Banjo Build

[S. Heslop]

Just going with this for the toolpost. Turned the thing by hand and it seems to be cutting okay.


For adjusting the depth of cut I can just use this threaded rod with a couple of nuts on it, and a bit of trig, to tap the tool forwards in the holder. I'll probably be moving the whole thing forward a few times to reduce the overhang when roughing this out.


Need to deal with the runaway speed control next though. My first thought was dimmer switches but they're surprisingly expensive and I can't find any nearby rated at over 400w. I guess i'd need about 800 for this drill.

Wonder if it'd be possible to take the speed controller out of another power tool...

[Will_D]

We use these to control out boilers in the homebrew world:

http://www.ebay.ie/sch/i.html?_from=R40&_trksid=p3984.m570.l1313.TR0.TRC0.H0.X3800w+dimmer.TRS0&_nkw=3800w+dimmer&_sacat=0

They really are rated at 3.8 kW and are pretty cheap (from China) or OK from the UK.

The heatsinks need to be earthed (to be sure to be sure) and will get hot if dimming  2.5 kW and above
[edit: fixed typo!]

HTH - Will

[philf]

Simon,

Years ago I built a basic drill speed controller which worked very well giving a large speed range with good torque at slow speed and had very few components.



I used to be involved in making BT151 thyristors and still have a few somewhere if you fancy building one.  The BT151 would need a small heatsink.

I can't find the control - I probably binned it as most drills come with speed control as standard.

Phil.

[S. Heslop]

Thanks for the offer and suggestion but I want to see if I can get this working with what i've currently got for now.


I thought that maybe the runaway speed was caused by vibrations so I tried removing the trigger and mounting it to a board and using a cam to push it in. Unfortunately the speed controller is just garbage, and while it can start slower (but still fast) it picks up speed slowly till it's running full speed.

I think next i'll try taking the speed controller out of the jigsaw.


Turned a little bit with the runaway speed though. It's hard work turning the crank all the way in and out. I might just try autofeeding with the cordless drill once I figure the speed problem out.

[krv3000]

Hi I like the switch mod

[vtsteam]

How about a belt to the outside of the banjo, since you've got that round? a vee belt will work on that much wood surface area without need for a vee. A small rod in wood bearings for an arbor (jackshaft) and a small vee-belt pulley between them at the other end. A couple of machinery bushings (stop collars) on the arbor to trap it. Chuck the arbor end. That takes the side load off the drill/motor.

[S. Heslop]

Hi I like the switch mod

Me too, shame it didn't work!

How about a belt to the outside of the banjo, since you've got that round? a vee belt will work on that much wood surface area without need for a vee. A small rod in wood bearings for an arbor (jackshaft) and a small vee-belt pulley between them at the other end. A couple of machinery bushings (stop collars) on the arbor to trap it. Chuck the arbor end. That takes the side load off the drill/motor.

The outside isn't really round yet. I was going to try cutting the outside first, but decided it'd probably be better to try get the thing more balanced by cutting the inside first. Plus it'd get in the way when I eventually get around to cutting the outside. I did consider a jack shaft to reduce the rpm but that was starting to get complicated.


I ended up pinching the speed controller from the jigsaw. If it doesn't catch fire I shouldn't have any trouble putting it back when i'm finished. The drill still has runaway speed, but it's much more adjustable than just with the trigger, so i'm able to fiddle with it and the speed of cut so that it maintains a steady and slower speed when working.


Using the battery drill to control the cut. Much easier than the crank and it doesn't sway the slide like when cranking it.


Managing a fairly heavy cut, about 3-4mm.


It's leaving a fairly okay cut. Bit hairy on where its cutting into the endgrain but it isn't tearing chunks out and its nothing a bit of sanding can't clean up.

Had a bit of a weird day today and didn't get alot done. I probably could've cut more but at the time I was thinking I might regrind the tool to reach all the way to the end. But I just realised that I probably wont need to cut it all the way since it can be cleaned up with one of those flush cutting router bits.

Also, just for the record, i'm switching the thing off when fiddling with the speed controller since it's all very exposed. If it was a more permanent setup i'd have built an enclosure for it.

[S. Heslop]

Damn thing seized up, then the nut the threaded rod was working against started spinning. Took it apart to see what the problem was.



That's alot of filings for just moving the thing in and out about 15 times. The nut was really fused onto the bar and I couldn't get it off with it in the vise. My guess is either that the filings built up and clogged the nut, or the nut wore so much that what was left of the threads swaged into each other.

I'm just going to try the same thing again but with some grease and see how that works.

[S. Heslop]

Tried John Mill's thing of squashing a nut into aluminium, but M8 was too much for my vise and it broke the screw. The screw had already broken at the tip and really I sort of anticipated this'd eventually happen, but i've been abusing the hell out of that vise since i've got a nicer one to replace it with. Problem with the nicer vise though is that I won't be able to abuse it so freely.


Plenty of grease. I'd expect the grease to just swell the wood, but with the layer of wax and not expecting this thing to last for long anyways I thought i'd try it.


But...


Inspecting the damage, it doesn't seem that the the tool moved in the toolpost as I'd imagined would happen, but the whole thing must've moved.

To be fair, before this happened the tool was digging in a fair bit. It was making unpleasant sounds, but I sort of just wanted to get the thing more or less to size. I think it was mostly the tool geometry at fault. I was being lazy in not wanting to regrind it into something more suitable. Although it's not a wholly rigid setup to begin with.

Shouldn't be a hassle to glue it back on at least.

[Will_D]

Simon,

I really admire your ability to take some pieces of plywood and build machine tools that work.

And boy do you test them to the limit!

Keep safe and keep up the good work

[RussellT]

Simon

I don't know how thick it has to end up but there seems to be a lot of material to remove.  Could you use a jig saw to take some of the wood out - if you can't find a long enough blade you could make an angled cut to remove some material.

The area glueing it to the face plate will become smaller as it gets thinner.  How about making some clamps around the edge of the faceplate, or even screwing the banjo on if the screw holes could be hidden later.

I'm following with interest.

Russell

[S. Heslop]

Simon,

I really admire your ability to take some pieces of plywood and build machine tools that work.

And boy do you test them to the limit!

Keep safe and keep up the good work

Continuing the theme, the conveyor motor on the thickness sander stopped working. Got a surprise looking at the electronics, but it seems that the mosfet got hot enough to melt the solder on the leads attached to it and they fell off. Surprised it could even still work at those temperatures. I remembered I forgot to put a diode over the motor and I wonder if it's just inductive currents from the motor causing that, but it could be what bluechip suggested in that thread with the power supply being too crappy to keep a stable voltage to the rest of the circuit.

Simon

I don't know how thick it has to end up but there seems to be a lot of material to remove.  Could you use a jig saw to take some of the wood out - if you can't find a long enough blade you could make an angled cut to remove some material.

The area glueing it to the face plate will become smaller as it gets thinner.  How about making some clamps around the edge of the faceplate, or even screwing the banjo on if the screw holes could be hidden later.

I'm following with interest.

Russell

My experience with jigsaws and deep cuts in the past is that the blade tends to wander. I did think about ways to rough out the material (was considering trying it with a router, as dangerous as that may be. But I couldn't think of any ways that wouldn't end up being more work than just turning it off.


Anyways I got the board glued back on and drying now. I was shy with the glue since I wanted it to be easy to separate later, but it was perhaps too easy to separate...

Still trying to think of a good tool geometry. Having a back rake seems like a bad idea since that's what I had, and a side rake might work better but i'm worried that with all the backlash (theres alot!) it'll risk pulling the slide forwards and digging the tool in. I could try a shallow rake (or no rake) but with wood that'd probably leave an awful cut and risk pulling chunks out anyways.

[mattinker]

Simon,

I really admire your ability to take some pieces of plywood and build machine tools that work.

And boy do you test them to the limit!

Keep safe and keep up the good work

Maybe Simon forgot to reply, Matias Wandell http://www.sentex.net/~mwandel/ has been an important influence, very interesting aproach to machines!
regards, Matthew

[Lew_Merrick_PE]

Simon,

You are now finding out why I use the circular lamination approach to banjo shells.  The outer guide is the female piece cut-off from the tool blank first using a bandsaw (OD then ID).  By making the tool blank in "halves," the ID of the outer guide may be "opened" to access a drilled hole from which the ID may be bandsaw cut (and finished on a spindle sander).  The inner core is a male piece that is easily cut with a bandsaw and finished on a disc sander.

A tire from a 6 inch hand-truck tire is the "clamp."  I carefully trim and place the outer (really nice wood) lamination to the ID of the outer guide, trim & fit the next layer in to that, remove the next layer, slather up the ID of the (first) layer, re-insert the next layer, insert the inner core (with inner tube), and inflate the inner tube as the clamp.  Repeat as required until your desired thickness is reached.

Another advantage of this approach is that only the inner and outer laminates need to be "matched" and "pretty."  The "filler layers" can be whatever wood is structurally sufficient and inexpensive.  A fascia of "pretty wood" covers up the "filler layers" and, itself, is hidden by an appropriate edge binding.

[S. Heslop]

The outer guide is the female piece cut-off from the tool blank first using a bandsaw (OD then ID).  By making the tool blank in "halves," the ID of the outer guide may be "opened" to access a drilled hole from which the ID may be bandsaw cut (and finished on a spindle sander).  The inner core is a male piece that is easily cut with a bandsaw and finished on a disc sander.

I don't fully understand what you mean. But either you've given me an idea or this is what you're saying; but I wonder if I could just make a plywood template for a router bit's bearing to follow, attached to the rough rim. Of course such a deep cut would be fairly dangerous (even if router bits of that length exist), especially with my crappy wooden router lift with questionable rigidity... Perhaps I could've routed each layer separate before glueing them together.

One thing that concerned me with laminate rims is the possibility of it warping. Since (as far as I know) there's no huge shaping after it's been glued up, and you've got a fairly damp thing soaked with glue, won't there be a fair bit of warpage as it dries? I assume that mustn't be that big of a problem since plenty of banjos are made that way, but it's still something I didn't want to risk having to tackle.


Anyways I ground up a tool bit without thinking too hard. I think I just ended up automatically making a bog standard turning tool with heavy rakes and reliefs.



I think trying to use that would be a bad idea since it's got more than a few ways it'd want to dig in.

So I figured that if I ground it again but with no (or very little) side relief, then that would prevent the tool from digging in too deep.

[Lew_Merrick_PE]

Simon -- I start off with two blocks laid up as 8 inches X 16 inches X 4 inches.  I create mortises in between the outer area (16 inch OD X 11 inch ID) and at the center of the (7 inch OD) inner core that I will use to screw together (and release) the "halves" of the mold.  I then bore a (usually ø1-1/2 inch) hole on the "inside tangent" of my 11 inch ID right on the (8 inch) "center mate."  I bore a (usually ø1/4 inch) hole at the center of my block set to use as a "pivot" when cutting on the bandsaw.

I cut the (16 inch) OD, take the pieces apart to insert the bandsaw blade into the (ø1-1/2 inch) clearance hole, reassemble the pieces, and cut my (11 inch) ID.  I can now unscrew the batten/inserts in the OD ring and remove it from the bandsaw.

It is now trivial to cut the 7 inch OD on the inner core.  I then drill a (ø1-1/2 inch) hole approximately 1 inch inside of the (7 inch) ID.  This is to access the filler stem of the inner tube.  I then bore (approximately centered on the 4 inch thickness) a (roughly) ø7/16 hole into that (ø1-1/2 inch) access hole for the filler stem itself.

Add a couple of "side plates" to cover the "gap" between the inner OD and the outer ID and you have a chamber that will support an inflated inner tube.  Lay-up the pieces as indicated in my previous missive.

When you are done, you can clamp the lay-up such that it extends a bit beyond the edge of the OD ring and hold that against a rip-fence of your table saw to get a nice, clean trim.

Does that clarify things?

[S. Heslop]

Got it all back together, which took a while since everything was going wrong. Forgot if I mentioned it but the faceplate got knocked when the thing exploded, so I had to remake that part since the hole was warped and it'd become fairly loose. Had to turn that flat by hand since I could only drill an undersized hole, and the shaft was crooked when I hammered it in (kind of wish I recorded it all but it seemed boring at the time).

Got all ready to cut though and the drill seized up.


The top hat part is stuck pretty solid. Too late to make alot of noise doing a postmortem, but I wonder if it managed to friction weld itself on. I'd actually made that top hat part myself since the original once (which appeared to be made from steel, but maybe it was a zinc alloy) was incredibly worn. I'll try find a new cheap drill to replace it at the boot sale on Sunday. Hopefully one with better speed control.


Anyways I had a go at cutting a bit turning the thing by hand, to see how the tool works. Cuts lovely and smooth going into the wood but was leaving some ragged stuff when going into endgrain. Perhaps it'll work better spinning at a higher speed.

[vtsteam]

Simon there's a silver colored heavy duty oilite sleeve bearing material that could be mistaken for steel. Maybe that was the material your first bearing was made from.

[S. Heslop]

Simon there's a silver colored heavy duty oilite sleeve bearing material that could be mistaken for steel. Maybe that was the material your first bearing was made from.

Yeah that seems likely. The shaft wasn't too worn if I remember right but the bearing was insanely so.

At the boot sale I managed to find a cheap drill that had the exact same problem with an extremely worn plain front bearing.



Lots of chips. The new tool geometry needed a slight change to flatten the tip of it so it'd not leave the cut strands still attached to the wood. I'll have to remember to get a close up photo of that. The new drill worked fairly well, the speed control built in is much more adjustable than the first and didn't require any fussing, but since the drill is running in reverse the chuck tended to unscrew itself since there's no second left hand screw to lock the chuck on. Some locktite more or less sorted that out though.



Got the inside turned true-ish and sanded. I made a wooden set of inside calipers to try gauge the diameters, and according to those measurements the inside should be fairly parallel. It'll of course be easier to measure the outside diameter (via the circumference) to get that closer to parallel, then I can see how well those calipers worked for the inside.



Got it partly set up for cutting the outside tomorrow. Too late now to continue making noise.

I'm also trying a new thing with also attaching the same photos for longevity. My usual image host went down for a bit a few days ago and it worried me that it might be gone for good.

[S. Heslop]

More problems. It's easy to blame all this on bad luck but it's probably more just carelessness.



I was getting a particularly rough finish on the outer diameter, and this probably should've tipped me off that something wasn't right. But after a few passes the tool dug in and stalled the drill, which was enough to cause something to start smoking inside of it. It wouldn't work at all after this.


So I opened it up to have a look inside, and it turns out that it was just the brushes overheating. They must've swollen or melted the plastic housings since they no longer fit, and the wires had detached from them. I had to shave down some oversized brushes with a file to get them to fit. It's working fine now but i'm gonna give it some time to cool down fully.

It also gives me a chance to think about tool geometry again. My guess is that the large relief for the inside diameter meant that not much of the tool was getting supported against the workpiece. I'll probably grind the other end of the tool similar to the first but give it no end relief.

I am a little concerned that I might've overdone the cut though. What I really need for this is a pair of long calipers. I made a sort of giant scissor caliper from wood but it's all too flexible to feel at all accurate. I might make some sort of caliper extender for the digital calipers. Hopefully I haven't already overdone the cut though.

[vtsteam]

Simon, just letting you know I'm continuing along with you here and reading every post. Not much to say as you're doing things I wouldn't do  , but I have a feeling you're going to be successful in the end, after whatever trials and tribulations occur. And I expect to learn as much as you do in the process! Be careful, stay safe -- other than that, it's all learning.

[awemawson]

Simon, yes just like Steve I'm following and silently encouraging you. What wall thickness are you aiming for? I imagine it'll get rather wobbly if too thin.

[S. Heslop]

Simon, yes just like Steve I'm following and silently encouraging you. What wall thickness are you aiming for? I imagine it'll get rather wobbly if too thin.

I'm going for around about 20mm, but ideally it'll be as thick as possible.

From what I understand, most of the force (from the strings) in a banjo is taken by the coordinating rod(s) that go through the rim. So you can get away with it being fairly flimsy.


I made that caliper extender. It's kind of goofy.


It would've been more rigid with the 'rods' running in line with the blades rather than perpendicular to them, but it's useable as is. I just didn't think to do it that way till after. Seems repeatable to about .1 of a mm at least and that's good enough for wood. Also with the rods not being in line it's not really reading millimeters but it doesn't really matter since I just want to get the sides parallel.


Here's the wooden thing I was trying to use before.


Also I appreciate people posting but don't feel you have to!

[awemawson]

I know zilch about banjo's, but I always imagined that the 'drum' was laminated up from thin veneers making in effect a custom plywood 'short tube'.

Now other string instruments that I have examined have deliberately thin walls ('cellos, violins and violas) so to me 20 mm seems extremely heavy - but then as I say I know zilch about banjos !

[S. Heslop]

I know zilch about banjo's, but I always imagined that the 'drum' was laminated up from thin veneers making in effect a custom plywood 'short tube'.

Now other string instruments that I have examined have deliberately thin walls ('cellos, violins and violas) so to me 20 mm seems extremely heavy - but then as I say I know zilch about banjos !

Most banjos are made that way but some are made from block rims like this. I don't think there's anything wrong with building them the other way, but I didn't have a drum sander at the time to easily make the veneers.

Violins and guitars do have fairly thin wood with bracing to strengthen it, because the vibrations from the strings are transmitted to those surfaces to make the sound (I guess the bigger surface just moves more air). Boy i'm having a hard time wording this elegantly. But in a banjo the skin head vibrates to move the air.

Guitars always amaze me with how they can stand the force of the steel strings considering how thin all the wood is.


Actually the main reason I wanted to build a banjo this way was because I thought it'd be nice to make some fancy patterns out of wood for the sides. The frame thing in this drawing was for a zither style banjo I was trying to make, although I thought i'd best make a regular old banjo before trying anything fancy.



If you look up segmented turning you can find all kinds of fancy patterns used mostly for vases and bowls. It requires some pretty accurate cuts and alot of work fitting it though, and with my crap tablesaw it'd be difficult to achieve. I couldn't even cut a six sided thing without large gaps.