micro-manipulator design

10 views
Skip to first unread message

daniel brastaviceanu

unread,
Jul 12, 2015, 3:44:25 PM7/12/15
to pv-characterizat...@googlegroups.com
https://3dwarehouse.sketchup.com/model.html?id=u741cce04-f44d-4dd2-93b7-8a16d74b671a
Hi guys!
I'm designing the micro-manipulator for 3D printing. I got the inspiration from the 3D printable micro-manipulator Mr. Pearce suggested, and I'm trying to make it as compact as possible while, in the same time, using easy to find parts like standard M3 screws (60mm, 40mm, etc) for the linear movement.
We (Tibi and I) are wondering if it's better to go with precision screws instead of M3, could you guys give your opinion on this?
The idea being that M3 screw has a pitch of 0.5mm so the linear movement with a full turn will be 0.5mm, half turn will be 0.25mm, quarter turn will be 0.125mm. Will that be precise enough for this application?
The other idea is that M3 screws have imperfections (unlike precision screws), so the linear movement will not be so linear.
Backlash: if you look at the model, I'm using 2 sockets in the middle of the screw, separated by a soft and thin washer to prevent the backlash. Precision screws have very little backlash.
M3 screws are available in every lab or hardware store, precision screws need to be ordered but they are not expensive (~7$ / screw+socket)

THEN, we are thinking of using set screws to compensate for 3D printing imperfection from using different printers. This will add some complexity so I will print one prototype to see if set screws are needed or not, if there is ware in the plastic from turning the screw, etc. Still need your opinion on using set screws or not.

Please participate in the design if you can, it would be nice to have as many brains as possible on this
Thanks!

daniel brastaviceanu

unread,
Jul 12, 2015, 4:02:47 PM7/12/15
to pv-characterizat...@googlegroups.com, daniel...@gmail.com
Notes: I made the rails like that for a reason: if the top surface of the rails was flat, the tool path of the printing head would make the exterior perimeters and then infill in a zigzag so the surface would come out very irregular. That would not be good for the sliding motion. By designing the last 3-4 layers only with perimeter walls, the top surface of those walls will come out almost perfect since the tool path is just linear.
Tibi suggests I should design the rails with an ''A''-shaped profile and the slots on the component sliding on those rails with the same profile. This makes sense because with square profile rails we need to adjust tolerances and parameters from printer to printer and from one material to another. With A grooves it it less of a worry.

Tiberius Brastaviceanu

unread,
Jul 13, 2015, 1:19:11 AM7/13/15
to pv-characterizat...@googlegroups.com, daniel...@gmail.com, Abran Khalid, Jim Anastassiou
Some feedback on the Micromanipulator.

From my understanding, we need this micromanipulator to have a very short hight. I think the vertical axis needs to be redesigned. We only need a very small travel in the vertical direction.



Here's a proposition, which is close to what we saw in the document Joshua sent to have an idea about the commercial system (see below)

This is what is used now in these instruments.


Here's what Joshua said about the micromanipulator

 I was picturing something like this  https://backyardbrains.com/products/micromanipulator  -except on a magnetic base so it is easy to move around macroscopically



For the precision screws, I would follow this feedback from Joshua

"For costs - keeping the costs down is important - but paying a little extra for the key components to make sure they are good is fine - realize the competition for a similar device on economic grounds is criminal - you will come in way under no matter what you do following your current concepts."

So I think we can spend a bit more money for precision screws.




On Sunday, July 12, 2015 at 3:44:25 PM UTC-4, daniel brastaviceanu wrote:

Tiberius Brastaviceanu

unread,
Jul 13, 2015, 1:22:33 AM7/13/15
to pv-characterizat...@googlegroups.com, tiberius.br...@gmail.com, jim.ana...@gmail.com, daniel...@gmail.com, abran...@gmail.com
I think Google Groups has a problem with pictures...

Tiberius Brastaviceanu

unread,
Jul 13, 2015, 2:11:17 AM7/13/15
to pv-characterizat...@googlegroups.com, daniel...@gmail.com
We should take into consideration that we can't control the precision of the 3D printer of someone who would want to reproduce the device. Therefore, we should design with a relaxed requirement on tolerance. To compensate, we need to use set screws or other techniques in order to compensate. Our experience in the lab already shows that tolerance in 3D printing is very hard to achieve. We had that problem a lot of times.

This makes me a bit worried about the rails too. A rail shaped as a "V" is better than one with a square top, because it is easier to fit a V into a V than a |_| into a |_|. since the walls of a |_| are straight, the distance between the smaller |_| that goes into the larger |_| needs to be very precise, in order not to have any lateral movement. We are talking about 500 microns tolerance, and that is not easy to achieve with a RepRap, on a distance of a few cm, the length of the rail. Especially when we don't know if these printers are well calibrated.

My idea is to use a metal rod rail, that we snap into a 3D printed casing. So I don't know if we really need to 3D print it, necessarily. We should 3D print as much as possible, but in a micromanipulator there are 2 critical parts, the rail and the screw.

We can make 2 types of micromanipulators, one that is more precise, with a metal rail, we have enough in the lab and you can find them in old printers and scanners AND a precision screw, that we'll sent to Queen's. The other one, the lower cost version, with 3D printed rails that someone can print if they want.

Back to Joshua's comment

"For costs - keeping the costs down is important - but paying a little extra for the key components to make sure they are good is fine - realize the competition for a similar device on economic grounds is criminal - you will come in way under no matter what you do following your current concepts."





Tiberius Brastaviceanu

unread,
Jul 13, 2015, 2:15:44 AM7/13/15
to pv-characterizat...@googlegroups.com, daniel...@gmail.com
Another feedback is the aesthetics. Daniel, I would try to make some round parts, try to find some nice shape... In the past you have designed beautiful objects
https://3dwarehouse.sketchup.com/model.html?id=51a498b4afd9ba49fb6303c1589ccbd 

On Sunday, July 12, 2015 at 3:44:25 PM UTC-4, daniel brastaviceanu wrote:
Reply all
Reply to author
Forward
0 new messages