two small grinding and milling spindels

[Joachim Steinke]

Hi

As the topic “compound rest mounted grinders” is discussed in the “question and answers” section at the moment I want to show you my answer to this affair. Some of you will have seen one of the grinding spindles on my already displayed cam grinder as well.

My first attempts at grinding on the lathe had also been with a Dremel like Proxxon miniature grinder which I had placed in a clamping fist on my tool post. But naturally the results turned out more or less disappointing, even after dressing the small grinding pens I had to fight with chatter and the surfaces where mainly rough and uneven. There is simply too little rigidity in the whole spindle bearings, okay, this small handheld grinders are not constructed for that purpose anyway.

So I had to look for another solution…..

For a better understanding of my approach to this (and other) self made tooling I have to remind you that I only have small machines and I’m building more or less tiny things with this equipment. So many of my self-constructed items have to fit (and work!) properly in this limited space, and that often leads to dimensions of nearly the size of watchmaker machinery.

People with full grown lathes can simply by a commercial tool post grinder, there is a wide range of offers from cheap to expensive ones with more ore less quality. But the commercial ones don’t fit on my PD360 in any way. Even if I’m able to mount them, there will be no chance to move around sufficient enough in all the limited areas, not to mention the heavy load on the small lathe support.

A still matching design could be borrowed from the traditional watchmaker milling- and grinding auxiliary spindles, but I don’t want to have a transmission above and beside the lathe. I wanted a small and compact but mostly efficient system which is carrying its own drive train. And I wanted the opportunity to carry out light milling jobs with the same equipment on demand.

So I turned on the CAD station and tried to design something that will fit my needs. And as a really small spindle is more convenient for fine jobs, otherwise a more powerful spindle is needed for some “bigger” tasks, I decided to build two versions.

This is the small one:

I established a diameter of 33mm for all my spindle housings, which fits barely in my machine dimensions but offers enough space for suitable bearings. For this spindle shaft I took a commercial ER11 collets shaft of 12mm diameter, the bearings are normal ball bearing types which get a small axial preload when assembling the rear drive pulley.

Powering comes from a 32V/75VA Escap DC motor which is driven by an adjustable power supply. For tensing the MXL tooth belt the motor got an eccentric mounting ring, so the pulley distance can be varied turning the whole unit in its clamping fist easily, that’s what you see on the drawing here:

The pulley train creates a drive ratio of 2:1 which enables a maximum spindle speed of 12thousand rpm. As I said, that is a small but very handy system for jobs like inner grinding jobs and also tiny carbide millers of no more than 2 ore 3mm diameter. This spindle runs very quite and without any vibrations, apartment sessions at midnight will not bother the sleeping neighbourhood…ha ha ha….


The design of the bigger version proved to be a little more complicated. I intended to get something about 150VA in the power unit and sufficient rigidity in the spindle bearings. Two angular contact ball bearings will be the first choice for this job. But the smallest ones are 30mm in outer diameter (my spindle housing has only 33mm) and they will have a bore of 10mm, no way to build a sense making spindle with this diameter.

In addition I wanted a collets system for the tooling interface and a hollow shaft with minimal 12.5mm inner bore. So I decided to take two double row angular contact ball bearings, *** 3803 (17/26/7). What is clearly a compromise as they can not take the same load than the bigger (one row) 7202 types, but it will work for my duty.

The bearings get their needed preload from a saucer spring supported by the belt pulley.

The hollow shaft is made from high carbon steel and got an ER20 interface with a tread for mini type collets nuts.

The collet system is highly universal and allows a wide range of tooling use, but it is no perfect support for larger grinding wheels. For that reason I designed an adaptor which uses the ER cone as a tool seat. The adaptors are made for grinding wheels with a standard inner bore of 20mm (solid corundum and compound dia/cbn wheels as well) and are fixed with a draw bar. The adaptors are much like the short cone SK25 tooling shafts we like to use on professional millers here in Germany, they offer high rigidity, a precise run out and some real fast wheel change.

Okay, that was done quite suiting until here, but how to get some serious 150VA of motor in my small dimensions?

I’m absolutely no fan of noisy machines, so normally brushed DC motors are no choice for me. But I didn’t want to disturb the whole neat design with a 3 phase motor of nearly 4 pounds weight which in addition will become as big as a drum (compared to the rest)…ha ha ha.

So I gave a different motor concept a chance, a brushless direct current servo motor, they have half the size of a comparable induction motor. But they won’t run on a normal power supply, you need a special amplifier that synthesises 3 rotating phases of direct current.

Finally I ordered a DECS 50/5 1-Q-EC amplifier from Maxon. With this controller the servo motor can by driven in a wide speed range and nearly full motor torque is available from about 800rpm up to full speed.

And installed it together with a 48V/7A switching power supply in a housing.

Oops…..someone must have finished the two grinding/milling spindles in the mean time as well…..ha ha ha….

The BLDC Motor comes from Nanotec, a DB42C01 type, only 42x42x100mm and 0.75kg, 48V/150VA, 6000 rpm max. speed range.

The drive train is realized with a flat belt, which provides a smooth running without any vibrations at high speed. The pulley combination can be altered easily from 1.4:1 to 2.4:1 by using the similar eccentric motor mount system than I had build for the smaller spindle.

As the servo motor goes with this Maxon controller up to 7000rpm (more than the 6000rpm from Nanotecs data sheet), the pulley combination enables maximum speeds of 10tousend and 17thousand rpm. That gives enough speed for all inner grinding jobs with small grinding pens as well. And the system is real quite too, only a small swoosh to hear.

And here the short cone adapters for use instead of the ER collet system.

To install the spindles in the lathe I made a special cartridge for my tool changer system, so changing goes real quick and the mounting is as near as possible to the other components. That provides rigidity and saves space.

As I generalized the outer diameter of my spindle liners both spindles will fit in all adapters I made for several purposes and for use on several machines.

At last here some pictures of the spindles positioned on the lathe.

And naturally I use both of them (mainly the bigger one) on my universal tool grinder as well now.

In the mean time both spindles had made a lot of jobs, inner cones for self made collet systems, some outer round grinding and small drilling and milling tasks on parts that could stay directly in the lathe chuck etc. etc….. and the complete work on the tool grinder.

The components proved to be very reliable and the whole concept is really convenient to me, let’s say, I don’t want to miss this little helpers anymore.

Achim

 

[gbritnell]

Hi Joachim,
Your design and build projects are works of art. I thoroughly enjoy reading your posts just to look at your work.
When do you find the time to build all of your accessory tooling while doing your other projects? By the way, what kind of engine projects do you have planned?
Thank you for your masterful contributions to this forum.
George

 

[aussie bruce]

Joachim what can i say but :bow: :bow: :bow:

STUNNING

Bruce

 

[RichD]

Extremely nice work Joachim!
Someday I hope to be able to do work of that quality. Wow!
Regards,
Rich

 

[dsquire]

Joachim

You certainly do have a masters touch when it comes to designing and machining new tools for your workshop. I thank you so much for sharing the details with us all on the forum. I will be watching and waiting for the next masterpiece that comes from Joachim. :bow: :bow:

Cheers

Don

 

[ozzie46]

Gorgeous, absolutely gorgeous!! :bow: :bow: :bow: :bow:

Ron

 

[cfellows]

Lovely work, Joachim. I was curious about the diamond grinding wheel. Is that a glass grinding wheel?

Chuck

 

[ghart3]

Very very nice. Good design and well done, Thanks for sharing.

 

[ttrikalin]

Apart from the design and the quality of the machining, your photos are superb and your writeup is a pleasure to read.

Please write a book - I'd buy it.

take care,

t

 

[Joachim Steinke]

Hi

Thank you so much for all that really nice replies.

Chuck, to your question concerning the diamond wheel:
Nothing special about that, this is a usual cup wheel with a resin bonded diamond covering. This one only has a very slender contour, it’s the kind of wheels normally made for sharpening carbide saw blades. But you can use it for all tiny and fiddly jobs on our mill and lathe tooling as well. Depending on the pointy v-shape profile it’s very suitable for getting into all slender edges.

And as George put a question to my actual engine projects ….okay, my design department is in a sate of confusion and indecisiveness at the moment …..ha ha ha….But I take the chance to give you a short survey of what’s going on inside my CAD system at the moment….

Well, actually I intend to build a 2 cyl.v-engine (2 x 5ccm) over the winter, much like Georges latest (and real excellent!) project, but a little bit smaller and with 50 or 45deg cylinder angle. The drawings, all done during summer, are nearly completed, only some minor details to be clarified left. But I didn’t get the right kick to begin with the machining somehow. And in October I started to experiment with some other valve gearings, and that was the beginning of the actual disaster….ha ha ha…

First I started to draw a one cylinder engine with an upright shaft, using bevel gears for the valve train.

But I’m not really happy with this design.

Bevel gears in the required dimension (and in high quality too) are hard to get. I want them really small fitting to the tiny dimension of the gear case and for the job in the valve train helical gears would be most suitable. But there is no good chance to get smaller than module 0.6 helical gears on the normal market. And I don’t get them with unequal transmission ratio either. That means I have to reduce the 1:2 valve timing completely “downstairs” in the lower gear case and the case becomes unproportional wide and long. And making your own bevel gears is not possible without a real gear hobbing machine, even the straight ones can not be made on a normal milling machine.

Okay, I half finished the construction anyway (mainly the upper head section is not done yet), the drawings show straight bevel gears in module 0.5 and 1:1 ratio which will be available in good quality from my suppliers here in Germany without any problems.

But as already said, I’m not really convinced about this version too…..

So I took the next idea, timing the valves with two push rods driven by eccentric shafts. This concept was realized in the 50ties by NSU in the MAX motorcycles. But with such a design you will get some trouble with the different elongation of cylinder, head and push rods during engine run. I calculated something about 0.08mm difference for this 5ccm engine when running really hot, so that will be too much for the bearing clearance of the rods

NSU had a clever solution for this, they simply placed a third but fixed push rod behind the two mains. This rod was connected to the cam shaft housing which could rotate to an axis which virtually travels trough the contact points between the rocker arms and the valve shafts. So the lesser (in comparison to cylinder and head) elongation of the third rod moves the whole cam shaft housing up a tiny little bit, but this has no real effect on the valve clearance, and that is the ingenious aspect of this construction.

I tried to replicate the NSU design, this here is only a rough layout because I quickly realized that this concept would by really hard to bring in my 1:4.5 scale.

The components get very small and thin and as you can see there is already no sufficient clearance for the rockers, the housing has to get even thinner. And it will only allow a single cam lobe, which is possible from kinematics. But you have no chance to change the cam timing later on, each new lobe shape requires different rocker lengths and new rocker axis positions, otherwise you have no influence on individual cam angle settings for start and end timing.

Okay, what will work if the original is hard to build in small scales?

Then, when driving on the motorway (that is always very good for contemplation …ha ha ha…), I had the idea to use a swinging gear train in the head.

This motion link, which allows the cylinder to move in relation to the push rods without changing bearing clearance, will have the desired effect as well. But the whole package of rods and gears need a lot of space, especially in the head area (and not much better in the lower gear case area), what leads to a really ugly and clumsy engine head design.

So that’s what I mean with not being that happy with the actual performance in my construction department…ha ha ha….but I won’t give up….if things will not develop to my wishes soon I will fetch the already finished two cylinder engine plans from my drawers and start with this project.

Achim

 

[cfellows]

Achim, helical gears are not at hard to make once you figure out the basic math, which isn't all that difficult either. Here is a link to some work I did making helical gears:

http://www.homemodelenginemachinist.com/index.php?topic=9916.0

Also, have you looked into helical gears to radio control model cars, etc.?

Chuck

 

[Lakc]

Excellent work Achim, as usual. You certainly set the quality bar very high! :bow:

 

[Noitoen]

Achim-

What about using a chain drive to power the camshafts?

Pat J

I would suppose that's an "ordinary" way to do it. He's looking for an extraordinary way to do it.

 

[BillTodd]

Reminds me of the Husqvarna ExCam System here:

http://members.chello.nl/~wgj.jansen/text/husqvarna.html

Bill

 

[Joachim Steinke]

Thank you Chuck

I already had a look at your article about helical gears some weeks ago.

I think I didn’t used the correct word for my desired gears (I better had a look at my English dictionary first…ha ha ha…), I didn’t meant helical but bevel gears with spiral shaped teeth. Like this kinds for example

By the way, they are still too big to my wishes, as you can see in comparison to the plug and a prototype camshaft. This is metric module 0.6 which should be nearly equal to a diametral pitch of 42 teeth per inch (for getting a better dimensional idea).

I prefer the spiral shaped ones because they run smoother and more quietly, especially when you have to use very small (small amount of teeth) gears that have to run at relatively high speed. But finally I will take the straight ones as well, it’s only a matter of excessive perfection…..ha ha ha….

Even straight bevel gears (the spiral ones even more) can not be manufactured precisely enough with standard tooling like module millers (and) on machines with only straight moving axes. In contrast to spur gears all bevel gears have no constant module, it varies from outer to inner gear diameter.

You can only approximate some semi suiting shape by making two extra milling passes on the already pre formed straight tooth path. For that you have to rotate the indexer main base in the amount of “n=i/(4*z)” by additionally shifting the gear blank in the amount of “n=i/z” parallel to the gear axis. This is to be done on the other (mirrored) tooth side as well.

But this will still form teeth with parallel head- and bottom lines, as I said it’s only an approximation and not suiting for high performance and fast running gear trains. And it becomes a really boring dividing job, especially if you have to do all the indexing without some nc based rotary table.

@Bill

Thanks for the Husqvarna link. This system is a little bit different to NSU as the camshafts are not revolving, they only sway around. But it is clearly an interesting variation too!

And Noitoen, you guessed it right, there is no chain drive and no tooth belt in the run as well….ha ha ha……

Achim

 

[Coda1]

Hello, I just joined this "forum" and ran into your threads, and I have to say, WOW! very nice. I was hoping to get inspired by some models I've seen on this forum and I was just impressed with your projects, Joachim. I would love to make a grinding wheel and was hoping if you can share basic dimensions of your smaller grinding wheel (with Escap motor), I have the same exact motor but I didn't know what I could use it for, until now. I visited your websit and as a few members stated here, they are a work of art. I would now like to build and use a grinding wheel on my mini-lathe. So, is it ok for me to ask for basic dimesions?