OK, OK, I give, I give!!! Thanks for all the insight. It seems that while using varnish to take advantage of 2.9.6 or 2.9.7 is theoretically possible, it is practically impossible. One would need an meticulously controlled production system with an equally proficient QA system to back it up, and even then it sounds like there may be agencies who don't like the idea.
Oh well, back I go, calipers in hand, to the uglier columns of Tables 3-6! This thread has illustrated once again one of the main benefits of this forum: the experience to pluck the practical reality from the maze of requirements. Thanks again, Jim Eichner Statpower Technologies Corporation [email protected] The opinions expressed are those of my invisible friend ---------- From: Peter@anetMHS (Peter Tarver){MHS:[email protected]} To: JEichner Subject: Re: Transformer requirement List-Post: [email protected] Date: Friday, July 12, 1996 12:43AM Reply to: RE>>Transformer requirements in IEC950 Jim Eichner wrote: >The one point of contention seems to be in what ways varnish impregnation is >helpful. I think there are three aspects to this question: >2. Varnish as a sealant per cl. 2.9.6 - As Mike Rains argues, a layer of >varnish surrounding the transformer windings creates a microenvironment that >is Pollution Degree 1, greatly reducing the creepages and in some cases >clearances as well. I note Mike said "vacuum impregnation", which is not >what we do at this time (we dip). I would think, though, that all we would >have to accomplish is an envelope, not the complete filling of internal >voids that vacuum impregnation is intended to achieve. Has anyone had >acceptance of this idea from an agency? There may be some case to claim vacuum impregnation can create a Pollution Deg ee 1 microenvironment for internal parts of a transformer out of an overall Po lution Degree 2 environment, but I suspect that a formal quality control progr m would need to be established at the factory to ensure continued compliance a d to get safety certification agencies to buy into the idea. Via my prior wor at UL, I have seen vacuum impregnation work well, giving good overall coverag in one sample, and work poorly with relatively large gaps in the impregnation in another, otherwise identical transformer, made using the same process, show ng at least some methods can provide inconsistent coverage. My experience is that dipping can be a very unreliable means of gaining the sa e Pollution Degree 1 considerations. The coating would have to be so thick as to act as encapsulation, rather than a sealant, and require lengthy drying tim s. The dipping resin may have to be more viscous than usually used for varnis dipping, implying a longer dip time (maybe higher resin temperature could red ce the time) to allow good adhesion and coverage, unless multiple dippings are used, with drying time in between. >3. Varnish as an encapsulant per cl. 2.9.7 - This clause mentions >impregnation specifically but not varnish, and gives inspection criteria in >addition to the test for sealed parts in 2.9.6. Again, I think this should >be valid, but I wonder how hard it is to eliminate all "significant" voids >in the sample, and what "significant" really means. Has anyone used this >idea and had an agency accept it? How did they judge the significant voids? I don't recall anyone envoking this subclause, while I was at UL. Their is a ost problem with cutting up inventory, both for type test evaluation and espec ally for follow-up factory testing. Since the requirements of 2.9.4 are used for this evaluation, I don't think a arnish would work well, but it might, if properly applied. Subclause 2.9.7 do s explicitly mention impregnation as a possible method, but I haven't yet seen a transformer with that much goop in it. The only encapsulated transformers I ve seen used a thermosetting resin, like epoxy, but that was during my househo d audio days. Another problem is method. How a transformer (especially a small one) can be ut to ensure the method doesn't miss a void or obliterate enough of a void tha the test result is inconclusive (requiring additional sample dissection) requ res careful study. Careful selection of the places to cut would help, if the slicing means has a very thin kerf. A diamond saw blade or an industrial lase could serve the purpose best, taking only the slightest amount of material in each pass. However, these represent no small investments in equipment and tra ning of personnel. "Significant" is a word that demands interpretation. The most obvious case is if a void completely exposes parts that don't meet the creepage or clearance r quirements in the void without the intervening material. Cases where a void d esn't expose these same parts is the toughest call to make. Basic and Operational insulation are evaluated by electric strength testing in 2.9.4, which may determine if a void is "significant," provided the parts aren t completely exposed, but I don't know for certain that this is the only appli able compliance criteria a safety certifier would use. But then, how could co sistency of the maximum gap size be demonstrated, if one is found, without cut ing open a large number of transformers? Since the through insulation requirements are applied, the 1 mm rule in Annex is useless for Reinforced and Supplementary insulation. Perhaps, for Reinfor ed and Supplementary insulation, the intent is that no continuous piece of the intervening insulation be thinner than 0.4 mm. Perhaps a void of 0.04 mm coul be tolerated. I don't know. The headaches the above could cause are reason enough to hope that such a meth d is not relied upon for compliance. >Regards, >Jim Eichner >Statpower Technologies Corporation >[email protected] >The opinions expressed are those of my invisible friend The above opinions are mine and do not necessarily reflect the opinions of Nor hern Telecom. Regards, Peter L. Tarver Northern Telecom [email protected]
