Thanks for the info Simpson, that stuff does look good but for my small scale needs the cost is too much I am afraid. As for using the air to cool the tool I use flood coolant so no air consumed at the moment.

Hood                                                           

I am planning to add flood myself and that was one of the attractions of flex shaft or air turbine  . . . immunity to water.  I know you have flood already, but you planning to use that for engraving? That seems like it would be a bit uncommon for engraving.

An economy approach to an engraving spindle would be an even less inexpensive (or used) die grinder mounted on your mill head. That's really the quick and dirty, but besides the end play issue you already mentioned, I do think that most consumer level products are only going to last a few hours in production.

I am making a permanent engraving head from a new surplus 30,000 RPM tool motor (ebay) to belt drive a straight shank ER16 coller holder running on hgh speed bearings in a simple casing made from DOM tubing and mounted on the mill head. This will satisfy three objectives in my application:

1) save the wear and tear on my poor much abused Dumore flex shaft grinder. (the replacement shaft costs more than many complete die grinders)

2) allow me to use 3/8" shank tool bits as well as the normal 1/4" and 1/8"

3) provide for easy ratio changes should I need more RPM or more torque for a particular job.

I have no doubt you could fab up someting similar if that sounds like it would be a useful approach. If you want to feally fast, a drop to ER-11 or 'universal' double taper would be mo' better

Use flood for everything I do Well except inserted cutters as they tend to like to cut  dry, especially  on the 316.

If this was going to be production I would definitely be buying something of high quality but its likely to be only a few a year so no point in splashing out, after all I am a Scot
Have been toying with the idea of making a small spindle and driving with one of the model aero DC motors but I have found out someone I talk to on Skype is about to do the same so I will wait and see how his turns out

Hood

Flood cooling at 22K RPM will make a big mess. Hope you have a full mill enclosure . . .    I have this funny mental image of the operator standing there in one of those huge yellow 'lighthouse keeper' type of rain suits.

Below is the 'spindle' I am thinking about. I already have a 1" shaft ER32 version that I bought some time ago and have never used, but that was not intended for engraving speeds. For 22K, you could use this, methinks:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=150477898066&ssPageName=STRK:MEWAX:IT

Also at 'only' 22K you can use a GT3 tooth belt, or a small v-belt or a flat belt. Belts are rated in linear FMP, so just keep the pulley diameter small and you'll be good to go.   

Not a full enclosure but the table has a guard round  three sides, the front having a removable perspex door, so no oil skins required

I have a DA pencil collet that I was contemplating using as a spindle but that ER is a great price.

Did make up a spindle speeder of sorts which was geared via a toothed belt but it got too hot to touch after about 3 mins running and the heat seemed to be coming from the belt friction on the pulleys, was thinking of directing some coolant through the body but not sure how the bearings would like it, would have to filter it and may give it a go.



Hood

A friend made something similar to what you show in the pic. He basicaly took the hand grinder added a pulley to it and used a small motor and was belt driven. Works fine, but would have been just as easy to use the flex shaft. I have run the flex shafted one on my end for 4 hours continously at one time
( say worst case ) and the handle just got warm, motor got hot though. My friend had the heat problem with his setup. I guess it all depends
on how roburst you want to make it and what your going to do with it. Some of the high speed expensive spindles have around 1000 hrs life max before they must be sent back and rebuilt and purchase and rebuild is not cheap.

As far as the using the ER, not worth making the holder unless you can grind the shaft if making it. Even if puchased you will have additional runout due to length. Again depends on what you intend to due with it.

RICH

Not a full enclosure but the table has a guard round  three sides, the front having a removable perspex door, so no oil skins required

I have a DA pencil collet that I was contemplating using as a spindle but that ER is a great price.

Hood, beware of these Chinese collet holders. Bought few BT30 holders (from another Chinese seller though) and accuracy is not something these were built for. The quoted 0.01mm runout appeared to be not the worst case like you'd expect, but the best tolerance they can achieve. Actual runout measured on 5 holders ranged from about 0.02mm to 0.25mm (I didn't miss a zero there)!!!

Did make up a spindle speeder of sorts which was geared via a toothed belt but it got too hot to touch after about 3 mins running and the heat seemed to be coming from the belt friction on the pulleys, was thinking of directing some coolant through the body but not sure how the bearings would like it, would have to filter it and may give it a go.



Hood

Round tooth profile belts perform better at higher speeds. They engage better with the pulley, with less friction, and thus less heat (and noise) is produced. But I assume this is what you already have... What speed are you running the driving pulley at? And what diameter is it? I have 75mm diameter pulley that would run for hours at 3500RPM without getting that hot. I am sure I could run it even at 5000RPM much more than 3 minutes and not get that hot. May be check the tension?

Anyway, the best choice for a high speed application is a V-belt, or any belt other than timing belts. Or put it another way, timing belts are least suitable for high speed applications because their teeth engagement and disengagement with the pulley teeth produce a lot of noise and heat at high speeds. You don't really need a timing belt there - no angular positioning needed and no big moments to be transferred - so may be put a V-belt or even an O-ring belt and see if it helps. It really should, unless there is another source for the heat.

Daniel

There is an amusing engineering expression: 'A Camel is a horse designed by a committee'.

In engineering a mechanical assembly, you should start with objectives and parameters and then design to satisfy your targets. For example, you don't make a soup spoon out of titanium just because titanium is 'better'. Running hot is not automatically bad, either. Heat is actually beneficial in many cases. Just don't exceed the specs.

If there is an unlimited budget, then the smart move here it to simply plunk down the cash and purchase a high speed spindle, electric or air, as desired. However, the primary parameter here has been stated by the 'customer' as cost, so living with the tolerances offered by 'cost effective' solutions is simply a necessary part of this particular project. It is unrealistic to expect .000000000001" runout for 25 bucks. 

A good example is the heat issue. Aluminum is obviously better at getting rid of the heat, so you would want to make the body from aluminum . . .  but only if you can tolerate the accuracy loss generated by the clearance you would need to design into the bearings to accommodate the expansion difference between the aluminum casing and the steel shaft within. If you are using high clearance ball bearings, probably not a problem over 4" distance. On the other hand, if you have a precision angular contact set, the expansion difference will erase your pre load . . not good.

Bearings have a temp spec. Hybrid ceramics can take a lot more heat and are not nearly as expensive as full ceramic. Simply using a cutoff tool to make  . . oh, say 20 or so radial grooves .100" x.100" cut into the casing OD may provide adequate cooling for hybrid ceramics, which run hot, but thats perfectly OK.  Clamping aluminum mounting blocks right at the bearing sites will draw off a great deal of heat as well. Then you could get fancy and cut cooling fins in the mounting blocks . . etc, etc.

A left a question unanswered some time back, so here is the answer: For my previous little X2 mill, in order to run the spindle at 7,500 RPM, I simply attached an old Pentium 3 aluminum heat sink to the side of the head. That was enough to keep the beargins in spec. Low tech, cheap, ugly, but completely effective.

A left a question unanswered some time back, so here is the answer: For my previous little X2 mill, in order to run the spindle at 7,500 RPM, I simply attached an old Pentium 3 aluminum heat sink to the side of the head. That was enough to keep the beargins in spec. Low tech, cheap, ugly, but completely effective.

Thanks Steve.

Daniel

May be worth a try:

http://cgi.ebay.com/Water-Cooled-CNC-Spindle-Motor-1-5KW-2HP-Engraving-ER11-/170513456929?pt=BI_Tool_Work_Holding