DIY Brushless Spindle

298 views
Skip to first unread message

Roland Mieslinger

unread,
Oct 24, 2016, 12:17:56 PM10/24/16
to Cyclone PCB Factory
Hi,

as this is my first post, I'd like to introduce myself. My name is Roland, I'm mostly a software guy, occasionally doing hardware work.

My goal is to have a better spindle than my Dremel 4000 is. This means at least 35k RPM, a better runout and and tool clamping.
I was unable to find a suitable spindle on eBay. It seems to me that the requirements for PCB milling (low power, high RPM) are somewhat exotic.

In the picture you can see my first version, built for the 52mm spindle gantry. I use a C8-ER11-100L as shaft, supported from 3 608 hybrid ceramic bearings (two at the bottom, one at the top end) that are spring loaded to remove any give. The motor is a 12V/3650kv brushless motor, usually used in RC cars. The coupling between the motor and the shaft is somewhat flexible to reduce the load on the motor bearings (as the motor shaft is never 100% precisely aligned with the spindle shaft). The parts are all printed in PLA (questionable as the bearings might get warm) .

So far I did only a few unloaded tests. The motor is able to run the spindle at about 40k RPM, without significant vibrations (better than my Dremel).

I'm still missing some details, e.g. proper mounting for the ESC, for a real milling test, but what concerns me the most is that my design leaves no room for errors. In other words, if anything fails, the spindle could explode right next to my head.

That's why I thought it might be a good time to share it. I'm wondering about other peoples ideas how to enhance the design and build something like a "safety net" around (e. g. steel case) to catch and contain flying parts in case of a catastrophic failure.

Thx

Roland
IMG_20161024_154631.jpg

Iceflow

unread,
Oct 25, 2016, 3:44:04 AM10/25/16
to cyclone-p...@googlegroups.com
The desing looks very nice and powerful, dont have experience with that setups... :( i am of no help, sorry. You can print a cover for the front part, al least to hide the bearing zone... also you can try some other plastic, like nylon, maybe ir will be more suitable for that pourpose, dont really know, the plastic parts will need to hold a lot of forces... you need to run a long test and hide away while it does hehehehe

--
You received this message because you are subscribed to the Google Groups "Cyclone PCB Factory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsub...@googlegroups.com.
To post to this group, send email to cyclone-pcb-factory@googlegroups.com.
Visit this group at https://groups.google.com/group/cyclone-pcb-factory.
To view this discussion on the web, visit https://groups.google.com/d/msgid/cyclone-pcb-factory/feb2b58a-1013-4b26-a809-d7eb1abea5bd%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Carlos García

unread,
Oct 25, 2016, 5:03:58 AM10/25/16
to cyclone-p...@googlegroups.com
Hi Roland, welcome!
I think your approach is much more suited than a Dremel tool, for milling PCBs and even harder materials.

But indeed it is necessary to prevent any damage caused by a sudden "explosion" of the mechanism in case PLA parts break down.
Maybe the most simple way to do this is by tighly attaching a steel net around the Z axis?

Even with that, I agree with Iceflow, you can leave it running "until it pops" (if it does) in order to find the weak spots.


Also, would you mind sharing a more detailed picture of the coupling mechanism? (or a diagram)
I don't understand it very well, in the picture it seems to have small metal pivots that could potentially fly away.

Cheers,
Carlos


On Tue, Oct 25, 2016 at 9:43 AM, Iceflow <ais...@gmail.com> wrote:
The desing looks very nice and powerful, dont have experience with that setups... :( i am of no help, sorry. You can print a cover for the front part, al least to hide the bearing zone... also you can try some other plastic, like nylon, maybe ir will be more suitable for that pourpose, dont really know, the plastic parts will need to hold a lot of forces... you need to run a long test and hide away while it does hehehehe
2016-10-24 18:17 GMT+02:00 Roland Mieslinger <rol...@gmail.com>:
Hi,

as this is my first post, I'd like to introduce myself. My name is Roland, I'm mostly a software guy, occasionally doing hardware work.

My goal is to have a better spindle than my Dremel 4000 is. This means at least 35k RPM, a better runout and and tool clamping.
I was unable to find a suitable spindle on eBay. It seems to me that the requirements for PCB milling (low power, high RPM) are somewhat exotic.

In the picture you can see my first version, built for the 52mm spindle gantry. I use a C8-ER11-100L as shaft, supported from 3 608 hybrid ceramic bearings (two at the bottom, one at the top end) that are spring loaded to remove any give. The motor is a 12V/3650kv brushless motor, usually used in RC cars. The coupling between the motor and the shaft is somewhat flexible to reduce the load on the motor bearings (as the motor shaft is never 100% precisely aligned with the spindle shaft). The parts are all printed in PLA (questionable as the bearings might get warm) .

So far I did only a few unloaded tests. The motor is able to run the spindle at about 40k RPM, without significant vibrations (better than my Dremel).

I'm still missing some details, e.g. proper mounting for the ESC, for a real milling test, but what concerns me the most is that my design leaves no room for errors. In other words, if anything fails, the spindle could explode right next to my head.

That's why I thought it might be a good time to share it. I'm wondering about other peoples ideas how to enhance the design and build something like a "safety net" around (e. g. steel case) to catch and contain flying parts in case of a catastrophic failure.

Thx

Roland

--
You received this message because you are subscribed to the Google Groups "Cyclone PCB Factory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsubscribe...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Cyclone PCB Factory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsub...@googlegroups.com.
To post to this group, send email to cyclone-pcb-factory@googlegroups.com.
Visit this group at https://groups.google.com/group/cyclone-pcb-factory.

Nando

unread,
Oct 25, 2016, 1:11:32 PM10/25/16
to Cyclone PCB Factory
Hi,

That is a great idea and something that I haven't lost hope to have.

Some time ago and trying to copy some commercial systems I made (just for testing) something that copies this idea:


But of course in PLA (it could be made in Nylon or HIPS to be stronger).
As you ask for ideas to enhance security, having a belt driven system may give some.
In case the mill get's stuck the belt may jump out or slide as well as you can have some 
multiplication allowing to use a lower KV motor.

Cheers
-Nando
Sent from my Sinclair ZX81

--
You received this message because you are subscribed to the Google Groups "Cyclone PCB Factory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsub...@googlegroups.com.
To post to this group, send email to cyclone-pcb-factory@googlegroups.com.
Visit this group at https://groups.google.com/group/cyclone-pcb-factory.

Roland Mieslinger

unread,
Oct 25, 2016, 1:55:03 PM10/25/16
to Cyclone PCB Factory


Am Dienstag, 25. Oktober 2016 11:03:58 UTC+2 schrieb Carlos Garcia Saura:
Hi Roland, welcome!
I think your approach is much more suited than a Dremel tool, for milling PCBs and even harder materials.

not sure about "harder materials". The joint between the motor and the shaft is intentionally weak.  
 

But indeed it is necessary to prevent any damage caused by a sudden "explosion" of the mechanism in case PLA parts break down.
Maybe the most simple way to do this is by tighly attaching a steel net around the Z axis?

Steel mesh sounds good. Maybe fixed to same bolts that hold the spindle.
 

Even with that, I agree with Iceflow, you can leave it running "until it pops" (if it does) in order to find the weak spots.

 
Also, would you mind sharing a more detailed picture of the coupling mechanism? (or a diagram)
I don't understand it very well, in the picture it seems to have small metal pivots that could potentially fly away.

Yes indeed, there are small metal pivots. For the moment, wrapping duct tape around the upper section should be good enough to contain them.
I'm not happy with the design, as it adds a lot of noise, but for now, I have no better idea (other than buying an expensive shaft coupler).


Cheers,
Carlos


On Tue, Oct 25, 2016 at 9:43 AM, Iceflow <ais...@gmail.com> wrote:
The desing looks very nice and powerful, dont have experience with that setups... :( i am of no help, sorry. You can print a cover for the front part, al least to hide the bearing zone... also you can try some other plastic, like nylon, maybe ir will be more suitable for that pourpose, dont really know, the plastic parts will need to hold a lot of forces... you need to run a long test and hide away while it does hehehehe
2016-10-24 18:17 GMT+02:00 Roland Mieslinger <rol...@gmail.com>:
Hi,

as this is my first post, I'd like to introduce myself. My name is Roland, I'm mostly a software guy, occasionally doing hardware work.

My goal is to have a better spindle than my Dremel 4000 is. This means at least 35k RPM, a better runout and and tool clamping.
I was unable to find a suitable spindle on eBay. It seems to me that the requirements for PCB milling (low power, high RPM) are somewhat exotic.

In the picture you can see my first version, built for the 52mm spindle gantry. I use a C8-ER11-100L as shaft, supported from 3 608 hybrid ceramic bearings (two at the bottom, one at the top end) that are spring loaded to remove any give. The motor is a 12V/3650kv brushless motor, usually used in RC cars. The coupling between the motor and the shaft is somewhat flexible to reduce the load on the motor bearings (as the motor shaft is never 100% precisely aligned with the spindle shaft). The parts are all printed in PLA (questionable as the bearings might get warm) .

So far I did only a few unloaded tests. The motor is able to run the spindle at about 40k RPM, without significant vibrations (better than my Dremel).

I'm still missing some details, e.g. proper mounting for the ESC, for a real milling test, but what concerns me the most is that my design leaves no room for errors. In other words, if anything fails, the spindle could explode right next to my head.

That's why I thought it might be a good time to share it. I'm wondering about other peoples ideas how to enhance the design and build something like a "safety net" around (e. g. steel case) to catch and contain flying parts in case of a catastrophic failure.

Thx

Roland

--
You received this message because you are subscribed to the Google Groups "Cyclone PCB Factory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsub...@googlegroups.com.
To post to this group, send email to cyclone-p...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Cyclone PCB Factory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsub...@googlegroups.com.
To post to this group, send email to cyclone-p...@googlegroups.com.
IMG_20161025_185058.jpg

Carlos García

unread,
Oct 25, 2016, 2:30:41 PM10/25/16
to cyclone-p...@googlegroups.com
Ahh I see, it is clear now, thank you very much for the explanation and for the picture :)

In that case, Nando's idea might be a safer way to go. Rubber belt coupling doesn't seem nearly as dangerous.



Cheers,
Carlos


To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsubscribe...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Cyclone PCB Factory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsubscribe...@googlegroups.com.
To post to this group, send email to cyclone-p...@googlegroups.com.
Visit this group at https://groups.google.com/group/cyclone-pcb-factory.

--
You received this message because you are subscribed to the Google Groups "Cyclone PCB Factory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsub...@googlegroups.com.
To post to this group, send email to cyclone-pcb-factory@googlegroups.com.

Roland Mieslinger

unread,
Oct 27, 2016, 5:53:42 AM10/27/16
to Cyclone PCB Factory
A belt driven spindle was what I had in mind first, but I found it impossible to find the right parts (belt and pulleys) on eBay. Maybe I used the wrong search terms :)

All I found was GT2 pulleys, but timing belts are considered bad, as they tend to vibrate and resonant. Do you know a good source for round belts and pulleys (8mm and 1/8")?

As of now, I printed a straight joint from somewhat flexible material (eSun Flex PLA). It didn't break immediately.

I'm now using a metal mesh, laminated in plastic to make working with it easier, as a "safety net". The zip ties are only temporary, the next iteration will have front and back plates over the mesh.


 
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsub...@googlegroups.com.
To post to this group, send email to cyclone-p...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Cyclone PCB Factory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsub...@googlegroups.com.
To post to this group, send email to cyclone-p...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Cyclone PCB Factory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsub...@googlegroups.com.
To post to this group, send email to cyclone-p...@googlegroups.com.
IMG_20161027_113817.jpg

Roland Mieslinger

unread,
Oct 27, 2016, 6:31:32 AM10/27/16
to Cyclone PCB Factory
Here is a short video of the spindle running, I estimate it is about 40k rpm.


The spindle is not clamped to the desk in any way, but not wandering around either.
Almost no vibrations noticeable.

Roland

Nando

unread,
Oct 27, 2016, 8:18:05 AM10/27/16
to Cyclone PCB Factory
Good point and I have to say I didn't knew that.
My first test indeed were made with a GT2 belt I had as spare from converting the cyclone X and Y transmission to belt driven system.
But I never tried it as a real spindle, only draw and assembled the parts and tested with low RPM.

I recall seeing those type of belp in old CD-Roms/DVD's so may be one cna be used for that, and even the pulleys as they should be
in delrin that is more wear resistance.
Even if the belt breaks damage should be minimal or null.



-Nando
Sent from my Sinclair ZX81

To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsubscribe...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Cyclone PCB Factory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsubscribe...@googlegroups.com.
To post to this group, send email to cyclone-p...@googlegroups.com.
Visit this group at https://groups.google.com/group/cyclone-pcb-factory.

--
You received this message because you are subscribed to the Google Groups "Cyclone PCB Factory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsubscribe...@googlegroups.com.
To post to this group, send email to cyclone-p...@googlegroups.com.
Visit this group at https://groups.google.com/group/cyclone-pcb-factory.

--
You received this message because you are subscribed to the Google Groups "Cyclone PCB Factory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cyclone-pcb-factory+unsubscribe...@googlegroups.com.
To post to this group, send email to cyclone-pcb-factory@googlegroups.com.

Roland Mieslinger

unread,
Nov 15, 2016, 3:39:58 AM11/15/16
to Cyclone PCB Factory

Time for an update, after a few tests of the spindle on the mill.

PLA seems to be the wrong material, at least for the bearing housing and the motor mount, as these parts are getting warm (>80 °C for the motor).
The motor needs a cooling fan.



On the upside, the joint between motor and spindle did exactly what I had hopped for. It broke before anything else. I think that a weak joint is a safety feature.

Here are a few pictures showing SMD (1206, 0805, 0603 and 0402), SO8 and QFP, SOT-23 footprints, etched at 300mm/min and 0.05mm depth. 

 


-- Roland


Reply all
Reply to author
Forward
0 new messages