DIY 40 tapper spindle For XQ6232A CNC converted knee mill

[epis]

Ok I made first spindle test run at 1500 rpm, I couldn't go higher because VFD is programmed for max speed 1500 rpm, will change that to 120Hz -> 3600 rpm when will install spindle on mill.

spindle was running smoothly with little vibration, barely detectable by hand, when touching spindle housing. so i guess it is success, at least vibrations are noticeably lower than swivel heads vibration at 1500 rpm. I could make some balancing, insert key way in pulley that could reduce some of vibration.

First I removed 2.2 kw motor from mill, and it had some pinion gear sprocket, and it didn't come off, so I made this thing to remove that gear without hammering and damaging shaft.


Then I made motor support pads, welded some scrap steel square bars, grinded, leveled, and drilled holes, I got lucky with hole drilling by hand and made it fit with first time, with belt tensioned, it is very hard to make something precise with hand drill, and I will need to make another support hole, then motor will be fixed with 3 bolts out of 4.


and this is how my DIY spindle looks from front and side. can compare it with milling machine swivel head.

[epis]

Installed spindle on milling machine and made first 3600 rpm -> 120Hz motor test run.


few pictures how it looks. I installed and aligned spindle housing with precision 0.01mm on 2 sides, actually I needed alignment on X axis side of rotation, where spindle box rotates, I made spindle box holes 1mm wider in direction of rotation => +0.5mm on each direction for this kind of rotary alignment, and all bolt holes fitted perfectly, had little problem of placing screws in that tight narrow screw cut out.
On Y axis rotation I didn't needed any adjustment it fitted perfectly, I guess I milled spindle box and spindle housing mounting surfaces with close enough tolerance.
only few steps are left, like turning taper surface, and making new drawbar screw with 15.8mm diameter road, I tried to hammer 16mm drawbar screw and it goes in, with hammering, but length is insufficient, and I need screw that goes freely, so i will make new 15.8mm screw.

[Azalin]

Sounds very nice.

[epis]

This weeks update.
I started to turn 40 taper cone, first I tried to make taper without any surface treatment, and I Failed, I got very poor surface finish here is picture. one of the reasons for poor surface finish was very chips strings, chips didn't brake and basically scratched surface while turning it.

then I decided to do welding heat treatment, I burned three 6003 diameter 3mm electrodes, and here is how it looked.

then I tried to make tapper cone and I did it.


I made video of how i cut first time and how it is cutting cone after heat treatment, the difference is very noticeable, boring bar was ringing, at lower cutting speed 100m/min, compared to first cut 250m/min, so surface was much harder. and chips didn't make strings, just small flakes.
then I made drawbar


and placed toolholder BT40 for the first time, and measured run out, and it was very poor 0.03mm, I noticed that toolholder marked few lines on holder, so I guess it wasn't connecting to all surface, I don't have this color contact marker, but I thought I should check how spindle cuts, and cut some aluminium, then I made drawbar for ISO40 holder, and runout was even worse, I guess it is because of drawbar misalignment 0.25mm, I made wider shaft drawbar hole from 16mm to 17.5mm 100 mm deep, and tried to make few extra passes to make surface more shiny remove thous waves, and made mistake in code, I removed 0.75mm from taper cone, by erasing first starting point coordinate y22.25, instead of that machine started at Y23 (rapid move distance) and did cone wider.
so I had to fix that, I bought surface hardening electrodes, and decided to try out this method,
thous electrodes give 59-60HRC surface, and they really do that, + surface is corrosion resistant , it has 8% chrome
here is pictures of weld.







I did not finished welding, I will make one more pass, but i can say now that this is the right solution, I should did it at first time, turning that weldment was very hard on machine, it chattered crazy, chips was sparkling in bright color, like in grinding machine and WNMG0804 insert edges burned out very fast, 2 passes and edge is gone, completely burned away, one pass was 0.5mm, so I started to use scrap inserts, with already damaged, burned edge, and I will also make special boring bar tapered toolholder, because I run boring bar just 40mm deep, to get max rigidity, less vibration, and I am glad I didn't put it 69mm out, like in the First and second taper turning time,

I made welding without removing shaft, and everything is fine, because I cooled shaft after I burned one electrode to 30-40 Celsius, with water emulsion, it took time to cool steel down but I saved lot of time by not disassembling spindle. also you can see that my 2 peace shaft made opening in middle, heat was so intense that it expanded diameter of outer ring, and when shaft cooled it left free space, i welded it up.
making this cone proved to be harder than I thought at the begging, on lathe it was so easy to make it, because i had coolant, and precision, but knee mill is not so precise,

[zmajmr]

interesting

[epis]

video update.
I combined 2 videos in one and cut out peaces of second video because it has out of focus and it is too long, speed was 1000 rpm and feed 0.05mm/rev so it was very slow process, and i couldn't increase speed because of chatter, you can hear high pitch sound, first video I coded 3 speeds of spindle to be close to 250m/min cutting speed.
so this is how it looks like.

[hanermo]

Great stuff !!!

[epis]

Update time.
did lot of experimentation with turning 60hrc harde surface I welded Up and here are best video I made,


and this is my grinding video, the grinding stone has no shape


and now few pictures.
first I want to show how is looking indexable boring bar WNMG0804 insert edge after one pass


and edge is completely melted away, and each time I made pass I rotated insert to new edge, and I used up few old worn out inserts, that weren't suitable for precision turning, and when I got rid of most of the metal, got to the finishing I used new fresh edge each finish pass, I did some 8 passes, each time new fresh edge, till I got pass so lite that edge held up and didn't wear. but still I couldn't get tapper cone the right surface finish, shape. I got runout 0.03mm on 30mm distance of BT40 holder and that was too big.
few pictures of inside taper looked like


Then I was once again disappointed in result and decided that I need grinding. bought grinding wheel, fitted it up on my 11000 rpm 110v DC 600w small spindle motor, i ran spindle on 80V DC so it goes to ~8000 rpm, here is picture


that is my spindle motor power supply
I don't know how to grind amd I think I made all mistakes I can, I started grinding with wheel right away, and soon noticed that it is very slow process and it removes so little metal, also I saw that wheel itself after one, two passes gets filled with metal dust, and becomes soft, I tried to make tapper cone shape to use more grinding surface in one pass, and reduce spindle RPM and feed to 10mm/min on 200rpm. and I ended up with wheel like this

I think you call it grinding wheel dressing, I did that with carbide tipped lathe cut off blade, before I tried to shape wheel with regular steel, by grinding it off, but it eroded wheel slowly, and i got wheel that looked like it is already filled up with dust, by using carbide I have clean stone, without any metal dust, ready to be used.
i think I will need to dress the wheel after few passes.
grinding gives me best surface finish i had so far, but I need to correct taper geometry, it is like oval, I think I made it with lathe tool, it is result of vibration, and it is very small 0.01-0.02mm, I think I am on the right path here. grinding takes very long and I have a good feeling that I will make it right this time, that wheel dressing was last thing I did last week, I made one light pass 0.003mm and it barely t ached surface in few spots, next week i hope to clear up all bumps waves, on taper face, to make that fit.

[epis]

Update.
I did lot of grinding last week, two days I tried grinding with that grinding stone in picture, and I got nowhere, it was too slow, then I tried different grinding stone white one it is D24 mm and it worked much better, grinded faster, and I got thous grinding sparks, and I grinded out all big bumps, and got to the micron level I made lot of 0.002mm passes and still didn't had perfect fit, there was spots that I couldn't clear, and One reason was i dressed wheel with hand , and carbide lathe tool, So at Thursday I tried out PCD tipped lathe tool for wheel dressing and that was a game changer, but the problem was that I couldn't get right geometry by hand, so on Friday I made boring bar attachment to spindle to hold PCD tool and made first CNC grinding wheel dressing. I got the right geometry of the wheel, and got much better results with new dressed wheel, but still not good enough.


At the end I failed to grind taper cone, i tried all I could and my conclusion is that my milling machines repeatability precision is not in few micron range, it is in 0.01mm and actual precision is even worse 0.02mm but I need at least 0.002mm repeatability, and the cause is cast iron table sticking, also i have C7 ball screws, I have linear scale with 5micron display resolution and I can see how numbers jump when i step jog with 1 micron increment table can stick and jump 20 microns at one instant on X axis, Y axis has more precision, and less sticking, but not good enough. So only thing I can do is to grind outer spindle flange, or turn with boring bar, to get it straight because after taper welding flange has 0.05mm elevations from thermal expansion. I will clear it up. and disassemble spindle, and put shaft in lathe, center it using flange as measuring surface and grind spindle on lathe.
I hope my lathe can do it, because i already bored perfect fit taper cone at the beginning when I made shaft, but after installation in spindle it was out of center by >0.2mm, and now i now where the thru spindle center is when all is assembled and I will try my best to get that precision. i want at least 5 micron runout.

[NeoTech]

Oki gotta ask.. why didnt you just try to hardened traditional way, a surface hardening with a blowtorch and oil quench would prob. have solved alot of your taper turning problems.?? You don't have a torch? Didn't know if the steel would take a surface harden?

[epis]

Update Time.
after taking some time off I renewed my work on spindle.
as I had failure in my previous taper grinding attempts I wanted to find a cause for such high inaccuracy. so I triede to grind spindle outer flange and get perfect run out free surface First I triede to turn surface by my boring bar with WNMG 0804 insert, and I got very big runout 0.04-0.05mm so that failed.

then I installed my grinder and tried once more and I got better surface finish but still the same run out.

then I decided to find cause of such a big outrun and I disassembled the spindle, and found out that shaft bearings was without any prelode, because inner shaft bearing spacer was too big preventing, any prelode, at the beggining when I assembled spindle I got feeling that bearings are preloded, but now all was loose, and my Idea of preloding bearings in assembled spindle also failed I made hole in spindle housing too far away, I will fix that in future.

this is "hammer" I used to press out and press in shaft. that is weary big shaft.

So I turned 0.5mm of the inner bearing spacer and and one more time assembled spindle. at the beginning when I was drawing inner bearing spacer I didn't know how far ,distance, bearings will go to make preload, and it turned out to be lot more than 0.05mm I was thinking of.
I upgraded my grinding stone and bought bigger one 50mm and made PCD tipped boring bar attachment for wheel dressing. to make perfectly flat grinding wheel, and got very good results by grinding, also I changed grinding spindle by old sherline spindle with original motor, I made motor attachment to put motor in opposite direction to fit that spindle with motor on my mill,

By grinding I got now best precision I ever had 12-14 micron runout, but that's still not enough. so where is problem ? why I can't get close to 0 micron runout. and I decided once again to search for answers by disassembling spindle.

and there is is spindle shaft after disassembling and there are friction marks on shaft where bearings sit, I measured bearings ways and results showed me where the problem was.
for big bearing 6012 I got 59.96-59.98 so the bearing way on shaft has runout itself. and on 7210 angular contact bearings that is next in row I got 49.97-49.983 and I think this is where I get constant runout on spindle in 12-14 micron range.
rest was not better. So i found the problem and last weed decided to try one more time and Fix the shaft.
started to rework shaft by welding it.


Then I decided to Upgrade my grinding wheel to 100mm size and needed to make New spindle adapter to accept 20mm ID hole grinding stone and give clearance for dressing inner part, I can place grinding wheel 30mm thick, I bought 20mm thick grinding wheel.


Installed grinder in lathe and I didn't know how to dress the wheel, and I placed boring bar in chuck and rotated it and did wheel dressing, disadvantage is very slow speed of dressing, but it can be done.
And finally I grinded shaft, I made tolerance 60.000-60.003mm for 6012 bering and for 7210 50.002-50.004 mm so it is tight Press fit, and zero play on shaft, roundness is also close to perfect, and I think I finally made it right.

[epis]

Here is video of Spindle shaft grinding.


I hope now I will be able to grind that tapper after assembling spindle and placing it on mill.

[NeoTech]

No to be rude or anything but i'm amazed how ****ty this looks and at the same time you really get **** done.. It is so contradictory but fun at the same time. =)

[epis]

Update Time.

I made some changes to shaft design, I placed spacer between 6012 bearing and 7210 inner race 8mm long ring with ID 56mm and OD 68mm to put that spacer I turned shaft 8.5mm out off big bearing seat place, so large bearing has 2 spacers lower and upper ring that pushes against smaller 7210 bearing, I needed this to ease, not damage bearings in assembly, and disassembly process, by pressing bearing, hammering on them, I will show my custom made press fit fixtures i made to assemble shaft later.


this is my shaft bearing pressing tool before welding, I used inner 7210 bearing spacer, tube, which I removed from shaft and now it doesn't have that, and as I remember I placed that ring to make more shaft inertia, so I used now it to make bearing press tool.

next I milled larger spindle housing bearing prelode adjustment hole, before Hole was too small to make any adjustments, and now you can see that it is possible to adjust lock nut bearing prelode.

and then after i assembled spindle I put it on mill, and grinded 4 days tapper, it had very large runout, after last welding rework, and this is result

this is how spindle look assembled with 50mm facemill that is most precise tool I can place it has 0.02-0.015 mm runout, others have 0.04-0.06mm on end depending on holder tool length, but as I remember face mill holder had its own runout when i bought it and it just sits in such position that it cancels out spindle runout.


So this is highest accuracy I can get by grinding with this grinding spindle setup, and it is still not enough rigid and at 3000 rpm speed is too small, for such a small grinding stone, so grinding was very slow, long process.
I made already some milling with different tools and it did quite good, by the sound of milling I can say that it is quieter cutting now compared to old gearbox milling head, most aggressive cut I made was with 50mm facemill 8mm dep 800 rpm feed 160mm and width of cut about 6-8mm I cleared edge of weldment, it was running good, I will make in future some tests, on normal steel and see if I can get those deep cuts with low cutting width that have those big power, torque ripple curves but with average power in 2.2kw range, to see if my spindle have enough inertia to make such cuts. old gearbox spindle had plenty of inertia for big deep cuts, and gear reduction also was available to increase torque.
I will post some videos later.

I still don't have that special paint, or how you call that paste to check for contact, clearance on surfaces in taper cone, I use marker but that gives me very limited information about contact surface, if someone knows what type of paste, is available in common household stores, or auto shops, can you please post it here, I already asked about that in few places where I bought grinding wheel, and bearings and they don't know, but as every one else they now that such paste exists, but can't figure it out either.

[nikn]

I still don't have that special paint, or how you call that paste to check for contact, clearance on surfaces in taper cone, I use marker but that gives me very limited information about contact surface, if someone knows what type of paste, is available in common household stores, or auto shops, can you please post it here, I already asked about that in few places where I bought grinding wheel, and bearings and they don't know, but as every one else they now that such paste exists, but can't figure it out either.

I think you're looking for "Dykem Hi-Spot Blue". It cost about 7$ for a tube of it here. Where are you located?

By the way, interesting and awesome thread.

[epis]

Thanks nikn
I ordered " Dykem 75182 Marking & Layout Ink Paste, Tube, 21g, Blue" I just need to wait month they don't have it in local warehouse.
Here is my First steel cutting test video I filmed this week, and edited it today, added subtitles of cutting data.


I did lot of stuff this week, I damaged bearing on 2.2 kw motor earlier by switching it from this spindle setup to old, and I replaced bearing, then I did spindle pulley wheel balancing by inserting key way, an I have slight vibration at 4500 rpm, but till 4000 rpm no noticeable vibration.
and i can run spindle now to 4000 rpm without problem.
run out I measured on endmill was 0.03mm and on facemill it was 0.02mm on shaft but actual indexable inserts has runout about 0.08-0.1mm it is very imprecise when I bought it, and it is very hard to measure,
I will post more stuff later.

[Azalin]

Looks like you made it.

[epis]

Azalin

Looks like you made it.

Yes but not quite as I wanted, I wanted runout of tool to be in 0.01mm range but I have 0.03-0.06mm depends on tool holder. so I will try and regrind taper in near future, when I will get my engineers blue paste, and when I will make my new Grinding spindle on which I started to work, today
I will use Straight Shank Collet Chuck Mini Type for ER20 Collets 20mm Dia 200mm Length I already ordered, and will press 2x 6204 bearings on shank put it all in spindle housing, place in front labyrinth seal, add HTD 5M tooth belt 9mm wide to get 10000 RPM
Today I made grinder spindle housing,


and labyrinth seal
that also will be bearing End Cap, now it is press fit but later I will drill 4 holes for M5 metric screws.


and Spindle housing with End cup


i am making this new grinder spindle because old setup wasn't rigid enough and had just 3000 rpm. it was better than first direct motor grinding setup with ER11 collet, with this grinding spindle based on ER20 collet Check shank I hope to get much more rigidity, less vibrations, and higher speed using old 6000 rpm DC 0.75 HP motor.
and i could also use this new ER20 spindle for milling with small diameter end mills, I will try that too sometime later, so this will be for grinding and maybe high RPM light milling.
I will post more info later.
For spindle housing I used stock OD 60mm ID31mm that was left from my 40 taper spindle shaft first design, When I wanted to make shaft from 2 tubes, I bought. I didn't wanted to drill that big hole, but changed mind and bought solid bar, so I use this thick wall tube to make housing, it was very easy and fast just few hours of machining (most of the time I did setup) i made bearing pockets with very good light press fit tolerance 46.98mm, i am glad that I didn't made any mistakes this time, I guess I learned my lesson in last build, actually I learned lot on new things, by making this project, and learning was one of the reasons i am doing this, i needed more experience in high precision machining field, and I got some.

[epis]

a few pictures of Tools I made for spindle bearing assembly that helped greatly
First I use this tool to press all 3 bearings on shaft

then I use this tool to press shaft and bearings in bearing housing, it is made out of square 22mm stock bar, and 10mm thick plate I had from different project left, but hole was too big so I welded up 4 peaces of steel and turned it on lathe to make right diameter.

these tools helped a lot, previously I did lot of hammering on spindle shaft and had lot of marks on shaft, and maybe even damaged some bearings, but now all is OK.
this is closeup on test peace steel stock I milled in video.


as usually I made those things out of scrap I had, I didn't wanted to waste money to buy thick tubes.

[epis]

Update on my project.
I made lot of things to machine here is video of how it looks now in production run.


photo of milling machine after painting, all partial enclosure and chip guard with Enamel paint, first layer was primer paint then enamel. to prevent corrosion when cutting, and I don't need to clean all machine table now just T-slot area and sliding surfaces.



I made side way Guards too short they are from 60mm wide sheet metal, and after first water test, coolant was dripping on floor from the table, so I made additional guard peace to collect water and channel ti right way.


and about the spindle.
It is running, I adjusted bearing preload, reduced it few times because after 5-8 minutes spindle heated up and made preload by itself it was harder to rotate, I guess that was thermal expansion that made extra prelod. so i adjusted it few times and in longest spindle run about 1hour 5 min. at 3000 rpm, and occasionally at 4000 rpm spindle housing temperature went to 52-53 degree Celsius and it sill had little preload, but after cooling to some 45C it rotated freely so made again very little adjustment, I think now it will rotate free without any thermal expansion preload, but on the other side I think i have spindle Play/ backlash in spindle something about 0.01-0.03 mm when cold, and ofcourse the more I push on spindle in Z direction more it goes to 0.06-0.08mm but it is without rotation, and when spindle is cold, I guess when it is rotating, and heated up it has very low play.

I do get vibraion at 4000 rpm, one of the source is 2.2 kw motor and its big pulley which is out of balance ass every thing else, I think I will need to change pulley ratio to something like 1:1.33 or 1.5 to bet 4000-4500 rpm with motor speed no more than 3000 rpm, in that case I will eliminate one source of vibration. but that is for future when I will have time to do some experimental stuff, for now i will run machine as Hard, as long as I can. I already had one issue with new coolant On milling operation, I burned my central lubrication pump, coolant got in and shorted it up. luckily I managed to fix by repairing damaged area



and interesting thing happened also potentially dangerous, X axis limit switch filled Up with coolant, I am lucky I noticed it before it shorted up and burned my electronic

I fixed it by drilling hole at the bottom for coolant to leak out of switch.