Size limits to fdm print jobs?

[Dave]

Have you priced single-start leadscrews in the size required for the FT-5? All you have to do is rotate one leadscrew by hand to see just how far the bed moves for each turn... (It's a surprising amount). With the stock screws, it really leans on the motor precision and holding torque at micro-steps to keep those layer heights steady... If single-start screws were affordable enough, I'd definitely consider it as an upgrade.

You could go pretty crazy with the pulley/belt ratios too if needed. For example, a 60T pulley on the leadscrew and a 16T pulley on the motor... (3.75:1). It looks like 60, 72, 80, 90, 100, and 120T are the common sizes.

[Ron]

Yes, you can get single start 8 mm diameter lead screws all day long.
If you don't mind a little shipping delay, here are a pair of them (including nuts) for $12.99 including the freight.

2 Sets T8 Lead Screw 2/8mm Trapezoidal Spindle with Brass Copper Nut 3D Printer | eBay

But they are also available domestically for only a little more.
It isn't much of a monetary investment.
In fact when I eventually do build a scrap box mini printer I will probably buy a pair of single start lead screws in the 500mm length and try them out in the FT-5 just for kicks.
Then if I want to I can put the old 4 start ones in the new printer (suitably reduced in length of course).

Be a bit careful on the pulley ratios though.
Remember that there are practical limits as to how fast steppers can be driven, even if you are sourcing them from 24 volts.

I suspect that the effect that you referred to on the website is more an artifact of the "springiness" of micro-stepped drive techniques than anything else.
If you set up your lead screws, pulley ratios and basic motor resolution so that you have a native full step resolution of 0.0025 mm as discussed above, it would naturally align on the motor's pole positions with almost any common layer height .
That might be just the ticket.

What is not generally appreciated is that microstepping is intended mainly to reduce noise and vibration (by making the motions more continuous).
It is not as useful for increasing resolution in any application where residual torque or load remains.
That would certainly be the case for the Z motion.

[PaulS]

you are the man! @Ron thank you much for breaking it down for me. exactly what i was after!

[Dave]

If you set up your lead screws, pulley ratios and basic motor resolution so that you have a native full step resolution of 0.0025 mm as discussed above, it would naturally align on the motor's pole positions with almost any common layer height .That might be just the ticket.

I have to say, that's appealing. With that much of a ratio, and with the harder-to-backdrive single-start screws, and with the motors being driven only in full-steps, you should have an enormous amount of static holding capacity. At a minimum, you could reduce the Z-motor current (or use the power saving when stationary modes of the stepper drivers)... but you could probably also get away with a smaller motor if you wanted.

I'm going to need to pull my Z-axis apart at some point in the future anyway to switch to a single motor / single belt and remove the top bearings..... hmmm...

[Ron]

Point well taken.
You actually could drive the motor(s) in full step mode.
But you might not want to because I imagine that it would become too noisy.
You can of course adjust the jumpers on the base board underneath the step stick from 1/16 steps to perhaps run 1/4 steps.
That way it should still remain reasonably quiet.

The anti-back-driving mechanical advantage is a potential plus too.
But some caution is required as it also means that there would be a lot of force available.
So it will become very critical that the Z home switch and/or other platform height sensor(s) be functional, preferably multiple redundant ones.
Otherwise things could actually be bent or broken.
Oh, and of course, keep wayward hands and fingers out of harm's way because pinch points will become much more of an issue.
However if the motor current was suitably reduced it might not be much if any less safe (self-limiting) than it is now.

Something to ponder.
I do want to experiment with this myself and am eyeing some 2mm pitch leadscrews as part of another group of parts I am putting together.

So we do not upset the original topic too much, perhaps whoever actually does it first should start another thread on this subject specifically?

[csorrows]

So we do not upset the original topic too much, perhaps whoever actually does it first should start another thread on this subject specifically?

I know I am a stickler for staying OT but I'm much more relaxed about it in the general category. I look at it as your guy's chat and just let it go wherever.

Carry on gents.

[Ron]

Thanks.
Much appreciated.

Actually I recently ordered some parts related to the discussion and should receive them this coming week.
I chose to take a slightly different path initially just to see what happens.
I will be putting 40 tooth pulleys and anti-backlash nuts on the original lead screws.
That will make the native full step vertical resolution 0.02 mm per step.
That works nicely with 0.1 mm layer heights (5 full steps), 0.2 mm (10 full steps) and so forth.
Even a 0.25 mm layer height aligns with half step increments neatly.
Half steps are fairly "rigid" as compared to microsteps and the motor speed, while doubled, should not be excessive.
The original resolution was of course 0.04 mm per full step which did not align as well with many commonly used layer heights such as 0.25 mm (6.25 full steps) or 0.35 mm (8.75 full steps).
The anti-backlash nuts should help "settle" the bed at each step (for lack of a better term), particularly with the small vertical motions that can be made within layers.
I will mount them with the spring-loaded portions toward the top so that they will assist gravity in pushing the bed downward against the lead screw.

It will be interesting to see if these changes make any discernable differences.
Then later on I will probably try the 2 mm pitch lead screws too (also with anti-backlash nuts).

Thanks again.

[csorrows]

I removed the "on topic" part of the description. I see no reason for that to be in there.

[Ron]

OK, I installed the 40 tooth lead screw pulleys and anti-backlash nuts last night.
I am still using the regular 4-lead (8mm pitch) lead screws.
This results in a resolution of 0.02mm per full step or 0.00125mm per microstep (for reference only).
I changed the firmware settings to 800 steps per mm for the Z axis of course, but changed nothing else.
I put a thin smear of THK grease on the lead screws as I have done previously.

First Impressions-
1. The Z motion is both smoother and quieter
2. A test print with 0.2mm layers turned out considerably nicer than I am accustomed to with very consistent layers
3. An ABS test print using 0.15mm layers turned surprisingly nicely and required almost no sanding and vapor polishing to become positively glossy (better than normal)

To put that into perspective, I have never had a great deal of luck using layers thinner than 0.2mm in the past.
Now I feel confident that I can probably run 0.1mm layers or even smaller and may try them later today or tomorrow.

So these appear to be a definite improvement as far as I am concerned.
How much of that is due to the new 2:1 pulley reduction ratio or to the anti-backlash nuts is a matter for conjecture, but regardless it is noticeably better.

FWIW.......

Update-

I did a 0.1mm layer height print over the weekend and it turned out excellent.
I may try going even lower in layer height.
I am not sure what the practical limit will be, but it was a tiny part (a connector clip) and the layers were as even as could be.

[PaulS]

just ordered almost $600 in linear rails for new machine. build volume will be right around 5 CUBIC FEET! Yes you read that right!!!

[Ron]

Whoa. Now that is a serious printer indeed.
Ooops, got a Pestogram again.
It tells me I am hogging the thread.
Please keep us posted.

[csorrows]

Ooops, got a Pestogram again.It tells me I am hogging the thread.

I just scrolled up. Yup, ya sure are. LOL

I am working on a secret project that includes about a 3 cubic foot build volume. I thought that was big... Keep us up to date!

[PaulS]

with all these great upgrade \ design idea floating around, i cant help but wonder when John, i mean @csorrows i mean John will come out with the next model printer...

[PaulS]

considering the length of the drive belts (11 ish feet) i will need, anyone know if something bigger or better than G2 belts and pulleys are available? with that length, i am worried about stretch and sag. any thoughts or have i over thought it already?

[csorrows]

Look into authentic Gates belts. Much higher quality and a lot less stretch than the Chinese crap in the kits.

Next kit is already in the works. This one will probably take awhile. It's somewhat uncommon at the price point FT sells their kits at. Unfortunately, I can't say anything except for it's in the very early stages of development.

[PaulS]

Ask a simple question get a complicated answer... LOL thanks, looks like i have tons of options, still waiting to get x\z gantry together so i have better estimate of lengths but great site there @csorrows like i said, we should have a list of things like this...

[PaulS]

ok so back to how the hell am i going to heat a 5 foot by 5 foot bed??? anyone got any more ideas? since bed will be stationary maybe radiant heating or something may be a good idea thanks for that btw!. I just plugged my ft5 with 750 watt gel pad into a 400 watt UPS and verified it could not handle the load. funny to watch alarm trigger every time heater came on LOL but this give me great concern about practical heating without having to have my own nuke plant to run the heat bed. I am guessing chamber may be necessary. anyone have any experience or ideas on this kind of build volume 5x5x5 feet... printing abs for sure and hopefully other stuff too.

[carefudundae]

Maybe you could use something like a floor heater? I installed electric heater wire (and thermistor) inside the thinset before laying our kitchen tile down. It's spaced out for much lower heat fluxes than you'd want (don't need our floor to reach 100C!), but it may work for your needs. I used a specific controller for keeping the current/temp regulated, but you could use an SSR. Could just lay it in thinset/concrete or even the floor leveling stuff that is very liquidy before setting up. If you build a form to dump it in, you could end up with a heated and possibly very flat bed. Just an idea...
Another option would be to use hot water tubing from your water heater, but that won't get to ABS bed temps...

[PaulS]

rails are in. big box is 1540 mm hiwin for x axis other boxes are 6 x 1500 mm (2xZ and 4XY) note yellow yard stick for comparison LOL

[csorrows]

OMG, those are huge. I bet those cost you a few dollars....