They have taps now that are made for drills and tap guns made especially for this situation. Dewalt and Metabo both have sets. Pro Consumer grade so probably pretty cheap. Same as what @Pat Kirk suggests but probably not as expensive.
10:15 What you probably want to do is run the motor in FOC_VOLTAGE mode, and implement a PID loop on top of that to control the motor velocity from hall sensor feedback. You will get the most low RPM torque out of voltage control in my experience, with the added benefit of 0 "wind-up". At some point you will reach the limit of what the motor can output until you can attach an external encoder to it, as the FOC observer/predictor can only guess so much about the rotor position at such slow speeds, and will begin to twitch. Edit: I should probably watch to the end, you figured it out basically 😅 28:30 Its slipping that much with such little force as taps are hardened, and the jaws in Jacobs chucks are also very hard. So even if you tighten down as hard as you can, they will always slip as its essentially like glass sliding on glass. 31:24 What I typically do when working with these HB mainboards is to put a jumper on the power-button JST port so that the buck-converter circuit and CPU is essentially always latched on. This of course empties any remaining charge in the bulk caps. I also typically keep the jumper connected and disable to power-button functionality in the firmware so that it is always on when power is applied. I do this for projects which have multiple hoverboard mainboards attached (4WD typically), so they all power on and off in unison. I used a precharge circuit & a 100A DC-rated MCB for power on.
Use all three positions in the chuck to get maximum tightness. Tighten at one position then move to the next, tighten, and finally tighten at the last.
That thing is awesome. by far one of the best repurpose projects i have seen in a long time. Below are some things, they are all unnecessary and wont give any better performance but add to the Safety of it. when fitting potentiometers to gear that has enough power to hurt you there are two main things that help increase the safety factor. -The first is to use two small resistors, one on each side of the power supply to the pot. then in software program the range of values of each resistor inhibits from the pots range as failure zones. For instance if you have a 5k pot and you add a 220ohm resistor to each side of the pot, and the controller is 256bit adc then you can assign bit range '0 to 10' and '246 to 256' as pot fail modes. that way if the wiring or connections to the pot fail and it becomes shorted or unplugged it will stop instead of going full speed. -The second is two use two pots in a inverted setup, one rising and one falling then in software compare the value of each and fail of the pot 'A' - Pot 'B' = greater than pot accuracy example 8%. The second is safer and even better when also combined with the first. The same fundamentals can be done to switches. adding a validity check resistor. Unfortunately most of the time this can only be done with a analog input or 3 well balanced digital inputs per switch state. the next thing is the belt. a correctly set drill press belt offers a primitive torque limit function. I some older drill presses the belt was never tight and would be hand liftable from one pulley to the other. Unlike a car fan belt or other drive belts that get adjusted to a tight tension.
You could put a limit switch on the depth indicator and screw it to the desired depth. That would rotate the tap backwards for a fixed duration until you're ready for the next hole.
Ya just gotta love the ability to weld steel together at home. Keep at it and you'll get it dialed in. Just from my very minimal observation and from one asshole's opinion......... It looks as if you may have her turned up to "obliterate". I'm assuming that you're in a nation that supplies 240 volts? A lot of my experience with the 120 volt MIG welders here in the states is that they HAVE to be turned up to "obliterate" just to weld anything that's substantial, but the 240 volt MIGs can get down on some serious thickness and blow holes through half inch steel like it's cool. Lol! Love this project!! I've got 2 of those exact hover boards and all the internal components and have been dying to use em on something worthwhile. So, thank you thank you thank you for sharing all the details that you have!! Keep em coming!!
I like your comment about sticking around toward the end of the video - to be honest this has been one of the more interesting posts of late and I didn't mind the length. I have to wonder if it isn't worth taking it all the way and automating the z-axis with the other motor. I am pretty sure you can do position control with FOC and the hall effect sensors. Come one! Any which way, nice work and always nice to see hands on work from a guy that is as industrious as you! :)
last year i was waiting for more videos then you said you were working on software for some thing... now I am getting what a marvel you have created in those times.... this is just amazing...
@@iforce2d man you remember my name... damn....... i am feeling special now... i was experimenting with "channel name" field of youtube.. it gives notification panic...
Great idea, looking forward for the rest of the build! Might copy your idea one day :) Btw, if you want to reduce the vibrations, instead of improving the alignment of your anti-torque bar, give it more degrees of freedom! (With an extra linkage between the motor and the steel support for example)
What a great exercise in problem solving. So many little problems solved along this DYI CNC journey. Thanks for documenting and sharing. Your programming examples with the sbns-mixer (universal programable controller?) continues to amaze me. You making a great case for this style of visual programming.
Wow, Cool project and great explanation of the process in the video. This kind of software control for motors is new to me and very interesting. I’ll subscribe to learn more on the subject. Thank you for sharing!
I made an overview about this recently, would be interested to hear any questions you might have www.iforce2d.net/tmp/swip/overview/overview220526.html
you can use a tapping head to do what you wont to do. in order to get to the bottom of a blind hole you have to use two types of taps a plug tap to start the the threads and a spiral flute bottoming tap to get to the bottom of the hole.
My forcefeedback motor is a industrial servo motor. Jep just a 15k step encoder on the back makes a regular motor work with very accurate positioning. The controller sees movement smaller than me moving the wheel 1 hair width. Forget stepper motors that can lose steps, slap an encoder on the back and know motor position at all times.
Knock the chuck out (it's a Morse Taper). Then refit the spigot an old 2-jaw or 4-jaw chuck instead. You can get them from old drill braces, or from a hand tapping thingy.
Magnetic drill press is what you are thinking about. the press you have as it sits is sufficient to do everything you just need to rip out the controller for the motor and install a more advanced digital controller
Love the idea. How about adding a depth bar to the column with a micro switch on the end of it? That way you can set the depth on the bar and once the micro switch touches the work piece it can enable reverse and retract? Keep up the great work!
You’ll probably need a function that takes the thread pitch and desired depth, so that you can count how deep the thread needs to go before a full retract. For instance you normally want at least 2 thread turns below a through hole so that the tap doesn’t have any taper. Take an m6 tap through a 10 mm stock through hole for instance they normally have 1mm pitch, so you can count your rotations, you should aim for 12 rotations (12mm depth), then command a full retract. Manual works too.
What would be really interesting would be a strain gauge sensor in the handles that allows you to turn them up and down to contact the tap to the work piece but when it contacts and you push against the work piece, it moves the motor ahead at the rpm you set it for and when you let of of it it stops. If you pull up on the handle while the tap is in the piece, the amount of reverse resistance would trigger the motor to run in reverse at the set rpm. Probably this approaches over engineering but interesting to think about.
Don't know if position is that important, but if it was, you can print 2,3 or more bands of stripes that can give you a binary position with resolution determined by the number of bands -basically a quadrature encoder. The strips can be read as they pass by some kind of optical detector that switches state depending on dark or light, or even better (and less influenced by ambient light), you can look for the transitions. Wouldn't be too hard to print the bands onto paper and adhere it to the rim of the motor. I suspect you can buy the led/detector in a package off the shelf and quite cheap. (I might be wrong about that).
since your controlling speed without belts why have the huge multi belt ? simly mount the hoverboard hub and put a small belt pinion on the spindle and use the tireless hub with belt to the spindle. better yet, take hub apart and from hub front facing face you can bolt a collet directly to it that fits the spindle shaft.. this would drop the motor right ontop the shaft much lower and bracketing would be much easier. we use this methos all the time when converting hoverboards to gocarts etc to mount larger rims.
Love the innovative approach and great video as always.... If the motor is close loop, could you use it for motor position control. Just out of curiosity you do know that you can get a pedestal drill tapping attachment
You mentioned early in the video you are tapping blind holes, using normal taps means you will have to change taps, and they also don't remove the swarf. You need to get some spiral taps, can tap to the bottom of a hole and push the chips out the top. And always use a tapping fluid or compound. Are you aware you can purchase power tapping attachments - they plug into the arbour of a drill and automatically come back out when you release the pressure on the handle, they also have a clutch which you can adjust so you don't break a tap. They are fast. But you need the spring return on the drill quill to do the return.
With the 40 or so volts sitting across the filter capacitors one way of discharging them is to permanently fit a bleed resistor perhaps in series with an led for visual indication. About 3.3kohm or more will be adequate. Although it would drain the battery if the battery is left in circuit in a standby state. Nice to hear a New Zealand accent on CZcams.
Cool little project, try gun taps instead... they are night and day better than hand taps. Also, not so sure about the way you held your motor. The rear part of the shaft should have some float to allow for misalignment. Check my video on the automatic door I made with one of these motors for more detail. The stress from misalignment will damage bearings, fatigue welds etc.
I think I'd just stick with the belt drive and eliminate all the issues that come from trying to centre a weird-shaped motor against some other piece of metal, against a shaft. You can lock the motor shaft into a plate screwed onto the top of the drill mount, chuck a belt on it, and you're done.
I have considered doing similar with an old drill press, but we now have a CNC mill with rigid tapping so I doubt I will do it. For me the issue was that my drill press doesn't have a reverse or low speed so I looked into using a small 3-phase motor with VFD, but your method looks good too. You might consider using an ER collet holder instead of a drill chuck. That would allow for some improved gripping strength on the tap and would allow you to use drill bits as well vs using a dedicated tap holder. Tightening is done with a spanner which I find to be easier than a chuck key. Regarding grip strength - We have an old Bridgeport that we were able to use a 3/4" NPT pipe tap going into thick aluminum using an ER40 collet if I recall correctly. I never had the pipe tap slip even with that large OD. Will you need to add some sort of depth feedback or torque feedback so that the motor can automatically reverse? Doing a through-hole like you did in the demo it would be fine to just control it manually, but a blind hole like you plan on doing in the future will cause the torque to go through the roof if the tap bottoms out and will be likely to break the tap I would think. Maybe sticking with the drill chuck will keep that from happening...
I'll just have to be very careful with the blind holes, especially as others have pointed out I should be using a spiral flute tap for that. Starting to think I should just make them thru-holes instead :)
So cool. Way beyond my pay grade. I'm sure you realize you could have tapped all the holes in your AU plate before you were half way through this build.... But where's the fun in that? Happy New Year!
Yea I probably could have, I tried not to think about it :) Hopefully this will make the job more enjoyable and will be available to use again in future projects.
A mag drill base might be better for this,although it wouldn't work on aluminum as easily, you'd have to clamp a steel plate on top and then magnetic base goes to that.
an easier option than trying to modify or replace your drill chuck is just grind 3 flats on the shank of the tap, as long as you're careful to space them evenly and take off the same amount of material in each position. maybe you could come up with a jig to help with the grinding
If it becomes an issue I might end up having to do something like that. So far on my three test holes it has been ok, although that was only 6mm plate.
nice work.. the only thing i didnt like was lowering the head to the point of inconvenience.. Theres not much axial load on the quill during power tapping , its self centering.. So you could have the drill head higher and the vise mounted to the table.( less hands involved ) . I power tap every day on the mill using a keyless chuck , which might be of some use here.. Yes , it occasionally slips , but thats similar to a clutch telling me its time to pull out , relube and make another pass. Anyway , cool project !
Right, for tapping it doesn't matter. The problem is drilling, where this press has awful runout, that gets worse as the quill extends. I tapped all the holes on the large plate yesterday and using the slipping as a cue to relube and make another pass is indeed how it ended up :)
If you want te have the motor free spin maby the easiest way is to put e switch that cuts off two of the motor wires. It can also serve as a emergency stop/ safety switch.
That would indeed avoid the braking. Problem is the hall encoder of the controller would detect a discrepancy between commanded movement and actual movement and enter an error state. Disconnecting the hall sensor also triggers error state. It would be better to modify the firmware to ignore any encoder discrepancy after the throttle has been in the middle position for say 3 seconds. Probably not something I'll bother with unless it does somehow turn out to be really annoying. A physical disconnect would be great as an emergency stop though.
I wiuld vacuum bag the MDF with epoxy, like used in carbon fiber, to make the MDF harder and stronger and it should work, it would water proof it also,
Great video..technically well beyond my level of understanding…just starting out re-purposing these motors!! Have recently built an electric barrow for the horse yard!!! Wish I had known about Low torque mode..I have installed just a generic co troller with no facility to adjust!! My next project is to make a remote golf trolley so was amazed to see your two motors running around the yard!😂 could you point me in the right direction regarding boards and what I would need. Apologies for being non tech!!
@@iforce2d Hi, did you have any idea how get more torque of that motors by changing some settings in feru foc firmware ? that hoverboard motors have too much rpm i need more torque not speed
I am sure I don't need to say this, but near the beginning you say you only need to hold the motor shaft from spinning. But you have built a rigid structure resulting in it being over constrained. You want either a flexible coupling between the motor and pulley, or replace the top structure with just a torque link. Now this has got me thinking. I wonder if this sort of motor would make a variable drive for my small lathe? 🤔
Actually a lathe was the first thing I daydreamed about using these for (at least in the power tools category). I watch a lot of turning videos on CZcams, but it seems like another huge rabbit-hole of learning that's a bit low on the priorities list for me to go down.
hi mate some wood turning shops sell a rubber ring that can be turned on a wood lathe to make custom jaws might be able to make coupler and dampener in one
One of the first things I noticed when I started watching BigClive, is he shorts the Caps EVERY - TIME with a little screwdriver . Everytime I see a capacitor , I see some fuck shit waiting to happen.
The cheaper drill presses usually have a terrible runout on the chuck - which is one reason they should be avoided. Clever reuse of it though :). I wonder if the hoverboard motor could do with either balancing or a radial load on the bearings (since that is what it is designed for).
Yeah this one is pretty terrible. Anything larger than a 6mm drill would cause it to bounce off the sides of the hole like crazy, eg. if there was a 8mm hole that you wanted to enlarge to 10mm, good luck with that.
@@iforce2d i have a drill press of the same design different brand. Mine also had the problem that i guess is generic to these with the wobbly spindle. I "fixed" it by cutting down the center of the front of the machine, and installing some bolts across the cut. Essentially it turns the front of the machine into a pipe clamp of sorts. Adjusting the tension of the bolts just right can get your a wobble free spindle that still travels freely up and down
Based on those results, it looks like you could cut down the length of the spring, and get the amount of force to resist gravity just barely. Fewer coils == less torque
I've been attaching a clamp to one of the three lever handles to provide a bit of weight. It lightly holds the toolbit up when I retract, and then lightly pushes it down when the weight comes forward, which works out ok.
@@iforce2d should have just left it how it was after you ton some tension off of it There is no reason that a bit of upward tension would hurt anything. In fact, it would be better if there was some upward tension.
You could get a cheap VFD and use the orginal motor on the drillpress, no need to change. Some VFD have tourqe controll and can be controlled by bus, voltage, 4-20mA, binary inputs and so on. Probably cheaper then your current setup.
That's a system I'm still working on. I will make detailed videos sometime, until then you can see other videos where I used it czcams.com/play/PLzidsatoEzeieb0l_EUxpR0p7ed_RUPD6.html
@@iforce2d Is the software propriety thats comes with the “sbusmixer” hardware? Or is it a standard software tool? And is the “sbus mixer” for sale? Where?
The firmware will be proprietary. The configuration side in the browser is mostly Javascript so that's essentially open source, but it would be a huge amount of work to adapt it to other microcontrollers. Between the browser and the firmware is a 'relay server' third component that allows the browser to access the serial port. I'm hoping to sell the boards someday, but the price of the MCU is extremely high right now (covid?) so nobody would want them. I'm told things will get back to normal after a while.
The motor has hall sensors to detect how fast it's rotating. Closed loop means it's using that sensor feedback to adjust the power, based on a specific target, instead of just changing the power blindly. For example if you're driving down the highway, you would adjust power with the throttle pedal, and 'closed loop' would be looking at the speedometer to guide your adjustments to a target speed, eg. 60mph. 'Open loop' would be simply a target pedal position, eg. 45% throttle, with no knowledge of the resulting speed.
@@iforce2d don't these type of 3 phase motor controllers rely upon the HE sensors to determine what frequency to output? Like if the motor is moving at 1 rpm and you tell it to go to 100%, it's only going to be able to ramp at a rate depending upon how fast the motor accelerates. I suppose you could just feed it a set frequency of 3 phase (a predetermined speed) and ignore the sensors. However I would think the rotation would be all chunky and shit until it got up to the necessary rpm, or even some factor of it. Maybe I'm misunderstanding your analogy to a car speedometer.
There are two types of 'frequency' at work. One is a simple duty cycle PWM that controls the overall power output, 0 to 100%. The other is a more complex phase 'switching' of which windings of the motor should get how much of that power, and which direction the current should flow. The switching depends on the HE sensors to know when to change to the next phase. The HE sensors can also be used to count how many switchings are occurring and from that infer a speed. czcams.com/video/QgcAJpRFj1U/video.html
Hello @iiforce2d, nice job… is the sbus mixer project available anywhere? I love the interface and was looking into how you build that. Cheers and happy 2022
cool! you realize you could use only the percentage dial to do the whole tapping operation right? just go to "1" to reverse the tap out of the hole. if i miss understood the function and is wrong i apologize. =) i really like your projects and i still wish we were neighbors! =D
Sure, it can even be used as is with the radio control, the question is how much convenience you want. I will probably leave out a couple of the dials I showed, since they will not change much in practice and can be determined well enough by twiddling with the 'virtual' dial.
Seems overly complicated to me, I think I would have used the complete wheel and tyre as a friction drive against the main spindle. Replace the pulley wheel with a knurled aluminium drum .???? Just an idea to reduce that out of centre “ wobble”…..
those motors are BLDC, they are basically stepper motors with limited pole numbers. Also, you can just control the position with a position sensor. With the right firmware you can do what you want, not with a hack with potentiometers...
Stepper motors are arranged so that energizing alternate coils will move a known distance, with no sensor feedback required. I wouldn't call these steppers because the method of control is quite different, and you can only roughly determine position based on the hall sensor feedback. But this is all beside the point because I wasn't really interested in controlling position, I was just showing how a rough position mode is available in the "Feru" firmware. The potentiometers I used are only inputs for adjusting speeds manually, they don't control the motors directly.
@@iforce2d you can control bldc motors position very finely by modulating winding current, which is what microstepping is. That is how gimbal motors are controlled.
That's a system I'm still working on. I will make detailed videos sometime, until then you can see other videos where I used it czcams.com/play/PLzidsatoEzeieb0l_EUxpR0p7ed_RUPD6.html The configuration interface runs in a browser as in the video, but it also requires a 'relay server' running in the background to allow access to the serial port.
Exactly, and that's all heating it would do too. There will always be a point that it settles on, and will increase in resistance further from that point. That is not the same as counteracting gravity with a constant force at any position.
That's a system I'm still working on. I will make detailed videos sometime, until then you can see other videos where I used it czcams.com/play/PLzidsatoEzeieb0l_EUxpR0p7ed_RUPD6.html
You mean the steel shaft inside the aluminum pulley, that's only about 2cm in diameter? Not sure how I would fix my MDF to that and maintain alignment, and I wouldn't be so confident that the MDF could handle the torque at such a small radius.
@@iforce2d yes... but dont you have a CNC table.. couldn't you make an adapter out of some plate :D, that aside you could still cut the belt pully down a buit so it wasnt so tall.
That's a system I'm still working on. I will make detailed videos sometime, until then you can see other videos where I used it czcams.com/play/PLzidsatoEzeieb0l_EUxpR0p7ed_RUPD6.html
step 1 : czcams.com/video/EckiOgHsuXg/video.html step 2: czcams.com/video/RZBWDAueUJg/video.html these modifications are great for ecomia and speed adjustments
That's a system I'm still working on. I will make detailed videos sometime, until then you can see other videos where I used it czcams.com/play/PLzidsatoEzeieb0l_EUxpR0p7ed_RUPD6.html
I am sorry What programming tool are you using at the end of the video is that a Arduino tool.I am sure you have mentioned it or covered it in other videos.
That's a system I'm still working on. I will make detailed videos sometime, until then you can see other videos where I used it czcams.com/play/PLzidsatoEzeieb0l_EUxpR0p7ed_RUPD6.html
If you don't drill through... you will continually collect crap in the holes. 32:36... Just weld an extension to the Key's Shaft. 36:37... since pulses are being sent to the motor you could count those pulse to get a rough idea as to the feed and retract distance.
Yeah it's looking like my plan to make blind holes will be a pain in the butt. I'm starting to think I'll just have to drill all the way through. The motor does not move by counting pulses, it's not a stepper motor. If it was a stepper though, yes we could do a lot of cool things.
@@iforce2d If I'm not mistaken... that is a brushless motor and the motor coils are being energized in sequences by the brushless motor ESC. Those are brushless motor pulses that could be used as a sensor/counter for direction and distance travelled by the motor shaft. Of course not as accurate as a stepper motor's pulses but i'm sure better than using time to get rotational distance travelled.
Yes you could run it as a stepper motor like that, but it would have only 15 steps for a full revolution. I suppose you could think of this a stepper motor that is doing very fine microstepping all the time, so 'pulses' are not the same as motor coils. I think the word 'pulses' is the cause of our confusion here. The firmware could be modified to count how many magnetic poles it has turned past, and stop at some desired point. That might be what you mean. But that's essentially what the 'speed' mode does anyway. It increments a number saying how many poles should have been passed, and the motor moves to match that (which is how it returns to the starting when I try to move it by hand). That number is being increased over time at a certain rate, eg. after 1 second it might increase by 5, then two seconds it will be 10, so the outcome you're talking about is the outcome it's doing already, just by setting it to a constant speed for some duration in seconds. The hall encoders are analog and not precise, so I think the way it is now is about the best we could expect.
If you use spiral flute taps you don't have to worry about breaking the chips. That's how most machine tapping works
The quick explanation czcams.com/video/y7G90eFYzxs/video.html
Or cheaper gun nose taps for through holes
They have taps now that are made for drills and tap guns made especially for this situation. Dewalt and Metabo both have sets. Pro Consumer grade so probably pretty cheap. Same as what @Pat Kirk suggests but probably not as expensive.
@@sevilnatas I have used the default 1/4-20 spiral drill tap and broken quite a few. You must have good bearings, a rigid set up and lube.
10:15 What you probably want to do is run the motor in FOC_VOLTAGE mode, and implement a PID loop on top of that to control the motor velocity from hall sensor feedback. You will get the most low RPM torque out of voltage control in my experience, with the added benefit of 0 "wind-up". At some point you will reach the limit of what the motor can output until you can attach an external encoder to it, as the FOC observer/predictor can only guess so much about the rotor position at such slow speeds, and will begin to twitch.
Edit: I should probably watch to the end, you figured it out basically 😅
28:30 Its slipping that much with such little force as taps are hardened, and the jaws in Jacobs chucks are also very hard. So even if you tighten down as hard as you can, they will always slip as its essentially like glass sliding on glass.
31:24 What I typically do when working with these HB mainboards is to put a jumper on the power-button JST port so that the buck-converter circuit and CPU is essentially always latched on. This of course empties any remaining charge in the bulk caps. I also typically keep the jumper connected and disable to power-button functionality in the firmware so that it is always on when power is applied. I do this for projects which have multiple hoverboard mainboards attached (4WD typically), so they all power on and off in unison. I used a precharge circuit & a 100A DC-rated MCB for power on.
Use all three positions in the chuck to get maximum tightness. Tighten at one position then move to the next, tighten, and finally tighten at the last.
If he made a longer key with a longer handle he could get to it easier and tighter
That thing is awesome. by far one of the best repurpose projects i have seen in a long time.
Below are some things, they are all unnecessary and wont give any better performance but add to the Safety of it.
when fitting potentiometers to gear that has enough power to hurt you there are two main things that help increase the safety factor.
-The first is to use two small resistors, one on each side of the power supply to the pot. then in software program the range of values of each resistor inhibits from the pots range as failure zones. For instance if you have a 5k pot and you add a 220ohm resistor to each side of the pot, and the controller is 256bit adc then you can assign bit range '0 to 10' and '246 to 256' as pot fail modes. that way if the wiring or connections to the pot fail and it becomes shorted or unplugged it will stop instead of going full speed.
-The second is two use two pots in a inverted setup, one rising and one falling then in software compare the value of each and fail of the pot 'A' - Pot 'B' = greater than pot accuracy example 8%.
The second is safer and even better when also combined with the first.
The same fundamentals can be done to switches. adding a validity check resistor. Unfortunately most of the time this can only be done with a analog input or 3 well balanced digital inputs per switch state.
the next thing is the belt. a correctly set drill press belt offers a primitive torque limit function. I some older drill presses the belt was never tight and would be hand liftable from one pulley to the other. Unlike a car fan belt or other drive belts that get adjusted to a tight tension.
Great job man! You've put a lot of work in it, thank you for sharing your experience with us! I am eager to see the next episode. Happy New Year!
You could put a limit switch on the depth indicator and screw it to the desired depth. That would rotate the tap backwards for a fixed duration until you're ready for the next hole.
I always get excited when you have new content for me to consume and you again didn't disappoint 🤓
Always love the effort you put into your videos.
Thank you for letting me see this video.
Happy and healthy 2022.
can't believe i watched the whole video 😭 you also sound like my digital electronics design teacher at Swinburne lol
Spiral bottoming taps and some sort of collet chuck for the taps and you’ll be golden!
Ya just gotta love the ability to weld steel together at home. Keep at it and you'll get it dialed in. Just from my very minimal observation and from one asshole's opinion......... It looks as if you may have her turned up to "obliterate". I'm assuming that you're in a nation that supplies 240 volts? A lot of my experience with the 120 volt MIG welders here in the states is that they HAVE to be turned up to "obliterate" just to weld anything that's substantial, but the 240 volt MIGs can get down on some serious thickness and blow holes through half inch steel like it's cool. Lol!
Love this project!! I've got 2 of those exact hover boards and all the internal components and have been dying to use em on something worthwhile. So, thank you thank you thank you for sharing all the details that you have!! Keep em coming!!
Great project, thanks for sharing the build
Very interesting, thanks for sharing. Cheers from Jacksonville Florida 🌞
the click click click killed me
This is great. I'm doing a similar project with those wheels. Such well built motors
I like your comment about sticking around toward the end of the video - to be honest this has been one of the more interesting posts of late and I didn't mind the length. I have to wonder if it isn't worth taking it all the way and automating the z-axis with the other motor. I am pretty sure you can do position control with FOC and the hall effect sensors. Come one! Any which way, nice work and always nice to see hands on work from a guy that is as industrious as you! :)
last year i was waiting for more videos then you said you were working on software for some thing... now I am getting what a marvel you have created in those times.... this is just amazing...
What happened to Shreyam?
@@iforce2d man you remember my name... damn....... i am feeling special now...
i was experimenting with "channel name" field of youtube.. it gives notification panic...
Great idea, looking forward for the rest of the build! Might copy your idea one day :)
Btw, if you want to reduce the vibrations, instead of improving the alignment of your anti-torque bar, give it more degrees of freedom! (With an extra linkage between the motor and the steel support for example)
Happy new year to you too
What a great exercise in problem solving. So many little problems solved along this DYI CNC journey. Thanks for documenting and sharing.
Your programming examples with the sbns-mixer (universal programable controller?) continues to amaze me. You making a great case for this style of visual programming.
It might be useful for some kind of sanding in the near future....
wtf. The CZcams page froze up just now and somehow that comment about sanding got posted to the wrong place.
@@iforce2d I've seen YT do similar.
Thought sanding comment was a hint to something in a upcoming video. LOL.
(no worries, all good)
Well you thought correctly, not sure if I'll actually do it though :)
Wow, Cool project and great explanation of the process in the video. This kind of software control for motors is new to me and very interesting. I’ll subscribe to learn more on the subject. Thank you for sharing!
I made an overview about this recently, would be interested to hear any questions you might have
www.iforce2d.net/tmp/swip/overview/overview220526.html
With all the machining you are doing now it might be time to invest in a lathe :)
Keep up the great work and happy New Year
you can use a tapping head to do what you wont to do. in order to get to the bottom of a blind hole you have to use two types of taps a plug tap to start the the threads and a spiral flute bottoming tap to get to the bottom of the hole.
My forcefeedback motor is a industrial servo motor. Jep just a 15k step encoder on the back makes a regular motor work with very accurate positioning. The controller sees movement smaller than me moving the wheel 1 hair width. Forget stepper motors that can lose steps, slap an encoder on the back and know motor position at all times.
I'd have been tempted to try and put the spindle clamp inside the square tubing, a bit more fiddly but looks like it would fit. Cheers
When you clearenced the internal bolts it unballanced it, if it becomes an issue for you try using car wheel weights to ballance it
Good job with this project. they make a drill chuck adapter to tap that would serve you well. Also a keyless chuck.
Nicely done I stumbled across this video and you have my subscription. I need to put some time aside to catch up to "speed" with your channel
You can also use a tap head instead of a drill chuck too! Could potentially save your taps in a tricky situation.
Knock the chuck out (it's a Morse Taper). Then refit the spigot an old 2-jaw or 4-jaw chuck instead. You can get them from old drill braces, or from a hand tapping thingy.
What can I say, everybody wants toys... your idea is awesome, nice project👍👍👍
Fantabulous 😊👍
Awesome and even pretty
Magnetic drill press is what you are thinking about. the press you have as it sits is sufficient to do everything you just need to rip out the controller for the motor and install a more advanced digital controller
Love the idea. How about adding a depth bar to the column with a micro switch on the end of it? That way you can set the depth on the bar and once the micro switch touches the work piece it can enable reverse and retract?
Keep up the great work!
you can weld an extension to the chuck key. And add a place you can store that key on the machine.
You’ll probably need a function that takes the thread pitch and desired depth, so that you can count how deep the thread needs to go before a full retract. For instance you normally want at least 2 thread turns below a through hole so that the tap doesn’t have any taper. Take an m6 tap through a 10 mm stock through hole for instance they normally have 1mm pitch, so you can count your rotations, you should aim for 12 rotations (12mm depth), then command a full retract.
Manual works too.
What would be really interesting would be a strain gauge sensor in the handles that allows you to turn them up and down to contact the tap to the work piece but when it contacts and you push against the work piece, it moves the motor ahead at the rpm you set it for and when you let of of it it stops. If you pull up on the handle while the tap is in the piece, the amount of reverse resistance would trigger the motor to run in reverse at the set rpm. Probably this approaches over engineering but interesting to think about.
You can get a mechanical attachment to do that, probably quite expensive though
czcams.com/video/3e8lfEbsLmg/video.html
Don't know if position is that important, but if it was, you can print 2,3 or more bands of stripes that can give you a binary position with resolution determined by the number of bands -basically a quadrature encoder. The strips can be read as they pass by some kind of optical detector that switches state depending on dark or light, or even better (and less influenced by ambient light), you can look for the transitions. Wouldn't be too hard to print the bands onto paper and adhere it to the rim of the motor. I suspect you can buy the led/detector in a package off the shelf and quite cheap. (I might be wrong about that).
since your controlling speed without belts why have the huge multi belt ? simly mount the hoverboard hub and put a small belt pinion on the spindle and use the tireless hub with belt to the spindle.
better yet, take hub apart and from hub front facing face you can bolt a collet directly to it that fits the spindle shaft.. this would drop the motor right ontop the shaft much lower and bracketing would be much easier.
we use this methos all the time when converting hoverboards to gocarts etc to mount larger rims.
Modify the chuck key by inserting an extension between the handle and the gear.
Love the innovative approach and great video as always.... If the motor is close loop, could you use it for motor position control.
Just out of curiosity you do know that you can get a pedestal drill tapping attachment
Yea it looks like the hacked firmware allready has rotational position values, might need to expose them. So you could get rid of the timers.
You mentioned early in the video you are tapping blind holes, using normal taps means you will have to change taps, and they also don't remove the swarf. You need to get some spiral taps, can tap to the bottom of a hole and push the chips out the top. And always use a tapping fluid or compound.
Are you aware you can purchase power tapping attachments - they plug into the arbour of a drill and automatically come back out when you release the pressure on the handle, they also have a clutch which you can adjust so you don't break a tap. They are fast. But you need the spring return on the drill quill to do the return.
With the 40 or so volts sitting across the filter capacitors one way of discharging them is to permanently fit a bleed resistor perhaps in series with an led for visual indication. About 3.3kohm or more will be adequate.
Although it would drain the battery if the battery is left in circuit in a standby state.
Nice to hear a New Zealand accent on CZcams.
Check out MartyT's channel. You will be occupied for hours.
Cool little project, try gun taps instead... they are night and day better than hand taps. Also, not so sure about the way you held your motor. The rear part of the shaft should have some float to allow for misalignment. Check my video on the automatic door I made with one of these motors for more detail. The stress from misalignment will damage bearings, fatigue welds etc.
Hole you mention is called countersink hole.
I think the drill chuck should work fine. To prevent slipping, tighten the chuck in all three holes instead of just one.
Fantastic!
I think I'd just stick with the belt drive and eliminate all the issues that come from trying to centre a weird-shaped motor against some other piece of metal, against a shaft. You can lock the motor shaft into a plate screwed onto the top of the drill mount, chuck a belt on it, and you're done.
you can use a "servo tester" to run your press instead of a transmitter/receiver
Love the comment "steel chips are hot and lube makes them stick to you" 😁
I have considered doing similar with an old drill press, but we now have a CNC mill with rigid tapping so I doubt I will do it. For me the issue was that my drill press doesn't have a reverse or low speed so I looked into using a small 3-phase motor with VFD, but your method looks good too.
You might consider using an ER collet holder instead of a drill chuck. That would allow for some improved gripping strength on the tap and would allow you to use drill bits as well vs using a dedicated tap holder. Tightening is done with a spanner which I find to be easier than a chuck key. Regarding grip strength - We have an old Bridgeport that we were able to use a 3/4" NPT pipe tap going into thick aluminum using an ER40 collet if I recall correctly. I never had the pipe tap slip even with that large OD.
Will you need to add some sort of depth feedback or torque feedback so that the motor can automatically reverse? Doing a through-hole like you did in the demo it would be fine to just control it manually, but a blind hole like you plan on doing in the future will cause the torque to go through the roof if the tap bottoms out and will be likely to break the tap I would think. Maybe sticking with the drill chuck will keep that from happening...
I'll just have to be very careful with the blind holes, especially as others have pointed out I should be using a spiral flute tap for that. Starting to think I should just make them thru-holes instead :)
So cool. Way beyond my pay grade. I'm sure you realize you could have tapped all the holes in your AU plate before you were half way through this build.... But where's the fun in that?
Happy New Year!
Yea I probably could have, I tried not to think about it :) Hopefully this will make the job more enjoyable and will be available to use again in future projects.
Happy new year 🥳
I think I would put a hinge o the base, so it can be tilted backwards for bit changes or tapping or drilling at 90°
Nice video
The allen "grub" screw with nut on it beside your quill feed, you can tighten screw just a bit and will hold 😉 any level where you put it, good luck
Using shaft of tap and not square allows tap to spin if bottoms out without breaking.
A mag drill base might be better for this,although it wouldn't work on aluminum as easily, you'd have to clamp a steel plate on top and then magnetic base goes to that.
an easier option than trying to modify or replace your drill chuck is just grind 3 flats on the shank of the tap, as long as you're careful to space them evenly and take off the same amount of material in each position. maybe you could come up with a jig to help with the grinding
If it becomes an issue I might end up having to do something like that. So far on my three test holes it has been ok, although that was only 6mm plate.
Most programmable transmitters have adjustable deadband.
nice work.. the only thing i didnt like was lowering the head to the point of inconvenience.. Theres not much axial load on the quill during power tapping , its self centering.. So you could have the drill head higher and the vise mounted to the table.( less hands involved ) . I power tap every day on the mill using a keyless chuck , which might be of some use here.. Yes , it occasionally slips , but thats similar to a clutch telling me its time to pull out , relube and make another pass. Anyway , cool project !
Right, for tapping it doesn't matter. The problem is drilling, where this press has awful runout, that gets worse as the quill extends. I tapped all the holes on the large plate yesterday and using the slipping as a cue to relube and make another pass is indeed how it ended up :)
You really need lots of torque. Not so much for Ally. Loads for steel etc. Drill chucks only just have enough grip. Even collets lack drive
If you want te have the motor free spin maby the easiest way is to put e switch that cuts off two of the motor wires. It can also serve as a emergency stop/ safety switch.
That would indeed avoid the braking. Problem is the hall encoder of the controller would detect a discrepancy between commanded movement and actual movement and enter an error state. Disconnecting the hall sensor also triggers error state. It would be better to modify the firmware to ignore any encoder discrepancy after the throttle has been in the middle position for say 3 seconds. Probably not something I'll bother with unless it does somehow turn out to be really annoying.
A physical disconnect would be great as an emergency stop though.
You can make a very nice 2X72 belt sander with that motor.
I wiuld vacuum bag the MDF with epoxy, like used in carbon fiber, to make the MDF harder and stronger and it should work, it would water proof it also,
Great video..technically well beyond my level of understanding…just starting out re-purposing these motors!! Have recently built an electric barrow for the horse yard!!! Wish I had known about Low torque mode..I have installed just a generic co troller with no facility to adjust!! My next project is to make a remote golf trolley so was amazed to see your two motors running around the yard!😂 could you point me in the right direction regarding boards and what I would need. Apologies for being non tech!!
links are in the description...
@@iforce2d Hi, did you have any idea how get more torque of that motors by changing some settings in feru foc firmware ? that hoverboard motors have too much rpm i need more torque not speed
great stuff, how would you go with a set of drill taps? only Aluminium your working with, if they work its 1 operation
I am sure I don't need to say this, but near the beginning you say you only need to hold the motor shaft from spinning. But you have built a rigid structure resulting in it being over constrained. You want either a flexible coupling between the motor and pulley, or replace the top structure with just a torque link. Now this has got me thinking. I wonder if this sort of motor would make a variable drive for my small lathe? 🤔
Actually a lathe was the first thing I daydreamed about using these for (at least in the power tools category). I watch a lot of turning videos on CZcams, but it seems like another huge rabbit-hole of learning that's a bit low on the priorities list for me to go down.
Have you thought a out using something similar to a drum barrel clamp instead of screws to attach the hub?
A gear puller will pull that shaft.
hi mate some wood turning shops sell a rubber ring that can be turned on a wood lathe to make custom jaws might be able to make coupler and dampener in one
One of the first things I noticed when I started watching BigClive, is he shorts the Caps EVERY - TIME with a little screwdriver . Everytime I see a capacitor , I see some fuck shit waiting to happen.
The cheaper drill presses usually have a terrible runout on the chuck - which is one reason they should be avoided. Clever reuse of it though :). I wonder if the hoverboard motor could do with either balancing or a radial load on the bearings (since that is what it is designed for).
Yeah this one is pretty terrible. Anything larger than a 6mm drill would cause it to bounce off the sides of the hole like crazy, eg. if there was a 8mm hole that you wanted to enlarge to 10mm, good luck with that.
@@iforce2d i have a drill press of the same design different brand. Mine also had the problem that i guess is generic to these with the wobbly spindle.
I "fixed" it by cutting down the center of the front of the machine, and installing some bolts across the cut. Essentially it turns the front of the machine into a pipe clamp of sorts. Adjusting the tension of the bolts just right can get your a wobble free spindle that still travels freely up and down
Based on those results, it looks like you could cut down the length of the spring, and get the amount of force to resist gravity just barely. Fewer coils == less torque
I've been attaching a clamp to one of the three lever handles to provide a bit of weight. It lightly holds the toolbit up when I retract, and then lightly pushes it down when the weight comes forward, which works out ok.
@@iforce2d should have just left it how it was after you ton some tension off of it There is no reason that a bit of upward tension would hurt anything. In fact, it would be better if there was some upward tension.
I wonder if that back and forth motion would be of any use for a drum sander...
It might be useful for some kind of sanding in the near future....
I am sure you could have glued the MDF rings to the hub.
I certainly could have. What type of glue would hold to smooth aluminium with the torque necessary is another question.
You could get a cheap VFD and use the orginal motor on the drillpress, no need to change. Some VFD have tourqe controll and can be controlled by bus, voltage, 4-20mA, binary inputs and so on. Probably cheaper then your current setup.
Not sure if the belt drive would handle the torque necessary. So far I've had slipping all over the place, in just about every connection.
@@iforce2d Metal gear probably better. But maybe you could add new, fresh belts in parallell? Seems like it is poissible.
Nice serie. I’m wondering what software you use in the last part of this video?
That's a system I'm still working on. I will make detailed videos sometime, until then you can see other videos where I used it czcams.com/play/PLzidsatoEzeieb0l_EUxpR0p7ed_RUPD6.html
@@iforce2d Is the software propriety thats comes with the “sbusmixer” hardware? Or is it a standard software tool? And is the “sbus mixer” for sale? Where?
The firmware will be proprietary. The configuration side in the browser is mostly Javascript so that's essentially open source, but it would be a huge amount of work to adapt it to other microcontrollers. Between the browser and the firmware is a 'relay server' third component that allows the browser to access the serial port.
I'm hoping to sell the boards someday, but the price of the MCU is extremely high right now (covid?) so nobody would want them. I'm told things will get back to normal after a while.
You may want to take a half mil off that pulley to get rid of the wobble, that may be where it's coming from
What do you mean by "closed loop mode" on that motor? My ebike direct drive motors have the problem of low torque unless rpm is fairly high.
The motor has hall sensors to detect how fast it's rotating. Closed loop means it's using that sensor feedback to adjust the power, based on a specific target, instead of just changing the power blindly. For example if you're driving down the highway, you would adjust power with the throttle pedal, and 'closed loop' would be looking at the speedometer to guide your adjustments to a target speed, eg. 60mph. 'Open loop' would be simply a target pedal position, eg. 45% throttle, with no knowledge of the resulting speed.
@@iforce2d don't these type of 3 phase motor controllers rely upon the HE sensors to determine what frequency to output?
Like if the motor is moving at 1 rpm and you tell it to go to 100%, it's only going to be able to ramp at a rate depending upon how fast the motor accelerates.
I suppose you could just feed it a set frequency of 3 phase (a predetermined speed) and ignore the sensors. However I would think the rotation would be all chunky and shit until it got up to the necessary rpm, or even some factor of it.
Maybe I'm misunderstanding your analogy to a car speedometer.
There are two types of 'frequency' at work. One is a simple duty cycle PWM that controls the overall power output, 0 to 100%. The other is a more complex phase 'switching' of which windings of the motor should get how much of that power, and which direction the current should flow. The switching depends on the HE sensors to know when to change to the next phase. The HE sensors can also be used to count how many switchings are occurring and from that infer a speed.
czcams.com/video/QgcAJpRFj1U/video.html
Hello @iiforce2d, nice job… is the sbus mixer project available anywhere? I love the interface and was looking into how you build that. Cheers and happy 2022
The main microcontroller for the sbusmixer increased in price by 16 times last year, so it will not be available until that cost goes down.
Should name it "product tap"
cool! you realize you could use only the percentage dial to do the whole tapping operation right? just go to "1" to reverse the tap out of the hole.
if i miss understood the function and is wrong i apologize. =)
i really like your projects and i still wish we were neighbors! =D
Sure, it can even be used as is with the radio control, the question is how much convenience you want. I will probably leave out a couple of the dials I showed, since they will not change much in practice and can be determined well enough by twiddling with the 'virtual' dial.
@@iforce2d i am all for ALL the dials. it was just something that i thought was cool.
Aren't you running the wrong way for tapping?
Nice project, how do you come up with these ideas?
No, the motor is upside down
@@scream221 You are dead right, I didn't notice that, I was wrong and I don't mind admitting it.
Seems overly complicated to me, I think I would have used the complete wheel and tyre as a friction drive against the main spindle. Replace the pulley wheel with a knurled aluminium drum .???? Just an idea to reduce that out of centre “ wobble”…..
those motors are BLDC, they are basically stepper motors with limited pole numbers. Also, you can just control the position with a position sensor.
With the right firmware you can do what you want, not with a hack with potentiometers...
Stepper motors are arranged so that energizing alternate coils will move a known distance, with no sensor feedback required. I wouldn't call these steppers because the method of control is quite different, and you can only roughly determine position based on the hall sensor feedback. But this is all beside the point because I wasn't really interested in controlling position, I was just showing how a rough position mode is available in the "Feru" firmware. The potentiometers I used are only inputs for adjusting speeds manually, they don't control the motors directly.
@@iforce2d you can control bldc motors position very finely by modulating winding current, which is what microstepping is.
That is how gimbal motors are controlled.
Yes, that's how the Feru firmware is doing FOC.
I'm not familiar with the software you are using at the end. I see it is local host so you are connected somehow? Thanks.
That's a system I'm still working on. I will make detailed videos sometime, until then you can see other videos where I used it czcams.com/play/PLzidsatoEzeieb0l_EUxpR0p7ed_RUPD6.html
The configuration interface runs in a browser as in the video, but it also requires a 'relay server' running in the background to allow access to the serial port.
Rap the spring back tight and heat it with a torch slowly until desired tension.
You can adjust the tension just by winding it less. The point is it will always move to a target point, rather than counteract gravity.
@@iforce2d lmao did we both watch the same video?
What's your point, am I wrong?
@@iforce2d yes… all you did is change how soon it hits the same rate of strength.
Exactly, and that's all heating it would do too. There will always be a point that it settles on, and will increase in resistance further from that point. That is not the same as counteracting gravity with a constant force at any position.
I may have missed it in the video but what is the software that was used for the flow-chart like programming/tweaking?
That's a system I'm still working on. I will make detailed videos sometime, until then you can see other videos where I used it czcams.com/play/PLzidsatoEzeieb0l_EUxpR0p7ed_RUPD6.html
i'm surprised you didnt just pull the spindle pulley off and mount the motor directly to the spindle.
You mean the steel shaft inside the aluminum pulley, that's only about 2cm in diameter? Not sure how I would fix my MDF to that and maintain alignment, and I wouldn't be so confident that the MDF could handle the torque at such a small radius.
@@iforce2d yes... but dont you have a CNC table.. couldn't you make an adapter out of some plate :D, that aside you could still cut the belt pully down a buit so it wasnt so tall.
35:13 hahah "Click. click. click click.. click" hahahah heheh
What's the node based programming language/environment you're using at the end?
That's a system I'm still working on. I will make detailed videos sometime, until then you can see other videos where I used it czcams.com/play/PLzidsatoEzeieb0l_EUxpR0p7ed_RUPD6.html
47:00 Still here! ;o)
step 1 : czcams.com/video/EckiOgHsuXg/video.html
step 2: czcams.com/video/RZBWDAueUJg/video.html
these modifications are great for ecomia and speed adjustments
What is that programming environment?
That's a system I'm still working on. I will make detailed videos sometime, until then you can see other videos where I used it czcams.com/play/PLzidsatoEzeieb0l_EUxpR0p7ed_RUPD6.html
I am sorry What programming tool are you using at the end of the video is that a Arduino tool.I am sure you have mentioned it or covered it in other videos.
That's a system I'm still working on. I will make detailed videos sometime, until then you can see other videos where I used it czcams.com/play/PLzidsatoEzeieb0l_EUxpR0p7ed_RUPD6.html
If you don't drill through... you will continually collect crap in the holes.
32:36... Just weld an extension to the Key's Shaft.
36:37... since pulses are being sent to the motor you could count those pulse to get a rough idea as to the feed and retract distance.
Yeah it's looking like my plan to make blind holes will be a pain in the butt. I'm starting to think I'll just have to drill all the way through.
The motor does not move by counting pulses, it's not a stepper motor. If it was a stepper though, yes we could do a lot of cool things.
@@iforce2d If I'm not mistaken... that is a brushless motor and the motor coils are being energized in sequences by the brushless motor ESC. Those are brushless motor pulses that could be used as a sensor/counter for direction and distance travelled by the motor shaft. Of course not as accurate as a stepper motor's pulses but i'm sure better than using time to get rotational distance travelled.
Yes you could run it as a stepper motor like that, but it would have only 15 steps for a full revolution. I suppose you could think of this a stepper motor that is doing very fine microstepping all the time, so 'pulses' are not the same as motor coils.
I think the word 'pulses' is the cause of our confusion here. The firmware could be modified to count how many magnetic poles it has turned past, and stop at some desired point. That might be what you mean. But that's essentially what the 'speed' mode does anyway. It increments a number saying how many poles should have been passed, and the motor moves to match that (which is how it returns to the starting when I try to move it by hand). That number is being increased over time at a certain rate, eg. after 1 second it might increase by 5, then two seconds it will be 10, so the outcome you're talking about is the outcome it's doing already, just by setting it to a constant speed for some duration in seconds. The hall encoders are analog and not precise, so I think the way it is now is about the best we could expect.
@@iforce2d That makes sense...
Lathe work