How to do a proper insulation test on an electric motor

[antisa555]

I want to check the insulation of the 3-phase 400VAC asynchronous squirrel cage motor.

Motor's stator windings are connected in delta.

I've read in some textbooks it is necessary to open motor's junction box and disconnect terminals and remove jumper bars (which are used to establish delta or star connection) and only then perform insulation test between:
a) each winding and earth and
b) each of the windings.

I think this might not be necessary and that it is enough to test the insulation from the motor's contactor output between any of the phases and earth.

This will let us know if there is an insulation problem between windings and earth.

However, testing this way doesn't allow us to measure the insulation resistance between each of the windings.

Questions:

1. How likely is it that there is a low insulation reading between windings and that this reading will not manifest itself when performing insulation test on the motor's contactor output between any of the phases and earth?

2. I am using Megger MIT330 and testing 3-phase asynchronous squirrel cage motor which operates at 400VAC.
Which voltage should I apply when testing and for how long should the voltage be applied?

3. If I apply 1000V for 20s on motor's contactor output between any of the phases and earth, I notice an increase in resistance from 50MΩ up to 250MΩ during the first 20 seconds and then it remains steady around 250MΩ. Why does this happen?

4. What is the acceptable insulation resistance when testing resistance on motor's contactor output between any of the phases and earth for the 400VAC motor?

5. Why is this considered acceptable insulation resistance?

 

[briwire]

No need to disconnect the links, just the supply cables, tighten terminals once removed. Set your megger on 500v, and one lead on the frame of the motor and take reading between that and each of the 3 now vacant terminals.
As a rule of thumb, anything above 1Mega ohm is ok.

 

[Fazz711]

The BS for IR test is to as you say remove links, supply and test each phase.
To earth and between phase.
A between phase fault won't necessarily show to earth but will still be U/S.
As stated anything above 1 meg is just about acceptable.
The test should be done for 1 minute and this will allow the reading to settle.
Some larger motors and DC motors you are effectively charging the coils so this is why the IR reading starts low and goes better the longer it is tested.
Hence the reason for testing for 1 minute

 

[alexcarnes]

The BS for IR test is to as you say remove links, supply and test each phase.
To earth and between phase.
A between phase fault won't necessarily show to earth but will still be U/S.
As stated anything above 1 meg is just about acceptable.
The test should be done for 1 minute and this will allow the reading to settle.
Some larger motors and DC motors you are effectively charging the coils so this is why the IR reading starts low and goes better the longer it is tested.
Hence the reason for testing for 1 minute

And yet only one of the three devices I own for insulation resistance testing has a one minute timer!

 

[Fazz711]

And yet only one of the three devices I own for insulation resistance testing has a one minute timer!

Non of the ones I have used have a timer.
You have to hold finger on button untill you are done.
Much better than when I started when you had wind up meger that you had to wind for a minute.
Felt like hand was going to fall off when you had a few to do LoL

 

[Moley]

I'm pretty sure out of all the ones we have it's only the Chauvin Arnoux 5kV jobbies that have a timer. All the others are usually a case of 'yup, that's a good enough reading' type test.

 

[alexcarnes]

Non of the ones I have used have a timer.
You have to hold finger on button untill you are done.
Much better than when I started when you had wind up meger that you had to wind for a minute.
Felt like hand was going to fall off when you had a few to do LoL

Yeah, just find it quite incredible that anyone would make an insulation resistance tester that won't perform a standard one minute test with you having to keep your finger on the button!

 

[alexcarnes]

I'm pretty sure out of all the ones we have it's only the Chauvin Arnoux 5kV jobbies that have a timer. All the others are usually a case of 'yup, that's a good enough reading' type test.

I guess that's why a lot of MFTs don't have a timer - once it's above 1MR that's it, job's a good 'un.

 

[Marvo]

For standard electrical testing and fault finding normal insulation tests using the next voltage range up from the supply voltage for a few seconds should be fine. Personally I work on if results are more than 5 Mohms it's not requiring any further investigation. Often the regs only require 1 Mohm but the regs are the absolute minimum safety requirement and with modern motors there's no reason they should measure less than 5 Megs IR to earth.

Using higher voltages and longer test times gets into the realms of dielectric testing or what we udes to call Hi-Pot tests. These types of tests are used to predict time to failure rather than determine safety..

 

[davesparks]

Yeah, just find it quite incredible that anyone would make an insulation resistance tester that won't perform a standard one minute test with you having to keep your finger on the button!

Most testers have some means of latching it on so you don't have to hold the button.

 

[Fazz711]

When working in the motor repairs industry you have to fill in test sheets and keep them for so long.
Doing that will do is ok for most things but not for QA record keeping.
A lot of large three phase motors and larger DC motors will start off a bit low on the IR then constantly get better as the coils charging up.