I LOVE a TOT video but seeing my mug (and its being tweaked) made it even better. Top job Tony as usual. Still the only other channel I'll watch the whole video regardless of length.
Hahaha, how did I know The Furzinator watched this one too? Outstanding. Did I really just say Furzinator? I think I need to get out more. Ideally to The Far Frozen North to have a look in Colin’s shed...
For future reference, when you need to cool your belt, a good option is to move to Greenland and only machine in winter with the garage door open. It's a little inconvenient, but you can't argue with the results.
@@Hopeless_and_Forlorn Belts I see on linear motion are quite loose. I'm sure you could either use positional accuracy as a means for tightening or temperature. At some point it might drop off.
It might be better to investigate direct-drive with no belt at all -- if Tony really needs speed on the Z-axis then an uprated servo on that axis only would be a better bet. Direct-drive would also mean he won't have to faff around to get the servo carrier plate out of the way of the knee if there's interference plus it removes another source of positional inaccuracy and maintenance when (not if) the belt wears, stretches or breaks.
My 7 year old was watching this at the start with me. When she saw the chisel "Magic" as she called it she actually watched seven more minutes of the video, full of questions and then totally lost interest. Point being, Thank you. Its amazing you got her interested in a part of the world that is of no interest to her . GG
What's more is, seven year olds of today are not like the seven year olds before the millenium ticked over. When i was 7, even though i was a boy and not a girl, i watched and helped in the workshop all the time. I operated machinery like excavators, dumpers, dozers, wood splitters, circular saws, tractors and a few other strange things. Had i been born seven years ago, all i would probably have learned by now is to stuff my face and record Tik Tok garbage and use snapchat. Funny thing is, i've kept myself updated with the times as well. So i can of course do recordings and all that garbage if i wanted to. I do programming as well as operate machinery and a whole host of other things pretty well. Meanwhile, kids these days will probably grow up to be a two trick pony pretty much. Getting by on social score and becoming e-famous.
@@RealCadde Please keep in mind that this is a generalization. Not all Gen Z people like TikTok, not all seven year olds sit on their iPhone X all day. Some of us do enjoy labor. I, as a Gen Z, detest TikTok, love fixing bikes, and do a bit of programming too.
AvE's the boss who stops talking to you when you ask him how many feet are in a yard. ToT's the boss who answers, "It depends how many people you're having over."
I remember the time when a young TOT could karate chop a hunk of aluminum into submission with one chop! It’s a sad day when you have to use a chisel to do it.
I repair machines in a forging facility with about 30 production turns and grinders. Servo's are the best thing since sliced bread! Black rubber belts may help with the heat. And NEVER underestimate having a "home" position to use. Batteries die all the time, crap happens.
No batteries involved, those servos have a 23bit absolute encoder, but a repeatable home position ist worth a lot when you do things like change the belt or screw up your programming, drive the machine against a mechanical block and your belt jumps a notch or two. Slightly related, the Maho came with glass scales, if the Mach 4 Software can use a secondary measurement system for positioning I would reinstall those too to compensate an possible backlash in the mechanics.
@@WoodmanX yes, but those 23 bits of resolution are applied to 360 degrees of rotation of the servo, so, someone or something making it take more than a full turn in a no power scenario could mess up the home position (I guess. Correct me if i am wrong)mk
@@lcsantos777 It seems 23bit singleturn encoders actually exist, I have no clue where one would need the 0,00004° in angular resolution so. I had simply assumed a 23bit encoder would be a multiturn encoder, that would split angular resolution and turn counts, those do run over at some point, but assuming a typical 12bit reolution for a single turn would leave you with 2048 turns to count, that seems plenty enough for a normal sized milling machine.
@@WoodmanX might not even be because of that. Might be simply because you can add 8bits for some other info, a parity bit, and then you have a nice little 32 bit data frame.
I read this as "5 year welding anniversary" I thought, wow, this guy is serious and commited to welding metal together. Almost brought a tear to my eye.....Then I thought, I should get some new reading glasses before I embarrass myself on the internet.
Zed is imperial. The only reason you have imperial is because you were in an empire. The British empire. The British still use Zed. So zed must be imperial.
Look into my eyes, you hear only my voice...you want a ball screw for the Z axis...you need a ball screw for the Z axis...you will be happy with the speed and cooler running with a ball screw for the Z axis. When I snap my fingers, you will awake, refreshed, at ease and very happy with your decision to replace that old, slow, clunky acme screw with a nice, new, sleek, carefully matched ball screw for the Z axis. Thank you for some of the most fun I have on RU-vid! Safe, productive days, and peaceful nights, good sir.
So in that case he'll need some kind of brakes (or holdin ax with the servo), without it all z axis will roll down on ballscrew under its own weight when motor is off
N T, yup that’s one of those complications that caused TOT to stay with the acme screw, while he was going back and forth across much of the project. BUT a spring sustained solenoid brake driven by a slaved relay eliminates that concern. In the end, he’ll be fine either way, but I can’t resist the chance to tease him. His are some of the consistently best videos on YouTub, IMO.
@@leslieaustin151 oh, silly me, not only I was using metrique hammer, I was using metrique anvil! It's a miracle it didn't braze additional metal in the hole!
keep your PLC / machine network separate (physically isolated) from your home network. preferably VLAN/subnetted and firewalled. because the last thing you want is that machine unexpectantly running/moving/changing/updating while your hands are in it. as a security analyst for machine/process networks....i'm just saying, keep it isolated and keep your hands.
That is typically not a concern for EtherCAT servos. EtherCAT in general does not support TCP/IP. (It is a rarely used extension, which requires special support in both the servo drives and the PC controller software) EtherCAT also don't play well with switches. It needs to be a direct connection to the PC. But I guess the PC could have second NIC. Which of course should not be online.
Yeap. I remember reading an article called "Why is this computer connected to the internet?" or something like that. The premise of the article is why do companies want every computer hooked up to the internet. A computer controlling a piece of machinery at a factory. "Why is this computer connected to the internet?" There is little to no reason that most computers are connected to the internet. The computers can be old, not secure, running old OS's and it doesn't matter.
@@vonnikon That is basically every industry system. EtherCat is just another protocol language like CanBUS, RS485 etc. They all work with RJ45 contacts today. Because it is easier to disconnect/re-connect etc. Doesn't require an electrician to assist with the set-up among other things. Not to mention, you can use CAT6 cables for the best magnetic resistance. To be able to handle electrical field distortions that can happen in certain Industrial enclosures. But conversing it to TCP/IP is fairly easy. Just a gateway that acts as a translator.
@@vintagespeed as a technician working on fairly black box equipment troubleshooting an intermittent critical failure, having access to the machine's event logs and settings would have saved me hours on the phone. Ultimately the problem was caused by condensation which they wouldn't be able to detect but it would have saved all the "have you set x to such and such value?" Some times being networked helps but it's entirely dependent on the individual equipment and who has to work on it.
Is it cool with you to say that your videos remind me of a feeling that I had as a kid watching Mr. Rogers Neighborhood? By the time you say thanks for hangin around the shop with me, I'm genuinely happier than before.
Seriously. Being furloughed, and wondering what the hell to do with myself to keep my house healthy and happy has been challenging for my brain health. When I finish any of Tony's latest uploads, I'm ready to get off my ass and go do something. He deserves more patrons than however many he has.
14:15 Its free to turn in one direction and not in the other. The nylon thing adds resistance in the downwards direction. Without it the machine can make the screw turn down due to vibration. It also gives you an equal amount of torque in both directions, this is much more practical for the machinist.
Not just that, but it also balances the load on the motor. When holding stationary, it's always best to try to get as close to no torque when possible to prevent premature failure of the servo.
I worked with a group of guys who. 30+ years ago, learned a lot about stepper motors "the hard way". These were small, very inexpensive steppers that had a gear box that converted the rotary motion in linear motion. We would drive the motor to one stop at power up and call that zero. Through the "school of hard knocks" we learned that if you hit either stop at max speed, the motor would "bounce" and start running at twice the speed in the WRONG DIRECTION ! We also discovered any time you stopped or changed direction there was a chance of loosing a step. Good thing was, there was "indirect" (but very slow) feedback in the total system, so we never got too far "lost" !
Hey Old Tony Great production and hilariously instructive as always. My partner brought me a coffee as I began this vid and stayed to watch. She was amazed to see such a clean cut with one strike of a masons chisel. I had to disabuse her of that misunderstanding or I would forever be hounded to try that instead of a hacksaw. Who woulda guesst A fine machinist and a cinematic genius.
1:28 confused me. I thought those were just random PNGs or something overlayed on screen then the reflections made me think "Thats good video editing." Then I finally realized its just the lighting and they are real. Also shout out to Brian for sending over the gear.
Nice job Tony!!! Servo control is amazing and fun - I have done high speed servo control design for several years. Fastest so far for me has been 10g of accel to 300ips max speed using a 3HP servo. Watch yourself, these motors can cause some major harm. I really like the accel I saw. Don't go Tim Taylor on us. I have broken several shear pins (among other things) with over-aggressive accels. A few other things to note: 1) From a pure motor standpoint, there is no difference between a servo motor and a stepper motor (other than an encoder which you can add to either). They are both just brushless DC motors. The difference is the control electronics. 2) Often a servo motor (like yours) will include some electronics (like for the encoder, resolver, etc) which a person who wants a stepper doesn't need. Also, servos usually market to a higher end task than steppers so the manufactures often put more $$ into them (like better magnets) so they often have better performance. 3) DC Motors deliver their most power at their highest rated RPM. It is best to gear them accordingly - but with some margin. Never run them at max RPM as the control loop needs room to operate and motor power drops rapidly after the motor reaches max RPM. 4) Servos DO NOT like lag/lash (non-linearity) between the shaft and the load. This can really mess with the control loop. Best to eliminate as much as you can. If you hear any buzzing when the servos are holding position, you need to either eliminate the lash or soften the tuning. The buzzing in not only annoying, it can do mechanical damage. 5) EtherCAT (Ethernet for Control Automation Technology) is cool. It allows the motion controller to talk to all the control electronics over the same wiring and there are a lot of compatible devices. 6) Don't completely give up on limit switches, they can be a really good friend. Control systems can sometime go haywire due to a SW or HW (or operator) failure and having a switch that tells the control system to "go no further" can save slamming several hundred pounds of metal at full speed into a hard stop. 7) Make sure to include a big easy to hit E-Stop (Emergency Stop) switch in your design so you can quickly kill power to all the motors. Not only is it a VERY GOOD idea, it is also the law.
13:55 I think the sprag clutch and plastic bushing combination is to balance out the resistance while you're moving the head up or down...the plastic bushing provides extra resistance when moving the head down while you get no extra resistance(besides gravity) when going up. It might even keep it from freewheeling down. Love your videos, take care!
I predict that Tony will be visited by the ghost of backlash future. The feedback is on the motor, so it doesn't measure where the actual machine is, and this is especially true for a z-axis belt-driven lead-screw.
It may just be that the sprag clutch and nylon bushing inside the jack shaft housing are your anti-gravity device. Seriously. You are hanging the weight of the X & Y axis from the Z axis. With the servo motor turned off, and no hydraulic (or any other type) brake, this weight creates a downward force that can (will???) turn the screw, and cause some drifting issues. With closed loop, you can certainly continue fighting this downward force by constantly readjusting the Z axis ... but that may be an additional variable in your Z axis positioning that will cause you issues. When fully assembled, you can adjust the nylon bushing for equal force required to raise and lower the Z axis ... knowing that based on the current condition of the sprag clutch outer race, this is a constant wear item that needs periodic adjustment. Counterbalance springs (flat spring stock wound like a clock main spring) are available, and don't have the constant wear, but require careful selection (and maybe some ballast) to get set correctly. The servos may also allow for different torque and speed adjustments in the tuning process to counter the torque change between raising and lowering without the sprag clutch ... but without an integral brake, you'll still get drift. If the servos have holding torque available, kindly ignore all the above! Thanks for sharing. R
it might just be to delete the feel of the weight of the thing so turning it up gives the same resistance n the crank as moving it down (where the weight helps with going down and if oiled enough might even freewheel all the way to the bottom )
@@watahyahknow I came here to make that suggestion, you wouldn't think that it matters, but having to use different ammount of force in one axis really makes the feeling wierd and working with is sucks
You will want ballscrews on the Z since its being driven through gears.Without a ballscrew there will be backlash in the gears, but if you use a ballscrew with a large lead then the drive will always be under tension in one direction as the weight of the knee and table will be enough to overcome the friction in the nut and it will always be wanting to move down.
just a small note to ya. your vids are one of my top five subscriptions. love the humor that you incorporate into the videography, not annoying like some, just perfect. i might be the only one on break on a construction job laughing my ass off meanwhile being inspired and learning at the same time. that you for the time that you put in to create such perfected coolness brother. much appreciation sir.
Tony! I’m a rookie mechanical engineer working in my first position where I need an in depth understanding of GD&T. You should start a mini-series going through it! You seem to have a great understanding of it, and your teaching of things with great visuals has been extremely helpful for me. I’d love to see something like that!
My friends dad had a machine shop in his home. He had two Bridgeports and a Boxford lathe. He worked for dream works. He use to do a lot of projects at home. I remember we used to go in there and start cranking away. Man he would get so red faced at us.
I am a trade school student and I have been watching these videos for awhile, I am taking classes on machining with both manual and CNC lathes and mills, since the pandemic started and the schools closed, I've been watching these videos to learn my trade and I've learned alot. Thank you.
TOT, You got Z axis to return to .0000. This indicates that you have no mechanical backlash in the Z axis drive train. You are driving via a timing belt and gears and using a lead screw. At some point this drive train will experience drive train backlash. Your position encoder is integral in the servo motor. So you are monitoring only the motor in your servo closed loop system. How is your retrofit going manage machine accuracy and repeatability since you have no machine position feedback? I repaired a large Mazak mill that we purchased used with non functioning linear scale position feedback by using only the motor encoder. A replacement linear scale feedback wasn't deemed cost effective. In this machine design the servo motor was direct coupled and it was naturally a ball screw system. My experience led me to believe that the ball nut and screw backlash and ball screw TAC bearings preload was reliable enough to risk running with only servo motor encoder feedback. We also had inspection equipment to periodically evaluation linear axis positioning accuracy. I don't like trusting any drivetrain that uses a timing belt. I realize that your closed loop system is a big improvement over your open loop stepper motor system. When I broke into the repair trades in the early 1980's, we had several open loop machines that occasionally gave us positioning errors. Your explanation of how this happens is spot on. I used to explain this to new technicians the same way you did. I only ever maintained one CNC machine that didn't employ OT switches and zero reference switches. All is fine until you lose battery backup power and then you lose absolute positioning. When this happened on this machine, the technician's life just got miserable. Life was always so much easier when even an absolute positioning machine could be simply referenced to re-establish lost machine position. I would certainly like to know how your servo motors maintain absolute positioning. Battery life keeps improving, but I would like to know where that battery is located and how long you can expect it to last. I applaud you for the servo retrofit. I no longer take much interest in the effort. That was the type of project for my employment years. I am interested in simpler tasks in my retirement.
In the industrial world we use flags for the axis limits. (Home and Max) Might be worth considering adding a mechanical flag with a EE-SX-671A OMRON photoelectric sensor. Omron also makes a wired one, like the EE-SX671-WR 1M. Url: www.digikey.com/product-detail/en/omron-automation-and-safety/EE-SX671A/OR612-ND/307730
Your metal cutting shenanigans have spoiled me. Every time I watch a video where someone scribes a cut line, I expect a karate chop or knife slice or something to happen
Yup, it contracts when it gets on the pully and releases that stored energy from streaching as heat into the pulley. Mind you, what heat scale was that, are we talking 0.001c or 50c??
No, it's just too heavy of a load for that belt. It will get worse when he adds the X and Y and the workpiece mass to the load. I wager TOT will eventually have to upgrade to a precision chain drive. Or a gear drive.
@@jeffvoight6586 I dunno if your gonna get much back lash with all that weight on it. I guesseing that sprag clutch was also setup to be a bit of a helper in someway in the up direction.
It's the kind of video that makes you want to go back to school, and that is what I would have done if I hadn't already gone back to school when you first got this thing! Guess what I'm trying to say is that: when you clear leaves out of the Gutter, you never know who's going to come knocking the next day, asking for a bottle of milk and some change.
With ToT's superhuman skillz at cutting metal, I figured he wouldn't even have to touch it. I figured he would just tell it to center itself and it would obey his command. :)
ToT you should be proud, your videos have made it to the big leagues. My professor used your lathe chuck and surface grinding videos in my 331 engineering course at NIU, I'd recognize those hands anywhere.
Or maybe the belt is not constructed/ incorrect material composition for the torque application....timing belt not drive belt? Reacon it is flexing/streching, just wait till the rest of the wieght goes on, might just fail, that sprag clutch was probably there to prevent the weight of the mill table combined with gravity from free spinning down when it was manually wound.....just my 20 cent bet ;)
@@AadityaKhare42 Every belt has a maximum force you can use it for. These belts look like T5 belts to me. I switched to T10 (that can handle bigger forces) on my robotmower because the T5 broke. If this T5 breaks TOT is in for a bang.
This channel is the perfect mix of knowledge, information, visuals, and absurd silliness. I am not or have ever been interested in machining, but I watch all your videos and find them fascinating and hilarious.
Imagine him hitting that chisel only for it to make a useless "thwack," the aluminum unharmed. Frustrated, he puts the hammer and chisel down, and takes the aluminum to the band saw.
When I refreshed youtube and saw this video, I made an audible "Ooh!". I'm excited to see progress on this, I don't really understand a lot of it, I don't have the brain smarts to, but I enjoy watching for sure!
I have been eagerly awaiting an update on the Maho and I am impressed! The time you have spent putting this video series together so people like me can understand the basics of how to upgrade a manual mill to CNC is greatly appreciated. I have learned something and gotten a good laugh from all of your videos. A sincere thanks to you.
TOT regarding the belts heating up: "I'm not really sure what to do" Us viewers: "what do you think it means?" Soft voice in the background: "Gear cutting video"
@@hackish1 Is there any reason not to use glass scales on the axes? Then backlash anywhere in the system shouldn't matter. I mean I suppose it would add extra cost....
@@blahblahblahblah2933 the sync time is problematic. When you command the motor to turn, there are a lot of things that happen before the axis completes its move. You can't stop the machine after every move. Using the encoder on the servo is a more reliable method.
"I also knew I would need a TON of torque. Maybe not 2000 pounds, but a TON of torque nonetheless." Is that a metric ton joke? Cause I am seriously impressed.
@@byronwatkins2565 usually rotary encoder on the motor and linear on axle, loads of advantages although a bit harder on software, don't know if those drives have the option though
@@Nuno.A Servo motor position (angle) is quite precise; however, gear backlash allows the table position to vary several thousandths. Clockwise motor rotation gives different table position than counterclockwise motor rotation.
@@byronwatkins2565 exactly true, and to avoid backlash, play, on the gears, belts, ball screw, coupling, etc, the table encoder is usually linear, has that give you a very precise position on the last possible moving part, on a vertical is less of a issue has gravity hides many of the issues, with that you can also do temperature compensation and correct backlash.. to a point, too much wear and the axles start shaking has the control goes in to constant overcorrection
If you are moving one object into a relative position to another then it's going to end up at **some** position and that position **could** be that close to where you wanted. I'd believe it's possible to happen to hit that mark......you just couldn't know without then measuring accurately to find out.
we recently added cnc to our e mill. we had already added steppers to X and Y so my dad's arms would stop hurting. then got a router, learned mach3 and cam softs, and transferred the brains to it. it was a project in itself to mount the z motor. I also used a timing belt reduction.
I am honestly not particularly interested in machining, but your videos are simply so great that I watch all of them, some twice. I really love the small quirks as well as the outstanding attention to detail in the production of both the videos and the actual stuff you make. Thanks so much.
With a lead screw, that may be true, but with a properly designed ball screw set up, with preload. The backlash becomes vanishingly small, certainly good enough for home. In very high end CNC, they switch to glass scales for basicaly insane accuracy.
@@teryk The short answer is it depends. But highquality timing belts can get into seriously lil backlash. Many actual machines use em ? Like CNC lathes or packaging equipment (what I design). Realistically you could eliminate everything and put a servo gearbox instead, the "common" servo gearboxs have like 0.05 degree of back lash (3 arc minutes) and you can find better if budget allows.
@@stickshaker101 this whole covid bs. Is going to usher in the nwo. Unless we RISE. It's so hard to see people so brainwashed by their tv. Turn it off, it's a weapon. Wearing a mask is submission RISE #WEDONOTCONSENT #WEDONOTCOMPLY
Awe, yes. ECAT. “Deterministic communications “. Welcome to my world! For industrial controls it is a smart choice. The daisy chaining brought to you through the “train” of data flowing in to modified and right back out to the next node. I had not tried to chop aluminum with my putty knife wow have I been missing out!