Professor Builds Huge Table Saw With Computer-controlled Fence and Blade Height

Here are some more technical details about my table saw:
Table size: 242 cm x 120 cm (95-3/16" x 47-1/4")
Maximum rip cut: 32 inches with the 1,000 mm lead screw used in the video. With the longest lead screw that I know of, which is 1,200 mm long, a rip cut of close to 40 inches should be possible. However, the speed of rotation of such a long lead screw would have to be reduced to maybe half of the current speed, to avoid wild oscillations of the lead screw.
Maximum blade extension above tabletop: 6 cm (2-3/8")
KEY COMMERCIAL COMPONENTS
1. OpenBuilds V-Slot NEMA 23 Linear Actuator Bundle (with 1,000 mm Lead Screw)
openbuildspart...
2. Teensy 4.1 USB-based microcontroller development board
www.pjrc.com/s...
3. RoboRemo App for Android phones
roboremo.app/
4. DEWALT Circular Saw DWE575SB
www.dewalt.com...
-------- Acknowledgement -----------
Special thanks to luni64 who wrote the fantastic Teensy 4 library "TeensyStep4." Without that library my application would not have been possible. Thank you for your great work!

[@mdude7778]

Beautiful work. Love the look of that table. Steppers accelerating and decelerating are music to my ears.

[@johannborenstein]

"Steppers accelerating and decelerating are music to my ears" - My sentiment exactly!

[@dale7561]

Most German table saw i have ever seen!
Amazing, I wanted to do something similar with my combination table saw and planner.

[@jonasgranlund4427]

Really exciting Project, I can imagine that it was a lot of work to get the software and hardware to play nice? I have worked with software development for the last 25 years and for a few short times I have developed against Raspberry Pi sensors/Actuators and every time I have struggles with firmware and/or communication and precision. This is a fantastic build! Hats off! :)

[@jonasgranlund4427]

I must also add that if you make the part that goes into the fence longer i.e 500 mm you could probbably adjust the fence + - 45° or so by only moving one side. It is already absolutely amazing but with some math and change in code I'm sure you could do this implementation :)

[@verybigheart]

Superb achievement. Well done. How long did it take to build?

[@johannborenstein]

Total build time including designing and building the table, designing and building the electronics, and writing the software was about 4.5 months with an average of 6 hours of work per day.

[@jonasgranlund4427]

Hi again, You really got me to start up my research and coding interest within sensors and stepper motors again, thank you very much for that! My project is a lot easier than yours, I will make the sawblade on my band saw mill to move into height position with a Nema 34 Motor combined with a PX86 Planetary Reducer, I think the motor will be strong enough to lift the saw-head if I use a 5:1 gear ratio but I need to do some more calculations since the Torque seem to lower quite a bit at higher RPM so I might need to go up towards 10:1. right now it is a manual handle where each revolution lift the sawblade 1/4 inch and it has 20 holes to choose from, so persison on the manual part is 1/80 inch steps. So I get close to mm and I often miss a hole here and there when I saw planks so they have a small size difference so I really look forward to handle this as my winter project til I go to cabin next spring :) If it is possible to buy the code for accelerate, decelerate the stepper motor from you I would love to buy that, it sounds absolutely fantastic when the fence start a little longer travel in how the increase and decrease of speed is handled. Thank you for the inspiration (and late nights) :)

[@johannborenstein]

You don't need to buy that code from me, my code uses a stepper motor library that someone with far better programming skills than myself posted. The library is specific for the Teensy 4 (as well as Teensy 4.1). Look at the bottom of the description page of my video where I thank that programmer who goes by "luni64." The library is called "TeensyStep4." I spent a great amount of time looking at all kinds of stepper motor libraries and this one is way, way better than libraries for Arduino microprocessors because it allows for much faster pulse rates than other stepper libraries.

[@jonasgranlund4427]

@@johannborenstein Thank you so much for tthe explanation, I have added that program to my solution but there is the setMaxSpeed and SetAccelleration on that library that I am unsure of how much it should be to make it work well, I have moved out some of these to a config stored on SD card as I thought now but Im not sure of what values to start with . But I wioll look through your video more to catch up with more details also and get more feeling for how fast you are running the motor, but the first 30 inch move just sounds so perfect in acceleration and speed :)
stepsPerMillimeter = (config.stepsPerRevolution * config.microstepping);
motor1.setMaxSpeed(config.maxSpeed * stepsPerMillimeter); // steps per second
motor1.setAcceleration(config.acceleration * stepsPerMillimeter); // steps per second^2

[@johannborenstein]

My settings for the blade stepper motor and the two fence stepper motors are these (including my in-line comments, for what they are worth):
blade.setMaxSpeed(8000); // worked fine with blade motor at 8000. // was 10'000
blade.setAcceleration(2000); // worked fine with blade motor at 2000. //was 50'000
fenceFront.setMaxSpeed(10000); // was 16,000 at 800 spr was 1500 at 200 spr worked fine with blade motor at 8000. // was 10'000
fenceFront.setAcceleration(4000); // was 1000 at 200 spr worked fine with blade motor at 2000. //was 50'000
fenceRear.setMaxSpeed(10000); // was 1500 at 200 spr worked fine with blade motor at 8000. // was 10'000
fenceRear.setAcceleration(4000); // was 1000 at 200 spr with NEMA 23 motors. worked fine with with rotateAsync(10000) command. //was 50'000
When my in-line comments say "was" it means that I had used that setting before.

[@jonasgranlund4427]

@@johannborenstein Hi Johann, thank you soo much for these values, it will make me get into a good place much much faster regarding the travel. I am still working on basic functionality and will get my Nema 34 with PX 86 5:1 reducer in 2 weeks to try some in the real world.

[@merk3574]

Is the software allowing for the blade kerf? It appeared when you measured the 30" with your tape you measured off the blade and not the teeth, it's possible that is why the first dado cut was a hair off.

[@johannborenstein]

In the scene where I demo the dado cut I make the first cut which leaves a dado of the width of the kerf, namely about 2 mm. Then I make 3 more cuts, each one offset from the previous cut by 2 mm. The result is a dado of 8 mm width, which is made up of the first cut (2 mm kerf) + 3 x 2 mm offset = 8 mm. The kerf is relevant only in that scenario of cutting a dado. Most other cutting tasks only call for measuring the distance between the side of the blade that faces the fence and the fence. The kerf is not of relevance in those cuts.