Home Made Bearpaw Spine Tester

dvd8n

Supporter
Supporter
AIUK Saviour
In this link:


Geophys2 told us about his Bearpaw arrow analyser:

50054-ArrowAnalyzer-Mood-680x330_copy.jpg

The idea intrigued me, and as detailed here:


I did some experiments to figure out how it worked; I thought that was the end of it but apparently not.

So here we have presented the next in my series of expensive and over-complicated solutions to simple problems.
 
Last edited:

dvd8n

Supporter
Supporter
AIUK Saviour
To recap, I collected together some test materials, including a set of kitchen scales, an ashtray, two blocks of wood, two pencils, some foreign coins and five pencil marks on the kitchen table, set up like so:

IMG_20210310_115958.jpg

Like the Bearpaw tester, you press the middle of the arrow down to the plinth which is set 0.5" below the bottom of the shaft:

IMG_20210310_142212.jpg

And read the number on the scales:

IMG_20210310_142230.jpg

Obviously it wasn't reading spine directly. I was going to develop a spreadsheet to work it out but in the end I decided that would be overkill as the equation wasn't complicated:

At 28": spine = 220,000 / weight on scale in grams
At 23": spine = 396,929 / weight on scale in grams

Given that it's an experiment / prototype, it seemed to work pretty well, what with hexagonal pencils as supports rather than frictionless bearings, a 0.5" plinth that was pretty much just eyeballed for height, kitchen scales with a dodgy battery that is only precise to 5g, and test arrows that I couldn't be bothered to defletch:

IMG_20210310_143734.jpg

The readings were not bad, and, happy that I knew more or less how the Bearpaw worked, I resolved to forget about it.
 

dvd8n

Supporter
Supporter
AIUK Saviour
But, life's not like that, is it? It really bugged me that I wasn't getting a reading in thou like a real spine tester would give. So I went and had a look in my box of bits in the attic and found this:

IMG_20210315_232411.jpg

It's an Arduino micro-controller and display that I bought in Maplin's closing down sale (remember Maplin?) a few years ago and never did anything with.
 
Last edited:

dvd8n

Supporter
Supporter
AIUK Saviour
I got myself a load cell and HX711 interface card from ebay and connected it up:

IMG_20210318_153540.jpg

A bit of work got the rig reading in grams:

IMG_20210318_153611.jpg

So, for the cost of a few day's work and a few tenner's worth of bits, I have implemented a £5 kitchen scale. This manages to be both heartening and dispiriting, both at the same time :rolleyes:
 
Last edited:

dvd8n

Supporter
Supporter
AIUK Saviour
It then wasn't much more effort to add the code to convert the force to a spine reading:

IMG_20210318_221531.jpg

As you can see. this tin of Carnation Condensed Milk has a spine of just under 500.

I guess that means that it should shoot quite well from my bow, although I may need to aim off just a little. Maybe I should take it to the practice butts tomorrow ;)
 

dvd8n

Supporter
Supporter
AIUK Saviour
First test assembly looks less shonky than my version with pencils, ashtrays and foreign coins:

IMG_20210319_132004.jpg

In position on the load cell on my table in front of the telly workbench:

IMG_20210321_101923.jpg

Still waiting for some bearing spacers and materials for less ugly plinths but it's good enough to try out.
 

dvd8n

Supporter
Supporter
AIUK Saviour
Well, I had to buy this from ebay:

IMG_20210328_121818.jpg

It's a one kilo reference weight.

Really, who could have predicted that using a jam jar with a badly fitting top filled with water as a reference weight (see post #4) might be a bad idea around electronics?

It's not a laboratory quality weight, but have you seen the price of those? It's definitely good enough for messing around in lockdown, though :rolleyes:
 

dvd8n

Supporter
Supporter
AIUK Saviour
Started making the hardware for the plinths: IMG_20210328_112746.jpg

There should be six aluminium plates but of course I didn't have quite enough aluminium - I was short by 4mm :rolleyes:

And of course I don't have enough bolts of the right sizes...
 
Last edited:

dvd8n

Supporter
Supporter
AIUK Saviour
The plates are big and weighty as I had the idea that the plinths and the central pillar could be separate pieces, making it easy to pack away. When I wanted to use it I could position them on a flat surface setting the spacing between the plinths each time. I could even choose the spacing, using 28" for ATA, 26" for AMO, 23" for short ATA or 500mm for Cartel, setting the height of the central plinth as appropriate for the measurement and spacing.

But a little testing revealed that this just wasn't going to work. The spacing of the supports seems to be critical, and the position of the central pillar even more so. In retrospect it should have been obvious. If the central pillar isn't bang on central then the forces are not equally distributed between the two end supports so you can't intuit the forces on the two pillars from those on just one (i.e you can't just use one load cell and multiply the weight by two). Plus, the height of the pillar is super critical. An error of one twentieth of an inch in the height of the pillar would be the difference between, say, a 450 spine and 500.

So, the idea of having separate components, throwing them on the ground and taking measurements is a non-starter. All the positioning is way too critical. It'll need to be a big monolithic structure, and it'll way over-sized as I've over-engineered the parts deliberately for heft that I don't need now. But never mind, at least it'll be hard to break.
 
Last edited:

dvd8n

Supporter
Supporter
AIUK Saviour
So, I finally received the last of the mechanical components, so I managed to get the mechanical part of the build completed:

IMG_20210402_144548.jpg

As I said earlier, the two arrow supports and the central pillar are massively over-engineered as the intention was to have them free-standing for easy adjustment, with lots of weight to keep them stable, but that idea had to be abandoned when it became evident how critical the positions and heights of all the parts would be.

So, plan B was to use the supports that I'd already made but use the top and bottom plates to clamp to wooden spars. The plates don't screw to the wood; bolts pass through oversized holes in the wood and just clamp it. The oversized holes allow a little adjustment so that the supports can be dead-on 28" apart with the pillar dead-on half way between.

The wood is Ash, 19x44mm. I could claim that ash was carefully chosen for its strength and durability but the truth is that I had some in the shed and it was about the right size.

Here it is with an arrow in position:

IMG_20210402_144621.jpg

It's much wider than it needs to be; if I were to be making it again then I'd probably build it into a hollow square aluminium extrusion. Which would, I guess, make it look a lot like the Bearpaw one. Also, I doubt that I'd bother with the bearings in the supports. I'm not convinced that the axles move much if at all when I press on the arrow.
 
Last edited:

dvd8n

Supporter
Supporter
AIUK Saviour
I did some tests and graphed them:

IMG_20210402_145121.jpg

All of the spines came out a tiny bit stiff (on the graph the line is the 'correct' spine). A little of that is possibly down to the arrow wraps on the rear of the test arrows that I had, but the root cause is almost certainly that the central support is a tiny bit low.
 

dvd8n

Supporter
Supporter
AIUK Saviour
Setting the height of the central pillar seems to be critical. Its really hard to set with a tool like calipers.

IMG_20210402_150056.jpg

The correct thing to do would be to set the spacing with a half inch slip gauge, but have you seen the prices for a set! You can sometimes buy an odd one on its own but they're still expensive.

Then I had a brainwave. Ball bearings are made really accurately and are really cheap. I got a 1/2" bearing +2um -0um for a couple of quid from Ebay.

This led to its own problems; the ball wouldn't stay balanced on top of the pillar and kept rolling off and getting lost on the floor. I solved this by gluing a skirt to the bottom of it to keep it in position.

IMG_20210402_150149.jpg

Its a bit ugly but it works.
 
Last edited:

4d4m

Active member
I love the ingenuity on this thread! I do like to make stuff, to save money, for the learning, or just for the satisfaction of doing it... but I think in this case I’d probably just buy the Bearpaw!

Have you considered that wood can expand and contract with humidity? I suppose you’ve still got the issue of thermal expansion with alloy strut.
 

dvd8n

Supporter
Supporter
AIUK Saviour
I reworked the electronics to use an Arduino Nano with remote buttons and a display on leads rather than the Arduino Uno with the daughter board.

IMG_20210406_114347.jpg

I did this as, good as the Uno is for prototyping, it can be a bit of a pain to mount. The tiny Nano gives much more flexibility in that respect.

It has given me a weird issue though; previously the display worked flawlessly, but now it intermittently won't show line 2. It'll work, then suddenly stop working for a while then mysteriously start again. It shows all the hallmarks of being a hardware issue, except that the demo display programs that come with the display always work without fail. Which makes me think that it has to be something really insidious in my software. It's a pain as I was ready to transfer the circuitry to Vero board, put it in a box and put my toys away.
 

dvd8n

Supporter
Supporter
AIUK Saviour
I love the ingenuity on this thread! I do like to make stuff, to save money, for the learning, or just for the satisfaction of doing it... but I think in this case I’d probably just buy the Bearpaw!

Have you considered that wood can expand and contract with humidity? I suppose you’ve still got the issue of thermal expansion with alloy strut.
Buying the Bearpaw is 100% the sensible thing to do. The cost of the components is probably less than the retail solution, but factoring in the cost of my time makes the Bearpaw an incredible bargain. It only makes sense if you do it for the fun and challenge.

As for the wood in the design, it was only done out of expediency - it's what I had to hand. It is ash, which is stable, and I'll seal it with something like teak oil, but any twisting or warpage will kill the accuracy. Pine would probably be a non-starter for the job - it warps if you look at it funny. If it does happen then I'll swap the wood for aluminium U channel. I don't think that expansion would be a big issue - the accuracy of the tool lengthways doesn't seem to be as critical as the height of the supports and central pillar.

A lot of the design comes from the early idea of building the supports and column as separate pieces. If I was starting again I'd probably build it on a length of V-Slot 2080.
2080 ANODISED-500x500.png
It'd be infinitely adjustable with T-nuts and very stable. But that will probably not happen - I don't need two spine testers.
 
Top