You didn't mention backlash, which is necessary for rack and pinon motion systems to work. You can eliminate whip by turning the ball nut rather than the heavy screw.
There's a reason he didn't. This was a puff piece meant to beef sales camouflaged as unbiased information. Almost everything he stated was borderline deceiptful. Like the statement that R&P can be more accurate than "Any ballscrew"... BS... If he had been honest, he could have said the best R&P can compete in accuracy with some rolled ball screws, possibly. Now, what he could have done is talk about how each platform will always be a compromise and the longer the router is, the more R&P starts making sense. And like you said, he could have mentioned rotating ball screws which help reduce whip and then he could have talked about how you need very thick and super expensive ball screws to counter whip. But they are so afraid to play that honest card so they try to make people believe that ballscrews have hardly any advantages. To me, that's just a bad strategy as the honest one is good enough.
Agreed 100%. On the large format industrial machines which cost 200k+ they use rack and pinions but they compensate for backlash using software. They would calibrate each section of the rack using a ball bar by moving the spindle in a circle. Then they store the backlash information for each section of the rack and the controller would compensate according to the stored values in real time. That is to say, each time the x or y reverse directions, the controller would add/subtract the backlash distance nearly instantaneously. They are able to achieve +- 00005" accuracy using this method. @@AntiVaganza
@@AntiVaganza Looking to build a 5x10 CNC router to make cabinets as a side gig. Would you recommend balls screw or this rack? Trying to keep it cheap for the start. Thanks
How could you have this discussion without talking about the main issue of a rack. Backlash. The helical gear contact doesn't change the clearance necessary for gears to roll. Also the drive is likely to have more backlash too as the rack doesn't have any inherent speed reduction like the screw does. So a multi stage reduction is more likely adding places for backlash and compliance to the drive where belts can stretch. Then the argument of ball screws whipping is easily solved by turning the nut instead of the screw. You might have multiple gears on the rack with an adjuster to compensate for backlash, and if so seems like it should have come up in here. I don't know the accuracy of your machines and they may be as you claim, but this video certainly didn't convince me that a rack has the accuracy of a ball screw.
It could happen because the company that sells these machinese which rely on R&P had their own sales engineer do the "discussion". So, it's flawed. They made a sales pitch but camou'ed as unbiased info. Which was plain silly as if they had been honest about it all and explained why as you go to larger machines R&P start making sense, especially from a cost standpoint. Instead, they tried to convince us it's the best thing since sliced bread... Old school tactic. People are smarter that that now (I hope).
With a ball screw you get a lot more linear force from the same torque motor. It means less acceleration on the axis. It is because one rotation of a screw is typically 5-10mm of linear movement. With rack and pinion it can be from 60-200mm. When you are using stepper motor set to 200 step per revolution you get from 0.33mm to 1mm per step! You can gear down a motor with a timing belt, but you loose the main advantage of rack and pinion - simplicity. In my opinion using rack and pinion is a reasonable choice only to drive a gantry in machines longer than 2m. If machine is smaller i would always choose ball screw over rack and pinion. Rack and pinion -Easier to use on bigger mashines -Require bigger steppers/servos to get good accelerations Ball screw -Dont require big motors -High forces and accelerations -More precise than any existing rack and pinion due to pre tensioning and zero backlash -Slower maximum axis speeds
Great video, thank you. Some suggestions, though: It'd be helpful to clarify pros and cons between both approaches when speaking of (1) pre-loading capability, (2) axial load capability, (3) contact angle aspects associated with rack-pinion approaches, (4) options to shield or seal the driving elements, to keep debris/contamination away, etc. I tend to agree that for 'long-travel' motion systems, helical rack should definitely be considered. However, I'm concerned that the video may be misinterpreted by some as rack and pinion is 'always' better than a ball-screw approach. I've read some of the comments below, and I believe that some people are misunderstanding such details. In any case, great video and reference knowledge. Thank you so much.
How would a straight rack perform when two motors/pinions are use? They then also engage at least two teeth, and maybe they can be placed so that it minimizes backlash.
"Superior" depends entirely on context - who is building it, how much money is being spent, how large is the axis, it is fully closed loop or not, etc etc. If you're talking about a very large axis and a massive CNC machine like those in the video, then sure you don't have much choice, but for smaller axis it's way more complicated.
Thats very interesting! Can helical racks work with zero backlash on change of direction like ballscrews? Are building your racks in-house or any preferred supplier? Thank you!
Crazy how youtube randomly offers up videos like this when I was just discussing this topic earlier today. I think on smaller mills a ball screw setup is perfectly fine. But when you start getting over 8' on a gantry mill you going to get whip on a fast moving machine. A helical rack can match them in accuracy and have a rock solid foundation. We have a 30 year old helical rack machine a work that respectfully runs at least 30 hours per week under heavy loads most of the time. I've just started to notice a slight ammount of play on the end of the table that gets the most use from running small parts. I doubt very seriously ball screws would have held up that long.
Great Video Mr You get what you pay for! ha,, one thing you did forget to mention is that most rack n pion set ups? your project STILL relies on THE MERCY of a BELT!!!!!! yeah its a long toothed steel track but those 3 pulleys that house in the motor are belted together! lol!!! hey but they are people out there running machines On Toothed timing belts on its own orbit of thought! So its really a double edged sword in a way, Another option are chains!!!! for the individual that feels that kevlar or steel belted rubber Either way you will still have to keep on eye on it! Rest assure you will still have to service and maintained the unit regardless of what you choose!!
rack and pinion is the cheapest and easiest way to create numerical machines, it reduces components and you can use fiber reinforced performance concrete instead of metal for the structure, so this makes cnc routers, printers, tapping affordable to anyone. also if you install a contact layer in the rack and pinion , you can replace it regularly after wear saving materials
Ball screws are easier to install. Also the accuracy of ball screws are very high. The price of ball screws is also less due to higher production volumes
@@arbjful i can make a precision fiber reinforced thermoplastic (even self lubricating) rack with wax milling and molds, however making a screw is only possible with metal rods and it has to be forged metal with plenty of treatment, and after this metal has to be precision milled in all the 360º degrees compared to 3 axis milling of a rack. precision is more about the pitch and lead of both systems
@@crapisnice we use 10mm pitch ball screws for our centering application. An open loop stepper motor easily handles our accuracy requirements, we used a timing belt system, but apart from speed, there were accuracy issues here
@@arbjful yes but rack and roller pinion is more reprap friendly, more accessible to be done diy and spreaded, that the reason i have chosen it anyway beside the bending stresses of screws, however fiber reinforce composite screws can be done as well (and applied in vertical z axis with no problem)
Yeah, I'm definitely starting to regret going for 2,7m long ballscrews in the Y axis of my CNC router, even if they are 25mm diameter they sag a fair bit, and is a pain to align :P
If rack and pinion is so superior then why don't you use it on your spindle axis? Backlash on R&P is insane - fine if all you are doing is making signs but otherwise, I'll stick with my ball screws.
Except you can get as close to zero backlash as you will ever get with one while the other requires that you build some in just to keep it from destroying itself in short order. Didn't need the marketing wank, its simple, want to move really fast or cover distances over about 6-8 feet then you probably want rack and pinion. Whip isn't even an issue, your screw is either bent or you're trying to spin it way faster than the ball nut bearings are going to tolerate for long anyway, turning the nut is a solution also. A lot of mechanical advantage is lost with a rack vs screw, so more expense in bigger motors to meet the demand but that also means a likely reduction on positional resolution, or you gear it down which is an expense that likely adds more backlash yet. No silver bullet here.
celiturbo said 'A lot of mechanical advantage is lost with a rack vs screw'. What are the disadvantages of increasing mechanical advantage at the pinion through gears or pulleys or some other means? .
Hmm it’s a ballscrew but not a conventional one.. but yeah the helical rack for these maschines makes sense as of there work room. Others go even further by either using magnet based rails or air cushions.
First of all, don't mix measurements. Either Imperial or metric. Also in machining 0.0002" is a large. I built a macro to l rail in my spare bedroom with a theoretical resolution of 0.0000118" (that 0.0003mm for places that aren't America). The next one in currently working on could potentially have an even higher resolution.
Что за бред?)))) Допустим мы используем ШВП с шагом винта 10мм. Получаем 1 оборот=10мм. Чтобы получить такое же соотношение на зубчатой рейке, на редуктор. Редуктор если не свехточного исполнения это люфт зубьев в 2х местах.... компенсация на реверсе это однозначно ударные нагрузки при смене направления движений...