I purchased my kit from the website around 2018. I’ve recently updated most of it to the new Primo version which included the T8 lead screw. My setup is 1.8 degree steppers, drv8825 drivers, and 2mm thread pitch T8 lead screw. This comes out to 3200 for 1/32 step and 1600 for 1/16 step.
The issue is when I take Z above 2600 the steppers makes a weird buzzing noise and seems to skip a bunch of steps resulting in the bit wanting to nose dive into the material (luckily I caught it by the time it did). I’m trying to understand why it’s doing this. I’m assuming it’s overloading the Arduino, but I’m not sure.
Is it a 1 start or a 4 start? The 4 start is the “standard” and even though it has 2mm pitch, it is 8mm in one rotation. That is why the firmwares have 400 and 800 steps/mm in the configs.
You didn’t mention what arduino/controller you are using. Is it a ramps/mega?
If you have a 4 start, then you need to reduce your steps per mm to 800/400.
If you have a 1 start, then you need to reduce the top speed to something like 4-5mm/s and increase it, while keeping your movements stable.
If you look at the end of the leadscrew or the nut. You can see if it has 1 or 4 starting threads. You can also tell by looking at the side and judging the slope of the threads, but you have to know what you’re looking for.
I think there is some standard name, like “T8x8 2mm pitch” for the 4 start, because it moves 8mm in one rotation. “T8x2 2mm pitch” for the 1 start.
It isn’t a problem if you get the wrong one. You just need to slow down the Z. A 24V PSU will help make the Z faster (steppers lose torque at higher RPMs, because they run out of usable voltage).
Be aware that every microstepping increment you do, you trade off torque.
Just by going 1/2 microstepping you will lose 30% of the holding torque and gets worse every division factor you increase.
That is incorrect. That site has some issues (at least for me. I can’t see the charts and the equation formatting is weird). But I think the table and the equations are confusing how to calculate how much current to send to each coil, and the increment is how much to change the current for each microstep. So the 5% for 1/32nd microstepping is saying that about 5% of the current moves from coil A to coil B at each 1/32nd microstep. Not that the torque at 1/32nd is 1/20th of full stepping.
Here is a better source. Written by Ed Nisley (who wrote about the MPCNC in the Digital Machinist magazine):
It makes a lot of sense that microstepping would not significantly reduce torque. When the stepper reaches the full steps, it is doing the same thing as a full step driver. In between, it is still sending as much (or slightly more) current to the coils, split between them.
TMC drivers split it even more, interpolating and providing 256 different levels of current. They definitely are not orders of magnitude weaker than drv8825s.
If you were arguing that the final tool will not be accurate within 1/32nd of a full step, and therefore it doesn’t have an improvement over 1/16th, then I would agree with you. But a full stepping driver also has more than 1/32nd step error in the same circumstances. It isn’t any stronger or more rigid to use full steps. There are a lot of very good reasons to go to at least 1/8th steps.
There are clear advantages of using microstepping. Vibration and noise are on the top.
Usually most people opt out for ultra resolution.
But from my personal experience I don’t go beyond the resolution I need. I usually stay within 1/4 to 1/16 microsteping. That allows me to push speeds up a bit more also without the risk of stalling.
I don’t recall well but there are also the problems of current and voltage getting off phase at higher step rates due to the slower rising of voltage.
I also think that’s why using higher voltages yelds better torque (I may be saying something completely wrong). I know “smartter” drivers do some magic to counter act this…I just don’t remeber what cause I studied that years ago.