Why use a 8 mm lead screw?

Why is everyone using a 8mm lead screw i.e. 8mm movement per rotation? The advantage 8mm lead is faster Z movement. But the huge drawback of a 8mm lead screw is that if the steppers are not powered, the z axis drops down. The mill bits can then break or damage the work piece. I am seeing a lot of parking frameworks just to get around this issue.

I am using a 2mm lead screw (2mm movement per rotation) for my lowrider2, and that works very well. The Z axis stays in place, even with unpowered steppers. And the Z movements are still fast enough.

So why use a 8mm lead?



That’s interesting. Is there a significant price difference?

You’ve got about 3200 steps/mm then? We used to use an allthread 5/16" screw for the mpcnc and it had about 4500 steps/mm and there were a lot of problems with the 8 bit micro crashing if it went too fast. Marlin has gotten a lot better in that respect. Also, at 12V, the steppers don’t have as much torque when going faster, ao you’re closer to losing steps. Those are my two guesses.

With a 24V or just limiting it more in Marlin, that seems like a very food option.

I use marlin on a ramps1.4 with A4988 stepper drivers in 16x microstep mode on 12v. It uses then 1600 steps per mm. The Z feed rate is limited to 15mm/sec maximal. 20mm/sec also worked. At 30mm/sec the stepper started loosing steps. Possible even that could have worked if I had turned the stepper current higher.

Because the movement is 2mm instead of 8mm per rotation, the torque required on the stepper for a 2 mm lead is thus only 1/4 of what it would have needed on 8mm lead. In other words the available force on the z axis is 4x higher.

At 15mm/sec I tried to stall the stepper by blocking the z movement, but that was almost impossible.

1 Like

The trouble is that without more voltage, the driver actually can’t reach the current limit before it steps again. There is a limit there, and it’s in rev/s on the motor, and if you exceed it, your torque will drop significantly.

You make a good point about the increased leverage though. IDK what at what speed with 12V the torque would be reduced by 4x.

I have chosen to run my steppers at 24v with the SKR 1.2 and it seems like lowering the lead on the a screws so that the gantry will not fall with the steppers off would prevent a whole host of problems such as the gantry falling and burying the collet in the work.

It seems like it would also let the z stay parallel without constantly having to retram or setup dual end stops on the z. My 3D printer gantry is a dual z lead screw and I rarely check to make sure the gantry is trammed to the printer base because the steppers just don’t drift apart.

You could use dual shaft steppers and put a hand wheel (finger wheel) on the top for easy adjustment and with 24v and the extra advantage from the finer pitch on the screw you might not need to give up any z speed. I’m not sure if a 4mm lead or a 2mm lead would be the way to go here. I’ll probably try a 2mm lead unless there is a good reason not to.

I think that is reasonable. 24V and a 1 start screw seems like it would work well. But I would want someone to do some “bench press” tests to see that it can still move as fast, or what the difference is.

I also think that the self-locking with a 2mm slope is high enough to keep the slide in z. I am with the
z axis of the lowrider not satisfied either, so ordered an 8x2 lead screw.

thanks for the suggestion

I have also ordered a pair of 8x2 lead screws and will report back my findings once I’ve got some time on them.

If you don’t have a 24V power supply, you will definitely need to reduce the top speed to about 3-4mm/s.

i ve it, thx :wink:

so, my lead screw Tr8x2 is there. works perfect.
no more sagging of the z-axis (if machine is off)

1 Like

I also changed to a Tr8x2 1start and it works perfekt for me, too. No more need to rest the gantry on blocks - what a win! :+1:

The only thing you have to mention is the loss of maximum travel speed of your Z axis. I’m using a 12V power supply and the DRV8825 driver is running on 1/4 step jumper.
At 480 mm/min it starts loosing steps, so I set the max. to 360 mm/min to be on the safe side.

My question is - would a 24V power supply enable higher speeds? If yes, could someone estimate how much in %?
For me, the Z axes doesn’t need to be faster, but it would be nice to know if there were possibilites for an upgrade in case.

1 Like

In theory, yes. Probably more than double. Sorry, this won’t be perfectly clear, and it probably isn’t perfectly accurate.

The steppers produce current when they move. You might have seen the machine come to life when you move the gantry quickly with the machine unplugged. They do this when powered too. That current goes against the power supply and can be thought of as a voltage.

If you have a 12V power supply, and there is some loss due to wiring and connectors and driver stuff, maybe 9V is reaching the coil.

If the motor is stationary, it isn’t creating any of that backward voltage, so 9V is causing a big current in the motor. Maybe 2A. The driver is measuring the current going to the motor, and it is set to 0.9A (or something). So it turns the voltage on and off quickly (by adjusting the duty cycle of the PWM) and it means that the 9V is maybe only going to the motors 40% of the time.

If the motor is moving (this is dependent on the rpms of the motor shaft, not anything that is attached to it), then it might be fighting that voltage with 6V. At that point, the effective voltage going to the coil from the power supply is only 9V-6V=3V. That would (in our imaginary motor) make the current (when on) 2A/3, or 0.6A. The driver then leaves the power on constantly, but it can’t make any more power, so the motor acts like the driver current setting is 0.6A. That results in a lot less torque, and the motor can skip steps. The voltage generated by the motors is doubled when wired in series vs. separate drivers.

If you had a 24V PSU, then you might get 21V at the coil when stationary, and 15V when moving at that same speed. So instead of the driver getting effectively 3V to run the motor, it has 15V.

That is pretty convincing, I know, but I should also point out that if the driver is working at a lower speed, and the voltage generated by the motor is smaller, then the driver can still reach the 0.9A set point, and there is absolutely no difference between the power supplies.

This is the mental model I have in my head. It leads me to these conclusions:

  • A higher voltage power supply does not lead to more power at low speeds. We need the most power when milling (not traveling). A 24V PSU is not useful for most CNC machines.
  • A lower voltage power supply will reduce torque at higher speeds.
  • A higher voltage power supply is about raising the ceiling on speeds where you can still have strong torque.

You can calculate the point where the drivers are in saturation. You need to find the number in the stepper motor spec sheet that shows how much voltage it generates per rpm. You need to calculate the voltage drop up to the point of the motors (or make a good guess). You need to determine the voltage needed to reach the set point current given the resistance of the motor. So if the motor is 5Ohms per coil, and your set point is 0.9A, then you will saturate at 0.9A*5Ohms=4.5V. So if your power supply is providing 9V to the coils, then you need 4.5V to drive the motors at full current, then you can have up to 4.5V feedback from the motors before you saturate. That will give you a max speed at the motor shaft, which you’ll have to convert to mm using the geometry of the machine.

It is also possible to drive the machine past the saturation point. You’ll get most of the torque you asked for. So if you’re skipping steps at 480mm/s, you might reach saturation at 360mm/s. I don’t really know.


Thanks Jeff for your very detailed answer! I highly appreciate that :+1:
Very good explanation. Even a mechanics guy as me understands that :grin:

I don’t have issues at lower speeds, only at high speeds…so it could really make sense to upgrade ro 24V if I want to go faster at G0 movements

1 Like

Just did this upgrade and it’s really a game-changer. 2mm lead screw and a 24v power supply… everything is so much smoother and the Z just sticks in place. I wish I’d done this from the start, as it would have saved a fair amount of time and frustration. IMO this should be the default configuration: let ppl opt for 8mm only if they want (theoretically) faster Z movement.


I did this upgrade (24VDC PSU, 1 start T8 leadscrew), and I agree with @jgunnar ! Makes a huge difference vs 1 start 12VDC PSU. I am running X50, Y50, and Z15 (mm/s) with no issues via M203 and saved to EEPROM. My “bench press” test involves G00 X0 Y0 Z70 F3500 / G00 X234 Y234 Z0 F3500 to max out all 3 axes and repeating 15 or 20 times via script. It works well, no skipping steps. Does anyone know of a better way of testing?

EDIT: Just noticed I posted in the LR section instead of MPCNC. My apologies…,

If you send an xyz command like that, the z will probably move slower than a pure z move. So testing would be more on the redline with a z only up and down and back type routine. Maybe do some reps with increasing z acceleration to find the limit.

1 Like

Yep. It will scale so that only one axis is maxed, unless you have all three scaled perfectly.

You should do it on each axis one at a time, but doing a multi axis test won’t hurt, it just may not be the end of the tests.

You should also tune the max speed first (and I like to stay away from the fastest to be conservative) and then tune acceleration. Jist make sure your speed tests are far enough and your accels are high enough to actually reach the speeds you are trying to test.

I ran a couple of short projects today without a hitch and no more dropping of the Z axis when the stepper is off so I am happy. 15mm/s on the Z is sweet icing on the cake (I had it set at 8mm/s in FW previously, based on what I read here. I can’t honestly say that I tried 15mm/s with the 4 start T8 and 12VDC). The only downside is that I can no longer raise or lower the Z by hand…a small sacrifice.

Anyway, I think a 1 start T8 with a 24VDC PSU is a great fix for dropping Zs and if Z will run faster to boot, all the better.

EDIT: While speed is not my goal, I did try to scale my “bench press”. At 50mm/s, XY travels 3x as fast as Z at 15mm/s, so for a 70mm travel on Z, in theory XY would need to travel 3x as far for all 3 axes to arrive at the same place at the same time…hence the X234 Y234 Z70 travel.