On Tue, 08 Jul 2014 22:38:27 -0400, clare wrote:

On Tue, 08 Jul 2014 19:52:15 -0500, Tim Wescott
wrote:

On Tue, 08 Jul 2014 20:26:30 -0400, clare wrote:

On Tue, 08 Jul 2014 15:50:31 -0500, Tim Wescott
wrote:

This is related to my dynamometer thread. We didn't find anything, so
I'm working on plan B, which is to roll my own.

The thing would need to work with an existing piece of equipment,
which has no solid external reference points to provide for good
alignment to the shaft on the device under test (or, for that matter,
a really easy way to get centered onto a shaft).

I'm visualizing getting something that can locate things to within an
inch, and within five or ten degrees, then having one of two things:

1: A foot-long shaft with U-joints (probably CV) at each end, to
couple the dynamometer to the device under test, or

2: A much shorter mechanical gizmo, which is currently labeled in my
mental sketches as the "magic coupler".

The important thing in all of this blathering is that when all is said
and done the coupling needs to transmit torque accurately -- it can't
just be efficient, it has to not have any severe differential gearing
effects that would mess up torque readings.

So, do you think #1 would work? Is there a name for #2, and could I
count on it, or should I just stick with #1?

As suggested by my foot-long shaft idea, space is not a huge issue,
particularly if the thing can be stowed away compactly when not in
use.
#2 is called a belt.

Not if I can't easily mount it concentric to the axis of rotation.
OK, spill the beans--- What is the application??? Picture or
diagram, or at least a good description?? How about a hunk of hose the
right size to fit the shaft, and a couple of hose clamps. Braided hose
if required. Even steel braid. You are only talking 80RPM from what I
remember - and what torque, again???

And I'm assuming you mean parallel to the axis of rotation.

It's for a customer, so I have to be all secretive. Which is a pain, I
know.

Things could, in theory, be mounted dead center on a shaft on the device
under test, but doing so would require some disassembly and reassembly to
do the test. I'll go there if I must, but I'd much rather have something
that'll work anywhere, anytime.

50 newton-meters, whatever that works out to in 'merican units.

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com