Welcome! Log In Create A New Profile

Advanced

Can a poorly designed cheap machine be 'corrected' by software ?

Posted by MKSA 
Can a poorly designed cheap machine be 'corrected' by software ?
April 29, 2019 10:42AM
An interesting analysis that bears many similitude with a lot of 3D printers.
OK, far less risky, eventually just burn your house ?

[medium.com]

Edited 1 time(s). Last edit at 04/29/2019 10:47AM by MKSA.


"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)
Re: Can a poorly designed cheap machine be 'corrected' by software ?
April 29, 2019 11:13AM
The Prusa I3 is a perfect example of poor hardware being compensated for by great firmware and tweaking the hell out of the slicer.


Son of MegaMax 3D printer: [www.instructables.com]
Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Re: Can a poorly designed cheap machine be 'corrected' by software ?
April 29, 2019 12:42PM
Thank you for a pointer to a very good and detailed background to a set of engineering compromises.

Engineering itself is very much the art of the compromise - but in business maximising profit is also the art of the compromise. When you have this mixture, already fraught with risk, mixed with other "arts of compromise" like government regulation, campaigning journalism etc., it is amazing how far into the 21st century we have made it.

On 3D printers and optimisation by computer, my own opinion is that a more can be done by using engineering principles - increasing stiffness and compliance in the right place to improve rigidity, reducing weight where it only contributes to moving mass, increasing compliance where this will absorb unwanted vibrations etc.. Once a good product has been achieved than getting it as close to perfection as possible with inexpensive computer power is a very good thing.

On the other hand, designing a product with little good engineering and then making it right with the same computer power is just "polishing a turd"

Mike
Re: Can a poorly designed cheap machine be 'corrected' by software ?
April 29, 2019 01:03PM
Within limits, we do this all of the time, but there are limits.

Mesh bed compensation is a prime example of this. We compensate for a build platform that isn't flat and level with software. We cam compensate for a machine that doesn't hold true to the X and Y axis not being perpendicular to each other with software.

Some things we can't do, or maybe we could for one-of cases, but it's cheaper and easier to fix the hardware for things like Z wobble.

I've seen CNC machines with a compensation routine for backlash on its axes, where it basically always makes sure that the axis has been pushed in the same direction last, no matter which direction it was moving before that. This could benefit more than a few cheaply made 3D printers as well.

Some things can be compensated for by software, where the results are consistent enough to work well if you always follow a set of rules to compensate for, but the less of these that we need, the better our final results can be.


MBot3D Printer
MakerBot clone Kit from Amazon
Added heated bed.

Leadscrew self-built printer (in progress)
Duet Wifi, Precision Piezo parts
Re: Can a poorly designed cheap machine be 'corrected' by software ?
April 29, 2019 02:25PM
One slightly worrying trend that I have noticed is that in some instances "polishing a turd" as Mike so aptly puts it, has not only become acceptable but some people are deliberately removing parts in order to turn machines into "turds" so that they can then be polished. confused smiley

I'm talking about sensor less homing. This is something that Mr Prusa introduced for one reason only, and that is save a few cents on the cost of production by not fitting a cheap as chips micro-switch. People are buying kits or cheap machine, and upgrading the electronics. There is nothing wrong with doing that, but what's crazy is that in some instances, they are removing or bypassing the already fitted switches, so that they can use firmware to badly do the job that the simple switch does well.

Instead of using a switch, let's just shove the gantry hard up against the frame and detect when the motor stalls. Why would you want to do that for any reason other than to save the cost of a switch? It can sort of be made to work if you carefully select the stepper motor, then fine tune the firmware to work with that particular stepper motor and driver. But it will never be as reliable, consistent, or precise as a simple switch. But many people want to do it, even when they already have the switches, for no other reason than that's how Prusa does it.


[somei3deas.wordpress.com]
Re: Can a poorly designed cheap machine be 'corrected' by software ?
May 03, 2019 03:02AM
Re: Can a poorly designed cheap machine be 'corrected' by software ?
June 27, 2019 07:36PM
Many people easily forget about the precious time they will have to waste.
Sorry, only registered users may post in this forum.

Click here to login