It is currently Sat Dec 14, 2019 6:15 pm




Post new topic Reply to topic  [ 146 posts ]  Go to page Previous  1, 2, 3, 4  Next
 Affordable low backlash worm gearing? 
Author Message

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Hi, Gerald,

:) Still a nice name, though!

The latest model (the 10mm hollow-shaft one) is 80mm in diameter and 42mm in height (not including the input flange - another 5mm). Haven't weighed it yet - will weigh it dry (with no lubricants) and let you know. About a timeline: mechanically, we're pretty much there. The things holding us back right now are the patent filing process and the funding for the first batch.

For even better radial and axial alignment, I am already working on a slightly bigger casing version (85mm diameter, same height) - other than that, same characteristics.

Eros

P.S. here's a promised update: the hollow-shaft version, grinding away on a lathe, in both directions, at 1500rpm input (180 degrees/second output):


Sat Jul 05, 2014 12:14 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Hi again, guys,

Now, that I'm finally beginning to settle in my design plans, one question still bugs me, to which each of you might have the right answer. The question is:

What reduction ratio / ratios do you find the most useful for a / your time-lapse moco rig?

(I know I'm going with the 1:50 for now, but am curious about your opinions/needs too!)

Thanks,
Eros


Sun Jul 06, 2014 7:38 am
Profile
User avatar

Joined: Wed Nov 11, 2009 10:45 am
Posts: 1685
Location: Merritt Island, Florida, Estates Unitas
Post Re: Affordable low backlash worm gearing?
100:1


Sun Jul 06, 2014 9:44 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
sciencelookers wrote:
100:1


Thanks, mate!

Anybody else? :-)


Mon Jul 07, 2014 12:45 pm
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
Hi Eros

All my heads have 50:1 or 63:1
I have tried 100:1 but it's just too slow - and no significant gain as far as smoothness goes.
It should be reserved only for ultra macro and telephoto shots.

With 50:1 if you really want it slower, or the motor is under powered, then I'd say use a 2:1 belt drive stage on the input.

BTW 80mm diameter is quite large - but trust you can make a smaller version in future... :)


Wed Jul 09, 2014 12:32 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Thanks, Gerald!

My target is also 50:1 in the "flexwave" + 2:1 with timing belt (since the best position for the motor, if you want to have a hollow shaft, is next to the reduction, not axially co-linear).

Yep, I know 80mm (in fact, 86 for the newer one) is a bit on the large side - that's given by the pitch of the internal gears, which right now is 1.5mm, resulting in 47.75 inner diameter for the gears. To make them sturdy enough, I made the external diameter of the gears, in this case, 70mm. The rest up to 80mm are casing details. I can also work with a 1mm pitch, which results in 31.83mm inner diameter and ~48mm for the external diameter for the gears. This way, the overall size of the reduction drops to ~60-62mm X ~38-40mm. But this will result in a reduction for lower output torques, obviously. For now, we're focusing on refining the last bits and details on this one. All these facts relate to the 1:50 reduction - a smaller factor is also possible (I see no problems down to 1:30) - which again results in a smaller body.

And speaking of the form factor, this is how the unit feels in my hand:



Cheers,
Eros


Wed Jul 09, 2014 2:00 am
Profile
User avatar

Joined: Wed Nov 11, 2009 10:45 am
Posts: 1685
Location: Merritt Island, Florida, Estates Unitas
Post Re: Affordable low backlash worm gearing?
Go with Gerald's suggested ratios. He has a lot more experience than me. My suggestion was based on wimpy NEMA 17 motors which need the reduction to get enough torque. I will probably switch to bigger motors.


Wed Jul 09, 2014 5:24 am
Profile

Joined: Wed Feb 15, 2012 3:36 pm
Posts: 408
Post Re: Affordable low backlash worm gearing?
I think, currently, my preference is a 100:1 ratio as a good compromise. I have a couple of little 50:1 HD's and I could see the motor vibrations doing close up photography, although admittedly i was using little Nema 17 stepper motors. 100:1 is fine if you use servos, as you generally get plenty of smooth speed, but then maybe so is 50:1 (if you use servos)


How about 80:1 :)

Apart from accuracy and robustness, size is also important to me. If they could be condensed to around 60mm dia and maybe 40mm or less tall (with a hollow shaft, that would be ideal)
.

Edward


Wed Jul 09, 2014 7:00 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
@sciencelookers - Yep, 50:1 fits my bill so far too, so I'm on the same page with Gerald there. Same NEMA17 motors, but amplified 2x or more.

@edward - Thanks, man. Your 100:1 also fits my bill (after the 2:1 reduction with belts from the motor) and also fits what Gerald was saying about the 100:1 as being more fit for close-up/macro. Any reduction size is basically doable - 80:1, 100:1 included - although they would definitely result larger in size. An 100:1, for instance, for macro (where output torque should be in the lower end of the scale) would be done in 1mm pitch, resulting in an inner diameter of 63.66 for the gears. That's INNER - so your 60mm target size would be impossible, in my version of a reduction, for 1:100. Just for the sake of math, an 80:1 in 1mm pitch would have a 50.92mm inner diameter for the gears, resulting, with a more robust ant thin casing, in the same ~80mm diameter as the current 50:1 in 1.5mm pitch. Simple maths. My version of a strain-wave reduction doesn't leave much room for pitches below 1mm, for now...

Cheers,
Eros


Wed Jul 09, 2014 7:35 am
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
Hi

Since the ratio effects the diameter then 50:1 still sounds optimum for starters.

And yes - Ed uses fast expensive servos with 100:1 so they work for him. But with steppers 100:1 is slow.

*SL The size of stepper required will depend on the no-load input torque requirement of the gear. With small HD's I can use a 40mm long nema 17 and get good performance. But if the gear input shaft feels viscous to turn by hand, then a larger motor might be needed. The drive and voltage also have an influence.

BTW - will the outer housing have mounting holes? Is that how the gear would be secured to a frame?


Wed Jul 09, 2014 4:25 pm
Profile
User avatar

Joined: Sun Jul 06, 2014 1:33 am
Posts: 7
Location: Milan, Italy
Post Re: Affordable low backlash worm gearing?
Any news? I'm planning to build a 3axis lightweight and accurate MoCo for timelapse photography and this topic has definitely caught my attention!

_________________
www.streetcredbmx.com


Mon Mar 23, 2015 3:32 pm
Profile
User avatar

Joined: Sun Jul 06, 2014 1:33 am
Posts: 7
Location: Milan, Italy
Post Re: Affordable low backlash worm gearing?
Any news @erosnicolau ?

_________________
www.streetcredbmx.com


Mon Mar 23, 2015 3:36 pm
Profile
User avatar

Joined: Sun Jul 06, 2014 1:33 am
Posts: 7
Location: Milan, Italy
Post Re: Affordable low backlash worm gearing?
For anyone still interested. I got in contact with Eros, who told me that he's still working on the project, but he's facing some issues due to the inability to find CNC machines with enough accuracy. He promised me he's keeping on working on it and he'll put some news as soon as possible.

_________________
www.streetcredbmx.com


Tue Mar 24, 2015 2:04 pm
Profile
User avatar

Joined: Wed Nov 11, 2009 10:45 am
Posts: 1685
Location: Merritt Island, Florida, Estates Unitas
Post Re: Affordable low backlash worm gearing?
What three axes are you interested in making? The world seems to be full of sliders now, all competing to sell for less. Its hard to build a one-off for less than you can buy. Maybe convert a manual video slider to motorized use if you want to make parts of it yourself. eMotimo rules the cheap pan-tilt market. If you can make something like it without backlash for less, I'll buy one.

When you have your three axes running, you'll find you want a focus motor. We make focus/zoom motors specifically for timelapse and stop motion animation. They also work for realtime video as well. Check out stepoutmoco.com We also make adapters that let you run stepper motors from Dynamic Perception's small, portable MX-2 and MX-3 controllers, and USB controllers for Dragonframe stop motion software


Wed Mar 25, 2015 5:53 am
Profile
User avatar

Joined: Sun Jul 06, 2014 1:33 am
Posts: 7
Location: Milan, Italy
Post Re: Affordable low backlash worm gearing?
I'm basically trying to follow both the Dynamic Perception Stage One+Stage R path, as well as the work this guy did:

https://www.youmagine.com/designs/ultra ... nformation

I need to build something as lightweight as possible. I'm going to use the NMX and three NEMA17 motors. Everything else is up to your suggestions :)

First of all: anyone could help me finding the most compact (shortest) NEMA 17 motor following this specs?:

- Bi-polar 4 wire stepper motors
- Recommended maximum 6V rating
- Recommended maximum 1A rating

A friend of mine gave me this head today:

https://www.servocity.com/html/pt785-s_ ... RMvQi55OQU

I think it could be a good starting point. Any hint on the best way to replace the servo with NEMA17? (position, gearing...)
Which should be the minimum output torque from the motor? Could Nema11 be enough? (I?m going to use a DSLR+photo lenses, all balanced on the fulcrum)

As for the focus motor: It's a bit too much for me now, but I'll annoy you for sure as soon as I make the next step :)

Thanks!

_________________
www.streetcredbmx.com


Wed Mar 25, 2015 2:01 pm
Profile

Joined: Sat Nov 01, 2014 11:45 am
Posts: 23
Post Re: Affordable low backlash worm gearing?
Try http://www.omc-stepperonline.com/
I use them for all my projects.


Wed Mar 25, 2015 11:42 pm
Profile
User avatar

Joined: Tue Nov 24, 2009 7:55 am
Posts: 951
Location: UK
Post Re: Affordable low backlash worm gearing?
danieledepiccoli wrote:
I'm basically trying to follow both the Dynamic Perception Stage One+Stage R path, as well as the work this guy did:
https://www.youmagine.com/designs/ultra ... nformation
I need to build something as lightweight as possible. I'm going to use the NMX and three NEMA17 motors. Everything else is up to your suggestions :)
First of all: anyone could help me finding the most compact (shortest) NEMA 17 motor following this specs?:
- Bi-polar 4 wire stepper motors
- Recommended maximum 6V rating
- Recommended maximum 1A rating
A friend of mine gave me this head today:
https://www.servocity.com/html/pt785-s_ ... RMvQi55OQU
I think it could be a good starting point. Any hint on the best way to replace the servo with NEMA17? (position, gearing...)
Which should be the minimum output torque from the motor? Could Nema11 be enough? (I?m going to use a DSLR+photo lenses, all balanced on the fulcrum)
As for the focus motor: It's a bit too much for me now, but I'll annoy you for sure as soon as I make the next step :)
Thanks!

Hi, not quite what your looking for (30Kg wieght may be more ;) ) , I adapted a Servocity rig to run stepper motors. I modified their rig that used belt driven motors rather than gears, modifying the existing clamp type mounts 32mm diameter to fit the 31mm diameter gearbox with a thin split collar.
Hope this may give you some ideas, maybe ask Servocity if they can supply belt drive in place of gear, then make your own motor mounts. I have found them quite willing to listen to requests for mods. See vimeo description + comments for a bit more info.

_________________
Smile, it gives your face something to do!

http://www.vimeo.com/user1214873/videos
http://www.youtube.com/user/MrReggub


Thu Mar 26, 2015 6:50 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Hi again, guys,
As Daniele reminded me, there's already been a 8 month gap in my updates. This was mainly due to the winter season (very non-productive here) and even more to the fact that, as @sciencelookers already knows from my stories, it's impossible to find precision metalworking here in my country - which forces me to try to invent and build all sorts of auxiliary machines meant to achieve something closer to the tolerances I am for. And this is so darn time-consuming, not as much in terms of man-hours, as in terms of days and weeks spent waiting for the X or Y guy that you're counting on for machining this or that part.

@Daniele: From my modest experience, in your search for a motor, you should first know what's the speed-range you're aiming for with that motor. I'm saying this because each individual model has a specific speed-torque curve. I, for one, in order to build myself a better picture of what's available out there in terms of torque-vs-motor-weight, saved, stacked and scaled all the speed-torque charts from the Chinese manufacturer WANTAI, and came up with a birds-eye view of their range of available torques, for each 1A step range (0-1A, 1-2A, 2-3A etc.). I'm attaching here a couple of comparative charts, for 0-1A and for 1-2A - as these are the most likely useable motors when you're on the go and carry as lightweight as possible batteries)

@MikeA: Very nice-looking rig you've got set up there! Since this is a backlash-focused thread, dare I ask about your measurements in that direction? Also, what software is running on that laptop, that you're using to control the rig?

Eros

Attachment:
wantai-motor-charts-0-1A.jpg
wantai-motor-charts-0-1A.jpg [ 106.5 KiB | Viewed 12666 times ]


Attachment:
wantai-motor-charts-1-2A.jpg
wantai-motor-charts-1-2A.jpg [ 56.88 KiB | Viewed 12666 times ]


P.S. For whoever might be interested, I can provide the PSD files for the 2 JPGs attached, so one can easily pick the motor model corresponding to the chosen speed-torque curve - just leave me a PM and I'll sort this out for you.


Thu Mar 26, 2015 7:26 am
Profile
User avatar

Joined: Tue Nov 24, 2009 7:55 am
Posts: 951
Location: UK
Post Re: Affordable low backlash worm gearing?
erosnicolau wrote:
Hi again, guys,

@MikeA: Very nice-looking rig you've got set up there! Since this is a backlash-focused thread, dare I ask about your measurements in that direction? Also, what software is running on that laptop, that you're using to control the rig?


Hi, backlash details in video description, a reason I posted + to be helpful to the other guy that posted here ;)

That is why I pointed to the description in my post.

Following with interest.

_________________
Smile, it gives your face something to do!

http://www.vimeo.com/user1214873/videos
http://www.youtube.com/user/MrReggub


Thu Mar 26, 2015 8:32 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
My bad :) I was so into watching the video presentation, I totally missed the description.

10 arc-minutes for a gearbox? That's WOW, man! In my book, anything 15 arc-minutes or below is worthy of attention. My 0.75 arc-minutes (let's round to 1) are there just because I'm a bit coo-coo that way ;)

What model is that? I'd love to see how much it sets you back price-wise and what output torques it's capable of holding (those, multiplied by 6, would then give your rig's individual axe torque values)

Also: how did you measure your backlash (I understand it's gearbox-only backlash, without taking into account the belt drive, so it's not final backlash. also, did you test it under load or not?)

Also: did you happen to check the post-belt (final) backlash? I'm asking because my previous 100:1 gearing setup (2 x 10:1, via MXL timing belts) had humongous amounts of backlash due not as much to some stretch in the belts themselves, as to the simple fact that timing belt teeth are themselves inherently elastic, allowing for play at the small pulleys which amounts to significant play accumulating in the next stages, and resulting in an utterly unusable gearing)

Thanks,
Eros

Image


Thu Mar 26, 2015 8:58 am
Profile

Joined: Sun Mar 06, 2011 5:27 am
Posts: 12
Post Re: Affordable low backlash worm gearing?
@Daniel:

Don't you want to open a new topic for your project? I'm also thinking in this direction, I have already the DP's Stage 1 slider with a DC motor + the cart + MX2 controller, but I want to change for stepper motor + build a pan/tilt head to use later the NMX controller with smart phone.

Till now I found this topic on dvxuser.com (if you register, you can see the pictures), he also modified a ServoCity head and I have the same question, how to do it. I live in Europe too, so it should be solved here somewhere.

Cheers


Fri Mar 27, 2015 1:34 am
Profile
User avatar

Joined: Tue Nov 24, 2009 7:55 am
Posts: 951
Location: UK
Post Re: Affordable low backlash worm gearing?
Hi,
Gearbox address:http://www.zappautomation.co.uk/power-transmission-products-306/precision-planetary-gearboxes/planetary-gearboxes-in-stock.html
Motor was 0.9 degree step angle.
I assumed manufactures figures were accurate for backlash calc. I suppose could set up a laser pointer and measure but for more practical purposes I did a couple of passes with 360 degree turns applied to pan, tilt, rotate with 100cm dolly move to produced the following video:
See Vimeo description for detail ;)

The unbalanced load (camera + zoom and focus systems) seems to pass through the tipping points and reversals of the move without creating any noticeable problems :D


Regards MikeA.

_________________
Smile, it gives your face something to do!

http://www.vimeo.com/user1214873/videos
http://www.youtube.com/user/MrReggub


Fri Mar 27, 2015 1:46 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Hi, Mike, and thanks for the link!
I checked the page, but it seems it's just a generic page, not for a specific product - meaning I can't figure out, from there, what are the characteristics / what is the code designation for the gearboxes you used.
What I did notice however, that struck me as odd, is that the backlash they're declaring is for basically no-load (a load a mere 2% of the rated torque) - where even the lube within the gearbox acts as a spring to reduce backlash. True, my tests are done in pretty much the same context, but I'm not using a commercial product there ;)
But even so, the results of your test are looking very good! However, what you measured there, as you yourself mentioned in the title, is not backlash, but repeatability. Different things. Even a sloppy gearbox will produce good repeatability under identical conditions / runs. So I'm still waiting for your product codes / tests to see what the actual backlash is there.
Take care,
Eros


Fri Mar 27, 2015 2:09 am
Profile
User avatar

Joined: Tue Nov 24, 2009 7:55 am
Posts: 951
Location: UK
Post Re: Affordable low backlash worm gearing?
erosnicolau wrote:
Hi, Mike, and thanks for the link!
I checked the page, but it seems it's just a generic page, not for a specific product - meaning I can't figure out, from there, what are the characteristics / what is the code designation for the gearboxes you used.
What I did notice however, that struck me as odd, is that the backlash they're declaring is for basically no-load (a load a mere 2% of the rated torque) - where even the lube within the gearbox acts as a spring to reduce backlash. True, my tests are done in pretty much the same context, but I'm not using a commercial product there ;)
But even so, the results of your test are looking very good! However, what you measured there, as you yourself mentioned in the title, is not backlash, but repeatability. Different things. Even a sloppy gearbox will produce good repeatability under identical conditions / runs. So I'm still waiting for your product codes / tests to see what the actual backlash is there.
Take care,
Eros


Hi, Code: RE34/100-SL-5X18STD-F49-10' and RE34/100-SL-5X18STD-F49-20'. Specification is contained in downloadable .pdf file - Link is towards bottom of the page.
Test will take a while too much less pleasurable work to do currently :( I have found when backlash is present eg. from standard Phidget motors it is easily detected in the moves created in the video at the tipping and reversing points of moves.
When checking by hand, no detectable movement found after the belt pulley stage.

_________________
Smile, it gives your face something to do!

http://www.vimeo.com/user1214873/videos
http://www.youtube.com/user/MrReggub


Fri Mar 27, 2015 3:13 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Yep, it looks like 10' and 20'. I know what you mean - testing isn't always fun and games ;)
Thanks for the codes, man.
Take care,
Eros


Fri Mar 27, 2015 4:41 am
Profile
User avatar

Joined: Wed Nov 11, 2009 10:45 am
Posts: 1685
Location: Merritt Island, Florida, Estates Unitas
Post Re: Affordable low backlash worm gearing?
Thats an impressive amount of black Mike! A few very minor things drift by every now and then, but basically perfect repeatability. I had looked at those Servocity pan/tilt heads back when they first came out. My first impression was that it cost a lot for what you get, especially given a new eMotimo can be had for less. Then I made a few P/T heads and now the Servocity kit seems a bit more reasonable. The hollow shafts are real convenient. The idea of adapting stepper motors and running Dragonframe was great. Sure can't argue with the results.

Nice to hear from you again Eros! The kind of precision you are going after is impressive. All sorts of shops around me make stuff for NASA, so the precision machines here are as good as they get but it costs so much to get those guys to use them, you need to tax a lot of peasants to pay for the work. Since I am one of the peasants, I don't get parts made there. I admire your dedication to the goal. Even with all these resources all around me, I moved on to things that are easier to do. If you can make the zero backlash gearing at prices I can afford, I'll buy some and try to sell some P/T heads that use them. Of course, thats the easy part. I wouldn't be surprised if you made the heads and sold them yourself. After all, the precise gearing is the hard part. On the other hand, Mike's gearboxes aren't bad at all and they are mass produced. Do you think they'd work for your robot?

Daniel, Mike's idea of using a split-ring spacer to adapt the existing motor mounts is a great one. Get some 0.5mm sheet and cut a little rectangle that wraps 95% of the way around the gearbox. Done. I usually make motor mounts out of aluminum angle. Drill holes for the shaft and mounting screws. The orbital gearboxes have threaded mounting holes on the face where the shaft is. Cut the angle a little long so screws to the side of the motor can hold it to the extrusions making the frame of the P/T.


Fri Mar 27, 2015 5:34 am
Profile
User avatar

Joined: Mon Nov 22, 2010 1:18 pm
Posts: 1294
Post Re: Affordable low backlash worm gearing?
Slow day for me got a bit of a hangover so laptop bound :( .
But on a positive note thought this thread may be interested in this . I was brushing up my CamBam skills when i came across this hypocycloidal gear generator .

http://www.cambam.co.uk/forum/index.php?topic=4640.0
http://www.cambam.co.uk/forum/index.php ... 9#msg22739

Will have a go at some point but not high on my list happy with the gearing as it is .

D1


Sun Mar 29, 2015 1:27 pm
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Hey, @sciencelookers,
Long time no... write :)As you can see, I haven't quit on the project - it's just going extremely slow. I originally built a 1:10 lever-type pantograph meant to reduce by 10 not only the size, but also the machining errors for that cam. But the linear grinder I used it with (you'll see in the video below) had some inherent flaws (places where any misalignment could frack-up the result), so right now I'm working on building a mini-grinding machine where the spindle isn't horizontal, like with normal linear grinding machines, but vertical - thus eliminating all those possible misalignments, and upgrading the pantograph from manual to motorized, to eliminate any force and speed variations induced by my hand... I'll put here a sketch of the machine we're building for machining the last precision component I was writing you about last year. Yes, you're right, the lack of (affordable) precision machining is daunting, but... I keep trying.

so... first, the old pantograph in action (excuse my dirty look... it's a dirty shop):



And here's the sketch of the upgrades I'm working on (vertical spindle, motorized pantograph):

Attachment:
pantograph.jpg
pantograph.jpg [ 55.22 KiB | Viewed 12611 times ]


Cheers,
Eros


Sun Mar 29, 2015 10:53 pm
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Thanks, @Displacement1,
But... :) this kind of gearing was already discussed earlier in this thread (I even built one, with improvements, for testing) and what nobody tells you is that this kind of gearing is the worst in terms of linearity.

Let me re-explain: Any gear-set has, mathematically, a certain non-linearity (if you put on a chart the input speed on the X axis and the output speed on the Y axis, you never get a perfect line, but a very slightly sine/wavy line). Most gear-sets have this non-linearity under check (very, VERY small, negligible, and the linearity affected by a single sine function). This makes the chart somewhat predictable and usable.

This arrangement's linearity chart, however, is subject to not one but two sine functions, with different frequencies (one for the 10 tooth stage and one for the 9 tooth stage), in such a way that the corresponding complex sine becomes a nightmare, virtually impossible to manage. And the more elastic the gear-set, the more visible the effect of that sine, in such a way that you get a really jaggedly output for a perfectly smooth input. Too bad, since it looked SO INTERESTINGLY SIMPLE :)

Cheers,
Eros


Sun Mar 29, 2015 10:54 pm
Profile
User avatar

Joined: Mon Nov 22, 2010 1:18 pm
Posts: 1294
Post Re: Affordable low backlash worm gearing?
Sorry my bad not really been following this thread . The hypocycloidal gear may still have its use for a cheap head though lets face it phidgets motor gearboxes not great but people use them .



Seem to work ok anyways this prob all been covered so ignore me . On another note have you thought of building you own CNC or at least adapting a existing mill it may help in your development work at least.
D1


Mon Mar 30, 2015 4:20 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Be cool, mate, you yourself told us you were a bit hungover ;) and this thread is already 3 pages long anyway ;)
As I can see the guy in the video stumbled upon the same problem as me, with the hypo-cycloidal: compound error. At least I was fortunate enough to discover it at brass-plastic, small size, stage prototyping - didn't have to spend lots of time and materials building a huge gearbox like the one on the video.
About the CNC - I also wrote about this earlier in the post: I DO have access to a decent, even if not perfect (0.15mm error on the Y axis) engraving (not heavy duty, but enough for our purpose) CNC

Attachment:
PICT2759.JPG
PICT2759.JPG [ 116.22 KiB | Viewed 12601 times ]


Cheers,
Eros


Mon Mar 30, 2015 4:48 am
Profile
User avatar

Joined: Mon Nov 22, 2010 1:18 pm
Posts: 1294
Post Re: Affordable low backlash worm gearing?
Yes was a little fragile :D . Nice CNC really must read full threads before i jump in ! If you get your hands on a decent metal working mill their not hard to convert to CnC . I have one of those little Chinses mills i upgraded works fine with ali but struggles with anything much harder .
Last mention of those gears but did find this so will print one out and have a play it may be possible to print a very cheap if basic full head will see .

Hope it all comes together for you.
D1


Mon Mar 30, 2015 6:13 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Hi, guys,

I thought I'd update you with some of my news (because long time, no write...). I'll start with the more off-topic items (1 and 2, below), then update you with the "harmonic drive - type" gearbox (3)

1. The cam grinding machine: I finished the planned changes (I adapted it to a dirty DIY vertical-spindle-grinding-something; I also gave it a cute little water pump, just because I'm newbie enough to learn that "grinding heats parts up" the hard way...) Here it is, in action (and the results are beyond any expectations, regardless of the deceiving looks):



2. After MUCH struggle, after seeing like a dozen "lathe masters" (being ironic here), I finally came across one able to hold the 0.02mm tolerances I required for the three aluminum rings that I prepared for a big diameter, small cross-section, "roll" action bearing. The bearing is made of aluminum and the balls are running on steel wire races. Here it is:



and, finally,

3. I was telling you in my previous posts about my search for perfectly executed internally toothed gears. I knew WEDM is the way to go for prototyping these, just didn't know that the standard ±2.5 micron tolerances that are so common abroad are impossible to achieve in my country (due to friggin' carlessness and zero machine maintenance, as standard practices). Anyway, I did find a shop way across the country, able to do the gears to spec, and I received them a couple of weeks ago. Not mentioning the pair cost me almost as much as a Harmonic Drive (500$ - still working on finding a gear shaping or a rotary broaching machine/service for making these parts). After this, it was all a matter of finding some free time, to put everything back together.

The results:
- The motion of the gearbox is MUCH smoother than before (what for no more eccentricities, better executed parts etc.). You can still feel some areas where you need just a bit more force to turn the input, but it's much better now.
- I put up (again) a rig to laser-test the unit. Same as last time (remote control of a stepper motor connected to the gearbox by a timing belt and pulleys pair). I put the gearbox in a separate room, to keep floor vibrations to a minimum (you'll see what somebody entering the room at the end of the first video does to the stability of the fixture). This first video is just to show the testing setup (you'll also get a taste of my native language, Romanian):



And this second, shorter and to the point video, shows our measurements on the scale:



The results appear to work out around 1mm - 1.5mm error for a 10m distance (1.5mm error at 10m stands for 0.0086 degrees, or 0.52 arcminutes of backlash).

The next step: coming up with an idea for testing the same thing, under load, to exacerbate the error under a known load (I don't know, maybe a brake, or something, attached to the output?)

Thank you for the patience...
Eros


Thu Apr 14, 2016 12:46 pm
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
Hi Nicolau That's impressive backlash control. Especially pleased to see you moved to position from both directions. You are a maniac... Defense contract for you next maybe?

I think the next experiment should be to test smoothness when running at various speeds. I have tried certain HD gears from China, and while the backlash was good, when loaded with a camera and long lens there were vibrations at resonant frequencies. Even genuine HD gears can have this issue if they are under-specified. I think you can understand if the camera load has some inertia, and there is even a small amount of flex in the structure, this can easily happen. This can also be caused or exaggerated by stepper vibrations, so a carefully tuned servo motor might be necessary...

Gerald


Thu Apr 14, 2016 10:59 pm
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Hi, Gerald,
Why do I sense a lot of sarcasm in your tone? This saddens me... I'm just a guy with limited resources trying to make the best of it...

About the smoothness: isn't this more of a concern with video, than with photo / timelapse? I mean, with timelapse, I suppose you never shoot while the rig is moving, right?

I'm struggling alot with this smoothness - even this last version, which reallly maxes out in terms of what I can realistically achieve in my country, still has small areas of slightly higher friction, when moving it at very low speed, by hand (so I can better feel it). Not a big deal, but I still can't feel in my hands the smoothness I feel when turning a similar, 1:50, 70mm diameter HarmonicDrive unit. Then again, you find exotic and elastic bearings in the HD, whereas in my unit instead of a bearing I use a flexible bronze-stainless-steel interface. Not as elegant, tribologically speaking, but hey, it works. No idea how to test for smooth running at higher speeds (are there special instruments for that?)

If by smoothness you mean (lack of) vibrations, though, I can tell you I didn't feel those anymore in the dangerously stupid "keep the gearset by hand in the lathe and run it" test. I mean, in previous versions, where these periodic friction areas were highly pronounced, heeping the gearbox by hand while spinning the input via the lathe used to be a very "vibratory" experience. None (or incomparably less) of that now...

About servos: I got in contact with an american guy who programs and sells programmable chips for steppers/servos and, based on his patented chips, and with the help of a local electronics engineer, we managed to prototype a complete hybrid stepper (and by complete I mean: stepper + encoder for closed loop feedback + driver + controller) run via a USB cable that can accomodate up to 32 concurrent units. It's basically a complete servo (with reduced noise, reduced power consumption, increased peak torque) just run by a stepper. And it's the unit you're seeing in the previous video, running the motor that drives the gearbox's input. The episodic vibrations in the system come from the servo jittering in search of the correct end position - a clear sign that the servo still needs to be fine-tuned.

Here, the same servo running an experiment for a slider:



And here, showing the servo "hold position" behaviour:



But that's another story, maybe for another thread...
Eros


Fri Apr 15, 2016 11:50 am
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
Hi Nicolau

My comment was not meant to be sarcastic. Perhaps it was lost in translation a little? If this gear is very good then it would certainly have applications in industry and military. I mention military because they usually have more money and require cutting edge standards... :)

Smoothness is naturally most important for live action. Another way to test smoothness is attach a small beam to the output with some weight on it - like a small boom arm. When it turns at a regular speed you can tell more easily if there is a problem. If it is extreme then you will see vibrations, but a more scientific method might be to attach an inertial sensor.


Fri Apr 15, 2016 4:35 pm
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Hi, guys,
I thought I'd come back with a little update.

I re-did all the backlash tests again (same as always, with a laser - this time at 13m away from the ruler).
This time I did the testing with and without an external load (for external load I used a power-off brake which is specified at 2Nm).

Image

The results:
- no load: backlash between 0.8 - 0.9 arc-minutes (between 3 - 3.5mm at 13m)
- 2Nm load: backlash between 23 - 31 arc-minutes

Since I don't have multiple brakes to test with, only this 2Nm one, I'm guesstimating a rigidity loss of ~12-15 arc-minutes/Nm (around 1.2-1.5 arc-minutes / Kg-cm). Not sure how this stacks up against real-world, professional gearboxes, but my main question for here (and I'd really appreciate your feedback) is: Is this kind of rigidity good enough for time-lapse purposes? My intuition tells me that for a well-balanced system, it's more than enough, but I'd really like to know what you think.

Thank you so much for following through,
Eros


Thu Jun 23, 2016 9:01 am
Profile
User avatar

Joined: Wed Nov 11, 2009 10:45 am
Posts: 1685
Location: Merritt Island, Florida, Estates Unitas
Post Re: Affordable low backlash worm gearing?
This has come a long way since your gearbox project started. You have made an amazing amount of progress. The tests with the laser were very impressive. Even under load, the precision is very good. Amazing actually. The finish on the parts in the last video is also very nice. Really beautiful. And the last two videos with the laser look more than accurate enough to me, although I admit to having no practical experience.

I think the most demanding application is for multipass, whether it is timelapse, stopmo, or realtime movement. The second and third passes have to be exactly like the first, or visible imperfections get introduced into the composited shot. A black test like Mike showed us is a good test to see how well the rig repeats the move. Maybe some of the guys who specialize in precision and multipass can help out here. Gerald, Doug, Mike, D1, Edward, and anyone else qualified to comment, can you please tell us something about the specs you consider good enough to do an acceptable multipass shot? Of course, the focal length of the lens is going to have its effect by magnifying the error as focal length increases, but for a normal 50mm on a full frame camera, what would the spec be approximately? This is something those of us aspiring to follow need to design for, but is never spoken of.

So bottom line, what is the acceptable amount of non-repeat-ability in an exactly repeated movement? This is the fundamental question.


Fri Jun 24, 2016 8:47 am
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
SL.

While it's nice to have absolute numbers, but ultimately a real world test is will tell you if the system is adequate. For instance there might technically be some backlash, yet repeatability is still be good enough.. First test individual axes in isolation. In the scene make sure you have some sharp horizontal and vertical lines so when you composite using a simple split screen you can see if there is misalignment or wobble. The level of accuracy is also driven by the way shots will be used in post. For instance if you shoot a background plate and then an isolated green screen foreground element, absolute alignment is less critical.

Also make sure the speed of the move is fast enough to clearly distinguish consecutive frames - if it is drifting languidly you will have more trouble seeing any issues. Shooting at reduced shutter angle will also help - ie. less motion blur.

Finally - the lens matters but is relative to the shooting distance. Even a wide angle lens can be very critical if the subject is close to the camera. In fact any wobble will be more evident on a close subject. Whereas a distant scene may not reveal the same wobble.


Fri Jun 24, 2016 6:40 pm
Profile
User avatar

Joined: Wed Nov 11, 2009 10:45 am
Posts: 1685
Location: Merritt Island, Florida, Estates Unitas
Post Re: Affordable low backlash worm gearing?
OK, So this is interesting. Not at all what I expected, but makes perfect sense.

Acceptable precision being defined by the visible effect on whatever shot you are attempting at the moment. This is very much the artist, DP viewpoint. Somehow worlds away from the engineering, machine building, laser point testing mindset I find myself in when I build something. Somehow I have a hard time thinking both ways at once. Probably why I get bogged down building so many things and find it hard to break away and go shooting with all this stuff. And when "on expedition", I find it hard to deal with machine shop stuff.

Eros, I think you're there already. Attempting to add more precision to get the perfect results even with a load may be torturing yourself over a margin which is not noticeable in the final composite shot. The laser pointer is a very impressive test, but try doing a black screen test like Mike did. If its all black, or mostly black with only the tiniest bit of non-overlap showing, its ready to be made into a pan-tilt head or some end-use device.


Fri Jun 24, 2016 7:05 pm
Profile
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 146 posts ]  Go to page Previous  1, 2, 3, 4  Next


Who is online

Users browsing this forum: No registered users and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB © phpBB Group.
Designed by Vjacheslav Trushkin for Free Forums/DivisionCore. pozycjonowanie