Megger vs Hipot

[Metaltomato]

I am working to create a manufacturing process to produce small brushless ac motor/generators.

Recently I equipped the final testing station with a GW Instek 4 in 1 electrical safety tester. With the help of the manufacturer and several other fine folks, I set up the Insulation Resist and AC Withstand stand tests and started making pre-production units.

I do not know the tests that well, and only have UL, IEEE, and other standards to go by. My question is this;

Why would my motor pass the Megger and impedance test, but fail the AC Withstand test? Wouldn't the megger and impedance tests show the fault?

Thank you in advance for any input you might have!

 

[waross]

The peak Hi-pot AC voltage may be higher than the Megger DC voltage.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[VLFit]

AC is not DC and DC is not AC. The stress imparted by one or the other is not at all the same. The failure mechanisms of insulation are very different under AC or DC stress. For example, DC voltage does not instigate partial discharge and cause the defect to fault like an AC over voltage test will. Using DC, how do you know the coil is good or bad. What megohm reading, or if DC hipoting what leakage current reading, is the magic number to tell you good vs. bad. The tests are very different. There are DC and AC tests that are valid but you must understand the differences in the physics of each. If you can afford to replace the winding if its bad, just AC over-voltage test it. If it holds, great. If it fails due to some defect that can not hold the test voltage, great too. You saved yourself an in service failure. I would suggest you consult the EASA website and start to gather information. They can help you greatly.

http://www.easa.com

Good Luck

 

[Metaltomato]

Thanks for the link.

As you may already know, if you are a neophyte, it is a TALL order to make the leap from gathering information to effectively testing.

I have spent the better part of a year trying to get my head around these two tests and what they mean in the manufacturing environment.

Of course the standards are there, and very helpful for setting up a pass/fail according to them. But how long it has taken me to understand that those standards are minimums and may not tell a manufacturer a thing.

Right now, I have found the right gear, know some of the the right test settings, and have been doing what I can to engage in conversations that will try their patience, but help me learn, and for those folks out there that see what I am doing, I am truly grateful. One of the biggest obstacles has been guys like you who would love to talk about the subject in detail, but understand the liability issues involved. I have talked to many companies to try and hire a motor manufacturing pro to come in and teach me, but the cost is always prohibitive.

I think the long and the short of it is that I simply lack experience to understand what the tests are telling me, and I will just have to keep asking questions.

Im meggering 500v (dont know if its ac or dc) with a lo-set of 20 megohms and hipotting according to ul (2x DUT rating + 1KvAC); So my hipot is set to 1.8KvAC with a hi-set of .500 milliamps for 1 second.

This leads me to another issue. Because I will hipot every unit that comes off the line, how much hipot is too much? We are now thinking because we cant find the root failure (passed megger, failed hipot), maybe we should do a low voltage hipot in the middle of the line, say before trickling, but we are afraid from the stuff we've read that the hipot is a destructive test.

 

[VLFit]

sorry for no reply. i ran out of time today. I'll get back to you tomorrow but it would help if you supplied some data, like motor size by watts or hp, voltage ratings, current input, the capacitance of the load tested, the dc leakage currents, what percentage are failing AC withstand, why the .5 mA trip setting, and anything else that may help. have you looked up other motor makers that build similar products? many have test and maintenance guidelines published or search on something like "AC brushless motor test" What is your go/no-go megohm criteria when you do the IR test?

catch you tomorrow.

 

[Metaltomato]

Thanks for the attention VLFit, please let me try to give you particulars. My comments in red;

like motor size by watts or hp, as a motor; 2.6hp, as a genny; 2kW (7.6 max)
voltage ratings200V
current input 7.5 Amps
the capacitance of the load tested 1kOhm 1%
the dc leakage currents ?
what percentage are failing AC withstand this was the first unit to make it all the way to the ACW
why the .5 mA trip setting an electrical engineer told me so...I asked him what the leakage expectations should be for a unit of this kind and thats the answer he gave me. I was not able to glean leakage expectations from the standards

and anything else that may help. have you looked up other motor makers that build similar products? I have looked, but our unit is unique. any suggestions in that direction would be helpful. many have test and maintenance guidelines published or search on something like "AC brushless motor test"

What is your go/no-go megohm criteria when you do the IR test? 20megohms. I think I got this from reading the book from megger on the subject. It's a pretty low value. My understanding of the IR test is that in many ways it is predictive rather than go/no go, and that an operator learns to apply the readings more intuitively over time. Perhaps this is correct, perhaps not.

The long and the short of it is that I was hired as an assembler and have become the manufacturing designer. This is not to say that I am completely hopeless, however, my electronics background is minimal and this particular subject has been one of the greatest challenges to me along with all of the purely mechanical and logistic puzzles. As I have mentioned, we have looked into all manner of solutions, however, the further into this I get, the more I am coming to value the fact that for what I am doing, there is no manual as I am the one writing it.

So far we have quotes from $4-6K for a "consultant" to come out and take a look. The best information on testing has come from GW Instek, UL and IEEE standards, Megger, and some visits to a local motor repair shop.

I will post as an attachment a hand out from Thor Power with some more details, dimensions, and specifications of the unit. In the meantime, thank you again.

 

[Metaltomato]

Im attaching a couple of docs; the first is a spec sheet, the second is a drawing of the unit.

 

[Metaltomato]

Here is a drawing of the unit

 

[Metaltomato]

Please forgive the multiple posts. I dont see a way to edit the posts, so the first post with attachments has the same doc as a docx and pdf....the pdf is has an old schematic...again, sorry for the confusion and redundancy.

 

[VLFit]

Good Morning,

many people can tell you things in this forum but to get the most and best information you should buy, or download for free if you become an Associate Member, the many sources of info from EASA.

http://www.easa.com.

Go to their site and you will see their latest versions of their test books for all types of motors. There is no better source for this info than from EASA. One point I will mention is that the trip setting on your AC hipot may be too low. I asked you for the capacitance of the winding tested and you gave me a resistance. When AC high voltage testing, it is the capacitance of the load that determines the current draw during test. You need an AC source with enough kVA to be able to test the capacitance of your load. There should be no trip setting. It is a go/no-go test. The coil holds or fails - flashes over at the defect location.

have fun.

 

[Metaltomato]

wilco on EASA. I went to their site and took a look at the costs which are low. I will peruse the sight.

I understand what you are saying about capacitance, and I will look into it.

Thank you very much for your cogent response!

 

[rhatcher]

I agree with VLFit, the AC hipot is a withstand test, meaning the insulation of the device under test either holds or it fails. If it fails, the insulation is punctured and the resulting flashover trips the hipot. By 'puncture', I mean that a hole is literally blown through the insulation, rendering it unserviceable.

The potential (no pun intended) to 'puncture' the insulation and render it unserviceable is the reason that an AC hipot is considered 'destructive'. However, the reasoning behind the test is that insulation that fails (is punctured) is not suitable for use and, since it could not (should not) be used anyway, there is no actual loss when it is rendered unserviceable (punctured).

In my experience, the hipot trip setting would be 5ma, not 0.5ma. However, if the expected insulation capacitance of the unit is known, a more precise (lower) setting can be selected as a means of quality control.

My question is; after failing the AC hipot, does the device still pass the DC meggar test? If so, is the measured DC insulation resistance lower than before the AC hipot test was conducted? If the answer to both of these questions is no, then the tripping of the AC hipot was not an indication of failure to withstand the AC test voltage and the trip setting is too low.