar min-max meter. A 20 second time period for the elevator cycle is plenty of time to read and do the min-max reset.

As I roughly calculated before the transformers have about an 800 A full load capability, and one might expect a 3% voltage drop at full load. 3% of 120 V is 3.6 V. 3.6*300/800 = 1.4 V out of 120 V is the voltage change that might be expected at the transformer for a 300 A transient load. This is to address some comments by others.

There is certainly a greater voltage drop on one phase vs the other two.

Some have mentioned the load is a delta, but from what I conclude from this thread the source is a wye and the sourcre being a wye is the important factor in looking at the problem.

Power input to the elevator is about 3*120*100 = 36 kW. Power factor won't be 1, but probably not 0.6 . About 10*36/3600 = 1/10 of a kWh is used to raise the elevator about 10 ft. Use half of that as an estimate, or 1/20 kWh. 40,000 ft-pounds of work equals about 0.015 kWh, So this systrem is likely quite inefficient.

.

Magnitude differences may very well be due to the data logger.
The finest grain we could get on the data logger was one second interval.
I'm going back in with meters on min max readings
we're planning to min max everything on Tuesday

 

[Ingenieur]

It's rated at 156a and elevator guy wouldn't set it lower than 200% which is where we are now

The controller is rated at 156? Rated 50-60 hp
and he set the limit ar 200% or 310A ?

the motor looks like 60-70 A
or 20-25 hp

 

[wyreman]

https://1drv.ms/f/s!AkuU8mYuf3eEgQ16WZ50_9CHPfkc

hydraulic pump motor
200[208]v
50 hp
148a
pf 90.2

siemens series E03 elevator starter
voltage 200
hp in delta 50
amps in delta 156

overload relay set to 156a max

 

[gar]

170225-2050 EST

I have previously shown that an AC induction motor has no significant inrush current. It does have what I describe as starting current. See

http://forums.mikeholt.com/showthread.php?t=174880

In particular see the post numbered #27. These three plots are for a 3 phase induction motor on a Bridgeport mill with no load except friction and inertia. By changing gear ratio and belt speed it was possible to provide some loading change. Initial current is essentially locked rotor current.

Initial current is the same for all three tests. Final current is close to the same meaning it is mostly motor windage, and motor spindle and associated high speed bearing friction. What is quite different is the time to reach full speed. In a 3 phase motor there are no starting windings or centrifigual switches as in most single phase motors.

In these 3 phase motor current plots you see no 1/4 or 1/2 cycle inrush current as you see in the turn on of a transformer.

My photo P6 at http://beta-a2.com/EE-photos.html shows a peak inrush current pulse for a transzformer.

I have no idea what the motor and load for this elevator application are and therefore any reason for the short current pulse. Sometimes about 300 A, and other times about 0 A. Instrumentation might be a reason for the short pulse variability and non-existence at times, but I don't think so. It woul;d be desirable to know the cause or source of this current spike.

With a 100 kVA transformer at 120 V we have a full load current of 833 A. A 3% impedance and therefore a 3% voltage drop at 833 A should produce a 3*300/833 = 1.1% drop at 300 A. This would be 120*0.011 = 1.3 V drop at the transformer.

.

 

[wyreman]

OCT 2016 THE MINIMUM VOLTAGE WAS 197V, [256A]
FEB 2017 MINIMUM VOLTAGE DOWN TO 185V [AMPS UP TO 335A]
perception of more dimming seems accurate

We have doubled the xfmr and the conductors
starting to look more like a motor issue

THE HANGUP IS THAT VOLTAGE DIP IS FOUND BEFORE THE METER AND THAT MAX AMPERAGE IS FAIRLY UNIFORM

Question is about those spikes - is it an instrument issue?
Data logger had sampling rate of 1 second so we are going back tuesday to read just the min/max and put the system thru its paces

 

[Ingenieur]

https://1drv.ms/f/s!AkuU8mYuf3eEgQ16WZ50_9CHPfkc

hydraulic pump motor
200[208]v
50 hp
148a
pf 90.2

siemens series E03 elevator starter
voltage 200
hp in delta 50
amps in delta 156

overload relay set to 156a max

thanks
how many motor leads?
says submersible, 3?
motor fla 156, yet draws 50-60?
something doesn't add up?

 

[wyreman]

I'll look when we go back, again the logger has 1s grain
I'll set it with a calibrated meter, but i think it just needs the power to lauch
and its been unloaded when we've tested

I'll make sure to have at least one passenger in there - I have just the guy in mind for that job

 

[Ingenieur]

Motor data sheet http://apps.motorboss.com/connections/908000.pdf
http://ecatalog.motorboss.com/Catalog/Motors/ALL/J228

 

[Ingenieur]

I'll look when we go back, again the logger has 1s grain
I'll set it with a calibrated meter, but i think it just needs the power to lauch
and its been unloaded when we've tested

I'll make sure to have at least one passenger in there - I have just the guy in mind for that job

Lol

take a pic of the controller inside
make sure it's wired in side delta
get the capacity from the car lbs/riders

 

[wyreman]

ok!

my old boss asking about

The flicker is a result of voltage changes at the main assuming that is where the elevator feed Is.
The voltage drop at the motor is reflected back to and measured at the main towards the elevator.
Are there any more over current devices between the elevator main and the meter that could cause an IR drop?

from the top down
pge parallel conductors
main switch
bus junction
8 meter bank
attaches to a sep house meter with house disconnect
which feeds an Elevator Shunt Trip Breaker
goes thru some funky splices
into a ???” diameter pipe
comes out 40' -50’ away in a skinny pipe
changes to a bigger sweep off a reducing LB
to the local elevator disco
and down to the terminal compartment

https://1drv.ms/f/s!AkuU8mYuf3eEgRHZIw53kGN6ds9J

 

[gar]

170226-0847 EST

wyerman:

I took a quick glance at the Siemens 72 manual and this is a delta control with SCR phase shift current control.

I would expect that there would be no random 0 to 300 A spikes during startup. Startup current variation should follow some reasonably smooth curve, but possibly not as smooth as I showed for my Bridgeport 3 phase motor. Thus, possibly a faulty starter. If the starter is the cause of the unexpected pulses, then there is still the question of whether the 100 A pulses that are expected are sufficient to cause undesriable light flicker.

.

 

[gar]

170227-0754 EST

As I have thought over your problem more I am rather sure you have a defective soft starter. That does not mean it is the only problem.

I have no experience with your starter or any other soft starter. But logic, limited information, and being remote where I can't make measurements still leads me to this conclusion.

With the thousands of members of this forum it would seem to me that somebody would have direct experience with this particular starter, or at least a similar starter, and have plots of expected current and voltage vs time. We need to see such information. I don't believe the 0 to 300 A random transient belongs in a correctly working starter of this electronic design. If this were an electromechanical starter, then some random pulses might exist (reference my old AC machinery textbook), but I don't know of what size those pulses were.

Ordinary electrical circuits are not just toggle switches, incandescent bulbs, simple transformers, and simple induction motors anymore. To troubleshoot today one has to have much more knowledge of how the devices they use work.

.