Any idea if there are small 200-500W spindles for ER11/ER11A collets that go beyond 12.000rpm?
My DIY machine is currently limited not by acceleration, speed or stiffness but by RPM in materials like Polyacetal.
My Kress 1050 runs up to 25k, I believe their 530FM and 800 FM run up to 29k and they both use ER11 collets, iirc
@Mike_Thornbury Bu you can’t control a kress via Modbus/RS485 or 0-10V PWM, can you?
I checked the the Kress are 89 dB(A) an 1.5Kg.
The reason I am using a 300W spindle and am looking in the 200-500W range is that they are very quiet and don’t have much mass to accelerate/bolt on securely. ( probably should have mentioned that)
It’s the rule of three: you can have powerful, quiet and cheap - pick two.
More speed with air-cooled tools means more noise, usually. My 1050 is not much louder than my water-cooled 2.5kw, and much quieter than my hand drill or pillar drill. It’s not intrusive to be in the workshop without ear protection while it’s running.
But, it doesn’t have pwm or rs485 speed control - it’s a dial on the top of the unit.
What you now need to specify is your budget. There will be a unit that fits your requirements, but it will not be cheap.
All the brushless Chinese spindles top out at 12k, But there is another option, to make your own out of an RC motor - they are light and you can get upwards of 50k rpm, with pwm control being intrinsic.
Maxon make high-speed brushed motors, but maybe hot the HP you are after, and they ain’t cheap. https://www.maxonmotor.com/maxon/view/configurator/BOM:DCX10S01EBSL593:::::?etcc_med=Navigator&etcc_cmp=Konfigurator+starten&etcc_ctv=Brushed+DC+motors
Below 1000eur I can at least think about it. Below 500eur I’ll definately consider it.
If it has a dial, it has a potentiometer and can thus be software-controlled via a PWM signal.
I have found some 250W chinese 40Krpm and 60Krpm brushless spindles for 380€ and 530€ and a 95W 60000rpm one for 150€ . They need a 1000Hz VFD for about 100€. All prices plus customs and 19% import sales tax Problem is that they have ER8 collets and I use 6mm tools.
I’m not sure if a brushed motor at that speed is a good idea?
You can achieve this using a bldc outrunner motor available from rc suppliers. If you get the right motor you can remove it’s center shaft and replace it with a shaft/collett. Should be below $300 even if you have to have the shaft customized.
@Josh_Rhodes I don’t think I have the skill to take apart a brushless motor and remove/insert the bearings on anything that shall run fast and true afterwards. I don’t dare to.
Let me look up the designs later, it’s really not very complicated.
I’ll keep these 3 chinese spindles, I found as an easy upgrade option. That’s probably my best bet.
In the end the machine itself is not my hobby. Making actual parts with it is.
@Marcus_Wolschon There are many problems to solve: You want PWM/RS485 control, light weight, low power, low noise, high accuracy/low runout, and you want to spend ~500 euro.
You haven’t given much information about many of those requirements - why do you need a low-weight spindle? Why is noise important? Why is PWM control important? Some are obvious - we all want high accuracy at a cheap price 
What are you making? What are you making it on?
From what I’ve read, Delrin/Polyacetal machines well at 12k rpm - that’s the recommended speed from Dupont using an 1/8"/6mm tool, so why do you want higher speed? Are you hitting 12k and wishing you had more? Have you thought about changing the type of tool you are using?
And, lastly, you are in the home of the high-speed accurate spindelmotor… but trying to find something in 0.5kw with PWM and 20,000+RPM under 2.5kg is proving somewhat difficult… the rule of threes, again - high speed, light, accurate - you can have two of those.
It may be that at the end of the day to achieve what you want is going to be difficult and that you may have to sacrifice one of your conditions - I would dump the weight requirement and either make the machine more robust or invest in a heavier-duty chassis that you can move all your electronics over to - but. that’s guesswork as I have no idea what you have or what your environment is.
If something like the little Kress 530 fits all your other needs, it’s simple enough to retro-fit PWM control. The SuperPID does exactly that: http://www.vhipe.com/product-private/SuperPID.htm
Machining time in finishing passes is quite long. That’s where I want to improve.
- sound: ability to run longer hours
- homing accuracy: ability to split longer hours into multiple days when surface finish matters
- rpm and weight: reduce the hours themself
Currently I’m improving my homing sensors and adding a simple electric tool-length-sensor, so I can pause while running an WOC=0.05mm finishing pass and resume another day without the point of resuming becoming visible or detectable by touch.
I can work longer hours if the machine does not get louder because with the current sound-proofing it can just barely be heard in the flats above the workshop. That’s why I use a 300W DC spindle while having a 2.2KW spindle and VFD around.
Obviously when incesting any money in a spindle, it should be a spindle and not just a handheld tool attached to an axis. start, stop and speed selection should be done in software and not ask for human intervention in case of E-Stop, pause, tool length meassurement,… I get that even in the cheapest of DC spindles, so I won’t accept such a basic limitation in a higher priced one.
Same goes for the simple fact that spindles as opposed to repurposed handheld routers are supposed to be mounted in a spindle-mount running for a good portion of their length. Handheld tools are mounted using a rather small 43mm ring at the tip, away from the center of mass.
I’m hitting 12.000rpm and thus can’t increase the feedrate even as the steppers can handle that. In finishing passes I already see about as much tool marks from the rotation itself as from the stepover.
However increasing the weight of the spindle that is accelerated would mean that (the steppers and) the stiffness of the axes becomes a limiting factor. That would prevent higher feedrates itself and thus risk the very point of why I want to change to another spindle.
I’m not all about accuracy, however when I take something apart, I no longer assume that I can spin it with 40K or 60K rpm for hours and hours where even a slight assymetry in weight distribution could lead to the thing flying apart or one of the bearings to wear quickly and fail. That’s why I reject to build my own spindle.
If I can’t find a ready-made one to buy or have to make sacrifices like being limits to 5mm tools due to ER8 collets, that’s a perfectly valid result for this investivation.
@Mike_Thornbury Dupont has a recommed RPM? That doesn’t make sense. You can only recommend an RPM/feedrate/tool-tip(not shank) diameter. These 3 always play together.
For a low feedrate, that RPM would incinerate your material. For a high feedrate it would break your tool.
@Marcus_Wolschon they produce a tool/feed/rpm sheet. I found it in about fifteen seconds (I don’t machine plastics - surprised you don’t have it).
I use HDMAdvisor as a calculator. It has a material database and takes the limits of my specific tools and of my machine into account.
The calculated cuts match real world results surprisingly well.
Also I’m not using The DuPont “Delrin”=The homopolymer but POM-C, the copolymer everyone else makes. Both materials are nearly identical but not quite.
I use highly tappered tools because of their rigidity. With cylindrical tools assumed in nearly all tables I would be well past the limit of what abuse my tool can take already. (0.5mm ball at the tip but a 6mm shank for example)