Brushless Makita Mod (unsupported)

Alright, any requests?

Running a reduced diamond coated 0.125 SF

6 Likes

I would really like to know how it handles low RPM milling.

1 Like

Yes low rpm 4-8000 would be interesting :face_with_monocle:

Well I went with a more comfortable fusion adaptive cut. Because Im running the Nomad I can’t push as much force as the Shapeoko with its Nema17s. An 8 pound max machine force feels about right. Ran a six minute toolpath, split into duty cycle and speed control settings on the VESC.

I like the speed control but it is a little rougher. PID settings are stock and seem to work pretty decently, rpm looked to be within 100 of target. If anyone has any tips on tweaking it I’m all ears.

All the vids are shaky cam because the DSLR was dropped last week and is in repair. I have the CSV log files as well, but playback on the VESC Tools app isn’t that great as its meant for an Eskate with gps data. Even running without a fan the temps of the router body were pretty good. Its cooling down now and I might have some stainless to test for low rpm.

Also would like some tips @gmack on how to setup the spindle/router table in the workbook for accurate estimates.

8 Likes

Nice! Have you tuned any of the settings since you sent me your config?

I might have changed a few things, not really sure. I’ll email the current motor confi

Since you guys are a hard bunch to impress. Chucked a Titanium plate into the Nomad and ran it.

According to @gmack workbook, measured K factor was 0.4, please let me know if I did that wrong. In hindsight I should have dropped the rpm to 4,000 to net a 262 sfm instead of the 1/4 running 5k and 327sfm. Little to no heat could be felt around the router body, would be nice to get that sensor.

AVG Sample rate - 25.45hz

9 Likes

Who said we were not impressed. LOL. I just wanted to see this thing sustain a lower than stock RPM while doing actual work. Ill be joining you soon. I have been making noobie mistakes making a waste board.

2 Likes

It looks like you did it right to me. HSMAdvisor apparently uses 0.5 cu-in/min/HP for it’s toughest grade of Titanium. But K-factors are likely dependent on cutting parameters, conditions, and tools. That’s why it makes sense to keep track of the details (like the workbook enables). It’s easy to add sheets to reflect details of your latest cuts.

What does Lakeshore recommend for that endmill? Guhring has some that will support 430 SFM with Titanium. As you know, higher speeds take less force. IMO a higher speed spindle on the Nomad is a good move!

Maybe you don’t need to water cool it after all (little power = little heat)
Can the software provide a “real time” display of motor torque? Is it in the .csv file? What all is in it? Nice to be able to see stuff in “real time” and get some more useful “real world” material K-factors! :star_struck:

2 Likes

I’m definitely going to have to learn how to use your workbook better and start logging things.

Lakeshore actual recommends a 100-250 sfm for soft Titanium and 80-220 for hard with a 0.0004-0.0009 chipload for that general purpose endmill. They also recommend coolant. Titanium is a game of heat management and when your doc is small then things get even tougher.

It will need some type of cooling with heavier cutting and more mrr. Fanless was just a way to force myself to make a custom body.

The Vesc Tool software has a 0-95% duty cycle gauge that updates realtime. That gives a pretty good idea of what kind of power you’re seeing. It would be nice to figure out how to smooth the data a bit. I’ll get back to you on how the csv file is configured but I do have it saved for that cut.

3 Likes

In case more data is needed…

Real time vibration monitoring for $13, sold

3 Likes

This might make development easier: SparkFun IMU Breakout - MPU-9250

Invensense says, “Note: The MPU9250 is Not Recommended for New Designs” That must mean that this model has a newer version.

2 Likes

The VESC has three different compatible external imu options and I liked that one because it has a nice plug for i2c connection.

The sample rate looks to be pretty high due to the fact that usually these are used for self balancing applications like a diy One Wheel.

1 Like

Wow, it’s a lot more than just an ESC.

1 Like

Case bottom revision 1:

Revision 2 is already printing.

3 Likes

So I just connected my arduino to the PWM pin and tried to read the duty cycle. The pulse length is coming up as everything from 0 microseconds to 82043 microseconds. When I take the same code and attach it to the PWM out of another arduino I get a solid 500 microseconds +/- 3 microseconds.

I am not sure if I am getting a shit ton of interference from somewhere or what.

Figured it out. Had the unshielded PWM wire running next to the X and Z stepper motors. They were giving off a shit ton of interference. Looks like I need to use shielded wire. Just means I need to rip some USB cables apart.

In other news the Carbide board’s duty cycle is not linear which means I need to go back to an RPM map.

So, that’s not a nice metric tonne of interference then? :wink:

2 Likes

Ugh… This is going to take MUCH longer than I thought. The stable version of the UART control library is for firmware version 3.40. The development version of the UART control library is for the firmware version 4.2. The version of VESC we have installed is version 5.1. There were major changes between them all.

I can implement PWM to PPM on the Arduino board, but that does not allow for setting the RPM precisely.

I can’t help but wonder what’s the difference in driving a vfd to accurate rpms. Can we do ppm to pid speed control?

At least we know it works well with USB connection.

Just picked up an Uno to play around with. This arduino stuff is pretty interesting.