Plenum Headaches
–Well let’s see where did I leave off? Well let’s go back to when I broke that #10 tap on the bolt circle for the vacuum cleaner attachment fitting: http://www.flickr.com/photos/steamboat_ed/4818676077/in/set-72157608281318568/ Upon closer examination it looks like I inadvertantly grabbed a plus-two tap; i.e. one that was .002” oversized, for tapping holes in aluminum that would later be anodozed. This meant all of the holes tapped with it in steel were sufficiently over nominal size to make the bolts very loose and wobbly, particularly in the relatively thin 1/8” steel wall of the plenum. So steel broke the tap and oversize tapped holes rendered the bolts incapable of holding back air. So my beautifully machined aluminum plug, complete with O-ring groove did a dandy job of sealing the large 1” hole in the plenum but the bolt holes themselves leaked like crazy! http://www.flickr.com/photos/steamboat_ed/4853310541/in/set-72157608281318568/ Trying to stop the leaks by dogging down the #10 screws extra tight resulted in two stripped holes; i.e. I was now in a real pickle. At this point my options were: Weld the threaded holes closed, then redrill and retap with the correct diameter tap and hope I’d get a tight seal or scrap the plenum altogether and start over. Neither of these struck me as terribly good ideas so I took a few hours off and noodled on it some more.
–I decided to find a way to increase the thread depth somehow. This would give the screw more bite and it would mean air would have to leak thru a longer, twistier path to escape. I thought about using nutserts but these don’t sit flush and I would have had to use gobs of sealing compound to make the connection air-tight. Another idea was to weld up the bad holes, ‘twist’ the hole pattern half the angle between holes and use a Formdrill to friction-drill the steel, stretching it so that a form tap would produce a longer thread. But I distrust the integrity and shallow thread depth produced by a form tap and so that was out. In the end I opted for Plan C: I drilled out every other hole to 3/8” dia and chamfered the edges a bit. Then I switched on the lathe and, using some 3/8” dia steel rod, I made six: 3/8” long threaded inserts. Next I made a ‘quicky’ bridge to hold each insert in the correct location, inside a drilled hole but flush with the surrounding steel and squarely so that fasteners wouldn’t go in crooked: http://www.flickr.com/photos/steamboat_ed/4853355009/in/set-72157608281318568/ This whole scheme worked very well and I think I might do it again intentionally, so to speak, should I ever again have need to attach something to thin wall tubing. Welding was, relatively speaking, a breeze and by the third one of the six I had it down. The mating parts had chamfers that permitted good penetration with a fusion weld and I used almost zero filler rod to connect the dots: http://www.flickr.com/photos/steamboat_ed/4853310601/in/set-72157608281318568/ End result: very little grinding needed and, after chasing the threads one last time with a plug tap and the right lubricant (Cool Tool II) I was able to snug up the six #10-32 socket heads well enough to get a good airtight seal using a bit of paper gasket material instead of the O-ring. As for the 6 remaining not-so-great bolt holes I added phenolic washers, the kind used for air logic fittings, to each bolt hole in the plug before gently snugging up each of these screws and that seems to have achieved an adequate seal.
–Thinking I might want to attach stuff later and not wanting to perforate the plenum’s integrity when the time came, I made six 1/4”-28 threaded bushings. Two were welded on top of the plenum, on centerline and at each end, to provide anchor points for a cable trunk which will run parallel to the plenum, along one side, to protect solenoid wiring. I welded the remaining four on the sides, near the bottom and close to the ends. These will provide attachment points for legs or whatever, to provide a bit of lateral stability as the contraption will be plenty top-heavy once all of the whistles and solenoid valves are fastened in place.
–Once all was set to rights I did a pressure test. Sure enough I found two massive leaks where for one reason or another I failed to complete a weld all the way around a 3/8” pipe fitting. Once I’d welded these two correctly test results were much better although not stunning. The big aluminum plug held great even at 60psi (3 times estimated working pressure). But it turns out one of the twelve fittings I used to plug the whistle attachment bushings leaked, despite being threaded in with two turns of good gas-rated Teflon tape. Well there’s always something, eh? Anyway once that was sorted the plenum seemed ready for paint: woohoo!
–I had intended to paint the plenum white but I couldn’t find the damned paint can anywhere. Bottom line: it’s blue, because it was handy. Looks pretty snazzy, I gotta admit http://www.flickr.com/photos/steamboat_ed/4853310567/in/set-72157608281318568/
–So today the moment came when I could finally test one of the cheapy solenoid valves I got from ebay. I mounted one valve with a whistle attached, I dropped shop air pressure down to 5psi, rounded up enough AA batteries to make a 12-v supply and twisted wires together. To my delight I actually got a note out of the thing! I increased air pressure to 10psi and this produced a much brighter tone. Jumping up to 20psi (what it used to take to get any sound at all out of calliope #4) the sound went all to hell with several overtones in the mix so it’s definite: the new valves let through a much greater air flow than the old sprinkler valves.
–There is an oddity that happens when the valve first opens: there’s a very loud pop and wheeze before the bell sounds and I’m guessing this has something to do with me supplying air to the (short but fat) plenum from a long skinny source, i.e. two lengths of coiled air hose at the end of a 25-ft string of overhead plumbing. Once the first note has wheezed into life though, subsequent rapid activations of the solenoid valve seem to be ‘normal’. Something to do with the elastic properties of gases methinks. I’m hoping the whole system is good to go, but not entirely trusting my own gut feeleings as to the nature of the wheeze I hope to get a second opinion later today from Rich Gibson as to the readiness of the system, as it stands, for the remaining solenoid valves prior to ordering same.