Mark, The pros and cons of various dielectrics (e.g. plastic films) have been discussed in the past on the list. I have attached a good sample. There are many others - search the archives. The bottom line is, go with polyethylene or polypropylene. Mylar is OK for DC and line frequency a.c., but at TC resonant frequencies, mylar is too lossy and will overheat and is likely to lead to premature cap failure. --Steve ---------- > From: Tesla List <tesla-at-pupman-dot-com> > To: tesla-at-pupman-dot-com > Subject: Plastic Capacitors > Date: Thursday, August 27, 1998 5:29 PM > > > From: Mark Finnis <mefinnis-at-medicine.adelaide.edu.au> > > Have built my first two caps with rolled polyethylene/alum flashing and > have been quite happy. > > Wandering through wire shop the other day buying some 1mm (19SWG) for my > new former, when I noticed some rolled sheets on the shelf.... > > This stuff was "slot insulation", various sizes & components, eg. > > 5/5/5 mil each of nomex/mylar/nomex > > nomex is actually polythene saturated & bonded each side > > effectively 7 layers bonded together, stable to 250 degrees C !!! > > rated at 17 kVAC !!! > > This would match stated figure for mylar from list at 7,500 kV / mil > puncture VDC. > > Thought ...... this would make a *great* cap ....... alas, $30 / m (very > dear). > > Looking at the plastics chart for puncture V versus dielectric K, mylar > looks very attractive: > > Polyethylene Terphthalate (Mylar) K = 3.0 - 3.1 Vp = 7500 V > > Available in rolls at about 2x cost of LDPE. But remember we get 50% > greater K and only have to use a couple of layers (5 mil), so gain even > more in the total C stakes. Actually doing the sums, this is barely more > expensive than LDPE, and is much higher grade / purity. > > OK, so why don't we use this stuff ???? > > Anyone tried it ?? > > I am contemplating when my current caps die (I am a realist, really) to > re-roll them with mylar. > > Basic assumptions: > > 1. For LDPE we use K = 2.0, not 2.2 to account for losses via "home-made" > OK, lets take mylar K = 2.7 (less 10%) > > 2. Reported puncture V for LDPE varies from 400-1200 V/mil. We use lowest > figure as "experience" shows TC use hard on dielectric. > OK, lets take Vp for mylar = 5000V > > Any comments welcomed ;-) > > > > > ___________________________________________________________ > > Mark Finnis Hm: 61 8 82895205 > Staff Specialist Wk: 61 8 82224000 > Intensive Care Unit Fax: 61 8 82236340 > Royal Adelaide Hospital Mbl: 041 23242268 > www.health.adelaide.edu.au/icu > ___________________________________________________________ > >
loss factors.eml (Internet E-Mail Message)