[gussdavey]

Without having to discuss off the topic things like the target machine base, x, y & z layout and gantries I would like knowledgeable feedback on a B/C Spindle head design.
The objectives are:

• Lightweight (target weight including servos and spindle <= 20kg)
• The cost within reach of a home CNC'er / Targeted at individuals and not corporates
• HF spindle 2 to 3 kW.
• Soft metals, wood, etc.
• Low manufacturing cost and as much as possible off the shelf parts.

Apparently, nothing meeting any of the above criteria on the market.

With the following 2 main components,

1. 1 x Teknomotor - 5160 DA
2. 2 x Harmonic Drives - FHA-17C Actuators

I came up with this:



I posted the actuator and spindle to GrabCAD if anyone want to check the dimensions.

https://grabcad.com/library/fha-17c-actuators-1
https://grabcad.com/library/teknomot...0-da-spindle-1

Weight 20kg (brackets, 2 x motors and spindle). This is the 2.2kW version of the 5160 DA. There are also a 3kW one, of wich the height is 20mm taller (but can be modified to be shorter)

This Technomotor spindle was the shortest one I could get (without having to go to Jaeger or stuff that cost the price of a farm)
It has ER25 collets on both sides. I opt to remove one just above the black end cap. (Does someone maybe have a photo of such a spindle to show the inside. Where is the bearing, inside the black collar or inside the square aluminium squirrel cage?

The B and C drivers are harmonic drives FHA-17C. There pricing is high, but within reach. Not sure about which gearing, but several options.
In addition, even though the spindle and C bracket can hang of the harmonic drives, I included a support shaft right through the harmonic drives hollow to support the opposite side. I use off the shelf FK15 blocks here. (Also fixed (not floating), with two angular contact bearings)

Here are my questions:

1. You think the design are sturdy enough for materials listed in objectives?
2. What about breaks. Will it be needed. Can Harmonic Drive Inc. FHA-17-C be ordered with breaks, or is it only the more expensive harmonic drives (like SHA20, not sure with ones now), that have the break option?
3. Why your thoughts on why there are not such products on the market?

In advance, thanks for the relevant feedback.
Guss

[slowlearner]

Interesting, whats the cost on this?

[gussdavey]

Interesting, whats the cost on this?

This is just ruff estimates:

1. On the harmonic drives, I had a quote of $600 before I decided to draw the part.
   http://www.alibaba.com/product-detail/HARMONIC-DRIVE-SYSTEMS-HA-655-2_1777994074.html?spm=a2700.7724838.0.0.AsslOO
2. On the Teknomotor, I do not have a price for that model, but if I look at the general pricing of some of their other product, this should be in the region of $1000.
3. Bearings at $150 per bearing block

So on the purchased parts is it ($600 * 2) + ($1000) + ($150 * 2) = $2500.
The brackets are machined at home.
+ 50% under estimate and creep costs ($1250)
= $3750

Still an expensive exercise, but what is the alternatives.
As soon as you want to cut down on the bulkiness you have to go with these flat servo drives with their harmonic gearbox.
Pancake motor does not have the power needed.
A servo or stepper with timing belts are a option (al cheapo route), but you end up with a head that look like some kind of monstrosity out of "Transformers".
The smallest pre-made 2 axis spindle head I could find was 120kg and the price I assume is x5 or x10 of the above cost.

$3750 is like a whole months salary for an average guy here where I stay. It still a load of money.
If you add the cost of the rest of the 3 axis system, I know, you end up with some expenditures that could have bought a second hand 6 axis KUKA robot.
Using the KUKA with RobotWorks, I can still meet my 5 axis machining objectives, but with a repeatability of only 0.1mm.
Going for a second hand High Precision (0.05mm) series KUKA, too expensive again.
Then there is the new light series KUKA's at about $12000, but their reach is on 600mm and 6kg payload (not enough to handle my spindle)

Maybe this is why its a better solution to have Axis A, B and or C on the planer surface like with a 2 axis table.
In principle, I have an issue with some of the layouts as one of the axis is actually only 1/2 an axis.

Maybe, and maybe, a better and more affordable layout will be having the C axis as a rotary table on the bed (so rotating around Z), and have an B axis on the head (like a dividing head that can swing left and right (looking at the spindle from the front along your Y axis)

Guss