> At the risk of re-opening this discussion. Besides coke cans > there is another readily available material that could be used. Less of that lying in the streets of Port au Prince than Coke' cans, I suspect :-) But: > For the price of two or three cokes you can buy 10 feet of > tubing used for built in household vacuum cleaners. In Canada > and the US available in most big box hardware stores. > > This tubing is thin walled light and easy to work with. I assume that is a plastic material and not metal, yes? Regardless, any idea of what the material is? > (I use it as > replacement tubes on bird feeders) It lasts quite well outside. > Some of the bird feeders are 7 or 8 years old Sounds good. What location is that in? Any plastic that maintains mechanical integrity outdoors for 8 years is respectable. In NZ the UV is generally more intense than anywhere more 'normal'. We were most surprised to find that across the Southern US (Ca, Ar, ?Mon) and all Europe and UK at the height of a hot summer we could literally stay out in the sun all day long for days on end with no sun screen and survive OK. In NZ you'd get maybe 2 to 4 hours and be severely affected. "Even" in Africa ABS products have a lifetime of "a few years" unless great care is taken with UV stabilisation. Even visible light is a significant issue (hence use of eg HALS stabilisers). So 8 years is a very benign climate, or not ABS or excellent stabilisation. Polycarbonate can last 20 years with due care with stabilisation. ___________ > A few months ago I attended a seminar that Microchip was > running on reference designs for putting power back into the > grid. Three things came out of that seminar. > > 1) Intermediate storage batteries are a bad idea. They don't > =A0 =A0last very long. Depends very much on chemistry, construction (some chemistries), cycle depth and cell temperatures and magic stuff. 1/Nth of depth in cycling generally gives more than N times cycle extension= .. Simplistically: LiIon is good for 300-500 full cycles plus has a calendar limit but responds quite well to shaving ends off cycle. Lead acid construction matters muchly plus magic plate constructions and formulations help heaps (deep cycle LA magic is much practiced) and temperature matters. Basic LA is very very poor for cycles (200?) but can be extended to many thousands with capacity reduction and magic. ? LiFePO4 is good for 2000 with minimal care and more with suitable effort. NimH is very variable. Maybe 300-600 std but out to 2000 with end of cycle shaving of say 20% of capacity and temperature matters much. NiCd lower than NimH notionally if abused but with modest care seems to go forever and ever and ever. NiFe is unseen almost nowadays (very low energy density) but is said and seems (N=3D1) to take brutal treatment indefinitely. Auld lang syne we had wooden crated metal bused "prismatic" NiFe with screwed on joining busbars that had been rescued from a rubbish dump and that seemed to withstand continual school-boy over discharge crash discharge and random charging excesses forever. But yes, putting a battery in the middle is bad if it can be avoided. 90% in/out transfer efficiency would be commendable and worse to much worse is easily achieved. Direct solar to end target conversion is good if viable. (My special interest is lighting where direct use tends to "not be favoured" :-). Solar to eg cellphone charging direct is of some interest but even that often tends to be a night time application.) > 2) Even small amounts of power can be practically fed into > =A0 =A0the grid even through a plug in the wall. Yes. No reason why not. Main issue at micro power levels is rate of return. 1 Watt continuous is worth about $2 at typical grid electricity charges. If grid is localised and/or high cost per kWh it may be economically justif= ied. > 3) The major design problem is a control loop that phase > =A0 =A0 locks to the line. The reference designers we mostly > =A0 =A0PIC32 based. I know people here whose specialty is energy recovery back o mains in systems often in the sub kW range. Russell McMahon --=20 http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .