Non magnetic linear displacement system for moving a Hall Sensor

[maxim]

TL;DR Summary

ideas on how to realize a 600-700 mm linear displacement system for a sensor in a strong magnetic field, controlled by a motor stepper

Dear mechanical expert

I have to realize a linear displacement system for a Hall sensor that has to slide along the central axis of a narrow cylinder and with which high intensity magnetic field measurements (1-14 Tesla) have to be made.

The field is produced by a commercial vertical magnet (Oxford 600/51 Superconducting Magnet).

The sensor is a Lakeshore HGCA-3020 (diameter: 6.4 mm, length: 5.1 mm)

The linear displacement is along the z axis of the magnetic field, and the maximum expected excursion is about 60-70 mm starting from the center of the magnetic field (highest field point, T = 14 Tesla), downward.

I'm looking for a compatible solution to this data, but I need some suggestions because:

1) all components must be absolutely non-magnetic in order not to interfere with the measurements and not to get stuck in the magnet.

2) All the mechanical moving parts must be contained in a cylinder 550 mm long and < 40 mm diameter

3) The Lakeshore HGCA-3020 sensor must move while remaining centered along the z axis of the cylinder (where B measurements will be made)

4) Each reading point must be associated with the sensor position along the z axis, in order to reconstruct the "profile" of the magnetic field B(z)

For point 1) one can use non-magnetic metals such as some steels, copper or bronze with which to build worm screws and, of course, all types of plastic, polymeric and / or Teflon material. In my opinion the main problem arises for the actuators, because they have to work in the presence of a strong magnetic field, of the order of 1-2 Tesla.

For point 2), on the other hand, I need some ideas both to create the supporting skeleton of the structure and to create the actual linear movement.
Since good precision in moving is required, is it better to use a stepper motor or a linear actuator?
And where can I find these components?

Point 3) suggests that a worm screw in the center of the cylinder cannot be used, because the Hall sensor must be located exactly in that seat. Thus I imagined that a solution based on linear actuators is more suitable.
The main question is: are there linear displacement actuators (linear step motors) capable of covering the distance of 60-70 cm?

Point 4) is fundamental: how can I keep track of the displacement during the measurement?

What I've found in the market:

TSM17Q-1AG motor+drive+encoder+controller

 

[jrmichler]

Welcome to PhysicsForums.

Some thoughts:

I'm familiar with two types of linear actuators:
A true linear actuator that uses electromagnets within the moving slide. Would not work in a high magnetic field.
A servomotor with a worm screw inside a housing. These have magnetic components inside.

The stepping motor in your link may or may not have the capability of maintaining absolute position when the power is cycled off and on. If not, it needs to be sent to a known position every time the power is turned on.

Servomotors, such as those made by Siemens, have encoders that maintain absolute position when power is cycled off and on. Here's a link: https://new.siemens.com/global/en/products/drives/electric-motors/motion-control-motors.html. Siemens products are popular in Europe, Allen-Bradley servo motion products in the U.S.

Point 3 can be met by using a pair of screws geared together, with the sensor between them. I have built motion control systems doing exactly that, and they worked very well.

If you use a servo or stepping motor driving some sort of screw, will the motor be in a magnetic field? If so, check with the motor manufacturer if the motor will work. Optical encoders should work in such a field, but I would check on that also.

I like the idea of a pair of screws, geared together, driven by a single servomotor, and with the sensor between the screws.

 

[Baluncore]

A stepped belt driven by a stepper motor could position the sensor.
There would be an idler pulley at the far end.
The B sensor would be attached to the belt with a folded loop of cable.
An optical marker on the belt would identify the index zero position.

Examples of belts and pulleys are here.
https://www.ebay.com.au/itm/2GT-Open-Ended-Timing-Belt-Motor-Drive-Belt-Width-6mm-10mm-Rubber-for-3D-Printer/264309633649?var=563895805506
https://www.ebay.com.au/itm/4-10mm-...y-Gear-Sprocket/232252368069?var=531481235926

 

[maxim]

Thank you.
What about a pneumatic stepper motor?

 

[maxim]

Hi jrmichler

thank you for your reply!

Point 3 can be met by using a pair of screws geared together, with the sensor between them. I have built motion control systems doing exactly that, and they worked very well.

That's interesting. Can you share with me a simple project, so I can see exactly your approach?

If you use a servo or stepping motor driving some sort of screw, will the motor be in a magnetic field? If so, check with the motor manufacturer if the motor will work. Optical encoders should work in such a field, but I would check on that also.

Yes, this is the main problem. The motor will work in a quite strong magnetic field of course, and I have no idea how it reacts. I just discovered the possibility of "pneumatic stepper motors" but I have no idea if they are already on sale and where to look for...

I like the idea of a pair of screws, geared together, driven by a single servomotor, and with the sensor between the screws.

A couple of pair screws ... interesting. It is not clear in my mind how to handle simultaneously two worm screws and avoiding instabilities ... Can you explain better to me this point?

 

[Baluncore]

What about a pneumatic stepper motor?

A pneumatic stepper may be unnecessary.

The belts can be as long as required with many idlers, so the stepper motor can be far from the field. I do not have enough detail to suggest how the sensor connections should be taken from the belt.

The only problem with long thin loops of stepped belt is that the two belts must be kept apart so the opposite belt teeth do not contact. That can be done with a plastic sheet between the belts.

The wires connected to the sensor will be in the field. Non-magnetic wires may reduce the field if they are wide traces on a flexible backing, like the print head on a line printer.

 

[jrmichler]

I like @Baluncore's idea of a timing belt drive more than the one I first proposed. This document shows a diagram on page 6: https://www.thomasnet.com/pdf.php?prid=101106. And McMaster-Carr is a good U.S. source of timing belts and pulleys: https://www.mcmaster.com/timing-belt-pulleys/l-series-timing-belt-pulleys-8/. These are common components, so should be readily available in Europe.

The wires and the sensor could be glued to the outside surface of the timing belt.

 

[Tom.G]

A few of thoughts:

1) A belt thru the bore of the device can be cut, fed thru, then spliced, thus avoiding self-interference and allowing room for a larger, more stable belt.

2) Mechanical drive could use a piezoelectric positioner. They are available in both rotary and linear versions. One popular supplier (there are many) is: https://www.pi-usa.us/en/tech-blog/positioning-capabilities-of-ultrasonic-motors/

3) Position readout can be a rotary encoder as part of a rotating drive, or a linear optical encoder for a linear drive. Both are available in relative (pulse counting) or absolute position readout (i.e. or encoder-internal multichannel codes).

3a) If relative linear position is acceptable, a ruler with appropriate resolution can be attached to the moving element, and a photosensor and counter connected.

Cheers,
Tom

 

[maxim]

Thank you Tom for your reply!

A few of thoughts:

1) A belt thru the bore of the device can be cut, fed thru, then spliced, thus avoiding self-interference and allowing room for a larger, more stable belt.

Honestly I do not understand the solution by using belts, probably I have no experience in micro-mechanics and I have no details on how to realize a such positioning system. May be obvious for all of you, not for me. There are some examples I can look into figure out the idea? I am a chemist-electronic more that mechanical guy...

2) Mechanical drive could use a piezoelectric positioner. They are available in both rotary and linear versions. One popular supplier (there are many) is: https://www.pi-usa.us/en/tech-blog/positioning-capabilities-of-ultrasonic-motors/

I had already thought about it some time ago, and the idea of using a linear positioner seemed successful. Unfortunately, however, the excursions produced by piezoelectric devices are very limited and I have not found anything suitable to cover a distance of 600-700 mm, just to understand each other.

The model N-310 NEXACT Miniature Linear Motor / Actuator is interesting:
https://www.pi-usa.us/en/products/p...-linear-motor-actuator-1000700/#specification

but the travel range is limited to 10 to 125 mm, nothing respect to my needs...

3) Position readout can be a rotary encoder as part of a rotating drive, or a linear optical encoder for a linear drive. Both are available in relative (pulse counting) or absolute position readout (i.e. or encoder-internal multichannel codes).

Kindly, could you please indicate me a commercial product for the linear type of optical encoder to be coupled, for example, to the aforementioned N-310?

3a) If relative linear position is acceptable, a ruler with appropriate resolution can be attached to the moving element, and a photosensor and counter connected.

Cheers,
Tom

I guess I need an absolute value rather than a relative ones.

Thank you so much for the idea; i will try to contact PI to ask more details about piezo-actuators.

Thanks to all participants, this discussion becomes very interesting for me...Hope to find the best solution for moving linearly a sensor in a magnetic field.

Max

 

[Tom.G]

Belt:
The belt approach suggested by @Baluncore is rather common. You will see one by looking inside an inkjet printer. Printers also have linear encoders to sense the head position. Often you will see a clear plastic strip mounted just behind the print head that has fine vertical lines printed on it. This is called a Timing Fence, and there are a couple of LEDs and photosensors on the print head to count the number of lines going by. Two sensors so the movement direction can be sensed and the head position counter incremented or decremented.

There is a very short video at:
The above, and many more, were found with:
https://www.google.com/search?&q=timing+belt+for+stepper+motor

Linear Motor:
A Google search found over 45 000 of them:
https://www.google.com/search?&q=linear+motor+actuator+750mm

Linear Encoder:
A Google search found over 300 000 hits:
https://www.google.com/search?&q=linear+encoder+700mm
You would want an optical encoder, not a magnetic one!

These days, it seems designing stuff has become finding the right words to put into a search engine!

Using a belt could be the simplest solution, but a minor drawback is it won't stay exactly centered during movement. All that means is after the sensor is moved, you may have to wait for the belt to stop vibrating before taking a measurement. If this is a low usage laboratory situation that wouldn't be a problem.

If the measurements must be taken during ordinary use, or for quality control during manufacturing, a linear positioner of some sort could give faster results.

In either case, you may need instrument grade bearings to guide each end of whatever carries the sensor thru your tube, although finding non-magnetic ones could be difficult. Air bearings with a plastic body come to mind as a possible solution if centering the sensor is critical.

Please keep us updated on how this project turns out!

Cheers,
Tom

 

[maxim]

Just a small update: I am looking at linear pneumatic steppers. This seems the way to go on in MRI applications.

 

[maxim]

Yes, the pneumatic stepper is the way.
I have just started working with a 3D printer. Hope to give my feedback here soon.