[RE-wrenches] To Megger or not to Megger

[Joel Davidson]

Hello Matt,
All UL modules or panels with accessible metal parts are rquired to be production line factory dielectric voltage-withstand tested for 1 minute without electrical breakdown at 2 times maximum system dc voltage plus 1,000 volts (UL 1703, Section 43).
Joel Davidson
  ----- Original Message ----- 
  From: Matt Lafferty 
  To: 'RE-wrenches' 
  Sent: Tuesday, April 28, 2009 12:48 PM
  Subject: Re: [RE-wrenches] To Megger or not to Megger


  Hi Bill,

  I so totally agree with your comment about the price of these being a non-issue, all things considered. As long as we're getting the features we need in the sub-$750 range, that is. When it comes down to it, this is meaningful safety and reliability stuff. No slam to the Bolinas crew, but the SunEye is a lettle less important than that. Unless, of course, you shade a certain type of module and the bypass diode fails and the thing ignites.... Then you could make a case for the SunEye being prudent for Disqualification of location, which would mean there wouldn't be a need for the megger on that non-job.

  Perhaps you could address a couple of aspects that I have questioned for a long time.

  First of all, is it or is it not true that megger testing an array at 1kV may result in flaky-bias diodes (my term) if polarity of test leads is incorrect relative the array and/or the array wiring is grounded/not grounded during the test? (The root of my policy to NOT megger arrays unless I have specific procedure provided by the mfr... Me no wanna get in a "you meggered it so you don't have a warranty" conversation if ya know of what I speak.)

  Which of the mainstream module manufacturers are publishing a procedure for megger testing their arrays for use by installers?

  Are these being published for wet or dry testing? Both?

  Which of these manufacturers are supplying this documentation in a publicly accessible manner? (Asked quite a few sales reps over the years and only 1 ever produced a document... I have been able to get some from engineers in the factory, but this was generally when I had a big stick to pound on somebody's desk.) 

  Why don't most installers and distributors know this documentation exists or understand the procedure at this point?

  Yeah, it's one of those thorny issues with me. Maybe we can get our heads together to steamroll a solution to the above. Happy to take the subject up off-list if you like.

  On to today's question regarding resolution.... 

  When you are measuring an array, why is the resolution so critical?
  What value or value range do you believe constitutes a Pass on a quality Csi product?
  How about Asi / CdTe with crappy glass?
  How about Asi with Tefcel?
  Wet? Dry?
  Does your answer change relative the capacity of the array portion under test?
  Does the answer change with temperature &/or relative humidity?
  Does the answer change relative test voltage compared to array VOC?

  What I'm getting at is, unless the minimum "Pass" value is a very low number for the test setup, say 200K - 2M, or has a very narrow "Pass" range, a super-fine resolution shouldn't be that critical. On the few occasions back in the day when I actually meggered arrays, I used an analog megger. Top end range @ 1kV was 250M. My recollection is that these things pulled 15M or better, but I can't really testify to that... Fuzzy memory on the actual values. I'd have to dig through some deep stacks of paperwork to find out if I have any archives of the tests, and I'm not gonna do that for this post. Had to submit the test results to SMUD (can't find them) and the mfrs. Nobody ever said, "Hey. Close your other eye and tell me what it says." When I asked what values would be appropriate, all I ever got was, "Just like you got here. These are good." Never could nail them down and, frankly, never got the idea that they were all-to-happy having installers owning meggers.

  I'm with you on testing the array and module interconnects for leakage. Really. But, if the arrays rings out as "faulted" when it's really "OK", why bother? What are you looking for? If the resistance is that low, maybe we should be using a "Kohmer" for the arrays and a Megger for the wiring & equpment bussing. If you're getting such a low resistance, aren't you picking it up with a "Static" test... Would the Earth Bond Resistance function of the 1507 & 1503 be sufficient for these situations?  http://us.fluke.com/usen/products/specifications.htm?cs_id=35391(FlukeProducts)&category=HMA(FlukeProducts)  This is basically a voltage shot like a megger, but measures resistance in the <20k range. Resolution very nice there! This tester has a gap between 20k & 100k (0.1M). 

  From a "reasonable resolution" perspective, I think the gap is more like 20k to 2M. Anything over 2M, the 0.1M resolution should be fine. If the array is that close to failing, fail it and tell the mfr to start making some decent glass. They might understand it better this way:  您吸

  Dunno. I think it comes down to figuring out what ranges equal "Pass" for the application and going from there. Will look forward to what you find in your side-by-sides later this week. Please do post back. With a basic description of the glass-type, mounting, wire-management, test results and findings, if you can.

  I'm gonna hang onto my 1520... If anybody wants to see what one looks like with a 100% Battery Level, let me know... I'll hold it up in front of my webcam for ya!

  No Ground Faults!

  Matt Lafferty

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.re-wrenches.org/pipermail/re-wrenches-re-wrenches.org/attachments/20090428/7f1820ed/attachment-0004.html>