Electrician Talk banner

21 - 32 of 32 Posts

·
Chief Flunky
Field Service Engineer
Joined
·
2,005 Posts
Megger testing tests the ground wall insulation of a motor, the thick piece of material that gets laid in the slot before inserting the coils. It is also sensitive to contamination and moisture. By the time problems show up here the motor is usually done for. I rarely catch motors that are borderline but if you have even slight amounts of surface voltage tracking or flash overs the VFD will freak out at it intermittently. That’s the reason that the shape of the curve is more important than the result. Insulation resistance is strongly affected by temperature. If you test a motor in summer then winter the winter test will be 3-4 times higher for the same insulation. That is why 5 Megaohms is not a safe reading unless taken at 40 C. I was testing a bunch of motors in a cold snap a month ago (in North Carolina) and the correction factor was 0.11...needed almost 50 Megaohms on a 460 V motor to pass. PI is an attempt to get around this issue. You keep testing for 10 minutes until hopefully most of the insulation is polarized. Then you take the ratio of the 10 and 1 minute readings. PI doesn’t care about temperature. You are looking for about 2-6 for a reading but low isn’t necessarily bad, very high Megger readings invalidate PI (over 5000), and high readings often indicate very dry but cracking/crazing insulation.

That’s all the test will tell you. It catches about 80% of motor failures and when Klein and Extech have Meggers on the market for under $200 there’s no excuse not to have a motor techs #1 go to tool.

But this is only half of the story. A motor shop does the full Megger/PI test. They also measure resistance coil to coil which can pick up shorts or opens before the Megger test. Think about a 460 V coil with 100 turns. The voltage turn to turn is only 4.6 V. That is why insulation on magnet wire is so thin it’s transparent. This catches things the Megger test can’t. They look for resistive imbalance under 1-2% which is a way of comparing coil to coil. You need a Kelvin bridge tester (4 wire milliohm tester) to measure a motor though because resistance is in the milliohm range. Even on small motors under 5 HP where it is in ohms I’ve tried using a Fluke 87 V and the accuracy just isn’t there. This needs to be the expensive tool.

They also run one of two tests. In the shop they will run the surge test for historical reasons. This takes a huge capacitor charged to near insulation limits (which is why it is a shop test) and smacks each coil with it. They measure the wave as it bounces back and forth between the coil and cap and then compare coil to coil giving you a bunch of numbers. The key one is PP-EAR which tells you how much different they are. The field test uses an LCR bridge. It measures inductance using a low voltage of around 100 V and a frequency of 60-1500 Hz (tester dependent) to measure inductance in millihenries then calculates inductive imbalance. Anything over about 15-20% or any big shift from a baseline reading is a bad motor. This test picks up on insulation problems before the resistance tests do but also gives bad readings with rotor air gap issues or rotor issues. One of the more common problems with drives is reflected waves that destroy the first couple turns. It shows up on this test first then eventually the resistance test. It does not typically appear at all on the Megger test except in severe cases. I had a customer destroying motors this way over and over and it failed Megger for them about 25% of the time. It was so bad in their case a multimeter would find it (measure each coil). Again pricey equipment that most electricians won’t have.

There are many other non-invasive field tests. Rotor Influence Testjng does the inductance test as you rotate the rotor a little at a time and is much more sensitive to rotor and air gap issues but requires an uncoupled motor with free access to the end of the shaft and disconnected from any drives. Partial Discharge testjng does the surge test but also looks for sudden jumps in current/voltage if you have microscopic arcing going on inside air bubbles in the insulation. This test really only applies to medium voltage motors. Another one are variations of the shaft voltage or common mode current test. These look for bearing discharges that cause bearing fluting from VFD common mode voltage but has nothing to do with your problem.

There are several that have fallen by the wayside. Stepped voltage testing or tip up is an old variation on Megger tests intended to find surge voltage issues but it shows very little and the modern digital surge and inductance tests make it obsolete.

A DC hi pot is one to watch out for. All reputable motor shops have discontinued it. The test standard is IEEE 400 and reading the introduction is very enlightening. Summary: IEEE 400 says DO NOT do DC Hi Pot. Two reasons. First it shows nothing of value. The surge/inductance and Megger/PI tests show everything that hi potting can detect. But more importantly it can damage perfectly good insulation since it does something funky to it that never happens in the real world (applying DC at the insulation limits), and the hi pot can give passing results but the damage that it caused makes the motor fail hours to weeks later. Every credible third party quality controlled shop has discontinued the test for that reason or does the test but won’t warranty the motor after doing it. The solitary exception is on brand new motors.

The offline tests will find about 95% of motor issues. Online testjng gets you the last 5% but none of those tests except obvious things like bad bearings will cause a drive trip. All three offline tests can be done by a motor shop with an advanced motor tester in about 20 minutes (drop drive leads, clip on tester, screw with software, run test). I’ve run thousands of these tests on just about every kind of motor or generator there is on different testers at times and seen almost every type of failure there is. The machines are self checking...I had a lead issue on a motor Thursday where a scrap of varnished cambric was hiding and the machine kept giving me a hassle (bad connection) over and over until I found it.

Moving on...how far apart (cable length) are the motor and VFD? Do you have more than one VFD in the same raceway? What is the raceway made out of? What type of insulation/cable are you using? Solid or stranded? Do you have an independent supplemental ground (green or bare wire) connected directly between the VFD and the motor frames? No cheating.

The trouble with overloads is that It can be instant like a flash over somewhere and VFDs are very picky about these or it can be a problem like a chain binding up and jumping on a sprocket tooth or bearings and gearing or a process problem. The big thing to remember too is that the VFD is just a computer. An algorithm is deciding on the kind of fault. It could be a ground fault but the drive reports overload. So often you know it has SOME problem but you don’t know what it is.

Don’t get hung up on condemning a motor. If a motor shop runs it through the tests I described it’s not the motor. You need to look at other components. For one thing about 90% of the time overload us a load problem, not a motor or drive issue. So look at the event log first. How much load/current was there? What was the speed? How much load? Lots of hints buried in that data. Get a good vibration/mechanical tech to check things. I usually just Megger the motor to placate the customer then start looking for mechanical or process issues.

If you suspect a drive problem the diode test only checks diodes. You sort of have to rely on the VFD for shorted SCRs or IGBTs. But a final test is that if you have another drive or another motor, even a tiny but portable motor, try hooking up the motor to a different drive or the VFD to another motor. We keep a few rental VFDs which makes it easy for us. I can also check them with an oscilloscope but again that’s typically outside of the budget.
 

·
Registered
Joined
·
163 Posts
How about a real simple question?
I've come across this issue several times over the years.
So simple know one believes that was the problem.
Have you checked the acceleration and deceleration settings on the drive?
I've experienced trips during shutdown on a chilled water pump.
If the deceleration is too low the motor may overload trying to slow the flow of liquid.
Is the drive programmed to ramp to a stop, or coast to a stop?
A typical centrifugal pump, without a VFD, will coast to a stop.
 

·
Registered
Scada Supervisor
Joined
·
2,966 Posts
What is the plant voltage on weekends????
You state it happens on weekends, could the taps be set to high for the plant and when POCO has no load or plant has no load Voltage goes up. I have seen line voltage raise 20+ volts on weekends.

Cowoby
 

·
Registered
Joined
·
5,463 Posts
I can't download that manual without registering and I don't want to because I'm not willing to give up my personal data to learn about it. We have no distributors selling that line of drives around here, so I never have to deal with it..

Is that OI as in O for Over and (capital) I for current? Or is that O for Over and (lower case) l for load?

Over Current may mean something in your drive is detecting a rapid increase in current, maybe too rapid for it to be picked up by the display. Drives often have hardware overrides for things like this that protect the components faster than the microprocessor can calculate things. Usually the display will tell you it is a Short Circuit, but again, I don't know this drive.

If so, then it could be failing motor lead insulation, capacitive coupling in the cables, water in the conduit (which changes the capacitive coupling issue and/or exacerbates leaking insulation), any number of things that create an intermittent high current situation.
 
  • Like
Reactions: Kevhuff26

·
Registered
Master Electrician, Industrial and Residential work
Joined
·
52 Posts
Discussion Starter · #25 ·
Megger testing tests the ground wall insulation of a motor, the thick piece of material that gets laid in the slot before inserting the coils. It is also sensitive to contamination and moisture. By the time problems show up here the motor is usually done for. I rarely catch motors that are borderline but if you have even slight amounts of surface voltage tracking or flash overs the VFD will freak out at it intermittently. That’s the reason that the shape of the curve is more important than the result. Insulation resistance is strongly affected by temperature. If you test a motor in summer then winter the winter test will be 3-4 times higher for the same insulation. That is why 5 Megaohms is not a safe reading unless taken at 40 C. I was testing a bunch of motors in a cold snap a month ago (in North Carolina) and the correction factor was 0.11...needed almost 50 Megaohms on a 460 V motor to pass. PI is an attempt to get around this issue. You keep testing for 10 minutes until hopefully most of the insulation is polarized. Then you take the ratio of the 10 and 1 minute readings. PI doesn’t care about temperature. You are looking for about 2-6 for a reading but low isn’t necessarily bad, very high Megger readings invalidate PI (over 5000), and high readings often indicate very dry but cracking/crazing insulation.

That’s all the test will tell you. It catches about 80% of motor failures and when Klein and Extech have Meggers on the market for under $200 there’s no excuse not to have a motor techs #1 go to tool.

But this is only half of the story. A motor shop does the full Megger/PI test. They also measure resistance coil to coil which can pick up shorts or opens before the Megger test. Think about a 460 V coil with 100 turns. The voltage turn to turn is only 4.6 V. That is why insulation on magnet wire is so thin it’s transparent. This catches things the Megger test can’t. They look for resistive imbalance under 1-2% which is a way of comparing coil to coil. You need a Kelvin bridge tester (4 wire milliohm tester) to measure a motor though because resistance is in the milliohm range. Even on small motors under 5 HP where it is in ohms I’ve tried using a Fluke 87 V and the accuracy just isn’t there. This needs to be the expensive tool.

They also run one of two tests. In the shop they will run the surge test for historical reasons. This takes a huge capacitor charged to near insulation limits (which is why it is a shop test) and smacks each coil with it. They measure the wave as it bounces back and forth between the coil and cap and then compare coil to coil giving you a bunch of numbers. The key one is PP-EAR which tells you how much different they are. The field test uses an LCR bridge. It measures inductance using a low voltage of around 100 V and a frequency of 60-1500 Hz (tester dependent) to measure inductance in millihenries then calculates inductive imbalance. Anything over about 15-20% or any big shift from a baseline reading is a bad motor. This test picks up on insulation problems before the resistance tests do but also gives bad readings with rotor air gap issues or rotor issues. One of the more common problems with drives is reflected waves that destroy the first couple turns. It shows up on this test first then eventually the resistance test. It does not typically appear at all on the Megger test except in severe cases. I had a customer destroying motors this way over and over and it failed Megger for them about 25% of the time. It was so bad in their case a multimeter would find it (measure each coil). Again pricey equipment that most electricians won’t have.

There are many other non-invasive field tests. Rotor Influence Testjng does the inductance test as you rotate the rotor a little at a time and is much more sensitive to rotor and air gap issues but requires an uncoupled motor with free access to the end of the shaft and disconnected from any drives. Partial Discharge testjng does the surge test but also looks for sudden jumps in current/voltage if you have microscopic arcing going on inside air bubbles in the insulation. This test really only applies to medium voltage motors. Another one are variations of the shaft voltage or common mode current test. These look for bearing discharges that cause bearing fluting from VFD common mode voltage but has nothing to do with your problem.

There are several that have fallen by the wayside. Stepped voltage testing or tip up is an old variation on Megger tests intended to find surge voltage issues but it shows very little and the modern digital surge and inductance tests make it obsolete.

A DC hi pot is one to watch out for. All reputable motor shops have discontinued it. The test standard is IEEE 400 and reading the introduction is very enlightening. Summary: IEEE 400 says DO NOT do DC Hi Pot. Two reasons. First it shows nothing of value. The surge/inductance and Megger/PI tests show everything that hi potting can detect. But more importantly it can damage perfectly good insulation since it does something funky to it that never happens in the real world (applying DC at the insulation limits), and the hi pot can give passing results but the damage that it caused makes the motor fail hours to weeks later. Every credible third party quality controlled shop has discontinued the test for that reason or does the test but won’t warranty the motor after doing it. The solitary exception is on brand new motors.

The offline tests will find about 95% of motor issues. Online testjng gets you the last 5% but none of those tests except obvious things like bad bearings will cause a drive trip. All three offline tests can be done by a motor shop with an advanced motor tester in about 20 minutes (drop drive leads, clip on tester, screw with software, run test). I’ve run thousands of these tests on just about every kind of motor or generator there is on different testers at times and seen almost every type of failure there is. The machines are self checking...I had a lead issue on a motor Thursday where a scrap of varnished cambric was hiding and the machine kept giving me a hassle (bad connection) over and over until I found it.

Moving on...how far apart (cable length) are the motor and VFD? Do you have more than one VFD in the same raceway? What is the raceway made out of? What type of insulation/cable are you using? Solid or stranded? Do you have an independent supplemental ground (green or bare wire) connected directly between the VFD and the motor frames? No cheating.

The trouble with overloads is that It can be instant like a flash over somewhere and VFDs are very picky about these or it can be a problem like a chain binding up and jumping on a sprocket tooth or bearings and gearing or a process problem. The big thing to remember too is that the VFD is just a computer. An algorithm is deciding on the kind of fault. It could be a ground fault but the drive reports overload. So often you know it has SOME problem but you don’t know what it is.

Don’t get hung up on condemning a motor. If a motor shop runs it through the tests I described it’s not the motor. You need to look at other components. For one thing about 90% of the time overload us a load problem, not a motor or drive issue. So look at the event log first. How much load/current was there? What was the speed? How much load? Lots of hints buried in that data. Get a good vibration/mechanical tech to check things. I usually just Megger the motor to placate the customer then start looking for mechanical or process issues.

If you suspect a drive problem the diode test only checks diodes. You sort of have to rely on the VFD for shorted SCRs or IGBTs. But a final test is that if you have another drive or another motor, even a tiny but portable motor, try hooking up the motor to a different drive or the VFD to another motor. We keep a few rental VFDs which makes it easy for us. I can also check them with an oscilloscope but again that’s typically outside of the budget.
Thank you so much. That was is a lot of valuable information. Now of-course the VFD is not tripping. We still have to investigate.

I need to take an actual measurement. However, I believe the motor is about 20 feet from the VFD. The VFD-motor cables there is 9 total 3 for each phase. There is three conduits from the VFD-motor. Each with three non-grounded conductors. Then there is three bare stranded equipment grounding conductors in each conduit. They all terminate in the motor junction box and at the VFD. The 9 non-grounded motor cables are stranded conductors. I am not sure the wire type and size. I will have to follow up with that.

I haven’t received the motor testing results yet. However, I know they used a Baker set? We are having a meeting next week to make a game plan to troubleshoot.
 

·
Chief Flunky
Field Service Engineer
Joined
·
2,005 Posts
Thank you so much. That was is a lot of valuable information. Now of-course the VFD is not tripping. We still have to investigate.

I need to take an actual measurement. However, I believe the motor is about 20 feet from the VFD. The VFD-motor cables there is 9 total 3 for each phase. There is three conduits from the VFD-motor. Each with three non-grounded conductors. Then there is three bare stranded equipment grounding conductors in each conduit. They all terminate in the motor junction box and at the VFD. The 9 non-grounded motor cables are stranded conductors. I am not sure the wire type and size. I will have to follow up with that.

I haven’t received the motor testing results yet. However, I know they used a Baker set? We are having a meeting next week to make a game plan to troubleshoot.
Unlikely the cabling is damaged but worth a shot. When you get up over 100 feet (30 meters) is when you will often have trouble.

Baker testers use the surge test. They can do hi pot but it’s a waste of time at best and can potentially damage motors. If the motor passes and cabling passes then it’s either a power or drive problem once you rule those out.
 

·
Registered
Master Electrician, Industrial and Residential work
Joined
·
52 Posts
Discussion Starter · #27 ·
How about a real simple question?
I've come across this issue several times over the years.
So simple know one believes that was the problem.
Have you checked the acceleration and deceleration settings on the drive?
I've experienced trips during shutdown on a chilled water pump.
If the deceleration is too low the motor may overload trying to slow the flow of liquid.
Is the drive programmed to ramp to a stop, or coast to a stop?
A typical centrifugal pump, without a VFD, will coast to a stop.
I would have to check. It does state to check the accel and decel rates. However, I believe we have checked these settings before and matched them to our other two drives.
 

·
Registered
Master Electrician, Industrial and Residential work
Joined
·
52 Posts
Discussion Starter · #28 ·
What is the plant voltage on weekends????
You state it happens on weekends, could the taps be set to high for the plant and when POCO has no load or plant has no load Voltage goes up. I have seen line voltage raise 20+ volts on weekends.

Cowoby
I'll pull a trend on our SCADA system. If I notice anything odd I will let you know.
 

·
Registered
Master Electrician, Industrial and Residential work
Joined
·
52 Posts
Discussion Starter · #29 ·
I can't download that manual without registering and I don't want to because I'm not willing to give up my personal data to learn about it. We have no distributors selling that line of drives around here, so I never have to deal with it..

Is that OI as in O for Over and (capital) I for current? Or is that O for Over and (lower case) l for load?

Over Current may mean something in your drive is detecting a rapid increase in current, maybe too rapid for it to be picked up by the display. Drives often have hardware overrides for things like this that protect the components faster than the microprocessor can calculate things. Usually the display will tell you it is a Short Circuit, but again, I don't know this drive.

If so, then it could be failing motor lead insulation, capacitive coupling in the cables, water in the conduit (which changes the capacitive coupling issue and/or exacerbates leaking insulation), any number of things that create an intermittent high current situation.
The OI ac in the manual is a instantaneous output over current detected.

Those are good points.
 

·
Registered
Master Electrician, Industrial and Residential work
Joined
·
52 Posts
Discussion Starter · #30 ·
Unlikely the cabling is damaged but worth a shot. When you get up over 100 feet (30 meters) is when you will often have trouble.

Baker testers use the surge test. They can do hi pot but it’s a waste of time at best and can potentially damage motors. If the motor passes and cabling passes then it’s either a power or drive problem once you rule those out.
Unlikely the cabling is damaged but worth a shot. When you get up over 100 feet (30 meters) is when you will often have trouble.

Baker testers use the surge test. They can do hi pot but it’s a waste of time at best and can potentially damage motors. If the motor passes and cabling passes then it’s either a power or drive problem once you rule those out.
We are going to test the cable again. To hopefully rule it out. Then as you said we can focus on the power or the drive side.
 

·
Chief Flunky
Field Service Engineer
Joined
·
2,005 Posts
I would have to check. It does state to check the accel and decel rates. However, I believe we have checked these settings before and matched them to our other two drives.
Acceleration time issues usually cause overcurrent or overload...Basically speeding up too fast. Deceleration usually causes bus overvoltages. With any VFD issue you have symptoms and causes and it's not a 1 to 1 relation. So you pick one cause and try to verify it. If for instance you think it's a motor issue you check the motor. If that's not it try something else. Some issues are process of elimination...you suspect something but you prove it by disproving everything else. For instance you cant directly test Ights buried in an IPM or MPM.

Sent from my SM-T350 using Tapatalk
 

·
Band Member
DIYer Extrodinaire
Joined
·
7,402 Posts
wow, I did not know that
You have 5 posts. Didn't say anything in all of them. I'd say you don't know anything.

Waiting for a link now, spammer.
 
21 - 32 of 32 Posts
Top