I think you miss the point of my idea. I was suggesting putting a full size rotary axis along the front of the machine at a height of say 4". The new head would fire the beam forward at a height of 4". The Y axis would move it towards the object to get it within focus range depending on its radius. The X axis would move along the length of the object.
Z would be the redundant axis that could possibly drive A.
On the Facebook Laser Owners group, someone has just ordered a Muse for US$6,500 + $300 delivery.
https://www.facebook.com/groups/905784596166200/permalink/1221346387943351/
Regarding Delivery, in the comments:
“All they said is there are a bunch of them in the warehouse ready to ship and that they will begin to ship this month.”
Once they receive they are planning to report back…
I see, thanks for the clearification.
@palmercr, now that’s an idea I hope makes it out of the hopper and into our glowforges sometimes in the not too distant future, even if it means buying a different head (swappable magnetic heads are looking real good right about now.) The lack of rotary has always bothered me a little, just 'cause, but I never thought to think along a different axis. Personally I vote this hopper idea of the year! (so far) 
I haven’t looked inside the GF so I don’t know if this is the case - but if it’s in front of the gantry wouldn’t that make the beam point toward the front of the machine? I know it’s shielded but on the off chance something bad happened that’d be an exposure. Could it be put behind the gantry instead and then if something goes wrong it’s pointing out the back of the machine and less likely to be a problem? (Still a potential but a lower one as folks would more likely place the back against a wall or something.)
Very true, but a shield could easily be included as part of the rotary attachment to beef up the shielding in the direction the laser is pointing. I think putting it in the back of the machine would be difficult just because you would have to get it past the head and gantry that way. Come to think of it, as a pro owner, an integrated shield would be necessary, would hate to see the laser shooting out of the passthrough slot 
, or would I 
?
The pass through slot will be level with the top of the honeycomb, so it will be well below the proposed axis level.
You’re assuming that the beam comes off the head exactly horizontal. Depending on all the clearances and the Z position of the incoming beam from the tube, it might make sense for the beam from a rotary-attachment head to not be exactly horizontal (as long as things were arranged so it still hit the surface of the material perpendicularly).
But all of this is likely moot, because it would be way easy to incorporate some kind of beam stop, and to have interlocks in the software so that nothing would work if the stop wasn’t there. (So you could put the sideways head in place, but it wouldn’t fire unless the rotary gizmo was there with something in it.)
I think what @palmercr and @takitus were trying to illustrate are as follows. Add a new stepper/rotary attachment, park the head at the back of the unit (maximum “y”), and alter the beam path to allow for the head to perform "x "direction only movement, and the new “a” axis rotates the part. The autofocus head can still be used to focus beam onto rotary part. I threw together a sketch with garbage for dimensions to illustrate:
AS-IS:
I think the idea would be to fix the Z axis and gantry movement, forcing manual placement of the obeject on the rotary into range.
Possibly a bypass plug could shift the position encoding from being sent to the y axis belt and redirect it to the rotary making the rotation the new y axis giving you a mapping of bed length to 360 degrees of rotation effectively unrolling your rotated surface to the bed. A few software changes to map onto the new cylindrical surface and to get the errors from no y-axis on the accelerometer and it should be very doable.
If the 7" guess is right that’s on rotation the equivalent of about a 22 inch bed so that’s more software changes. and I guess new X limitations since you need the external attachment.
Yes! Exactly…good solution. This is technically doable, but significant engineering costs. Much easier if you can simply take out the original head and swap in a side firing one, then plug in the rotary device…software senses head type and rotary device and you are set. Not that hard to do, but not this year… - Rich
Bypassing motors, adding shields, disabling sensors… this is turning into quite the project. At least the head isn’t attached with screws!
DUDE AWSOME draw up !!!
A new 90 degree head bracket that attaches on the existing magnets and provides a new horizontal position for the existing head sound doable.
The main issue is potentially controlling the rotation. Not sure that a constant slow rotation will be sufficient even for simple jobs. Maybe a some of of light detection that triggers the next rotational step…
One could rig an accelerometer to an arduio attached to the head. 1 stepper motor increment per head pass. I’m looking forward to an engineered version from the Glow team…
theres another thread linked around here somewhere that has someone made for another machine that works by just letting the gantry push the rig (it sits on wheels). As the rig rolls it also turns the object on the rotary device. Pretty ingenious creation! Just gotta make sure you account for the amount of space it takes up in the bed
I saw that thread. If I had a glowforge in my hand… I’m pretty sure I could have that assembly made in a day or so. The main issue is the amount of clearance under the head. Need to alter the beam path so we can cut a bigger part. Its a pretty good idea tho.
yeah, thats the same thought I had. Im wondering if a nice 45 deg mirror on the bottom of the GF head will work to trick the GF and allow for a horizontal beam to engrave on the side instead of the top
The main issue is figuring out the ratio of the Y axis steppers and translating that to a rotational axis.
If possible its not going to be easy with the space allotted. Especially if a reduction drive is needed.
Yes, that is the real problem with a rotational fixture. Turning mirror…no problem. 1/2" focus …great. X-axis movement…some translation there, but not difficult. This seems like something that Glowforge can offer as an accessory in the not too distant future. Go 
- Rich
