I am suddenly losing transformers at one facility. We moved about 50 machines last year from one plant to another and had to modify each machine to accept the existing 480v 3phase supply bus. We decided on Dry Type Distribution trans, 9KVA, 480 Delta to 208Y/120. This choice accomodated all the various requirements of different machines. After 6-10 months in service I have 3 transformers go bad in two weeks (different machines). I have a fused disc on primary w/ 15A RK fuses and 3 phase 30A breaker on the secondary. What's burning them up and why?
Control Transformer woes
- Thread starter Billian
- Start date
- Status
winnie
Senior Member
- Location
- Springfield, MA, USA
- Occupation
- Electric motor research
There is no way that you've provided enough information to get an answer, but if you are lucky the various experts here will barrage you with questions, and these questions will point you at things to look at until you do find the problem.
My questions:
What sort of '480V 3 phase supply bus' do you have? Solidly grounded wye? Corner grounded delta? Ungrounded delta?
In what way do the transformers fail? General overheating? Single point failures? Evidence of arcing in the transformer?
-Jon
My questions:
What sort of '480V 3 phase supply bus' do you have? Solidly grounded wye? Corner grounded delta? Ungrounded delta?
In what way do the transformers fail? General overheating? Single point failures? Evidence of arcing in the transformer?
-Jon
Cold Fusion
Senior Member
- Location
- way north
I don't know. Did you check the current, incoming voltage? Transformers running hot, restricted ventilation, cheap transformers loaded to the limits?
CF's theory on burning up electrical equipment: The OCP does not protect the device. The design protects the device. The OCP is there to put out the fire when the device or the conductors fail.
cf
quick answers
quick answers
The incoming supply is from a Grounded WYE (277v to gnd on each leg). The two that I have checked out were blowing two of the "primary" fuses and the primary legs read <300 ohms to ground leg (pole). I haven't got my hands on a megger yet but it looks like the insulation broke down internally. I see no edidence of overheating or arcing but the maint people say they smelled smoke. After replacing the trans, I took readings of < 8 Amps on primary side under full load. The secondary peaked at around 14 A.
The load side shouldn't come close to taxing the transformer. I have the power company monitoring the incoming power today. Checking for spikes or harmonics and such. I've just never seen this type transformer go bad.
quick answers
The incoming supply is from a Grounded WYE (277v to gnd on each leg). The two that I have checked out were blowing two of the "primary" fuses and the primary legs read <300 ohms to ground leg (pole). I haven't got my hands on a megger yet but it looks like the insulation broke down internally. I see no edidence of overheating or arcing but the maint people say they smelled smoke. After replacing the trans, I took readings of < 8 Amps on primary side under full load. The secondary peaked at around 14 A.
The load side shouldn't come close to taxing the transformer. I have the power company monitoring the incoming power today. Checking for spikes or harmonics and such. I've just never seen this type transformer go bad.
TxEngr
Senior Member
- Location
- North Florida
I had a similar problem where we were blowing Control Power and Lighting transformers at a high rate. When the machine was running, the voltage into the plant was fine (575VAC) since the taps were adjusted for that load. When we shut all of the equipment down, the voltage spiked (no load - about 615VAC) and the transformers couldn't handle it. Take a look at your operating conditions and your voltages at each condition.
SG-1
Senior Member
- Location
- Ware Shoals, South Carolina
Make sure your ammeter is a true RMS meter ( best ) or a peak reading ( meter reads the peak current & calculates the RMS, the reading will be high & cause more investigation with harmonics ). Do not use the common averaging type meter ( this type will read low with harmonics present ).
Harmonics would cause overheating in the transformers.
Harmonics would cause overheating in the transformers.
- Location
- San Francisco Bay Area, CA, USA
- Occupation
- Electrical Engineer
Also, define what you mean by using the term "Control Transformer". There is a difference between a "Control Power Transformer" and a standard "Dry Type Distribution Transformer". 3 phase transformers are generally Distribution Transformers, not CPTs. The difference is that CPTs are designed to be capable of delivering short duration high peak currents seen when energizing induction coils such as those found in contactors, solenoid valves etc. without saturating. If you use a DT on control loads such as that, you have to oversize the transformer much more generously than what is usually done for CPTs. If you don't, you run the risk if causing the transformer to saturate as it tries to supply the sharp increase in load, which, through a complex series of events*, causes a corresponding extreme current spike, which will incrementally damage the insulation. A 9kVA DT is capable of 16.6A FLC at 208V, you are using 14A, therefore I think if you are that loaded and turn something else on that has a coil, you are putting that transformer at risk for failure.
* That "complex series of events":
Current would increase as the transformer tries to deliver the required peak power. This would cause the core to saturate when the magnetic flux density reaches its maximum value. Once saturated, the inductive properties of the windings change. The windings don't act inductive anymore and start to look resistive. The current, which was previously controlled by the mutual induction of the windings, is now only limited by the copper wire resistance. The current would exponentially increase, overheat the windings and damage the insulation. CPTs are designed to handle this, standard 3 phase DTs are not.
* That "complex series of events":
Current would increase as the transformer tries to deliver the required peak power. This would cause the core to saturate when the magnetic flux density reaches its maximum value. Once saturated, the inductive properties of the windings change. The windings don't act inductive anymore and start to look resistive. The current, which was previously controlled by the mutual induction of the windings, is now only limited by the copper wire resistance. The current would exponentially increase, overheat the windings and damage the insulation. CPTs are designed to handle this, standard 3 phase DTs are not.
Last edited:
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
100419-1656 EST
Billian:
I do not follow your discussion.
Post 1. You have a 480 three phase bus. Is it delta or wye. From post 4 it appears that it is a wye.
To this source you added "480 Delta to 208Y/120". Was this at each machine or just one transformer.
Are these CNC machines or what, how much current and at what voltages now? What were the original voltages and configurations?
The title of the thread is "control transformers". What are their input voltages and I assume they are single phase? Or are you calling the "480 Delta to 208Y/120" transformer a control transformer?
Have you measured the primary voltage on the primaries of the transformers that replaced the failed transformers? And have you monitor these voltages with recorder of some sort?
From post #4. "The two that I have checked out were blowing two of the "primary" fuses and the primary legs read <300 ohms to ground leg (pole)."
Does this mean that the two transformers that failed read less than 300 ohms with a low voltage ohmmeter, like a Fluke 27 or 87, between either primary lead and the frame of the transformer with no primary lead connected to anything except the ohmmeter? If so, then you do not need a megger to know that you have low resistance to ground.
Either the transformers were defective or you have had large voltage transients relative to ground that broke down the insulation. I suspect the hipot test on control transformers with input capability of 480 will be at least 4500 V if not more. Were these control transformers, I am assuming here single phase with 277 applied to the primary, connected between a line and neutral? Or were these connected 480 and connected line to line? Or were the primaries wired for 120 input?
Other possibility is primary was overloaded. In this case the primary probably reads an open circuit.
.
Billian:
I do not follow your discussion.
Post 1. You have a 480 three phase bus. Is it delta or wye. From post 4 it appears that it is a wye.
To this source you added "480 Delta to 208Y/120". Was this at each machine or just one transformer.
Are these CNC machines or what, how much current and at what voltages now? What were the original voltages and configurations?
The title of the thread is "control transformers". What are their input voltages and I assume they are single phase? Or are you calling the "480 Delta to 208Y/120" transformer a control transformer?
Have you measured the primary voltage on the primaries of the transformers that replaced the failed transformers? And have you monitor these voltages with recorder of some sort?
From post #4. "The two that I have checked out were blowing two of the "primary" fuses and the primary legs read <300 ohms to ground leg (pole)."
Does this mean that the two transformers that failed read less than 300 ohms with a low voltage ohmmeter, like a Fluke 27 or 87, between either primary lead and the frame of the transformer with no primary lead connected to anything except the ohmmeter? If so, then you do not need a megger to know that you have low resistance to ground.
Either the transformers were defective or you have had large voltage transients relative to ground that broke down the insulation. I suspect the hipot test on control transformers with input capability of 480 will be at least 4500 V if not more. Were these control transformers, I am assuming here single phase with 277 applied to the primary, connected between a line and neutral? Or were these connected 480 and connected line to line? Or were the primaries wired for 120 input?
Other possibility is primary was overloaded. In this case the primary probably reads an open circuit.
.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
100419-2053 EST
Jraef:
I do not believe that load current causes core saturation. The volt-time integral has not changed, actually it is probably a little less. Over voltage on the primary will cause saturation.
The flux relating to load current automatically balances between primary and secondary with one canceling the other. With high leakage inductance the conditions are not as simple, but transformers for power transfer in general are designed with low leakage inductance.
You do not see peaking of the primary current as the load current from a resistive load increases. Rather you see less peaking of the current waveform because the peak from the magnetizing current is less important than the load current. Put a scope on the primary current of a transformer with no load first. Next put 2 times rated maximum current load from a resistor for a short time. Observe the results.
See my site http://beta-a2.com/EE-photos.html and look at photo p8. This is the magnetizing current to transformer with a 120 V primary and rated at 175 VA. Load this so the primary current is 1.5 A and it will be hard to see this 200 MA pulse in a 2000 MA sine wave (2 A peak of the waveform).
.
Jraef:
I do not believe that load current causes core saturation. The volt-time integral has not changed, actually it is probably a little less. Over voltage on the primary will cause saturation.
The flux relating to load current automatically balances between primary and secondary with one canceling the other. With high leakage inductance the conditions are not as simple, but transformers for power transfer in general are designed with low leakage inductance.
You do not see peaking of the primary current as the load current from a resistive load increases. Rather you see less peaking of the current waveform because the peak from the magnetizing current is less important than the load current. Put a scope on the primary current of a transformer with no load first. Next put 2 times rated maximum current load from a resistor for a short time. Observe the results.
See my site http://beta-a2.com/EE-photos.html and look at photo p8. This is the magnetizing current to transformer with a 120 V primary and rated at 175 VA. Load this so the primary current is 1.5 A and it will be hard to see this 200 MA pulse in a 2000 MA sine wave (2 A peak of the waveform).
.
kingpb
Senior Member
- Location
- SE USA as far as you can go
- Occupation
- Engineer, Registered
As others have pointed out; more info is needed.
I might take a look at the load current and plot it against the transformer protection curves, making sure you plot the transformer thermal damage curve.
This could be an instance where protection is set above thermal damage curve, becasue the NEC maximum transformer protection tables were used.
I might take a look at the load current and plot it against the transformer protection curves, making sure you plot the transformer thermal damage curve.
This could be an instance where protection is set above thermal damage curve, becasue the NEC maximum transformer protection tables were used.
Control vs Distribution
Control vs Distribution
I always thought a "Control Transformer" was so named because of its usage.
So it appears that we are using Distribution Transformers for the wrong job. These xfmrs were spec'd by someone else and I neglected to verify them.
We spent $86K on transformers and material to move the machines from one plant to another. The old plant had varing supplies from 208V, 220V, and 480V. The new plant has a 3-wire Bus Bar with a GND lug and I am assuming the Wye config because of the 277V to GND from each leg.
I initially recommended installing a 500KVA+ transformer and adding new 208V bus bar.
My meter is a Fluke 787 ProcessMeter that has served me well for over 15 years as is indeed reads True RMS as well as many other features.
My experience is from the actual Controls of industrial equipment (PLC's, Servos, Robots and such) and never spent much effort on transformer specs.
This is my first post and I am impressed with the response. I just learned more about something I always considered simplistic. Thanks
Control vs Distribution
I always thought a "Control Transformer" was so named because of its usage.
So it appears that we are using Distribution Transformers for the wrong job. These xfmrs were spec'd by someone else and I neglected to verify them.
We spent $86K on transformers and material to move the machines from one plant to another. The old plant had varing supplies from 208V, 220V, and 480V. The new plant has a 3-wire Bus Bar with a GND lug and I am assuming the Wye config because of the 277V to GND from each leg.
I initially recommended installing a 500KVA+ transformer and adding new 208V bus bar.
My meter is a Fluke 787 ProcessMeter that has served me well for over 15 years as is indeed reads True RMS as well as many other features.
My experience is from the actual Controls of industrial equipment (PLC's, Servos, Robots and such) and never spent much effort on transformer specs.
This is my first post and I am impressed with the response. I just learned more about something I always considered simplistic. Thanks
- Status