I’m fairly new too, but I don’t see it as overstepping. By asking questions about potential improvements, we also learn the long road of things overcome to get here
On that note, what about magnet sensors? Let’s say you have a clip or some fixing device to hold 4 magnets to near the 4 corners of the project area. When the sensor magnet attached to the sled finds the home of the magnetic position, it can verify the borders? I know of a few devices that use magnets to define the exact position for the motor to stop, again not sure what level of accuracy that may provide. Of the examples I can think of, they usually only move in one axis so the magnet is always on a vector to meet flush, so maybe it isn’t as beneficial when x and y values are changing simultaneously.
This is entirely open to us figuring out what the best way to do it is. Right now I extend all four belts one at a time and then connect the to the frame (usually top ones first, then the lower ones). There is no reason that we can’t extend more than one at a time, or extend the top ones first, then connect them, then extend the lower ones. Generally I extend more length than is needed to connect to the frame and then
This is exactly how I’ve been doing it. Basically if I don’t want to run calibration and just use the stored one, the belts pull tight to take up the extra slack and do a single point measurement to check to make sure that the calibration is good. It’s just a sanity check to make sure that nothing has gone wrong.
It’s actually quite gentle. You can stop the belt from retracting with a gentle tug. The way it works is that we ramp the motor speed up slowly and using a moving average filter to keep an eye on the current going to the motor. As soon as the current starts to ramp up quickly it stops. The length of the belt which is currently coiled on the spool works like a spring to absorb the energy when it gets to the end so there isn’t really a sudden stop. For example if you got a finger stuck between the belt end and the arm as it was retracting it wouldn’t hurt.
Thanks! It also works the other way around. When the belts are extending it acts like a compliant mechanism even though the motor can’t be back driven (because the gear ratio is too high).
Basically it simulates being compliant in software by monitoring the encoder, when you start to pull the wound up belt compresses and it extends very slightly, when the firmware sees that it turns on the motor and extends more belt. The result is that the belt extends when you pull on it and stops when you stop pulling so it feels like you are pulling out the belt by hand, but really it wouldn’t work if the power was off.
The ability to increase the packing of the coiled belt on a spool is in essence a clutch- the savior of winches everywhere- but using the outboard encoder to identify a free-spooling state and initiate a powered extend is yet more ingeniousness.
If this powered extension were initiated as described and then the tension on the belt were then removed, would the belt pile up on the spool- loosening the packing and not moving through the encoder- and the motion stop, or is it able to push out through the encoder and continue?
A drum with known diameter and an imprinted pattern being rolled over the work area by the maslow - inked… That way you get exact dimensions marked by the machine itself as the drum determines the marks - not the maslow
you have a catch 22, until you calibrate the maslow, you don’t have the ability
to reliably move it in a straight line, so you will be pulling your drum
sideways.
once it’s reliably calibrated, you don’t need the markings.
Hi, I saw the video on YouTube of the horizontal maslow. I have a question, is it normal for the router to oscillate as seen at minute 2:28? Or is it my impression?
Do you mean the wiggle in the motors? We changed the design to support both sides of the motor shaft so they won’t wobble anymore, but I only replaced one on the machine. The motor on the back left has different color power cords and doesn’t wobble, but the other ones still do. I didn’t replace all of them because I didn’t want reprint all the parts and also I’ve been running these motors for about a year now and I want to keep pushing them to see how long they last.
Is that what we’re talking about? I’m not 100% sure I answered the right question
It seems that the router and all the mechanics wobble on the slide when it changes direction, and that the cutter tip moves late with respect to the position of the router.
If you are referring to the wobble of the sled and mechanics, Bar addressed that around 2:30 in the video you linked as well as in some previous updates. It’s a result of the sled being a SLS print instead of polycarbonate like the production sleds will be.
