Printable View
Far better than mine...wow has it been that long since I last looked at this....been too busy spending money on RC helis.
Bought a small boring bar and today in lunch bored the inside of the bearing holder out. Tomorrow will face off (parting tool not big enough) the part to the correct dims. Then just need enough Bellville washers to fill a 12mm gap...that offer of some washers still on Mark? ;)
Last bit is the cap then fit it all together...will take pics later!
itll cost you some pictures of where you are up to and how it works out with those 208 AC bearingsQuote:
Originally Posted by njhussey
PM your address and the gap you have between your bearings
Attachment 6009Attachment 6010Attachment 6011Attachment 6012Here you go Mark...pics from lunchtime's machining. The Ali has been in the bottom of a box of all my CNC bits from when I moved 2 months ago so is looking a bit worse for wear. If I can be bothered I'll skim the surfaces once it's done to tidy it up!
Neil,
Looks like it's coming along!
I've had a shake up here as well, lots of my toys are still in boxes. Was laid off, found a job, moved from California to Michigan (~2300 miles!), bought a house... phew, that makes me tired all over again thinking about it!
Here are some pics of my copycat spindle design, I think I'm in the same spot of the build as you are Neil.
Attachment 6032Attachment 6033Attachment 6034Attachment 6035
Attachment 6036
I'm using a Turnigy G110 motor, two 8x22x7 angular contact bearings, 4 belleville washers, and an er11 collet with 8mm shank. I'm mostly going to use this spindle for light work: plastics, engraving, PCB cutting, etc.
I'm a little concerned the bearings will heat up, any thoughts on this?
Also, did you re-install the e-clip ring onto the new shaft?
Quote:
Originally Posted by Villacherman
Looking good, I couldn't get the holding nuts in the housing as the PCD of them is too close to the 22mm centre hole. I'm going to have a real session on it this weekend as for once I'm in Hereford and not got anything planned (although if it continues to be sunny like predicted I'll be out up the field with my RC helicopters :nevreness: ) and as June will be a write off as I'm out every weekend :thumbdown:
I think the bearings will get warm, what speed are you going to be running it at? blackburn mark or Jonathan will be the ones to ask as they use theirs! If it gets too hot I'll use the router it to cut a water cooling jacket for it :nevreness:
I'm not going to bother turning a groove for the circlip as I reckon that the grub screw (I'll drill a small recess in the shaft) will hold it as well as some locktite :)
Looknig forward to seeing pics of the finished spindle!
Unless I'm missing something obvious the collet chuck shaft isn't actually fixed axially to the angular contact bearings?Quote:
Originally Posted by Villacherman
The outer rings of the angular contact bearings are fully constrained by the housing and axial force from the Belleville washers, however neither of the inner rings are attached to the shaft, so the two angular contact bearings are effectively acting as just a standard radial bearing with zero clearance. Currently when the shaft is subjected to an axial force it would move, except for the grubscrew and bearings in the motor - so it is those standard bearings in the motor that are actually resisting the axial force which defeats the object of using angular contact bearings.
You need to somehow fix the inner ring of the angular contact bearings to the shaft - best way is probably to machine a fine thread on the shaft and use a locknut (rather like on ballscrew mounts). You then have both bearings facing the other way round and place the Belleville washers between the nut and bearings and use the nut to apply a known preload by measuring the compression distance of the springs. This is how I've done it on a ballscrew... same principle applies. The problem is the shaft will be hard to machine so someone will suggest you simply locktight the lower bearing to the shaft, but that's rather crude.
The preload required must be greater than the maximum axial cutting force and greater than the minimum load specified by the following formula:
http://www.skf.com/portal/skf/home/p...newlink=1_3_10
This avoids sliding friction within the bearing which will lead to excessive power dissipation and thus heat build-up. However equally you don't want to use more preload than necessary as clearly that will also cause the bearing to heat more. You could machine slots into the housing with a parting tool to make fins so the greater surface area increases heat dissipation. Even better, make a channel in the housing (or mount) and pump water through it.
If the fit of the bearings on the ER11 shaft aren't a press fit then there would be a problem. I've just discovered my bearings slide up and down the shaft!Quote:
Originally Posted by Jonathan
Isn't the spindle up pressing up against the inner race of the lower bearing? So wouldn't this mean that any axial load is applied to the inner race of the lower bearing and the Bellville washers would transmit the load to the upper bearing?
If the bearings were a press fit the belleville washers wouldn't do anything - you must have a sliding fit so that the springs can apply a preload.Quote:
Originally Posted by njhussey
Yes for an 'upwards' axial force it's constrained - but not downwards. If it wasn't for the normal bearings and grubscrew the shaft could just fall out.Quote:
Originally Posted by njhussey
Thanks for the replies...
How about if I add a shaft coupling to the stop with a couple of belleville washers? Then using the collet head pressed up against the lower bearing inner race and the shaft coupling, you would get your axial constraint.
I would also load the angular contact bearing in the same orientation (both loaded with "upward" force). This would make the bearings strong in plunge operations, but weaker in operations where the bit is moving up out of the work piece. Would this be a problem though? Most operations are pressing into (down) the work piece.
Attachment 6037
Ok, yes I can see that.Quote:
Originally Posted by Jonathan
How does the motor work with a prop on it in a plane then? There is a circlip to stop the shaft pulling out if the motor is used as a pusher or reverse mounted but nothing to stop it being pulled out apart from the bearings and the grub screw as you say. This seems to be enough for a model so I'd have thought it would have been enough?Quote:
Originally Posted by Jonathan
Maybe if I machine 2 steel collars 8mm ID and 10mm OD and secure them with grub screws (locktited in of course!!) one above the top bearing and one below the bottom then that would stop the movement? Would the grub screws be strong enough?
I'll do a drawing later as on the mobile...
[QUOTE=njhussey;30976
How does the motor work with a prop on it in a plane then? There is a circlip to stop the shaft pulling out if the motor is used as a pusher or reverse mounted but nothing to stop it being pulled out apart from the bearings and the grub screw as you say. This seems to be enough for a model so I'd have thought it would have been enough?[/QUOTE]
On a plane the magnitude axial force (thrust) from the motor is much smaller than in our application, plus accuracy (end-float, radial clearance etc) is pretty much irrelevant for just spinning a propeller. The bearings will wear out much faster than if they used angular contact beraings, but the cost isn't justified. If sufficient axial force is applied to a deep groove bearing the balls will ride up the sides of the rings outside of the superfinished races on to the rougher (relatively speaking) area of the ring, causing increased wear.
I'll reply to the rest when you've done a drawing...
i personaly wouldnt get to bogged down in all the rocket science... make sure you get some tension in your balls and make sure your shaft cant easily rotate in the bearing bore (loctight the nose bearing if its a tad slack) and all should be fine
any slack will result in balls skidding or a sypathetic rolling rotation/vibration between your shaft and your bearing bores
(a very short life span for either and poor finishes)
iv used super cheap skate bearing on my er11 with aprox 2kg of tension between the two and its still going strong
if you get any high pitched squeals/squeaks etc when the spindle runs up or down you need to fettle
the beauty of these spindle is they are cheap n simple
iv seen threads on diy spindles that are works of art and that have followed the theory to the letter, however, i get good finishes in acetal and alli so seen no point in getting all posh if the aim of the game is to get yourself up and cutting :)
PS: your bellville washers are bagged n ready neil, i just need to get off my arse and post them :(
on most occasions your tool would be pulling down as you are cutting but as i have said above, get some tension in those balls and and loctight your nose bearing and you should be fineQuote:
Originally Posted by Villacherman
bearing heat may be an issue if your running above 10000rpm (keep an eye on the temp though as it may take a while for your bearings to settle in) if your cutting plastics and pcb's and your not in a crazy arsed rush then 10000rpm is a tad fast... i cut acetal at between 2000rpm and 4000rpm (6mm 3mm tooling) im sure i read somewhere that anything up to 110 degree C and your ok .... mine dont get anything like that hot
the small engraving spindle i made gets pretty hot when i run it above 30000rpm so i tend to run it closer to 20000rpm , i dont run that one very often so its hard to say how long the bearings will last?
Well I've had a look at the design and tewaked it a bit. Will probably do a MK2 spindle using this method or I might tweak this one I've already machined. this one is only really going to be cutting up to 6mm balsa and 3mm ply so I can't imagine the forces as being huge. I might stretch to some delrin/acetal and even try some Ali for a laugh :smile:Quote:
I'll reply to the rest when you've done a drawing...
Attachment 6038
This shows the top collet (but I haven't bothered showing the grub screws) on the 8mm shaft which is pre-loading theinner races against the bottom collet. I'm guessing I could get away with not securing the bottom collet (locktite) as it will be trapped between the bottom bearing inner housing and the ER16 collet.
Attachment 6039
This shows the assembled unit (minus the bolts but we all know what they look like!!)
Attachment 6040
The bearings in the above are not spaced the correct distance, they will be a further 10mm apart so I can get some Bellville washers in between.
This design has the inner races locked to the 8mm shaft and the outer races are constrained by the endcaps...............comments?
Cheers Mark, no huge rush as I've still got lots to do on the router (everything apart from the base!!) which I'm hoping to break the back of over the weekend.Quote:
Originally Posted by blackburn mark
Neil,
I like the design. Is there enough wall thickness of the collar (8mm ID X 10mm OD) to accommodate a small screw?
What about a collar that has a thicker wall and a little standoff, like this... gives you more meat to tap into.
Attachment 6047
This could also allow for a more tortured path on the bottom collar to keep dust out.
My last comment is to not forget to think about how you will assemble this... go through the steps in your mind, sometimes I design something and find out I can't assemble it because there are inaccessible features! :(
Adam
I just figured out grub screws are 4-40 set screws... I'm learning everyday!
Neil, was the shaft of the motor hard to swap out for the collet shank? Anything I should not do? I want to give it a try this weekend. I'll do all the usual, keep things straight and support, remove the circlip, etc...
I'd be using 2 off M2 x 3 grub screws and locktite in each collar (will prob make it 3mm thick for extra purchase of the grub screws) as per the attached photo.Attachment 6048Quote:
Originally Posted by Villacherman
I removed the circlip and pulled the bell off the motor complete with the shaft. Then put the bell on the vice so the shaft was between the jaws but the bell resting on top (obviously with a cloth on the jaws to stop marks) and using a 6mm piece of steel I had lying around gently tapped the shaft out of the bell (obviously remove the retaining grub screw first!!) slid out fine!Quote:
Neil, was the shaft of the motor hard to swap out for the collet shank? Anything I should not do? I want to give it a try this weekend. I'll do all the usual, keep things straight and support, remove the circlip, etc...
I've not put the collet shank in the housing properly yet (although I did fit it slightly on to see how it spum :smug:) as I'm waiting to make the bearing housing first. Been far too busy at work this week to do any machining in my lunch :grumpy: but hoping to get it done next week. This weekend I'll be making my router gantry sides and trying to get that sorted although I keep changing my mind on how to do it!!
Hi blackburn mark,
Looks very interesting concept! :-)
I just have 1 question so, How do you get the original shaft out of the motor?
Is it fitted with an interference fit to the bearings?
Ta,
RNR
Rnr107,
I just removed the shaft yesterday on my Turnigy G110 outrunner motor, so this is specific to my motor, others will vary...
1) remove the 4 M3 screws from the propeller mount.
2) remove the e-clip (circlip, retaining clip, however you want to call it) from the side of the motor with the shaft sticking out.
3) pull the shaft out by hand (I've heard you may need more force on other motors)
That's it! I really like the G110 for its simplicity in this manner, though it is a little pricier.
Adam
Hi rnr...most are sliding fits but as with all Chinese mass produced manufacturing will vary from motor to motor...mine is a medium sliding fit :-)Quote:
Originally Posted by rnr107
Thanks Adam , Neil! I think I will give it a go, It looks very interesting! :-)
Laurent
I want to change out my wood router for a brushless spindle as I’m having to replace the brushes all too often.
What you guys have done here is very interesting but I’m not sure how to size the motor / what motor to get.
Therefore I have a few questions:
Q1, if the RPM/v is rated at 295 does this mean for every volt it will increase by 295 revs, i.e. for 20v the speed would be 5,900 revs?
Q2, would you read the power output in the same context as you would a wood router, i.e. if both were 1000w would both be deemed to have the same output?
Q3, Any recommendations for motor to cut wood (Oak) ? At the moment I do a lot of 3D profile work thus cut at around 2.5m per min.
Q4, should I give any consideration to the speed control other than it meets the spec of the motor, ie is there a rule of thumb for the control to have say a 10% high power rating?
Many thanks in advance Paul
1) Yes that's right.
2) Hard to say - the ratings for the brushless motors are a bit optimistic since they assume excellent cooling from the very high airflow present in a model plane. What power rating the motor is will be dictated by the speed rating since 'hobby' brushless motors with a similar kv specification tend to have similar power ratings. That's unless you introduce a belt into the system...or couple multiple motors on one shaft, but that's probably taking it too far for machining wood.
3) Clearly depends on the tool diameter, but I'm guessing you'll want about 12000rpm for oak? So look for a motor with around that speed rating. Ideally you don't want to have to use 'high voltage' (i.e. above 25V) ESCs as they're very expensive. So roughly 12000rpm/25v=480kv, or more if run from a lower voltage...
http://www.hobbyking.com/hobbyking/s...ner_Motor.html
Your next problem is the shaft diameter - the one above is only 6mm which makes it a bit useless unless you have a lathe to modify/make a collet chuck...
5) Easier to advise once the motor is chosen, but it's sensible to get a bigger one than rated to be sure. They're pretty cheap from the site I linked to above.
Jonathon, thanks for the speedy reply.
I’ve just order from Hongkong the 100a motor control plus a drive module so that’s dealt with. I’m not too concerned about the shaft / collet holder / bearings as I will hand that off to my brother to sort (although he does not know it yet!) and he’s tooled up for the job.
As to the power supply, seems your suggesting a 24v, how about the amps? I’m not sure how the PWM supply to the motor is translated for the source.
Regards Paul
The motor linked to above is rated to 70A which should match OK with the 100A SC. In theory this would give 1920watts (24v x 70A) which sounds enough to machine wood. The actual amps pulled (for a given motor) will depend on the cutting load, but I'm not sure how easy this is to calculate. If you pull too many amps during the cut the SC will probably pop.
For this reason myself and other model aircraft users often use an ammeter to check the current draw on the ground for a given prop to make sure all is well before the first flight. If the SC gives up in the air, since it sometime also powers the reciever you loose the plane.
You could do the same thing for the first cuts to check your safety margin.
Yep power output would easily do the job. But, given the supply will be PWM then this shall have a resultant affect on the current drawn, dependant on speed required / length of pulse. What I hoping for is the input from those members who are now using these brushless motors to cut wood to help by letting us know from their experience, motors used – power supplies etc. I will poach with pride!! I’m happy to be told information resulting from experience as it does not always relate to theory...
Regards Paul
Power supply is tricky (unless you have a large budget) since 70 amps is a pretty large current:
http://www.ebay.co.uk/itm/HOT-Univer...ht_2890wt_1037
£40 for 20A .. not great, but there's another way. If you place a battery which can deliver the peak load in parallel with the ESC&PSU then the PSU only needs to supply the average current. This is where it's easiest to just measure how much current the motor draws when cutting and get the PSU to match. A standard multimeter wont measure 70 amps since they tend to be fused at 10-20A. Something like this is excellent as you can measure the current and it records the peak current:
http://www.hobbyking.com/hobbyking/s...Analyzer_.html
Or the cheap way is to use a multimeter to measure voltage across shunt resistor, which can just be a short length of wire of known resistance. I used one of the following batteries, with a 0-50V, 0-20A lab PSU in parallel on mine:
http://www.hobbyking.com/hobbyking/s..._Rex_500_.html
That battery is rated for 60A constant, or 90A for 10 seconds which is plenty. Since my PSU has variable voltage output and current limiting I can set it to just below the fully charged voltage of the battery (i.e. <22.2*4.2) and it will never overcharge the battery, which is very important since overcharging a lipo battery is a good way to start a fire. So whatever PSU you get if you decide to use in combination with a battery will need to have good stable voltage regulation below 25.2V. Two computer ATX PSUs in series is nice and cheap and will give very close to 24V, however it is CRITICAL to ensure they are electrically isolated before putting them in series. The battery would never fully charge, but that's a good thing if anything as it increases it's life expectancy. You could get the battery first, use it to power the motor to measure the current then buy a PSU to suit.
We can estimate what power you require; what cutter diameter and depth of cut do you generally use (at 2.5m/min presumably)?
For example:
12mm cutter, 2.5m/min and 60% stepover at 6mm depth of cut is 108cm^3/min material removal rate (12*0.6*2.5*6=108). That's about 740W required...and a pretty rigid machine to sustain that. So you'd ideally want a PSU rated for a bit more than that to be safe. Also it's a good idea to put better capacitors on the ESC, especially with such long wires, to reduce the impedance...
Hopefully it'll turn out the motor draws a lot less current and you can just use a standard PSU.
Jonathon this is really good stuff, please keep it coming.
So if I’ve understood this correctly, you are running a 70A motor supplied by a 20A power supply supported by a battery for peak current demands? What voltage do you have the supply set at?
I trust you are controlling the motor via a speed controller, I’ve ordered this one today “100A Brushless Motor Speed Controller” but just realised it has an input voltage of 22.5, which may now give me an issue.
Speed controller purcahsed from ebay: http://www.ebay.co.uk/itm/1606311667...84.m1497.l2649
I don’t understand your translation from the material removal rate of 108cm^3/min to the 740w motor requirement. Can you please explain or point to a source of information. Frankly I’m surprised a 740w motor will cut that much, I actually cut with a 6mm cutter and no deeper than 2mm with a 500w router (anything else frightens to much!!), so what would the motor requirement be for 18cm^3/min (6*0.6*2.5*2=18).
Thanks paul
I'm actually running a 120A, 130kV, 6.5kW rated (yeah right) motor from a 100A ESC, currently on my micro lathe but I have used it on the router - somewhere in this thread. This motor is good up to 50V, so I should use a HV ESC to get the full speed but I've not got round to buying one. The rigidity of the micro CNC lathe limits the power draw somewhat as it's not rigid enough, so it will only draw a few hundred watts. If I use the motor at a low speed with a big drill it could be more. I leave the PSU at about 24.5V, but it's really not critical so long as it's not above 4.2 volts per cell (i.e. 6*4.2=25.2V). It's nice to have the lab PSU, but it's not worth buying one just to power a spindle. Down to about 23V would be fine as far as the battery's concerned.
The listing for the ESC states 22.2V - this is because they expect people to use Lipo batteries, which are nominally 3.7 volts per cell so a 6 cell battery is 22.2V. However when fully charged the same battery will be 4.2*6=25.2V, so the ESC will withstand 4.2V. Since it's a cheap one it would be prudent to run a little below that.
I read somewhere that hardwoods take about 7W/cc/min to cut, so 108*7=756W. That makes 18cc/min about 18*7=126W. Bear in mind since this is motor output power we need to include the efficiency of the motor.
The motor I linked to was based on 12000rpm, but with a 6mm cutter you could be using almost twice that? I've cut some hardwood, not sure what it was, with a 6mm single flute carbide cutter at 6m/min and 6mm depth of cut and 20000rpm. However if you're machine can only manage 2.5m/min then a lower spindle speed is fine, but perhaps aim for 18000rpm? Although having said that the motor is most efficient when at maximum speed.
Anyway, if it's 126W and even if we double it to make sure, at 250W then you've got plenty of options for the PSU. At that power the battery is hardly necessary, but it would make a very robust system if you did have one. I'd mount the battery and ESC on the Z-axis as close as you can to the motor to keep all the wires short, then have long wires from the PSU to the motor. This will make the ESC run more efficiently, i.e. cooler and in extreme cases can make the difference between it surviving and not.
Jonathan, have you ever had any problems with cell balance using a LiPo?
I'm just wondering how actually using a multicell pack without balancing it works out longer term, as my current project needs a 2cell pack, and I'm trying to decide if I really need to include a balancer circuit.
Not had a problem, except when the battery is already on it's way out. With a 2 cell battery I doubt you need to bother with balancing.Quote:
Originally Posted by m_c
It's one of these things that gets mentioned alot, but not sure how much difference it's going to make in practice. I'm probably going to at least add the circuitry to the PCB as I've spare ADC pins on the uC anyway, and might built a couple prototypes with and without the extra components just to see what happens. I'm aiming for longetivity, so it may be a case of every little bit helps!Quote:
Originally Posted by Jonathan
Thanks Jonathan for this very helpful information. I hadn't realized that it was possible to connect ATX PSUs together in series to increase voltage -- just assumed they were regulated in a way that would prevent this. Makes sense really, when I think about it. Battery cells have nominal voltages and can be series connected into a battery of higher voltage. Somehow I got hung up on the regulated part of it.Quote:
That battery is rated for 60A constant, or 90A for 10 seconds which is plenty. Since my PSU has variable voltage output and current limiting I can set it to just below the fully charged voltage of the battery (i.e. <22.2*4.2) and it will never overcharge the battery, which is very important since overcharging a lipo battery is a good way to start a fire. So whatever PSU you get if you decide to use in combination with a battery will need to have good stable voltage regulation below 25.2V. Two computer ATX PSUs in series is nice and cheap and will give very close to 24V, however it is CRITICAL to ensure they are electrically isolated before putting them in series. The battery would never fully charge, but that's a good thing if anything as it increases it's life expectancy. You could get the battery first, use it to power the motor to measure the current then buy a PSU to suit.
Anyway, I've gone ahead and ordered two ATX supplies with 12V 30A outputs, and looking forward to making a 24V supply with them for a spindle PS. I will isolate the negative buss from the case, and I know about setting up the switch line and adding a starting load resistor on the 5V lines to keep the supplies running.
I build and fly R/C planes, so I'm familiar with the rest of the motor side components. I have two 5055 1500W rated motors -- one 400 kV and one 540 (I believe) -- both have 8mm shaft dia. I have an 8mm ER-11 collet spindle, and collets.
ps. I also wanted to post a possible correction to the above -- I think the bolded text above should read "(i.e. <6*4.2)"
Thanks for your PSU info!
(deleted and prior post corrected)
Made some progress on this after concentrating on helis for the last goodness knows how long. The amount I've spent on helis recently I probably could have built a reallt top notch router by now.....oh well still having fun with the helis so not too worried :D
Anyway....below is the progress I've made with the spindle housing, will dig out the bits from the box in the cupboard tonight and see if I can put it together tonight (minus the bellville washers as I've forgotten to order them!!)
Attachment 7095Attachment 7096Attachment 7097Attachment 7098
Here it is roughly assembled....I know the shaft is sticking down by 12mm ish as I've not put it all the way in the motor bell, see last pic! I'm waiting on getting some Bellville washers then I can finish it off :)
Attachment 7102
Attachment 7103
Attachment 7104
Brilliant work :)
What is your PSU?
i could have sworn i sent you someQuote:
I'm waiting on getting some Bellville washers then I can finish it off :)
I can't take any credit as I followed what blackburnmark did, ok I adapted it a bit but the principle was the same :)Quote:
Originally Posted by Tenson
I'm going to be using one of the cheap 40A 24V psu's from eBay...