Thanks 😄

He means the YT video - the cnc is okey'ish for dyi.. 😜

@@TheStuartstardust Yes, I actually just built it to have content for the video .. 😄👍🏼

@@EngineeringSpareTime hahaha 😄,

Learn us to how to make it again. Where is the tutorisk?

Awesome comment :D thank you - I appreciate it! Yes, there was quite a bit of research and testing going into the design that’s not discussed in this video. Iterations I had to make, because I wasn’t pleased with the result… At the moment it will be difficult for me to find the time to do an additional video about this machine, since I‘m working on different things I am going to share in a few weeks. Until I finished those I plan to do a more detailed video about the design and especially the ATC, since your not the only one that seems to be interested in this :)

Thanks! At the beginning I didn‘t know how it‘s going to look like at the end, because intentionally I just planed to upgrade my former CNC ... this plan changed a bit. That‘s why I haven’t token any further videos, just pictures of the building process and a slide show is probably not this exciting...

@@EngineeringSpareTime bro can you share the cad file of this little monster? please aka.vandalo@gmail.com (or well, just the spindle would help me a lot)

Thanks a lot! I still have some older projects I could share (mostly slide shows.. I mainly took pictures) and I‘m currently working on a 6 DOF robot arm, this will take some time to finish though :) I could release some “early stages“ of the build, if this is of interest..

@@EngineeringSpareTime you really should! I believe there are a lot of people who enjoy watching the process!

Thank you!

I'm planning to do something about it, since several people have already asked for it. You‘re right, not on KZfaq.. :)

Thank you :)

Thank you 🙏🏼

Thanks - I appreciate it! CNCZone was great. It was actually part of the motivation to do this

Thank you! This was probably close to the limit of the NEMA 17 steppers that I used back then :)

I’m glad to hear that you enjoyed it and appreciate your words :) When I choose what I’m doing, being patience is natural - in my opinion. Thank you, wish you the best too!

Thanks a lot! Yes, I totally agree. It‘s way more work then you‘d expect before - so many small things to solve and to decide. But this is what makes it‘s so rewarding at the “end“. Just keep doing it! I‘m also building a „“big“ one together with a friend of mine, for the last four years. We‘re getting there.. 😄

Thank you!

Thank you, I appreciate it. I still have some older projects to show, which might be interesting as well as a current one. I‘m in the process of building a 6 axis robot arm from scratch (aluminium construction) including the driving system (servo driven) and programming. The mechanics of the first four axis are basically done. When I find the time, I‘ll make some videos in the next weeks. I checked out your channel, looks like a really nice machine! Yes, such projects involve a big effort, but totally pay of in terms of learning and fun - in my opinion.

I really appreciate your words and can totally relate! Making components yourself first gives you an understanding of what is important during development. Of course I need longer or even more attempts for the components I make, but in a professional context a much more productive level arises on which you can exchange ideas with the experts in the field - like you - and find a solution together. We can all learn from each other and that is, in my opinion, the fun of it all. Thank you!

@@EngineeringSpareTime I agree, working together to solve problems is fun and the experience gained and shared keeps our world moving forward. I am Happy for you and your workmates, knowing your future will be rewarding.

Thank you :)

Thanks :)

Thank you! Yes, it‘s definitely a huge amount of work - totally worth it though:)

Thank you! A lot can be done - one way to find out is just trying :)

Thank you :)

Danke! Meine erste Fräsmaschine war auch aus Holz - es geht ja um die Begeisterung :)

@@EngineeringSpareTime Ja, so ist es....😄👍

Thank you 🙏🏼

Thanks! :)

Thank you!

Overdoing - I don’t know what you‘re talking about :) Danke! Begeisterung übersteigt manchmal den Nutzen und da beginnt der Spaß! Ebenso, schöne Maschine und schicke Spindel ;)

Thank you, I appreciate it! Yes, Piotr‘s machine is a nice one. There are some other really nice ones too. It weighs about 80 to 90 kg - I haven’t weighed it in a while. I run an old version of Mach 3, cause I was used to it from a older project. For the upcoming portal machine, I‘ll use LinuxCNC.

Thank you!

Thanks! I “just“ had the lathe available, no milling machine was used. The lathe is the only actual tool I‘ve used. You could check if someone close to you lets you use their lathe ;) At some point you need some turned parts. Building a machine like this without a lathe is nearly impossible - except you buy all the components finished...

Thanks!

Thank you, I appreciate it! When I started the machine I was doing an apprenticeship as industrial mechanic. I “finished” it mid of my mechanical engineering bachelors :)

@@EngineeringSpareTime thanks, I'm currently doing mechatronics in HS and wanted to do a mechatronics and automation systems specialization instead of a degree

That’s a good idea 👍🏼 I could also move faster, hit the trigger ones, retract and repeat it with less speed - not that save though :)

Thank you, appreciate it!

Thanks 🙏🏼

even gave you a sub to see some future vids

Thanks! :)

This depends on the stainless, but 4140 for instance would work just fine. The steel shown in the video is O2 tool steel - for comparison.

Haha, thanks. I've literally been doing this my whole life and all the time - all in all, that's a lot of spare time :-) Your experience will influence your approach, decisions and confidence in each project, thus you “just” have to start, try, fail, redo and you’ll get somewhere with it. If you are trying to build a robot, build it the way you are confident with it, add 30% uncertainty and go. Trying to build what other people build is not really an approach I’d take - as inspiration for sure

@@EngineeringSpareTime thank you man. Good luck with your spare time😂

Thank you! 1) For what its is I am quite happy how it turned out. I would have done a few things differently, would have designed things differently but for the time I was building the machine this is what I was able to do with my resources - knowledge, tools and budged.. For example: I would have liked to nitride the spindle shaft, use even more high quality bearings and linear rails (the spindle once and the ball screws are top quality..) and so on. 2) Yes, this was in the early stage and actually the first test of the ATC - so I started slowly (this was at about 300 mm/min, now it’s capable of 14m/min - I should make a video about it) :) 3) So far the clear path servos doing great. The implementation was straight forward, the performance is way more than enough (they actually run at 40% torque limit, otherwise they could rip everything apart). The tracking error and disturbance rejection is also good. They are more expensive (double the price compared to e.g. JMC), this is still cheap in comparison to a “actual“/ industrial grade servo drive (not the same “quality“ obviously).. For the next CNC I would order the free programmable version to be able to implement my own control algorithms to enable model based temperature compensations and stuff like this, its not really necessary though.

Thank you! Unfortunately, at the moment I can’t build you another one, because I couldn’t offer it to a reasonable price.. With my small lathe it takes a lot of time to machine and to get it working perfectly. Thanks for you interest though! Maybe I’m able to produce them more efficient in the future - we will see :)

Thank you! The bom is kinda short: - disc springs - 7007 and 7006 from NSK -> spindle bearings - 3.5mm ball bearings The rest is custom and made from stock material :)

Thats cool, but what about collects? The spindle interface for the collects is what makes ATC spindles difficult to build.

Thank you! The ratio is about 2.3:1 -> Max. 11 000 rpm. I did this to increase the torque at lower speed. Otherwise the torque would have been to small for milling high strength materials (e.g. carbon steel..) and the bearing lubrication of the main spindle is critical at higher speed with this diameter.. for the tool changer it’s not a problem, especially not if you build it yourself :-) For the bearings: Good point! I machined an adapter and shrink fitted it on the machined end of the ball screw (it was 8mm OD) to achieve a proper shoulder for the bearing

Thanks! Someone else already asked about this. I plan to do a video about this.

I would also be interested

Thanks! It weights about 80 kg.

Thank you! :) the overall build excluding the ATC the spindle motor and the servo system (those are exchangeable..) was about 1.2k - building it like this decreases the price, increases the involved work though.

Thanks! This layout has some advantages in terms of stability and stiffness, the effective working area is smaller compared to e.g. a portal though. It’s like a Industrial VMC

More typical for a 3 axis machine to be configured this way, great job man, I wish I had the resources and I'd make one too

Hi, thank you! The spindle head is made of 6mm steel plates, all dimensions were cut with an angle grinder and than hand filed to precision :)

@@EngineeringSpareTime makes sense, thank you!

Thanks! The build evolved over the time, so it‘s probably not the most efficient way to build it like this (e.g. more than 300 screws :D). That’s why I‘m not sure if someone would like to build it like this or should.. I‘d have to refine the plans and CAD for that to quite an extant. Due to my current projects and studies, I have to delay this.. But nice to hear that you would be interested in something like this, I’ll try to figure something out.

@@EngineeringSpareTime Many of us interesting in taking inspiration from your amazing build :)

Danke dir! Das Gießen war nicht all zu schwierig - lässt sich mit der richtigen Vorbereitung gut verarbeiten. Ich habe den UHPC von moertelshop.com bezogen. Dieser wird dort „TEGNO“ genannt. Bei den Motoren handelt es sich um ClearPath Servo Motoren. Die Controller sind bei diesen direkt im Servo integriert. Ich habe zunächst Schrittmotoren verwendet. Als alles soweit funktioniert hat, habe ich mich dann dazu entschlossen die Servos zu kaufen (definitiv mit Abstand das teuerste an der ganzen Maschine (etwa 50%), das war es mir am Ende aber wert :) ). Diese sind ebenso durch Step/ Dir anzusteuern, was die Implementierung damals vereinfacht hat.. Ich konnte so meine Steuerkarte beibehalten. Als Steuerkarte verwende ich einen SmoothStepper mit BreakOut Board für die IOs. Bezüglich der Genauigkeit, war das gezeigt natürlich auf den relativ kleinen Bearbeitungsbereich zurück zu früheren. Insgesamt spielt sich die Genauigkeit vereinfacht angenommen, mit entsprechenden Schnittparametern und Verfahrwegen, im Bereich eines Hundertstels ab - womit ich sehr zufrieden bin :) Wobei man nicht vergessen darf, dass Genauigkeit ein weit zu fassender Begriff ist.. Stichwort: „Maßverkörperung“

@@EngineeringSpareTime ok danke für die Infos! :) gibt es eine deutsche Bezugsquelle für die Clearpath Servos? Warum hast du nicht (wie viele anderen auch) die iHSV57 Servos genommen?

@@andysen1987 Gerne. Eine deutsche Bezugsquelle ist mir nicht bekannt. Ich habe meine direkt aus den USA bezogen. Die iHSV57 waren vor 2,5 Jahren, als ich die Servos gekauft habe, noch nicht all zu verbreitet und auf dem Stand. Zudem hat mir das Gesamtsystem der Clearpath besser gefallen: Passendes Netzteil, Encoder bleiben auch beim Notaus unter Spannung (Topologie Frage selbstredend..), Leistungsfähiger, AutoTune Fähigkeit mit vernünftiger Software, mit dem richtigen Stecker und Schaftdichtung eine vernünftige IP Klasse und vor allem es gibt einen Support. Der Preis ist allerdings der doppelte.. Die JMC haben sich in den letzten Jahren verbessert in Hinsicht auf Doku und Funktionalität, sodass man diese mit Sicherheit in einem Projekt gut verwenden kann. Der Preis von den JMC ist natürlich Top. Letztlich ist es wie so oft eine Einstellungssache :) Ich hätte auch welche von Siemens genommen wenn der Faktor 4 (min.) nicht wäre :D Übrigens verwendet Tormach ClearPath Servos in ihrer X Serie, das hat sich auch in diesesem Zeitraum erst aufgetan.

@@EngineeringSpareTime haha ja ok das sind alles argumente :)

Thanks for your comment. I am using an ESS SmoothStepper and additional break out boards for GPIO. The version I am using is most likely obsolete, since I have ordered back in 2015. Yes, I'm planningto use macros and a ModBus conntection to operate the ATC. Unfortunately, I couldn't find the time to fully utilize it yet..

Thanks :)

Hallo und danke! Ja, da ist was dran. Viel gibt es zu dem Thema nicht. Ich habe bereits eine 5-Achs-Erweiterung mit einem Bearbeitungsraum von 100mm (im Würfel) konstruiert. Ich muss nur noch die Zeit finden die erste Einheit zu bauen und zu testen. Außerdem plane ich an einem Bau einer kleinen möglichst erschwinglichen integrierten 5-Achs-Maschine (die ganze Maschine sollte deutlich unter 10k liegen..) - das wird sich allerdings noch etwas ziehen.. Möglich ist der Bau auf jeden Fall auch die Implementierung mit Linux CNC wird funktionieren. Soweit ich weiß arbeiten Alex CNC und talla83 schon länger an einer 5-Achs-Erweiterung. Diese wird mit Sicherheit mit Linux gesteuert. Welchen Bearbeitungsraum sollte eine 5-Achs-Erweiterung bzw. kleine Maschine deiner Meinung nach haben?

@@EngineeringSpareTime Hallo, das 5-Achs-Projekt von Alex und Peter habe ich gesehen. Ich glaube, da kommt aber nichts mehr...? Alex baut jetzt ja an seiner großen Fräse. Ein sehr großes Problem sind auch die HARMONIC Drive Getriebe >> teuer, schwer zu beschaffen. Aber, ich habe gesehen Du beschäftigst dich auch mit Getrieben. Ich selbst mache Rotorköpfe für Modellhubschrauber. Dafür wäre ein 5-Achs-System schon super ! 100mm Bearbeitungsraum wäre für mich persönlich vielleicht etwas zu klein. 150mm wäre besser. Außerdem darf man ja nicht vergessen, das da auch noch ein 5-Achs-Schraubstock rauf muss. Würdest Du mir Deine 5-Achs-Erweiterung zeigen...? Vielleicht per Teamviewer oder mail.... LG aus HH Achim

It’s triggered by the Emergency Stop Button with cross-circuit detection and a optional switch for e.g. an enclosure door. There is a 24V signal for e-stop that’s connected to the sps, a contractor for the spindle motor, a signal to the inverter and Servo drives. The servo encoders and the mcu are always powered. I hope this answers your question

The motor is a water cooled China 2,2 kW spindle, reduced by a belt drive -> max. spindle rpm is around 12000 rpm. The spindle cartridge with the tool changer is diy. For even higher torque I’ll upgrade the inverter to a vector driven one

@@EngineeringSpareTime Can you see how much power is actually being drawn by the spindle? i am currently upgrading a machine and have a 1.2kW DC motor to fit its max rpm is only 4200 so i was thinking of gearing for at least 8000.

Luc Davenport It’s possible to see check the power in the inverter ui, I didn’t check it though. How much power is actual needed, always depends on the cut you take. Lighter cut -> less power.. 1,2 kW should be enough for most operations up to a diameter of 6 mm.

They are made from S355 construction steel - laser plates.. The idea of the cut outs is to reduce surface deformation in X direction due to vibrations close to the eigenfrequencies (changing the vibration modes...), this shape is still stiff in Y and Z direction though. And it looks way cooler like this ;)

@@EngineeringSpareTime how did you go about evaluating eigen frequencies and whether they would be significant?

@@ewildgoose I just roughly calculated the nominal frequencies of the spindle motor, spindle shaft and according cutting frequencies. Some of the spectrum equals the vibration mode of the plate without the cut outs, so I decided to add them. If this was necessary is questionable, the plates are not vibrating or ringing though..

Thanks! It’s a nice one: Wabeco D6000 :)

Engineering Spare Time ah. Yes indeed a nice one!

X and Y are MGWR09, Z is MGN15HZ1HM

@@EngineeringSpareTime thank you. Are they rigid enough?

@@john1987742 This is what I was able to get at the time - I would have otherwise used a size larger. In retrospect I can’t complain about them.. they might not be this long lasting, we will see though. And according the rigidity, the video might give an impression of it.

Thank you! It actually did not cost a fortune, but quite a lot of effort/time. I manufactured all the parts in my 4 sqm workshop, using handtools, a drill press and the benchtop lathe shown in the video. Most of the standard components are purchased from eBay - saved a lot. For the construction including the spindle cartridge, I spend about 1.5k. When everything was basically working I decided to add the servo system. That almost doubled the spend amount. So with enough effort it is possible to get it reasonable cheap, but yes some sort of machinery - in my case the benchtop lathe - is needed (i also did some mill work on it). And yes the spindle design needed some iterations :-) I‘m sure you can, there are a bunch of different ways to built one..

Thanks! I kinda used the “best“ stuff which I was able to get and not 100% what I would have liked to use (due to cost). But anyway: The Linear rails X and Y: HIWIN MGWR09 + carriage WGW09HZ1HM (sounds small, C_0 is 5.9 kN though..). For the Z axis: HIWIN MGN15HZ1HM. Both types with light preload (Z1). For the ball screws - I was lucky - I could get my hands on some ground 12mm P5 ball screws from STAR (Rexroth) check the video 10:43 .. normally those would have cost more then the rest of the machine (without the servos and spindle motor). For the ball screw bearing arrangement, I used precision spindle bearings from SKF 7201 (good bearing are important! Everything else is not worth the time..)

@@EngineeringSpareTime Thanks for the quick reply! Yes very impressive. It's amazing how much performance you can get from relatively "light weight" linear motion components. I see a lot of over-built machines that don't perform that well. Any chance you have more info on that home built spindle cartridge? Even a video with some pictures and dimensions. I know a lot of us would be very grateful, as it's a bit of a stumbling block (for me anyways). Cheers!

Devin Moodley At the moment I‘m building a big CNC Machine, that is not light weight anymore - I could built this machine with the new one so much more efficient :) This small bench top machine is very practical in a sense of space, weight, speed and capability (it’s quite precise..) though. The working area in X, Y, Z is enough for most of the stuff you do. I didn’t take to many pictures of the spindle, cause it took for ever to machine it with this small lathe and I forgot about it. I have to think about this, it was definitely no straight forward to build the spindle cartridge completely from scratch (incl. the pneum. cylinder), I learned a lot by doing this, and would do it a bit differently next time - I might not be happy with releasing/ showing this “80% spindle design“

@@EngineeringSpareTime I'd love to see a video discussing what you would have done differently with the spindle. There are so many "My First Cnc Router!" videos out there and not nearly enough about precise and rigid machines like this!

BuySomeApplez If there is interest in this, I‘ll think about making a video about the thought process and construction of such a spindle cartridge. “My first CNC router“ is a nice way to say it - “Machining metals“ has sometimes different meanings :) Yeah, there are really just a view - mostly bigger machines. This was actually part of the motivation of making a a video about my machine..

Thanks! I used a SmoothStepper with expansion cards. I’m running it via Mach3

In one end two of 7201 in O-arrangement. On the other end you want to use a “floating“ bearing that allows lateral movement to compensate for temperature elongation - e.g. a needle roller or a regular 6000, 6200.. You can also get of the shelf bearing blocks - easier to use if the implementation is possible.