Discussion on the Nimble

No, we are not worried.

First of all, correct there is no way to adjust the clamping force. None is needed. There are two reasons for that.

1 We found a sweet spot in regards to pressure and hobbed wheel shape.
2 We engineered a certain amount of spring into the shape itself. It handles variations in thickness and type of material.
We worked hard to avoid any flex in the system, as we want to manage the forces accurately.

The only thing we don't know yet, is how well the breech block will stand up over a longer period of time. We have had the current breech block now for 3 months and have put it through its paces without any degradation noticeable. We choose the material with an eye on longevity, but we have not yet tested it for 3 years.

It is a small part, so it is easy to replace. We have a plan in place to evaluate and know just what to do if it becomes an issue. But we are confident it won't.

Lykle

Quote
Lykle
First of all, correct there is no way to adjust the clamping force. None is needed.

(emphasis mine)

That strikes me as unlikely to be true in the general case. I find idler pressure to be an important tunable, and use very different pressures for PLA and soft TPE. I suspect that once the extruder is in use across a wide range of hotends, nozzle sizes and filament types you will find customer's wanting different idler pressures. I don't see that as a problem, you could easily ship firm and light pressure idler clamps, but I'd start working on it sooner rather than later.

Unless of course you have invented a magic drive gear that is nothing like anything I've ever used, but from what I can see in the photos it looks fairly standard.

Edited 1 time(s). Last edit at 11/02/2016 08:40AM by JamesK.

I'm not saying it's right or wrong but Dyze Design don't have the facility to adjust tension their (expensive) extruders. I queried this on another forum and the maker claims that they have the optimum tension for all scenarios and adjustment isn't necessary.

Quote
deckingman
I'm not saying it's right or wrong but Dyze Design don't have the facility to adjust tension their (expensive) extruders. I queried this on another forum and the maker claims that they have the optimum tension for all scenarios and adjustment isn't necessary.

What else would you expect the maker to say? They assess the problem accurately on their website:

Quote
Dyze
Applying too little pressure causes the filament to grind before the motor reaches its full potential. However, applying too much pressure causes the motor struggles against friction, inducing filament deformation,

There is no single pressure that meets those requirements when the filaments have wildly different hardnesses. One interesting approach that I haven't tried is a rigidly fixed idler wheel in a non-compliant housing (i.e. steel and aluminum). In this case the fixed quantity is the distance between the idler and the drive gear, which hopefully translates to a fixed bite depth of the teeth into the filament (not entirely true if the filament deforms significantly). The stiffness of the housing means that the pressure experienced by the filament will vary greatly depending on the hardness of the filament, so where as a spring is (nearly) constant pressure, this would be constant displacement. The Dyze is clearly sprung, so doesn't fit this model - their claim that pressure does not need to be adjusted for different filaments does not match my experience. Nimble doesn't look stiff enough to achieve constant displacement, but possibly it's something of a hybrid between the two.

I must have missed how it adapts to 1.75 / 3mm filaments.

Edit - JamesK

Yes, I hear what you are saying and I'm not arguing with you, but if the teeth bite deeper into soft filaments so effectively changing the diameter of the hobbed bolt, then wouldn't changing the extrusion multiplier for that filament achieve the same result as adjusting the idler tension? I generally test for variations in filament diameter for each new spool and adjust the extrusion multiplier accordingly, so why not use the same variable, rather than change the extrusion multiplier as well as the spring tension? Just a thought.........

Edited 1 time(s). Last edit at 11/03/2016 05:11AM by deckingman.

Quote
deckingman
...if the teeth bite deeper into soft filaments so effectively changing the diameter of the hobbed bolt, then wouldn't changing the extrusion multiplier for that filament achieve the same result as adjusting the idler tension?

Yes, differences in bite depth do change the extrusion calibration, so you have to compensate for that, either through steps/mm, filament diameter or extrusion multiplier. All three work, so which you use is largely down to practicalities and personal preference. steps/mm would be more logically correct, but with 8 bit software it's tedious to change for every filament. I typically use the extrusion multiplier in Slic3r as a way of coding the per-filament compensation required, and keep the diameter as the actual measured diameter of the filament.

However, I don't think that adjusting extrusion gives the same result as changing the idler pressure. The primary purpose of changing the pressure is to get the optimum grip on the filament so as to ensure the most reliable extrusion. As a result of changing the pressure you may have to change the extrusion multiplier in order to get accurate calibration.

Interesting discussion!

In the mean time, I went and made a video. The Nimble printing 4 different materials in the same setting. The only thing I changed is the temperature and in the case of the Ninjaflex I also changed the speed. So none of the settings were optimised for the different materials, only temp was changed.
In fact is some of the prints you can actually see the previous filament in the first layer of the next print. I didn't flush the filament out after each change.

I am not saying that this is absolute proof that 1 breech block will handle all materials. What I am saying is that this breech block will handle a wide range of materials. So for general use it is perfect. Will we offer different breech blocks optimised for different materials. Probably. As James said, easy to do. We will probably even offer some very special ones. Hmm, long fiber insertion while printing, I wonder if that will work. Probably not, but if it does, we can make a breech block for it.

Once again, the Nimble is designed to be the core of a system. Add-ons, adapters, extra brackets, extra breech blocks would fit right in that strategy.

Oops, almost forgot, here is the Youtube link: [youtu.be] of the 4 materials being printed, remember?

Lykle

Great demonstration Lykle, guess I'll have to eat my words

I'm not bought yet, how do we know it is the optimum for each material? To know it is the optimum, you'd need to see retraction tests of each material as well as extrusion tests to know it is pushing what it should through the extruder. The latter is hard to test, but the former a bit easier to (just show voronoi prints of each material).

Quote
Origamib
I'm not bought yet, how do we know it is the optimum for each material?

Does it have to be optimal, or just good enough to work? Not having to adjust tension is pretty appealing, one less thing to forget, one less source of variability. You're right though, the good old test cube isn't a particularly hard test for extruders.

Still, given that it was at least working for that test, and I don't think my extruders would without changing the tension, it is interesting to wonder why there is that difference. One thing about these flex drives is that they use a much greater reduction gearing than a typical extruder, so have more torque available. Perhaps they are running at the high end of the tension range, and while the softer filaments may distort, there's enough torque available to overcome the increased friction. Other differences with my setup are that I'm running 3mm filament and using a nema14 stepper with 5:1 gearing, so I'm definitely running very close to the limits of my available torque.

Interesting discussion about the tension. I hardly every change the tension on my exturders. Only do if there is a feed issue, which is very seldom and normally caused by a clogged nozzle or being too close to the build plate.

The object of the exercise here was to change as little as possible just to show it works straight up.
Sure it was an easy test object. It is my default object to do a quick test if I change anything.

At one point I had about 2 kilo's of these little blocks in all different materials. Threw them all away. Pity in hindsight. I could have made a wall mosaic with them.

Quote
Lykle
At one point I had about 2 kilo's of these little blocks in all different materials.

Yes, my desk is littered with them too. For some reason I find it hard to throw even failed prints away. I guess the sense of magic and wonder persists

I have 2 blocks half printed, doubt i will do any more favoring functional parts over calibration pieces(yes they are functional), but even a failed print of functional part is more useful than a cube...though I'm sure they have some use, more use if they had holes in them.
How do you (Stress) test one of these cubes? Stare at it with an accusatory face joking aside
at least with a functional part, it could throw up some other errors either in design or machine.
There are more complicated calibration pieces out there, maybe its time to see some of them.

Edited 2 time(s). Last edit at 11/06/2016 08:19AM by MechaBits.

Sure, functional parts are the point after all. But there is value in something that is quick to print that you have printed many times in different filaments. You get used to the way it comes out and it gives a quick sanity check that the basics are still working the way they should be. The main check points on a test cube are first layer adhesion, corner shape, surface finish/layer alignment, infill overlap and the top surface bridging. As Origamib pointed out though, it doesn't test retractions and the extrusion rate is fairly high throughout, so it doesn't tend to stress PLA clogging issues and filament stripping.

I noticed with the prints the first layer looks like its putting down a river of molten plastic, hot bed glass...but looks to be pouring out.
am i mistaken?

Quote
JamesK

Quote
Origamib
I'm not bought yet, how do we know it is the optimum for each material?

Does it have to be optimal, or just good enough to work? Not having to adjust tension is pretty appealing, one less thing to forget, one less source of variability. You're right though, the good old test cube isn't a particularly hard test for extruders.

Still, given that it was at least working for that test, and I don't think my extruders would without changing the tension, it is interesting to wonder why there is that difference. One thing about these flex drives is that they use a much greater reduction gearing than a typical extruder, so have more torque available. Perhaps they are running at the high end of the tension range, and while the softer filaments may distort, there's enough torque available to overcome the increased friction. Other differences with my setup are that I'm running 3mm filament and using a nema14 stepper with 5:1 gearing, so I'm definitely running very close to the limits of my available torque.

The difference between optimal and 'just good enough' could mean failure on a long challenging print, which is not acceptable. I rarely change my tension, but it is something that has to be done for certain filaments I find. I run a pancake Stepper on a titan extruder.

New tensioner could be bought, and at the right price (ie, cheap) then I could be persuaded as it's a simple change but I don't want to buy an extruder that has extra costs to it, when a 5g tensioner mechanism could be made.

I have to add though, this is all just discussion and it shouldn't detract people from backing I hope. Tension will likely be a minor issue.

River of molten plastic, what a lovely way to put it.
Could be. Again, I was simply switching filaments and had only paid attention to the first layer when setting up for Ninjaflex. I did not use different profiles for the different prints.

As to more challenging prints, besides the lovely blocks. I printed a Voronoi 3D Benchy, right after I printed the normal one.



Lykle

I never seen a voronoi benchy, looks knitted but looking good...time to print a Nimble and see how it holds up.

What, you want the stl?

Hmm, sorry can't do it. The license of the 3D Benchy does not allow derivatives of it.

So yo will have to make it yourself at the Voronator site


Lykle

Edited 1 time(s). Last edit at 11/06/2016 09:54AM by Lykle.

Quote
Lykle
As to more challenging prints, besides the lovely blocks. I printed a Voronoi 3D Benchy, right after I printed the normal one.

Did you use any supports when printing that?

No need for the model but a handy link, I'll be trying some time.

@ElmoC, no supports used. Just print.

@MechaBits, yes I like it as well. Thinking of interesting models to mutilate with it. I have a female body somewhere that might be good.
Perhaps I will do it to a Nimble stl and see what happens there. Using a Nimble to prove that the Nimble is nimble in retraction by printing a Nimble.

Lykle

Talking about a female body, I had a torso I modeled a while ago.

Decided to see how smooth I can print it. There is the link to the photo's
Hesitant to show more because of the nature of the object in the images.

Lykle

Well, that was a pity, the Kickstarter campaign did not reach the goal.

So we went to plan B. All the critical parts are custom made and the body is going to be made using SLS.
This allows us to be a little bit more flexible in the production of the units and gives a little more freedom in the shaping of it.
Attached is a top view of the current body of the Nimble.

It might also mean we will be able to deliver a little faster. Not too sure about that yet, but it looks that way.

I do feel very happy with the support we did get and lots of people want a Nimble. All we have to do now is get it done.
Simples

Lykle

Edited 1 time(s). Last edit at 11/22/2016 09:56AM by Lykle.

Sorry to hear the kickstarter didn't work out, but I think you have generated a lot of extra interest in the technology. Perhaps the chance to continue without all the pressure that kickstarter seems to create is not such a bad thing. You have a great looking product and I'm sure I speak for many in wishing you success in bringing it to market.

Thanks James.
We will get there. Just by a different route.
Lykle

A bit of an opportunity in your neck of the woods [www.3ders.org]

Yeah, had a look at that.
I feel for the guy, he really tried hard and is failing.

The problem with that situations is that all you are paying for, is the debt. That is what he needs to money for. He is doing the right thing by insisting that all the debts are paid. Honorable man. I have been in that situation and it is tough.

Would love to know how much he wants for it all, but do not want to contact him and get his hopes up. Would be a cruel thing to do, just to satisfy my curiosity.
The designs look pretty solid and he has done a lot of work on the software side. That is probably what killed it. I agree with his statement that cheap clones are hindering the business, now more than ever.