It is currently Fri Nov 15, 2019 8:17 am




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?
sciencelookers wrote:
Awesome! Can't wait to see what you come up with. It would be nice to see something other than the standard dolly-pan-tilt. Send me a PM when you have some reducers you can part with.


:) my quest is for a dolly-pan-tilt-roll-zoom-focus rig complete with programmable camera controls... so yep, a bit more ambitious

about the reductions: will sure do


Mon Mar 25, 2013 5:41 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Coming back with some updates...

1. The third prototype of the reduction is ready. Improvements: I replaced the outer pins with miniature ball bearings, for reducing any grinding friction. Here's a video of the thing working at 2000rpm input:


2. The bad news: it looks like the mathematical model of these types of reductions has a non-linear output. In other words, instead of:
Output speed = Input speed * Ratio
we have
Output speed = Input speed * Ratio * Complex sine function

Noticing this odd behavior on my double stage hypocycloidal, I set upon finding more info about this. Didn't find much on the hypocycloidals, but did find quite enough about the harmonic drive - where the only mathematical differences lay with the fact that they're one stage as opposed to two stages, and that they have double contact/grinding point, as opposed to one:
- a scientific paper here
- a thread on a hobbyist/machinist forum - see post #2
- an MIT guy re-stating the same problem

In the case of the harmonic drive, these sines come from a single stage, and are - in theory - somewhat predictible / accountable for, for adjustments.
In the case of the two stage hypocycloidal however, we have ... 2 stages ... so predicting the combined sine from 2 separate out-of-phase functions is more than I can feel comfortable taking into account.

So, for now, in order to recover some of the time we spent on developing this drive prototype, we're putting the rest of the project in high gear - literally with timing pulleys and belts ;) - hoping to come back and study some more the hypocycloidal proposition sometime later...


Sun May 19, 2013 1:35 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?
Sorry to hear about the complex behavior of the hypocycloid. An affordable low backlash speed reducer would be a huge benefit to everyone building motion control rigs. You'll sell a lot of them if you can make something work. How large are the deviations from linear motion? Is it really noticeable in a test run while timelapsing?

If you want something quick and easy, there are these stepper motors with attached gearboxes. This link is to the 100:1 reduction but they have similar ones with 30:1 and 5:1.

http://www.phidgets.com/products.php?ca ... _id=3319_1

They should be fine for a pan/tilt head used indoors. If you need to use it outside, consider adding a little drag to the drive train somewhere. It would be easy to make an adjustable brake so you can increase the drag if wind is moving the head within the backlash limits. The drag prevents the wind from moving the head. The motor overwhelms the drag when its time to move.

It doesn't solve the basic problem but its cheap and there are probably dozens of ways to implement it.


Sun May 19, 2013 5:34 pm
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Hi,
Yep, imagine how disappointed I am... Well, failure is part of any design process, so I welcome any experience that teaches me something.
About the phidgets gizmo: I know about it, it was "on the table" from the very beginning - except it has a huge backlash - 1.8*, a whole step and it's not auto-braking...
We'll be using custom timing pulleys (17teeth with 170 teeth) and because these sizes aren't commercially available, we're building them in-house. In theory, timing belts, while not quite so elegant, have almost zero backlash, which can be always adjusted very easy by tensioning the belts with idlers.
Plus, we'll be including a passive brake on the last gear (output) which will permanently stabilize the whole thing while stationary.
I'll keep you guys posted with our progress ;)
Eros


Mon May 20, 2013 12:45 am
Profile

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

I haven't been sitting on my arse since my last reply. For now, my attention is focused on a cheaper way of producing a strain wave gearing system, similar to the infamous and very expensive HarmonicDrive, but a bit more complex (I know, I love simplicity more, but can't afford it) and easier to build on a CNC.

My current prototype, just as compact/lightweight, provides indeed near zero backlash AND features none - if minimal - of the irregular output speed "featured" by the hypocycloidal one - its only drawbacks so far are a lower efficiently (around 30%) and not such a great output torque capability (around 20Kg-cm). I'm already working on alternative meshing profiles for another, sturdier prototype, but in the meanwhile here's a video of this first prototype running in/grinding away, mounted on a on a lathe.



Attachment:
File comment: mounted on a stepper motor
strainwave.jpg
strainwave.jpg [ 119.53 KiB | Viewed 20212 times ]


Fri Jan 10, 2014 7:08 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?
Looks like you've got a working gearbox. Are you still planning on using these for the three motion control robots? It seems like a lot has happened in the motion control arena since you started this.

Given the time you've had to think about this, what do you want to do with the design now? Are you planning on just making the ones you need for the robots or are you going to optimize the design for CNC or some other production method so you can sell gearboxes? eMotimo still has its monopoly on reasonably priced pan/tilt heads, in case you have ambitions of manufacturing. There are still no good, affordable pan/tilt heads out there aside form that one that I know of.

I'd be interested in incorporating these into a P/T head but there needs to be some room for a profit on making the rest of the head. We are in the process of developing our own products and the website to sell them. If you still want to make a pan/tilt/roll head with these, we can maybe work out an arrangement to sell them on the site if you're interested. I've spent most of my time designing and building countless variations on focus/zoom motors but have now settled on two awesome models. Coincidentally, I was starting to get back to a P/T head because, to my dismay, none have emerged to take on the low-end consumer MOCO demand. Would be happy to add yours to our product line if the economics allow us to both profit.

Dynamic Perception had announced their plans for a universal gear reduction block to work with their quick-change motor mount, but fell silent on that topic a while back. If you can manufacture these economically, you might contact Jay and ask whats the status of their project and if they'd make your head a new DP product. They are in an interesting situation since Meade quit producing the Merlin head, depriving them of a P/T head for their MOCO systems. eMotimo is making great traction getting people to convert DP dollies to stepper motor drive and using TB-3's as the controller or as a Dragonframe bridge.

A big opening here for anyone with a competitive pan/tilt head, is what I'm trying to say. Wait too long and I'll be making one of my own.

Any interest in focus motors to go with this stuff?

Our site;
aonenet.net
The link on the second line that says "shop" is active, and you can view or buy any of the things listed right now.


Fri Jan 10, 2014 9:40 am
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
Lots of hard work here with these prototypes. Do you know harmonic drive actually had an offshoot company that made plastic HD gears for relatively cheap? Unfortunately last time I checked, they were not responding to emails... so I've never gotten to try them out.

There are two companies that make industrial versions of the cycloidal gears. Sumitomo (retailed as Vigo Drive) and Spinea from Slovakia. I have samples of both. The Vigo drives are mostly quite large, hence not suitable for pan tilt, but Spinea make some small drives which are excellent. Albeit really expensive!

While it's commendable to make your own gears, I think to achieve the level of low backlash and linearity of output that you'd really want - it's impossible without highly accurate machining such as the Japanese and Slovakians are capable of!

Also - if you look on Ebay you'll often find surplus HD gear sets for reasonable prices. So why bother? - besides for the heck of it I guess.. :)

Second - if you want a low backlash worm gear it's actually not very hard. Just use spring loading. The drive worm does need to be hinged at one end, so the motor drive input needs some flexibility, but it's not impossible. I have a a pan tilt head made using this technique and it works very well. You just have to take care with heavy cameras to balance them well - or else uneven forces can unload the gear, depending on the tension of the spring.

In the 90's in the US, a company called Sorenson made very good motion control pan tilt heads for big film cameras using the same idea. In fact a big advantage is that if you release the worm enough, you can free-wheel the axis, which can be really handy to reposition the pan or tilt - rather than laboriously inching the motor.

Here's a still of one of the worm gear heads I have. The gear sets have 37:1 ratio. This is probably the minimum you would want.

Image

The inching knob at the front can be pushed down to release the worm. In this case the motor is hard mounted to the worm shaft - so kind of floats with it. If the shaft is not perfectly straight you can see the motor wobble little at high speed, but in practice this has not been a problem. A better design would be to use an offset belt drive to prevent this.

If you want more details I'll make a drawing...

For many people making this kind of head is not an option because of the amount of machining required. But I think in your case (erosnicolau) it should not be a problem?


Fri Jan 24, 2014 6:36 pm
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
@sciencelookers:
Hi, man. Yep, that's what we're experimenting with for - our robots. Should we eventually find ourselves with a cool, working, manufacturable product, we'll be more than happy to start some product sales talks. But for the time being, my primary focus is on making sure these babies actually work as supposed to, reliably. We've still got some road ahead before that happens (improving on the efficiency by bettering the use of solid and flexible bearings, improving on the actual output torque by using actual steel instead of the easily machinable but soft brass etc.) but the direction is set: forward :)
About the other market players you mentioned to me, here are my newbie thoughts:
- eMotimo: cute toys for the beginner. PROs: KISS - oriented, which is great; very lightweight, which is also great. CONs: resolution/repeatability is not even mentioned - see those gimmicky plastic gears; no parallax centering; limited to only lightweight cameras.
- Dynamic Perception: definitely more design-aware, very nice looking pieces of kit. PROs: good looking, more serious stuff; more robust. CONs: still no word on resolution/repeatability (actually, the very nice VX1 blocks you were talking about are simple worm-screw drives, end even at that, with not such a great gear ratio: only 1:10) - however, a thumbs up for their constant thinking in terms of snap-on simplicity; also, no parallax centering (actually: am I the only one freaking out about parallax centering?); another CON: the pricing ($400 for a 1:10 gear that you don't even know what backlash it has, seems unrealistic). I would love to talk to Jay - thanks for the idea - but only when I know I can present a solid gear :)
About your products:
- we're also trying to make our own focus motor assemblies - but now that you told me, I also keep an eye open for yours, as a possible snap-on option ;)
- also: very interesting that you can and do design custom electronics for this field. we had to work on ours completely deprived of any electronics help, which on one hand limited us in quite a few choices, but on the other hand forced us up this learning curve and made us come up with the creative solutions we're currently using. However, I'm definitely interested in what kind of electronics you might help us with, in the next stages. Our still open problem: a lightweight and powerful power supply (my money would be on the LIPOs, but they require a good charging electronics kit which is still an unknown for us)
About your site: very nice looking, good job! - too bad though that it still feels a bit under construction, in terms in content - but I'm sure your focus is on development right now :)
Thanks for your encouraging words, I'll be back with news as soon as I have them.


Sat Jan 25, 2014 12:04 pm
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
@geraldft:
Hi, there! No, I didn't know about HD's plastic drives. Sounds very lightweight, but not too torquey.

About the near-zero backlash market:
1. besides HD (japanese guys, with a foot in Europe via Germany and a foot in the US - nice guys, I've been discussing with some of them), there's also a Chinese factory, TOTEL, independent from them, looks-like, which manufacture the same HDs, but with sadly the same price ranges (quite surprisingly, when you consider the Chinese are actually much cheaper in many fields - but alas, not here). I've also been considering getting HD/Totel component sets (just the 3 elements you see in the ads) and to build the rest of the gearbox in house - but for now the manufacturing costs for the gearbox would be even pricey-er than purchasing the complete gearboxes - so no.
2. Sumitomo indeed produce extremely large hypocycloidal combo drives, quite off-target here. I've been having some talks with the Spinea guys too - but their smallest current drive weighs in a hefty 4.2Kg, at a prohibitive 650eur, which to me and my needs makes absolutely no sense at all (since for that money I can get a 1Kg HD gearbox). Besides these two, there's also Standa, from Lithuania, with an entire array of high-precision gears and indexing tables - but with the same unrealistically extremely high prices and, as a "minus", with very low output torques for the reasonably small units. Basically, for their market (optics manipulation), output torque is completely irrelevant.
3. I've also been considering the double-enveloping worm-screw slewing drives of KinematicsMFG - another bunch of great and very helpful guys - but alas their currently smallest unit weighs in a more than hefty 18Kg.

About my precision target: I aim to achieve a max. of 3-5 arcminutes of backlash. With my current 2-stage timing-belt drive, I'm there. I hope to also get there with my strain-wave pancake drive too.

This brings me to your talking about the spring-loaded anti-backlash worm-screw gears (BTW, some other guys I saw around here use, as an alternative, 2 counter-rotating driving servos, for the same purpose of pre-loading for backlash). Any such preloaded system has two major drawbacks, in my opinion: 1. the output torque is limited to the power of the conter-force device (be it spring or secondary motor) and 2. pre-loading equals massive efficiency loss, which calls for much stronger motors, or for double motors in the other case, which results in a bulkier or a more energy-hungry system. When you use this near a 220V socket, that's not an issue - but I'm designing for the outdoors.

Why do I insist on high output torques: because I am the primary beneficiary of my robot, and I plan on taking it up the mountain with me. It currently weighs in a hefty 13Kg without battery - any additional counterweights would only add up to dead mass to squish down on my spine. Why is my system potentially unbalanced: because I am designing it to be parallax-centered, from 8mm lenses up to 200mm lenses. Simple math tells me that a parallax-centered 200mm lens, with a body at its end, generates, for TILT, a minimum of 20Kg-cm. Add to this the motor and gearing for ROLL (another easy 30Kg-cm), the 3-way sliders for parallax-centering (another ~5Kg-cm) and you come up with pretty high torques. Of course, this is the extreme case of 200mm (at 8mm the thing is a bit more balanced out) - but I know what I want and unfortunately what I want is pretty demanding in terms of torque. At 8mm I can't use counterweights because they'd enter the frame, but at 200mm this wouldn't be a problem - it's just that my back aches only when thinking about additional dead weights. Your P/T head has the TILT almost perfectly balanced - that's because it's not parallax-centered and because you don't have the ROLL component - that's why you can very well get away with a preloaded antibacklash drive...

Just to give you a potentially useful idea for a perfectly ZERO backlash drive - if you are using a very low torque, perfectly balanced setup: you can use what the astro-photographers consider the best solution: well hardened, perfectly machined, spring-pressed, direct metal-to-metal, interior-tangent, smooth wheels. The direct contact leaves no room for any backlash and, if you have the machining capabilities for this, it's the absolute best in terms of backlash - zero, as with direct drives. However, as I told you, this is only for perfectly balanced systems.

Just so you get a better picture of my current setup, here's a rendering, below, with (currently 2-stage timing belt) drives, enclosed in boxes: 1. DOLLY, 2. PAN, 3. TILT, 4. ROLL
(this is sketched with the extreme load: the biggest DSLR body out there, loaded with the parallax-centered 200mm lens)
Thank you for your thorough and thoughtful feedback!
Eros

Attachment:
File comment: current, incomplete, rendering
robot-boxes.jpg
robot-boxes.jpg [ 71.52 KiB | Viewed 20153 times ]


Sat Jan 25, 2014 1:01 pm
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
Hi Not sure who you spoke to at Spinea, unless they stopped making them? I have two of the M-series.

http://www.spinea.sk/products/twinspin/m-series/

They weight 500gm and have very high output torque for their size. Inside they are built like Swiss watches - you can't make something like this at home... but they do cost a lot of money, so only the military can afford them. I did justify a pair because I use them for professional work and they have paid for themselves by now...

Vigo's are bigger for sure - but smaller ones could be used to rotate the base of a medium size robot arm. In fact I use one to rotate a boom arm crane.

Plastic harmonic drives would work well for a well balanced head. Though your design is extremely unbalanced - so much more torque would be required, especially on the tilt axis. That pretty much demands a worm gear I think - such as used on the VoloCrane head. Backlash would also be irrelevant on that axis due to the constant loading...

As for objections about preloading worm gears. No.. in practice the loading does not make a significant problem for the motors. First you control the pressure so it is enough to hold the load but not too strong. Second - Steppers have lots of torque. The head I use works very well and has plenty of speed. The one in the picture has nema 23 motors but I have another that I changed to nema17 and it works equally as well. Third - if the load is a bit unbalanced, the spring only releases in one direction if the force is too great, so you can offset the loading a bit to this side if needed.

What is the problem with making the head balanced? It's really not important to have an exactly nodally centered camera for cinematography. Yes for the roll axis, and pan - but tilt does not need it. I suspect maybe you are creating unnecessary problems?

G


Sat Jan 25, 2014 5:49 pm
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Hi, Gerald,
Now that you prompted me to search through my emails, I can clearly see that I made a confusion: the M-series they told me about is 0.47Kg, not 4.2Kg as I erroneously remembered. Still the price tag is correct (650eur/piece - same as with the harmonic drives) and the reduction ratio is 1:65 (and my TOC fine traces push me to try and work with 1:100) - but thank you for making me realize my mistake!
I know Spinea's gearboxes are extremely complex and minute inside - this is visible from their product presentations - and I wouldn't even attempt to create such pieces of jewelry - my attempts are based on much simpler mechanics and machining methods.
The guys at Nabtesco (home of the Vigo) didn't even bother to reply to my info requests - so my guess was they only work with industrial partners, and I couldn't make an idea about the sizes or pricing of their products.

You're right: the only potentially badly unbalanced joint in my design is the tilt one - all the others could just as well be designed to be balanced out. Thing is: I don't plan on using this particular design for cinematography, but for landscape and night-sky time-lapse - so perhaps I am over-complicating here (I usually go by "always aim for more, to get what you really want"), and when you're tracking the night sky, you might quite often want to tilt up and then continue the movement down on the other side - meaning that backlash does come into play even there, for such situations.
What I realize from talking with you guys is that perhaps my standardized approach is wrong - and maybe I should only use higher torque gears for TILT - and smaller, lower torque gears for all the others - that's food for thought ;)

About the stepper's torques, here's my humble 2 cents: you're thankfully working with pretty balanced setups and perhaps that's why you might have not come across the problems encountered when pushing the steppers to their limits (but my guess is I'm wrong and you did, since I'm the newb here). Anyway, I did come across these problems, and I learned the hard way initially, then the theoretical way, that steppers actually have 3 speed-torque charts, not just one - if any (A.: no acceleration needed, no applied torque, B.: no acceleration needed, applied torque and C.: acceleration needed, applied torque), of which only a limited (B.) area is relevant for our use - and that area is generally at around half if not less than half of the quoted "holding torque". And, by the way, the quoted "holding torque" is just a veeery rough approximation of the peak of such a speed-torque chart, where you only get this max at a very specific speed, amperage and voltage combo, and get much less on any other config. As with cameras, it's all just a big marketing bull/bluff, unless you hit hard brick and start chewing information up. The simplest way to get around this is to over-specify the motors - the other way being to improve on the efficiency and the balancing of the system.

These days I'm working on finishing the assembly of the first robot prototype (the one driven by timing belts - with no focus&zoom yet) - so I hope I'm able to come up with decent pictures soon.

Thanks again for your thoughts,
Eros


Sun Jan 26, 2014 4:34 am
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
Best of luck. Choosing motors and drives takes a bit of experience and trial and error - the numbers don't really tell the story so well.

Holding torque is just that - holding torque - even so it's an indication of the motor power. Sometimes you will see torque vs rpm graphs - but manufacturers avoid them because the motor performance can vary so much according to the drive and voltage used.

The next critical setting is current - this effects the holding torque, but too much overheats the motor.

Next the most critical factor for maximum speed is the operating voltage. Higher voltage will allow higher rpm - in other words the torque will not roll off so quickly.
Again motor heating may increase - but motors with coils optimised for higher voltages will heat less. This means the inductance and resistance of the coils is a little higher...

Even so steppers in general have the most torque at low rpm, so this means to best utilise their power you may need to adjust gear ratios.

Finally - very small sized motors like nema17 will rev the fastest. They can easily do over 2,000 rpm with suitable voltage and driver. Nema 23 can run to 1500 or more with no problem.

Big motors like nema34 will run slower.... in the Mantis section there is a discussion on motors and drives - you might also find something there...

On your rig the issue will not just be torque but making the support bearings strong enough. For instance pan doesn't need very high torque, but needs to support the load well. Actually it looks like a point of some stress if the pan tries to go too fast. But maybe you are only shooting slowly?

Have fun and we look forward to more results... g


Sun Jan 26, 2014 7:27 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Thanks! And: you're right, trial and error did prove necessary steps in my learning curve, so far - guessing it'll improve with time.

Actually that you brought about the subject: my current voltage/amps/drivers setup (12V / 2.2A / Pololu DRV8825) only lets me achieve as far as 410rpm, irrespective of the motor or stepping mode (tested with multiple configs of NEMA14 and NEMA17 steppers) - so for now I'm not aiming for high speeds anyway. Hopefully @sciencelookers might help in terms of electronics for faster speeds, in the next stage, when we'll move on to another, more real-time-tracking setup.

My research for higher speeds initially pointed me in the direction of "lowest inductance possible" and "supplied voltage 4 times higher than the rated one" - and I did try that too, but with the same results. My money would ultimately be on the Arduino/Pololu combo frequency being too slow for this - although many people seem to achieve in excess of 1500-2000rpm with Arduino/Pololu, and Arduino, at least in theory, IS capable of quite great frequency pulses... It's still a mistery to me. (Others point to another formula: max speed would be given by a voltage equal to 32 times the square root of the inductance value - which results in ridiculously high voltages - not tried as of yet)

Speaking of low RPM: I noticed that "low RPM", as low as you might expect it to be, is actually not that low: generally, as far as I observed experimentally, below 200rpm all the motors are completely torque-dead (massively losing steps under torque - that "A." zone I was telling you about earlier).

P.S. Indeed, the panning joint is very sensitive in terms of lateral forces - that's why for that I'm using a one-piece, 10mm diameter, specially hardened steel shaft, fixed in solid blocks set apart for improved leverage.


Sun Jan 26, 2014 7:53 am
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
Hi - no I've never heard of motors losing steps below 200rpm as a matter of course. They might lose steps if there is really bad resonance (vibration) that you sometimes get at certain key frequencies. This is only a problem if you use very cheap drives which have little or no anti-resonance control. If you are losing steps then something is wrong.

High speed limits might be caused by the drive having a maximum pulse input frequency, or by the controller not being able to produce pulses fast enough.
You might recognise this by the motor hitting a wall at certain rpm.

As for voltage - of course the maximum depends on what the drive is designed for - so check the specs for that.

Ok - and if the motor drive has no microstepping ability the motor will run very coarsely. About 10 microsteps per step is ok which gives 2,000 steps per rev for a typical 1.8 degree motor.

If you only have Pollulu drives I would strongly suggest you try one of the smaller Leadshine drives like the DM422. Then you will have a baseline for how your motors should work... :)

G


Sun Jan 26, 2014 8:16 am
Profile

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

When I talk about step loss at low rpms I am referring to torque-load conditions - otherwise of course the steppers run as supposed to.

About the drivers: there are so many options to choose from, starting with these cheapest Pololu drivers, going through the much more reliable DM422 form factor (of which there are so many on the market), the Trinamic's, the Gecko's etc. Since I am designing a 6 motors robot, the costs add up very quickly and so does the bulk of the electronics - so both cost and form factor came into play when going for the Pololu's drivers.

About the speed limitations: the specs of both Arduino and Pololu allow, in theory, speeds in excess of 4000rpm, WITH micro-stepping - so hitting a brick wall at ~400rpm, even without micro-stepping, is somewhat of an enigma for me - even more so when you take into account that the REPRAP community is based on this config, running at very high speeds. You can check out this guy http://www.youtube.com/watch?v=LciTbXMnSz4 - using the same driver he reaches 4200rpm without breaking a sweat.

About resonance - I read a lot about it, but I've yet to encounter it - perhaps it occurs at speeds higher than my current max...


Sun Jan 26, 2014 11:57 am
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
Hi It's a bit of mystery... Clearly something is wrong though. Are you using microstepping? Maybe your motors are a bit weird? Maybe your power supply is not the best? What is generating the control pulses? Maybe put a frequency meter or oscilloscope on that and check.

If you don't know what resonance is, then the motors must be running quite rough already. If you accelerate them very slowly you'll find it at around 1 rps in a narrow band. Also slightly less at two other bands maybe.


Sun Jan 26, 2014 4:10 pm
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Just a quick update: I finally got a chance to tweak the reduction in the video above - finally completely eliminating any trace of braking/blocking points (now it runs as smooth and nice as any other reduction) and strapping it onto a NEMA42 motor like the one in the picture above - and to check for backlash, without load.

Here's how I did the checking (will follow up with a demonstration video very soon):
1. I fixed the motorized reduction onto a vice, for stability, 8.5m away from a white wall
2. I strapped a laser beam mini-flash-light onto an arm fixed in turn to the output shaft of the reduction and oriented the beam to touch the wall at reachable height.
3. I commanded the motor to rotate 500 steps CCW, then 500 steps CW
4. I marked the exact position of the beam on to the wall
5. I commanded the motor to rotate 500 steps CW, then 500 steps CCW
6. I marked the new exact position of the beam on to the wall
- I repeated steps 3 through 6 multiple times, to check for consistency

The results: the difference from the positions of the beam coming from one direction and the positions of the beam coming from the other direction come up consistently to an average of 0.5-1mm
At a distance of 8.5 meters, this results, for 0.5mm, in an angular error of 0.00337 degrees (or 0.202 arc-minutes, or 12.133 arc-seconds). I'd be safe to say then that my reduction's error varies between 12 and 24 arc-seconds - either way, well below one arc-minute. Can't tell you guys how exciting this result is for me!

Next step: making a demo video and testing for torque performance of the current model. Next step after that: I already have under development a new, stronger, more bearings-equipped model, which should be even more stable. I'll keep you posted :)


Wed Feb 12, 2014 1:11 pm
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
Sounds very encouraging. I presume you solved your stepper problems?

The second step will be to drive a load (camera) and check if you have any vibrations. Third - testing over time to test reliability. Then if all that works you will have something...

I wonder why this hasn't been done if it is relatively cheap to build? The principles are well established after all... (careful in case there is a patent...)


Wed Feb 12, 2014 4:05 pm
Profile

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

No, the steppers still run at relatively low-medium speeds (around 420rpm) but they run, so I could check for backlash precisely by using a motor. About vibrations: I'm sure there might be some vibrations when running - even the motor generates a small amount of them. Not too concerned about this for now, though, since I don't plan on using these babies for live action - not yet. Yes, the load tests are next :)

About why hasn't anybody done the strain-wave generator my way: no idea. Fact is, I looked for ideas far and wide, browsed through hundred of patents (and I met dozens of strain-wave patents, with differences between them down to just the shape of the tooth - my differences are bigger than that), looking for ideas, but none of those I found - and I found quite a lot of models - looked feasible relatively cheaply, in a CNC shop, and none looks like what we came up with. Even I had my fair share of doubts about its precision (if you remember, I was aiming for ~1..5 arc-minutes) - you're bound to have doubts, when you see machining and installing tolerances required by the standard harmonic drives & co (to within 5 microns) - but apparently the no-load results give me positive signs.

I'll look for some specialized intellectual property attorney help, to try and see, at least in broad terms, if we're infringing on some patents or if there is actually some new value to our idea.

Eros


Wed Feb 12, 2014 11:29 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?
WOW! This has come a long way since I saw the thread last. Sorry I missed the call a while back. I've used the Allegro drivers and the TI 8825/8824's. I also tried some of the conventional sized TB6600 drivers.

The big ones sure are nice. You set the DIP switches to set the current and microstepping and thats it. No screwing around. The motors seem to develop more torque and stay cooler. No problems. They're kind of big to be putting them in the P/T head or on the trolley though. I ran a wire chain alongside the dolly track and ran the motor wires through it. A bracket holds the drivers at the end of the dolly track.

All the little drivers seem to be finicky and need some adjustment if you change anything significant, like dollying uphill. For some reason, the Allegro seems hardest to get adjusted just so. The low current 8824 is easier to adjust for small amp motors, at least partially because you're not always trying to make itsy bitsy adjustments within the lower 20 percent, which would be the case if using the 8825 on a small motor. I can almost always hear the motor making a different sound going through resonance as its just starting out a jog or just slowing to a stop. It comes just a second after the motor first gets going and long before its up to speed. its just in a narrow speed range way down at the slow end. Not hearing it with the Toshiba 6600's.

Your speed reducer project sounds more encouraging now than it did a while back. Congrats on solving the problems along the way. Any idea when I can buy a copy of the production units? I'm in the process of designing a P/T head for hobby use. I'd design a new, more expensive one for advanced or indie shooters if I can just buy drives that don't break the bank.


Thu Feb 13, 2014 12:31 pm
Profile

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

This is just a quick update (video + photo) with the second prototype, more precisely and seriously built (closed, reduced axis wobble, also working on a tolerance adjustable cam).
The graphite-enriched grease you see - is oozing out of the holes used for attaching the reduction to the robot's arm (once attached, those holes are plugged by the screws and alignment pins, so that oozing is of no concern in this sample video test)

Will come back soon with performance test results.
Eros



Attachment:
2014-03-27-21.01.jpg
2014-03-27-21.01.jpg [ 74.32 KiB | Viewed 19898 times ]


Fri Apr 04, 2014 3:52 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?
Looking awesome! You've put a lot of work into it. Good to see it working. Are you planning to make a lot of these or just the ones for your robot?


Fri Apr 04, 2014 8:08 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Thanks, man! I put the whole robot on halt for some months now, because I felt I need to focus all my energy an time on this reducer before stepping anything forward. That said, I plan on building them for our robots first but, more importantly, I'm already looking for OEMs and looking to kickstart the production of a first 1000 units batch in sync with filing for a patent for our idea, should I get the financial backing.

Oh, and one more update: I powered up my motors with 23V instead of 12V (my current electronics board, Megatronics, is limited to 24V) and I got both almost double speed and almost double power. So, my problem with the motors appears to have been somewhat related to the voltage supplied, too.

Eros


Fri Apr 04, 2014 8:23 am
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
Looking good. A couple of tips to make it more flexible and simplify construction of pan tilt heads...

Use a decent bearing on the output shaft which can support offset camera weight etc. If you don't have two internal bearings to support the output flange then something like a crossed roller is good. Output should be optionally a flange with threaded holes or a shaft.

Input should also have option of direct motor mounting or a shaft so the motor can be offset and drive via a timing belt. This helps to keep the profile low.

Best of luck with the concept.


Mon Apr 07, 2014 10:29 pm
Profile

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

Here I am again, with some updates.

First of all, I re-built some of the internal gear components (replacing the initial copper ones with hardened steed, fine-grade wire-cut ones). This eliminated my previous backlash (and concerns :D ) which proved to be due to the copper parts deformations.

Second, I upgraded the inner bearings, which brought the whole thing to the next level in itself (all the previous wobble of the output is now eliminated)

Third, we now have a 3d-printer in the house so we printed a testing stand (with an additional 1:6 reduction from the motor, just to increase a bit the motor's power and to test the axial load capacity of the input, when used under belt tension) with an also 3d-printed output dial - to be able to attach the same laser torch to the output, for re-testing the whole thing.

You can see the testing stand in the video below, with everything working (you might turn down your speakers, as the video is quite noisy):



Here's the stand with the laser torch attached:

Attachment:
20140624_175435_.jpg
20140624_175435_.jpg [ 114.69 KiB | Viewed 19316 times ]


And here's the promised backlash test video, below (please ignore us talking in the background)



The laser is pointing at a remote wall, exactly 9m away, and you'll notice that the resulted backlash repeatedly measures up to exactly 2mm on that wall ruler - which, at this distance, corresponds to an angle of 0.0127 degrees, or 0.762 arcminutes. Not too bad - I'm really excited about this result! :D

@Gerald: taking on your advice, the next version is already under construction, featuring:
- a 8mm wide hollow-shaft
- both input and output flanges
- massively improved internal bearings system, yielding even better output stability under load
- all these resulted in a slightly thicker body (still 80mm diameter, but now at a 42mm width)

Hope to post a video of this one too, soon!


Tue Jun 24, 2014 6:59 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?
Looking awesome! How are your plans for a large run coming? Do you plan on Kickstarter or other crowdfunding to finance the run? Is 1,000 units still the goal?


Tue Jun 24, 2014 12:20 pm
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
sciencelookers wrote:
Looking awesome! How are your plans for a large run coming? Do you plan on Kickstarter or other crowdfunding to finance the run? Is 1,000 units still the goal?


Thanks again, man!

Yep, we've come a long way and we're planning going into production - and for this we need - you guessed it - funding. Yes, the plan is still to go for an initial 1000 batch - just because anything less generally costs much more per piece to manufacture.

We already tested and assessed indiegogo a bit - and it's a big disappointment (full of mostly crazy projects that do a better job of making your project look ridiculous in the context, than help you get funded)

Our first choice would've been kickstarter, of course - the most serious and active community - but we're Romanians and kickstarter only accepts projects from US, Canada, Australia, New Zealand, UK and Netherlands. Pretty exclusive. To start on kickstarter we'd have to find a reliable partner in one of those countries (UK/Netherlands for being much closer or US for being a much more flexible market) and completely transfer the project abroad, which is a logistic pain in itself.

So, right now, we're still focusing on further developing the initial concept (as you might have read above: with hollow-shaft, input/output flanges etc.) while keeping an eye out for a chance/idea of funding the project.

How are your projects coming along?

Cheers,
Eros


Tue Jun 24, 2014 10:49 pm
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Another question, for all of you guys and gurls involved in designing/using a speed reduction:

How much of a deal-breaker is a higher needed torque at the input?

The worm gears, for instance, have a pretty high mechanical efficiency (you need the smallest amount of torque at the input to drive the output), whereas my speed reduction needs a higher torque. On the bright side, this also means, for me, that the reduction becomes less (if at all) reversible, which for me is a big plus, since it eliminates the need for an additional power-off brake. Also, since I plan on using these babies with a side-by-side motor (hence, with a certain additional torque amplification by means of a timing belt and pulleys), this higher needed force at the input becomes a non-issue for me. But I'm curious about what you guys think about this...

Thanks,
Eros


Tue Jun 24, 2014 11:47 pm
Profile

Joined: Wed Feb 15, 2012 3:36 pm
Posts: 408
Post Re: Affordable low backlash worm gearing?
A lot depends on the type of motor you are using. As an example, some of the bigger harmonic drives do need quite a big starting torque to get them going, but they are OK afterwards. If you are using servo motors, at low revs they don't have the same strong torque that a comparative stepper has, so sometimes you need to use a bigger motor than you would wish.

BTW, even larger harmonic drives can roll on their own when the motor power is disconnected, for instance, if the load is more balanced to one side than the other, like in an unbalanced crane arm.


Edward


Wed Jun 25, 2014 4:06 am
Profile

Joined: Wed Feb 15, 2012 3:36 pm
Posts: 408
Post Re: Affordable low backlash worm gearing?
I'm just holding an equivalent harmonic drive and it measures 74mm in diameter and 52.7mm thick including the hollow shaft that sticks out 6mm on one side of the motor.

However, the hollow shaft internal bore is 14mm.I think the wider the bore, the better. 8mm is going to be pretty tight to pass cables, bearing in mind that this will be further reduced if you put an internal tube inside to avoid chafing the cables.


So I would say, try and get that hollow shaft as wide as possible.

Edward


PS, the starting torque on these little HD's is pretty small, any decent motor will cope with it.


Wed Jun 25, 2014 4:22 am
Profile

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

First of all, without presuming on knowing better but just drawing on my year and a half of learning this new stuff, you call them "harmonic drives", I call them "strain wave generator (SWG) reductions" (basically, elastic Chinese bird cages). "HarmonicDrive" is just a brand name, it's like calling any copier a "Xerox". I know HD is the main manufacturer, but they're not the only ones out there (see TOTEL, in China, for instance). But yeah, I know what you mean with those: those reductions' mechanical efficiency (which HD quote at around 76%) is great, mostly due to them using that incredibly difficult to manufacture deformable bearing - that makes all the difference in terms of low starting torque.

My reduction is in the same family with the HDs (meaning, it's also a "strain wave generator reduction") but using none of the elastic components used by the HD principle. My reduction works with a brass-steel bearing interface, instead of a deformable bearing - and this amounts for the extra-torque needed at start. This, and the fact that, unlike the HDs, which only use around 30% of the teeth in contact, my design uses over 55% of the teeth in permanent contact (more than half, almost double as the HDs)

All "strain wave generators" are reversible to a degree, no matter the size (the bigger the gear, the bigger the torque needed to reverse it, it comes natural). Mine is more self-braking just as a happy downside ;)

About the hollow shaft diameter: the bigger, the better - that's my bet too. My current limitations are given initially by the final target size of the gizmo (smaller size, smaller hollow shaft - but could scale up at any time, when needed) and then, once settled within a certain size, by the kind of internal bearings used (I'm using standard sizes here, pretty bulky, and which are limited in their variety but are so affordable - but again, I could always adapt the design for more exotic or even custom bearing sizes, at an evident cost).

Since the whole project (the time-lapse moco and, as a spin-off, the SWG reduction) is currently being financed out of our own DIY pockets, and since my own personal goal with this SWG is to bring an "under one arc-minute reduction that doesn't cost an arm and a leg" to the market, we're going ahead with the standard bearing sizes and with the 8mm hollow shaft (by the way, you just prompted me to check, and I DO have room for 10mm :D will try that instead)

Cheers and thank you again for your feedback!
Eros


Wed Jun 25, 2014 5:00 am
Profile

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

I think they are called Harmonic Drive because that company basically developed them. Any others from China are knockoffs and in my experience inferior. From this I deduce that it's pretty hard to make them well. So we look forward to a cheaper alternative that matches up with the original... :)


Thu Jun 26, 2014 3:40 am
Profile

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

Looking back over my previous message it does read as if I had some kind of negative attitude towards the guys at HD - nothing further from the truth. They took an old American inventor's concept (the first/original patent dates back to 1895 if I'm not mistaken) and ended up developing a kick-ass elegant piece of machinery that fulfilled NASA's wildest dreams in terms of low-footprint precise gearing. I'm in awe of their work! Not so much in awe of their prices, though :D

Not sure what you mean about the Chinese TOTEL "HDs" - in what way are those inferior? Higher backlash? Shorter working life? I'm asking because, after various talks with their sales rep, TOTEL's prices are on par with HD's. You'd expect a knock-off to be much cheaper, but TOTEL seem to feel as confident as to directly rival HD in terms of price.

I'm aware that my design is not as elegant as HD's, that my reductions may potentially yield a different output torque while needing a higher input torque - but so far, leaving these aspects behind, I feel very confident that, once we sort out the financing for this, we can come up with a much more affordable reduction on the market, at very similar backlash performances.

Cheers,
Eros

P.S. Yes, the HDs are extremely difficult to manufacture. First, in terms of tooling needed (custom miniature fly-cutting tools for the circular spline, that are extremely expensive and not quite so durable), then pretty exotic metal alloys for the flex-spline, again pretty exotic welding of tough steel for the teeth with elastic steel for the body of the same flex-spline, not to mention those exotic shape-shifting bearings - plus downright pervert tolerances for the rest of the body's components, or the custom cross-roller bearings for overall support). All those tight tolerances because of the nature of the teeth meshing together. Swiss clockwork stuff! In looking at THOSE problems did I come up with my alternative way of getting a similar effect eliminating some of the components and completely replacing them with other types of components, much easier to manufacture. I just hope my insight is right and that we'll be able to come up with a commercial version of this, sooner rather than later.


Thu Jun 26, 2014 3:58 am
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
Can't wait. I also can't imagine how you can make a similar device without the R&D HD put in over the years.

BTW I use the term HD as synonymous with "strain wave" - it has a better connotation I think... :)

Re the Chinese - I bought a couple from a company in Beijing a couple of years ago. They seemed ok on the surface and the web site had lots of shots of rockets and other high tech goodies to entice, and the prices were very good. So I gave it a shot - As it turned out backlash was acceptable but they had rather serious vibration issues above moderate rpm. It didn't seem too bad initially until I put a camera on the head. Then there was an increased resonant vibration that made the image judder. The company also suggested I use a larger model and even sent one for trial. But it was little better and would have made the head like a tank...

On that note have you tested your unit at realistic realtime panning speeds with a camera head attached? It would be important to know what output speed can be achieved with what inertial mass. Max speeds should be in the range 30 to 60 deg per second for normal work, but for high speed (slow motion) photography it's not uncommon to require 90 - 180 deg per second.


Thu Jun 26, 2014 3:37 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?
I agree. Strain wave invokes images of politicians straining to smile and wave at the peasants whose vote they need but for whom they feel nothing but disdain. You don't want that guy, but the other guy is the same.

Gerald's MOCO for highspeed is awesome. If I didn't know it was real, I'd assume it was sped up in post because the movement looks impossibly fast and precise. Gearing for that stuff needs to be very precise and very strong.

Does it help to consider that most photo applications need only light duty hardware? Most P/T heads for advanced amateurs and even many pros only make a 90 or 180 degree motion every now and then, and that will be at fairly low speeds. Its not going to be run continuously at high speeds like your robot or a moon buggy. Would this low speed, low duty cycle application allow use of different materials and manufacturing methods? 3-D printers with 30 micron accuracy are common. Some of the parts they make are fairly durable. Even better plastics can be cast from simple molds made from the 3-D printed parts. Wax models can be made in the same molds, and brass parts can be forged from them using small, desktop jeweler's spin casting machines or vacuum assisted casting machines. Given the large number of them you want to make, it might be economical to make many of the parts yourself.

I can make some precision 3-D parts if you'd like to try them. I also have CNC milling and other machines if it needs harder metals and if you'd consider making them here.


Sun Jun 29, 2014 9:05 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Hi again, guys, and thanks again for your kind feedback!

@geraldft: you're right, I guess there are no shortcuts to coming up with a "similar" product in less than a year (I started thinking about my moco in January last year, started the research and design in February and started effectively working on various parts and concepts in March of last year. It's just this January (5-6 months ago) that I came up with the idea for this reduction/gearing concept, so the time dedicated to developing it so far is extremely short - 5 months). The keyword here is "similar": we don't have access yet to high-tech measuring apparatus, so we can't give vibration or precise torque data yet - so yes, there are differences between what we've achieved so far and the finished jewels of HD. I wouldn't presume to brag about merits that are not proven yet. The only thing that's given so far is the backlash - and that's where my entire focus has been. About your specified speeds: my previous video where I was grinding the reduction on a lathe showed it working with an input speed of 1500rpm, equivalent to 180 degrees/second at the output. However, my focus so far is not on speed (as time-lapse rigs rarely need speed). I'm gathering you're using your rigs for real-time filming, or even high-speed-motion filming - for this purpose your speeds more than make sense, and I hope we'll be able to come up with precise enough work to satisfy those requirements too, in further stages of our work. We haven't conducted inertial tests so far - but we're getting there too, pretty soon, and as always I'll keep you guys posted :D

@sciencelookers: as always, thank you for the kind words. I found other posts where you're giving feedback to others and you always have encouraging words for everybody - or is this just my impression? I know "strain-wave" is no poetic wording by any measure, I just don't have a name yet for my principle and since it's definitely not a HD... I'm lost for words :D
Again, about Gerald's speeds: we have been grinding the reduction at the max speed specified by Gerald, with no problems (again, we haven't got the means for measuring vibrations yet, but I've yet to feel any resonant vibration, and we've managed to eliminate all the axial wobbles from the initial versions - will test further, anyway). About your CNC help offering (thank you!): I've also received your private message, and we'll move that part of the discussion in private.

In the meanwhile, to my lady's distress, I've forfeited my weekend to finally building the number 4 prototype, which incorporates some of Gerald's suggestions (the stronger bearings, the hollow-shaft, the input and output flanges). Since I finished working it and assembling it at 11PM, I didn't care much anymore for setting up a pretty-looking test, so I just propped the new reduction in a bench vise and turned it by hand with a wrench, just to see how it behaves and what kind noises it makes ;)

Here's the video, below (please excuse the dirty looking setup)



Sun Jun 29, 2014 1:57 pm
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
Congratulations - "powering on"... maybe something like FlexWave - or do you have something better in your own language? That might sound more exotic...?

I imagine eventually you might need a couple of sizes. One small and light for portable DSLR's and one for more serious loads?

Speaking of high speed - I've been testing for a project using a Phantom Flex at 1,000fps or so. (body weight with lens 6.75kg ) Tilting speed attained is about 275 deg/sec but more critical is the acceleration needed. To tilt through 30 degrees from standing start to stop in 0.18 seconds the peak acceleration is about 4,800 deg/sec/sec. Needless to say this rig needs to be bolted to a solid base or the whole thing does shake... :)


Mon Jun 30, 2014 5:51 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?
Very smooth, as would be expected. Nice job! What are the six spiral rows of holes for? were they for removing material to balance the part? or are they a way to attach it to the next part in the drivetrain? Your laser backlash test with the previous model really impressed some people here. I second Gerald's opinion that several different sizes should be made at some point (that topic came up here as well), but understand everything needs to start somewhere.


Mon Jun 30, 2014 8:08 am
Profile

Joined: Fri Feb 22, 2013 9:31 am
Posts: 70
Post Re: Affordable low backlash worm gearing?
Hi again, guys! You make this forum alive and such a refreshing place!

@Geraldft: very nice name, "FlexWave"! Is it copyrighted? ;)
About different sizes: of course everything is scalable, once you have a proven concept. The reason we're working with this particular size is related to our overall needed torque guestimations and to the fact that I do want my moco to result as lightweight as possible in the end, since I dream of wild time-lapses up on the mountain tops. This isn't to say that bigger, stronger versions wouldn't be possible. And speaking of honest modesty: I don't have the elegant advantage of HD, to cram a variety of gear ratios in the same unit by simply changing two components. But that doesn't dishearten nor does it discourage me. Trade-offs are part of life.
About higher speeds and accelerations: they eventually come down to simply higher needed torques (able to cover for the additional inertial torque) - so, if I'm not wrong here, higher speeds/accelerations simply require higher torque abilities and - as you mentioned - a lot of concrete foundation :D. Not in my target for my moco, for now, but definitely in the books for the future.

@sciencelookers: About the different sizes: see above. About the holes: they are simply sets of 3 threaded mounting holes each, set at various circle diameters, for convenience, and only set in a spiral form because if you set them in a straight radial line you get a structurally weaker flange. And because it looks cooler too! ;)
I'm so happy that my tests had a good reception over there. It's one more confirmation for me. The principle (strainwavegenerator) is not new, but I'm happy to have found a more accessible way of achieving it :D

Cheers,
Eros


Mon Jun 30, 2014 10:20 am
Profile

Joined: Wed Jan 20, 2010 4:50 pm
Posts: 561
Post Re: Affordable low backlash worm gearing?
A quick check shows "FlexWave" as the name of a range of guitar FX units. Not sure if it's registered...

BTW what dimensions and weight do you estimate for the first units you build? Any timeline yet?


Fri Jul 04, 2014 7:10 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 5 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