Yahoo Groups archive

Homebrew PCBs

Index last updated: 2026-04-28 23:05 UTC

Message

Re: [Homebrew_PCBs] Minimalist drill press

2005-11-11 by Kenneth Long

can you post photos? maybe on a web site? 


--- Mike Young <mikewhy@...> wrote:

> Well, the tiny little drill bits arrived this
> afternoon, reminding me that I 
> hadn't figured out yet how to put the RotoZip to
> work. I also wanted to do 
> this in a way that every man can do for himself in
> his garage, sans 
> Bridgeport. So, eschewing the heavy machinery, I
> headed out to Lowe's for 
> common hardware. The short story made long follows,
> but the end results are 
> cute little holes, down to 24 mil, through 1/2" MDF,
> and not a single broken 
> bit.
> 
> I cheated. The whole affair is unreproduceable
> unless you, like me until 
> this evening, happen to have a spare 6" linear rail
> sitting around unused in 
> your part box. (THK HSR-12RM, 150 mm long, single
> truck; likely useless for 
> anything and everything else, but perfect for just
> this and only this. I 
> sniped it for well under a sawbuck on eBay last
> year, apparently planning 
> ahead for this moment.) Anyway, its attributes are
> slippery smooth, play 
> free motion along the rail axis. They're usually
> used in pairs: two rails in 
> parallel to absorb lateral torques, and two or more
> trucks on each rail to 
> absorb longitudinal moments. The lone truck is more
> than enough for the 
> machining forces a tiny hair strand of spinning
> carbide can generate. 
> There's still the spindle's wind-up torque to
> consider; 0 to 30k rpm in 
> under a second produces a pretty hefty jerk when it
> starts up. I'm just not 
> overly concerned, since it runs steadystate rather
> than on-off when in use. 
> In short, the HSR-12 is a perfect solution to an
> otherwise sticky problem. 
> (Pun unintentional, but very appropriate.)
> 
> The structure is made from two chunks of 16 ga mild
> steel. I wanted .10" 
> 5051 aluminum sheet, but wouldn't you know it?
> Lowe's doesn't stock it. 
> You'll probably want someone to cut the steel for
> you; tin snips just won't 
> work. One piece about 8" x 8"; a second piece 3.5" x
> 6.5". I cheated, and 
> stopped to visit a friend who happened to have a
> foot shear and press brake. 
> Odd coincidence, eh? The alternative is a chopsaw if
> you have one, or an 
> abrasive disk in the tablesaw, if you have one of
> those. I saw 10" chopsaw 
> blades for under $5 there. It'll make an unholy mess
> of sparks, but can't be 
> too bad for the three or four straight cuts.
> 
> Bend the larger piece into an 8" long L, 2" x 6".
> This is the post for the 
> drill. More about that in a moment. Bending heavy
> gauge steel with the right 
> tools on hand is absolutely unremarkable. Without
> the tools, it's a project 
> in itself. I tried bending a scrap piece in a Record
> (metal) woodworker's 
> vise, beating the free leg down with a ball peen
> hammer. <grins! what fun!> 
> The first test turned out not too terrible, but it
> was a short piece, about 
> 4" long. Bending an 8" long piece was a different
> story. The best try was 
> with a 1/2" x 3" bar behind it and lots of muscle.
> It's do-able with a bit 
> of care. The 2" leg is the important one to protect;
> clamp that in the vise 
> to make sure it stays flat. You'll be mounting the
> linear guide rail to 
> this. I decided to leave the somewhat straight
> sweeping radius rather than 
> deface it with the hammer. It's quite usable in this
> form.
> 
> The "post" is mounted to a chunk of whatever you
> have on hand. I squared up 
> a small piece of 6/4 rock maple on the jointer, and
> screwed the 6" post leg 
> to one face. The block turned out about 1-1/2" x 4 x
> 6 by happy 
> circumstance. The 6" length is the height, and the
> 4" face sets back a 
> little from the post face, giving a couple extra
> inches of throat under the 
> post. I mounted the post angle 1/2" up from the
> bottom of the, umm, post so 
> the pcb can make use of the extra room underneath.
> 
> The real piece-de-resistance, the real jewel of a
> find at Lowe's, was a 
> 1-1/4" to 2" copper grounding clamp for $3.50. This
> is used in real life to 
> connect a ground wire to copper pipe, but for me,
> it's the perfect clamp for 
> the RotoZip. The RotoZip's throat is 1.70" diameter
> (sorta egg shaped, but I 
> doubt that was by design). Toss the tapped through
> back half, and keep the 
> slotted eared front half. A pair of 5/16" ceiling
> fan mount bolts holds this 
> to a birch ply clamp body. These have lag bolt
> threads on one end, and 5/16" 
> UNC threads on the other. The only really nerve
> wracking operation in the 
> whole project was ramming the lag bolt into the edge
> drilled ply. As careful 
> as I was, and as tightly clamped as it was, I still
> heard it delaminate. I 
> couldn't see a crack, but I know I didn't imagine
> it. Maybe I did; it holds 
> just fine. Oh yeah... the clamp body is 3/4" birch
> ply, 3.5" wide, and 
> somewhat longish. I used a 1-3/4" holesaw to cut the
> hole, 2.310" from the 
> edge (the sole critical dimension in the project,
> and depends on how the 
> rear lug is mounted), and then split it on the
> diameter, leaving two flat 
> surfaces for the clamp bolts. In hindsight, MDF or
> good dense maple will 
> work better because of the edge drilling.
> 
> That's about it. The rear mounting lug of the
> Rotozip mounts to small L with 
> a 45 deg notch to clear the body. Two 4mm bolts
> holds this to the 3.5" 16 ga 
> sheet, and four 4mm bolts hold that to the guide's
> truck. Three #10 
> woodscrews hold the clamp body to the front plate.
> Three 1/4" lag bolts hold 
> the post holder post to a 3/4" MDF work table. It'll
> need some rubber feet 
> under it clear the lag bolt heads, but that's a side
> trip for tomorrow. I 
> think I might route two short T-slots down the depth
> of the table, to hold a 
> guide bar for gridded hole patterns. And by
> fortuitous chance, the grounding 
> clamp has a 1/4" ID lug where the ground wire would
> be screwed. This will 
> make a handy mount for an LED "headlight", much
> needed.
> 
> Testing was anti-climatic. The spindle glides nicely
> under its own weight. I 
> push up to move things around, and guide it down to
> make a little mound of 
> dust. A just-right chunk from the scrap box props
> the head up to change 
> bits. Changing bits took a few tries to figure out
> but it's getting almost 
> second nature already. I hold the depth ring with
> the tip of the forefinger, 
> hit the spindle lock with the thumb, and twist the
> collet lock nut with the 
> other hand. All by feel, of course, because there's
> no room underneath to 
> see. I expected to have to re-align the kajillion
> and one bolted 
> connections, but it cuts just fine right off. Not
> sure how I'll get an 
> indicator under the collet to swing it, but it
> doesn't look to be an 
> emergency this moment.
> 
> I started out thinking I would build this as a
> temporary measure, only until 
> I could draw up something proper, but it works so
> well I'm loathe to change 
> it. The secret ingredient, of course, was the linear
> guide. It's hard to go 
> wrong with all that precision ground everything
> working for you. There are 
> some things that need addressing. For one, the L
> post is too flexible; it 
> tweaks a few degrees when the Rotozip starts up, and
> stays there until it 
> shuts off. No side-loads when drilling with hair, so
> no emergency. The whole 
> post needs rethinking; maybe build up an MDO box
> girder instead. For two, 
> the linear rail is bolted directly on the
> non-precision sheet metal. I'll 
> bed it in powdered metal epoxy at some point, but
> there's no harm leaving it 
> for the short term. Time enough tomorrow... Speaking
> of which, UPS tells me 
> I have a box arriving from Circuit Specialists. Just
> in 
=== message truncated ===



		
__________________________________ 
Start your day with Yahoo! - Make it your home page! 
http://www.yahoo.com/r/hs

Attachments

Move to quarantaine

This moves the raw source file on disk only. The archive index is not changed automatically, so you still need to run a manual refresh afterward.