precision planetary gearbox, NEMA 17 Voltage: 2.7V

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Adalbert
Posts: 2402
Joined: Mon Nov 30, 2015 1:09 pm

precision planetary gearbox, NEMA 17 Voltage: 2.7V

Post by Adalbert »

Hello everybody,

Does anybody know if I can use the 12V power supply and the TB6560A controller for this motor?

https://www.omc-stepperonline.com/nema- ... 84S-HG5%20

Does the “Voltage: 2.7V” mean the voltage between the phases?

BTW, the backlash of this gear box should be smaller than with the GT2 belt, shouldn’t it?

BR, ADi

DoppelHans
Posts: 9
Joined: Fri Feb 07, 2014 10:27 am

Post by DoppelHans »

with a step angle of 0.36°, it looks quite interesting. It is not a lightweight motor with 0,68 kg! Normally those motors run with 9 to 28 V. Therefore the given voltage of 2,7V is strange. I can't find it in the datasheet again.

Hans

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

For our macro use I believe the fewer moving mechanical interfaces between the camera sensor and subject the better. The reason is mechanical interfaces that must have some degrees of movement and thus a potential source of unwanted movements, thus the more gears and such not the better!

A good 400 step motor, direct coupled with a quality 1mm pitch rail (THK for example) gives a basic 2.5 um step, add a few good controller micro-steps (Tic-500 for example) and you should able to cover most of the macro range magnification to 20X (maybe 50X). Beyond this range a proper microscope base probably should be considered.

Stepper motor voltage is specified as the rated current times the internal resistance within a single phase, not between phases. The phases are completely independent (galvanic isolation) electrically. Some of the modern controllers/drivers utilize a much higher supply voltage to allow for the motor inductive effect (L*di/dt) to add to the motor I*R voltage, thus the "need" for a higher supply to be able to "force" the desired current into the opposing motor inductive effect. These same modern controllers (Tic-500 for example) utilize a complex arrangement of pulse width modulation (PWM) and decay modes (allows motor currents to recirculate) to achieve micro-steps. Some controllers have reported difficultly with certain motors (usually low inductance, fast time constant (L*R)), and some (Ti DRV8825 for example) have reported severe non-linear micro step behavior.

I've spend a lot of time, $, and effort experimenting/evaluating various stepper controllers and found the Pololu Tic-500 was the best choice for my needs with an experimental stack and stitch system I'm developing. This controller also employs recirculating some of the motor current back to the motor supply to help reduce supply current demands.

This controller (Tic-500) should work with the motor parameters indicated, it's worked fine with every motor I have with a 12 volt supply. I did not evaluate the Toshiba TB6560 (although I think I have one of these drivers on a board), here's a data sheet that might help. Based upon your motor parameters you may need to adjust the timing for the decay modes, this is outlined in the data sheet.

https://www.google.com/url?sa=t&rct=j&q ... cJYnMqK4ZT

Anyway, hope this helps.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Adalbert
Posts: 2402
Joined: Mon Nov 30, 2015 1:09 pm

Post by Adalbert »

Hello Mike,

Thank you for your comments!
For our macro use I believe the fewer moving mechanical interfaces between the camera sensor and subject the better. The reason is mechanical interfaces that must have some degrees of movement and thus a potential source of unwanted movements, thus the more gears and such not the better!
I have already measured (by Zerene) the backlash of my rail with the gear-box 100:1. The backlash was big but only at the first step after the changing of the direction of the movement. But I stack only in one direction.

If you don’t like gear-boxes, what do you think about the belt-mount e.g. GT2 ?


I've spend a lot of time, $, and effort experimenting/evaluating various stepper controllers and found the Pololu Tic-500 was the best choice for my needs with an experimental stack and stitch system I'm developing.
All my rails are Arduino driven, so I don’t need a controller like Tic T500 with the USB interface.
http://www.photomacrography.net/forum/v ... c&start=45
TB6560 has a cooling for 3A, supports 1/16 steps and costs $5. OK, Tic T825 supports 1/32 steps.
So, for me the best one is TB6560 :-)

My original problem was the different description of the motors on ebay.
Sometimes the supply voltage (e.g. 12V – 24V) , sometimes the coil voltage (e.g. 2.7; 2.8 ; 3,6V) is used :-(

BR, ADi

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Adalbert wrote:Hello Mike,

Thank you for your comments!
For our macro use I believe the fewer moving mechanical interfaces between the camera sensor and subject the better. The reason is mechanical interfaces that must have some degrees of movement and thus a potential source of unwanted movements, thus the more gears and such not the better!
I have already measured (by Zerene) the backlash of my rail with the gear-box 100:1. The backlash was big but only at the first step after the changing of the direction of the movement. But I stack only in one direction.

If you don’t like gear-boxes, what do you think about the belt-mount e.g. GT2 ?


I've spend a lot of time, $, and effort experimenting/evaluating various stepper controllers and found the Pololu Tic-500 was the best choice for my needs with an experimental stack and stitch system I'm developing.
All my rails are Arduino driven, so I don’t need a controller like Tic T500 with the USB interface.
http://www.photomacrography.net/forum/v ... c&start=45
TB6560 has a cooling for 3A, supports 1/16 steps and costs $5. OK, Tic T825 supports 1/32 steps.
So, for me the best one is TB6560 :-)

My original problem was the different description of the motors on ebay.
Sometimes the supply voltage (e.g. 12V – 24V) , sometimes the coil voltage (e.g. 2.7; 2.8 ; 3,6V) is used :-(

BR, ADi
Hi Adi,

Backlash is only one of the possible undesirable effects, another is non-uniform (non-linear) or even non-monotonic (highly non-linear) movements. I would think belts might be even worse than additional gears, since they have addition mating surfaces also that must move and probably have a much looser fit with the teeth & belt than a metal gear to metal gear. Also the potential for stretching, aging, temperature expansion/contraction, humidity and so on, seem worse than for other methods.

Why do you need more than a 400 step motor and 1mm screw thread can provide direct coupled for macro work, especially if you include a few micro-steps?

Pololu has a MP6500 based driver board, this is the same driver chip I am using that's part of the Tic-500. BTW the Tic-500 has USB, I2C, RC, and a few other interfaces, I just use USB since it's an easy interface with the Raspberry Pi. The Toshiba chip should work, but may require some timing adjustments for different motors and supply voltages. The MP6500 automatically sets these up and so far hasn't caused any problems with a 12 volt supply and the motors I have (various 200 & 400 step NEMA 17, and 200 step NEMA 11). I don't think expecting anything beyond 0.5 micron accuracy is reasonable, even with the THK rails. Sure you can get 1/16, 1/32, 1/64 stepper controllers, but I'm sure you can't reliably position to those levels with the setups we are discussing, and they have progressively weaker torque, so it's like "empty magnification" IMO. The only benefit might be quieter operation, but I don't know as I haven't tried any of these deeper micro-steps.

I stayed away from the Tic-825 which uses the Ti chips (have a bunch of DVR8825 don't use), as these had shown severe micro-step misbehavior.

Hope this helps.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Adalbert
Posts: 2402
Joined: Mon Nov 30, 2015 1:09 pm

Post by Adalbert »

Hello Mike,
Many thanks for your hints!
Why do you need more than a 400 step motor and 1mm screw thread can provide direct coupled for macro work, especially if you include a few micro-steps?
OK, I don’t need that accuracy, no more :-)
My first rail was designed for the Reichert Fluor 100x with the NA=0.95 and I wanted to use 7 steps a DOF, so the step size of 0,00009mm (90 nano) was needed.
I wanted to take a photograph of the human hair (the last one in the following thread):
http://www.photomacrography.net/forum/v ... highlight=
But I’ll never use this lens again, because of the small working distance.

The MP6500 automatically sets these up and so far hasn't caused any problems with a 12 volt supply and the motors I have (various 200 & 400 step NEMA 17, and 200 step NEMA 11)
OK, I have understood, I use only NEMA 17 400 steps, so TB is good enough for me :-)

BR,
ADi

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Adalbert wrote:Hello Mike,
Many thanks for your hints!
Why do you need more than a 400 step motor and 1mm screw thread can provide direct coupled for macro work, especially if you include a few micro-steps?
OK, I don’t need that accuracy, no more :-)
My first rail was designed for the Reichert Fluor 100x with the NA=0.95 and I wanted to use 7 steps a DOF, so the step size of 0,00009mm (90 nano) was needed.
I wanted to take a photograph of the human hair (the last one in the following thread):
http://www.photomacrography.net/forum/v ... highlight=
But I’ll never use this lens again, because of the small working distance.

The MP6500 automatically sets these up and so far hasn't caused any problems with a 12 volt supply and the motors I have (various 200 & 400 step NEMA 17, and 200 step NEMA 11)
OK, I have understood, I use only NEMA 17 400 steps, so TB is good enough for me :-)

BR,
ADi
Very nice hair braid image!!

Can you explain how you arrived at needing 90nm for a step length?

BTW I have the newer Toshiba TB67 motor driver version (but haven't tested it yet).
https://www.pololu.com/file/0J1523/TB67 ... 170818.pdf

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Adalbert
Posts: 2402
Joined: Mon Nov 30, 2015 1:09 pm

Post by Adalbert »

Hi Mike,
Can you explain how you arrived at needing 90nm for a step length?
Please take a look at that:
http://www.photomacrography.net/forum/v ... highlight=
8.) The gear-box 100:1 enables step-size in nano-meters
8.1.) NEMA 17 step angle of 1.8 degree ==> 200 steps for the 360 degrees
8.2.) TB6560 16 micro steps, 1.4 A
8.3.) Gear-box (99 + 1044/2057): 1 ratio
8.4.) Full rotation 318424 (200 * 16 * 99.5075352455031599) micro-steps
8.5.) Mitutoyo screw: scale 25 positions a 0.0254 mm
8.6.) One full rotation ==> movement of the rail = 0,635 mm (25 * 0.0254)
8.7.) Full rotation of the micrometer screw shifts the carriage 635 micro-meters.
BR, ADi

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Adalbert wrote:Hi Mike,
Can you explain how you arrived at needing 90nm for a step length?
Please take a look at that:
http://www.photomacrography.net/forum/v ... highlight=
8.) The gear-box 100:1 enables step-size in nano-meters
8.1.) NEMA 17 step angle of 1.8 degree ==> 200 steps for the 360 degrees
8.2.) TB6560 16 micro steps, 1.4 A
8.3.) Gear-box (99 + 1044/2057): 1 ratio
8.4.) Full rotation 318424 (200 * 16 * 99.5075352455031599) micro-steps
8.5.) Mitutoyo screw: scale 25 positions a 0.0254 mm
8.6.) One full rotation ==> movement of the rail = 0,635 mm (25 * 0.0254)
8.7.) Full rotation of the micrometer screw shifts the carriage 635 micro-meters.
BR, ADi
Hi ADi,

Interesting, but don't think that thread explains why you need 90nm steps.

Using this simple relationship for DoF of lambda (green)/NA^2 (think this is right, at office now and don't have my notes) gives ~609nm. So a 400 step motor directly coupled to a 1mm pitch thread rail with 4 micro steps "theoretically" gives 625nm and 8 micro steps gives 313nm. "Theoretically" implying sometimes you may achieve this step size in resolution with a very careful crafted and extremely stable setup, but not with accuracy/repeatability. The thread length (assuming SS 316) with change by about ~1600nm for a 100mm long thread with a 1 degree C temperature change, while the aluminum base will change ~ 2200nm.

BTW the Toshiba TB67 series apparently is a significant improvement over the older TB65 series driver chips. They have incorporated some automatic decay modes like the MP6500 has to improve use with various motors and supply voltages.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

genera
Posts: 127
Joined: Wed Oct 05, 2011 5:05 pm
Location: California, USA

Re: precision planetary gearbox, NEMA 17 Voltage: 2.7V

Post by genera »

Adalbert wrote:Hello everybody,
. . .
BTW, the backlash of this gear box should be smaller than with the GT2 belt, shouldn’t it? BR, ADi
No, there's no reason that should be so if you've properly sized the belt to the load. Beyond that, even if one choice or the other has absolutely zero backlash, what does it matter? You still need to compensate for backlash in the bearings and rails, the leadscrew, and the leadscrew end bearings, same as always.
-Gene

rjlittlefield
Site Admin
Posts: 23543
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

mawyatt wrote:Interesting, but don't think that thread explains why you need 90nm steps.

Using this simple relationship for DoF of lambda (green)/NA^2 (think this is right, at office now and don't have my notes) gives ~609nm.
That's a simplified formula which is pretty accurate as a 1/4-lambda rule for small NA. There's a more complicated formula for large NA, which if I recall correctly gives a smaller value.

But I would not trust any of the 1/4-lambda formulas to give a "safe" step size at large NA. Those formulas are derived from analyzing a planar subject with perfectly uniform illumination. They completely miss all of the nasty "squirming around" effects that appear with 3D subjects and non-uniform illumination. To get a clean stack with those added problems can require a step size that is several times smaller than the number that comes from a 1/4-lambda calculation.

So, for use with an NA 0.95 objective, I would strongly support ADi's original desire to have 90 micron steps, maybe even smaller.

In that case I would also strongly recommend using a microscope focus block instead of trying to re-create that functionality using other parts. But that's a different issue.

--Rik

Edited to add:
The more accurate formula can be found at http://www.photomacrography.net/forum/v ... 955#215955 .
The predicted two-sided DOF for lambda=550nm and NA 0.95 is 400 nm.
The calculation is =2*lambda/(4*RI*(1-SQRT(1-(NA/RI)^2))) , where RI is refractive index, value 1 for air.
Last edited by rjlittlefield on Tue Dec 18, 2018 12:52 pm, edited 1 time in total.

Pau
Site Admin
Posts: 6038
Joined: Wed Jan 20, 2010 8:57 am
Location: Valencia, Spain

Post by Pau »

With my Zeiss WL I can easily get 50nm steps with the Wemacro Micromate coupled to the fine focus knob. With a better controller it can be even smaller, but I doubt I would need it.
Pau

Adalbert
Posts: 2402
Joined: Mon Nov 30, 2015 1:09 pm

Post by Adalbert »

Hi Mike,
Interesting, but don't think that thread explains why you need 90nm steps.
My rail calculates the step size based on the NA, which is entered.
The default value for the number of the movements/pictures is 3 a DOF but for NA=0.95 I set 7 steps a DOF.
The DOF is calculated as follows:
DOF = 0,00055 / (NA * NA)
STEP-SIZE = DOF / 7

0,0006mm / 7 =~ 0,00009mm

BR, ADI

Adalbert
Posts: 2402
Joined: Mon Nov 30, 2015 1:09 pm

Post by Adalbert »

Hello Rik,
Using this simple relationship for DoF of lambda (green)/NA^2 (think this is right, at office now and don't have my notes) gives ~609nm.
Yes, I use this simple calculation :-)
BR, ADi

Adalbert
Posts: 2402
Joined: Mon Nov 30, 2015 1:09 pm

Post by Adalbert »

Hello Gene,
Could you please show your solution with the belt-mounting?
BR, ADi

Post Reply Previous topicNext topic