Homebrew PCBs

Stefan Trethan wrote:
> 
[cut]

> 
> a) a good bearing. the really cheap devices only put a short shaft with a 
> thread on a standard motor.
> thus only using the bearings of the motor. please make your own opinion on 
> that ;-).
> 
> b) PRECISION work of shaft and collars.
>  the cheap grinders i have here are all no precision work.
> the bit is not centered very good, thus it will break more easily.

> the proxxon precision is fascinating.
> the collars are very strong material (hardened - very springy).
> they are ground very exactly, also the shaft is ground.
> they are shaped so that they are always centered.
> proxxon knows this, they advertise this a lot....
> 
> i have inspected other tools (dremel etc) at the stores here also.
> they have even aluminium collars, far from precision..
>


The collets on the industrial tools I've seen are all precision 
ground hardened steel, exactly as you describe. I haven't 
measured the round out, but its definitely not visible.
This is the standard way collets are built. Collets in the Dremal 
are a cheap and nasty version. I'm just glad the store I bought 
the Dremal from gave me my money back when I returned it.

I'm not sure about double bearings on the end of the motor shaft. 
I think the grinder I bought has one ball bearing. Its designed 
for lots of side force so I'm sure the engineers thought one 
bearing is enough.

> i also think the thick throat is no help in using it (the proxxon has a 
> longer narrower).
> for me it looks like a chopped of angle grinder ;-).
> but sure for pcb it would be fine too.
> 
> 
> (the disadvantage of powerful motors (more than 100 watts) would be 
> noise... you don't need much power in pcm drilling)

Yes, its oversized for the job, but that was all I could find at 
the time when I was shopping. If I were buying another spindle 
I'd definitely look at the Proxxon, but what I got works well and 
no need to spend more money. I always wear ear protection when 
using a machines like this. But that doesn't bother me.

> *** 2) adams new press:
> 
> hey adam what are you doing?
> i really thought you are telling me for weeks now that a triangular arm 
> would be better.


Sorry, but I did some more thinking and thought that taper roller 
bearings would be more stable. I was ready to buy the rod ends 
when I learned about tapered roller bearings in car wheels.

> i thought of placing the back bearing on the two outer edges. this should 
> prevent the unit
> from sideways clearance at the front.
> i thought this because your first unit needed  a guidance bearing in the 
> middle of the arm to stabilize.
Yes, it did. As you say, if the two thrust bearings were much 
farther apart and the length of the arm decreased then the 
guidance bearings would not be required.

> it there no clearng with the new tapered bearings??

yes, the tapered bearings are forced towards each other like 
thrust bearings, except the tapered design also gives zero play 
both axially and radialy. That's the benefit. You could still go 
for triangle frame and place the tapered bearings and a far 
distance. The only important thing is the two bearings must be 
very parallel to each other or they will not rotate smoothly and 
try to warpping the frame/mounting blocks.

> maybe i will try it anyways with the thrust bearings and triangular setup.
> i found two of them finally now when i searched for tapered roller bearings 
> ;-).
> i had them at home all the time.....

You should be able to buy two new car wheel bearings from a 
bearing supplier, fairly cheap. They seem to come in industry 
standard sizes, known as SET 1, SET2,...ect  The bearing 
dimensions are neither imperial nor metric.


> maybe it would get even more stable if i add two normal ball bearings on 
> the sides...
> so each direction is blocked from any slight clearance.

you could do that too, but it might be complicated to build. 
normal bearings are not designed for large axial force, but for 
your application the forces are not that great.

> have you tested it already?
> is the lenght of the arm still no problem? (difference between tangent 
> movement and straight)
> 

During testing I drilled 50 or so 0.6mm holes and everything 
looked ok. I also drilled many 1.7mm holes and that works ok too.
The 350 mm radius of the arc seems like enough. I have a 
sensitive dial indicator that I can use to measure roundout, and 
side movements. I'll try to do these measurements and report the 
results on my web page.

Adam.