On Friday 04 September 2015 06:59:58 Erik Christiansen wrote: > On 03.09.15 15:28, Gene Heskett wrote: > > Mid afternoon PS: The LM317T is in and working, but that 1.75x3.5" > > bit of 1/8" alu for a heat sink is running just under 150F after > > about 20 minutes running. The draw of a 3rd fan will also make it > > run even hotter, so there is a point of vanishing returns to such > > thoughts. I may have to add a 3rd fan, but have no clue what to hang > > it from. It will no doubt get more air with the lid closed, spillage > > from the end of the X motor driver. > > > > The theory was when I built that box was to have enough air moving > > inside it to carry the heat to the outside walls for normal radiant > > cooling. > > If you add > Rth:LM317junction-case-insulation-heatsink-air-enclosure-air, it's a > big pile of K/W, even with an internal fan to help with the first leap > across air. A big finned HS for the device, with the fan blowing along > the fins, is a necessity if there are lots of watts to shift. That > small scrap of Al is too skimpy if it's running quite hot. > > After that, a fan on the outside would further boost heat transfer, > especially if a few lengths of Al channel or angle from the junkbox > were screwed on inside and out, aligned with airflow, and gooped > underneath with a smear of thermal compound. > > ************** > But mounting the (TO3?) regulator on a block of Al screwed and gooped > to the inside of the Al enclosure wall, would be nearly as good as a > heat-pipe, eliminating the less effective huffing and puffing with a > fan. If a silicone insulation sheet, twenty times the size of the one > now under the T03 case, insulates the Al block instead, Rth for that > step drops by 95%. OK, that calls for nylon bolts or bolt insulation - > both doable. And if a bit of milling of the block is called for ... > ************** > > Alternatively, blowing cool air into the enclosure, through a fine > dust filter would also exclude swarf - just not fine spray mist. (And > would wreck the current concept, I fully admit. ;-) > > A light spray of flat black paint on the outside will boost radiation > a bit, while not helping convection at all. I'd still plaster vertical > fins on the outside to give some chance of lowering Rth without an > external fan. > > If there's no room for an adequate heatsink in the LM317 locality,
Its fairly crowded. > and > moving it to the enclosure wall isn't an option, The T0-220's tab is hot. > then a heat-pipe will > pipe the heat half a foot and through the enclosure wall, for an Rth > of as little as 30 millidegrees/W. (They're out there for not many > dollars a piece, but watch those from RS; the one I looked at used > water as working fluid, and needed 70°C to get up to speed.) A > complete finned heatsink & double heatpipe assembly from a PC is cheap > on fleabay. Whack that through the case wall, and you're done. Or just wait till the buck regulators get here, whack the mounting block of white ash back off the wall and switch it out for the buck regulator. 1.5 amps at 90%+ efficiency. Unk is if any mounting holes for those can be grounded to the case. Switching both out will lop off at least 20 watts of heat in there. With the loss of a local rat shack, standoffs are in short supply unless house made. I do have some teflon scrap that could be machined for such use if push came to shove. But I've not found anything that can glue that stuff in 40 years. But cutting off a strip 1/4" wide of the 1/2" stock I have, and milling a slot for the edge of a board, and drilling & tapping it for 4-40 & bolting it to the case that way is possible. But last night after I traced the encoder cabling and got that to work, its obvious that I need to switch the solid wire cable to the spindle control out for a stranded cable, its too easily broken off. The signals are going down the cable, but now are not getting to the servo amp to run the spindle. I am also convinced I'll need another lookalike motor as I believe this one has damaged bearings and a bent shaft after one of those stop from 700 rpm in 5 degrees of spindle turn stops. If I was smart, I'd junk the idea of using a motor with a screw on pulley/fan/flywheel & get one with a normal shaft, put a high speed brushless fan so it had cooling at any rpms & put an XL belt and sprockets to spin the jackshaft. Or a narrower, it might even slip before it destroys the motor, polygroove belt. Unfortunately, anything on ebay with a plain shaft is 56 frame, and $200+ bucks. But I did just now find another ICON motor, possibly a slightly weaker one, and did the buy it now dance. Saw the mount brackets off to match what I did to this motor and it should interchange in the same mount. Rated 3/4 HP CCS at 95 volts, 130 IACS & 1.5 HP for treadmill use. And I'll have to glue the flywheel onto this one too. Would it do to drill & tap about a 10-32 in the gap line between the flywheel and the shaft to lock the two together for bi-dir use? That would, if the screw was fairly hard, make a key to assure it never unscrewed itself and yet the screw could be removed in the event I had to replace the BBLB bearings in it with decent ones. The trick will be in getting it mounted on the mill rigidly enough to pull that off. That will probably lead to a new definition of shade tree machining. ;-) And I will take a picture or 3. I hope the GO704 has enough overhead room. A tap, mounted in the drill chuck uses up half its Z travel. :( I'll have to measure the shank on my taps, but have not yet found an R8 collet that will take the odd sized shanks on the smaller taps. Do they make them odd sizes to prevent that highly usable fit? Now, I'd better get to replacing that easily broken cable, I've got work to do with it even if the motor is grumbling. Thanks Erik Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) Genes Web page <http://geneslinuxbox.net:6309/gene> ------------------------------------------------------------------------------ _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
