Welcome! Log In Create A New Profile

Advanced

Just how easily should the axes move...

Posted by davidgoodenough 
Just how easily should the axes move...
October 13, 2008 01:48PM
I have a Darwin built using the original Bits from Bytes laser cut parts.

I am trying to run the reprap host tools working volume probe, and sometimes the motors slip. I have tried to rub a little oil onto the rods, and that has helped a bit. I have also filed the shaft a bit to give a flat surface to grip the grub screws to, but they are still slipping.

Also the stepper motor drivers are getting very warm (the chip with the heatsink on it), rather too warm for my comfort.

So I tried moving the axes by hand, and I can do it (with the motor disconnected from the driver) but not easily. The think is I do not know how easy it should be.

Anyone got any thoughts?

David
Re: Just how easily should the axes move...
October 13, 2008 03:03PM
I can't say anything about the BfB version, but the standard design moves easily by twisting the pulleys between thumb and forefinger. I have put a couple of old fashioned radio knobs on the unused motor shafts to make manual positioning easy.

It is normal for the heatsinks to get too hot to touch.


[www.hydraraptor.blogspot.com]
Re: Just how easily should the axes move...
October 14, 2008 03:28AM
Sounds kind of like the problem I the chips on the stepper driver were overheating. Before I found that out I thought it was getting stuck as well. I took apart the X and Y axis at least twice and rebuilt them. that never changed a thing. Practical dipped the whole thing in vat of Wd-40. I want to say with about 80% certainty that you are having the same problem as me. If you want to check measure the temperature of the heatsink and integrated circuit. The IC is only rated to 130C so if you are getting temps around 127 and up it is overheating.(in my case when idle, powered on but not printing, it was at 129C. For me when I told it to print and it started skipping all I had to do was blow air lightly on the heatsink and it would suddenly stop skipping.....only to start up 30 seconds later.

For me I kind of went overboard. [objects.reprap.org] You may only need some thermal paste or just a heatsink with more surface area. I was kind of pissed off at having the problem plague me for over 2 week. ( It was a long 2 weeks of trying everything I could think of but checking for overheating.)
Re: Just how easily should the axes move...
October 15, 2008 02:56AM
It should be possible to push the extruder carriage with your finger in both the X and Y direction with the stepper motors connected without needing to hold the main frame to stop it moving (motors not powered!) i.e. very easy to move. If they are not setup like this you will struggle to get good movement.

We have just put up some more instructions that give details on ensuring the X carriage is built correctly. It takes upto 3 hours to build a perfectly aligned X carriage. But it is worth the effort as you get a rigid machine with very smooth operation. I will put up some more info soon on doing the same with the Y axis but it is all down to acheiving good alignment and getting the parts square.


Ian
[www.bitsfrombytes.com]
Re: Just how easily should the axes move...
October 21, 2008 10:39PM
Guys

Is this design just a little too hard?

After managing to build it I now have to unbuild it to get the alignment right.
and then rebuild it again. With no guarantee of success.

It's enough to break your heart.

Surely there is an easier simpler way?

Stephen
Re: Just how easily should the axes move...
October 22, 2008 07:59AM
Accurate alignment and tight tolerances are a must for accurate printing.

That being said, there is a very high probability that we can improve on the method.

perhaps the instructions should be more "aggressive" when they specify tolerances, for example:

"It is absolutely critical to have this distance accurate to within 0.5mm, otherwise you will have to re-do it later on"

or something of that sort.

-Leav
Re: Just how easily should the axes move...
October 22, 2008 08:14AM
With the original Darwin design I don't think there are any critical measurements. That is the clever thing about it.

I just used a piece of the diagonal studding with two diagonal tie brackets on it to make sure all the distance that should be equal, are equal. Everything moved freely on first build. I did change the z-axis design though as I have heard people have problems wth that aspect.


[www.hydraraptor.blogspot.com]
Re: Just how easily should the axes move...
October 22, 2008 08:33AM
The Z axis is the one we are currently having problems with. When I do not have the problem I have described in the electronics forum where all the motors fire together, I can get the X and Y axis to move properly, but the Z axis just sits there stuttering. I can move it by hand, although getting a grip on the bit just above the motor is a bit difficult due to its position.

I have seen the change you made to the Z end stops on your blog, but I am not sure I went back far enough - I could not find that you had changed or whether it was relevant to the original BfB laser cut parts.
Re: Just how easily should the axes move...
October 22, 2008 08:50AM
My deviation is here: [hydraraptor.blogspot.com]

It use four smaller motors instead of one big one. Not relevant to your current problems.

Try unplugging the X and Y motors so that the software/firmware thinks it is driving all three. If z still stutters it would imply a software or firmware bug. Otherwise it fit runs it is definitely a hardware problem.

Perhaps the step rate is a bit high, the current a but low, or your PSU can't drive all three.


[www.hydraraptor.blogspot.com]
Re: Just how easily should the axes move...
October 22, 2008 09:47AM
> "It is absolutely critical to have this distance accurate to within 0.5mm, otherwise you will have to re-do it later on"

Where there are dimensions they tend to have +/-0.Nmm after them. I found for the longer distances, such as the diagonals, a tolerance of 0.2mm was very hard to achieve. In fact as I don't know anyone who has a 600+mm caliper, I can't really be sure I did.

What I could do though was concentrate on precision [en.wikipedia.org] rather than absolute accuracy. I did this by making a story stick that only just fit between the polls when they were the right distance. The fit needs to be tight enough to produce a certain amount of friction, this ensures there wasn't still a gap and that the stick isn't being crushed.

In the end I can be confident the diagonals are all to within +/-0.5 and are probably very near the +/-0.2 specified, but importantly I know they are the same length to +/-0.1 or less. With the side diagonals the result is that the top may be slightly higher or lower than it should be but it is at least as parallel to the bottom as was intended.

>After managing to build it I now have to unbuild it to get the alignment right.
>and then rebuild it again. With no guarantee of success.

Some modular tests would go a long way, e.g. the X carriage should start to slide when one end of the Y carriage is raised Nmm above the other. It's currently possible to make a mistake right at the start but not find it until the end.

For mine; I've found why it wasn't as free as it should be, everything moved just not easily. My damn set square isn't square! Ironically I purchased it forthe building because I was worried my old set square was too battered and might not be square.
Re: Just how easily should the axes move...
October 22, 2008 11:15AM
Your mention of the current limiter was the clue, we had the potentiometer set wrongly. Adjusted and how it works.

Now to work out why the three motors pulse together when you tell only one to run.
Re: Just how easily should the axes move...
October 23, 2008 10:56PM
Thanks peteredworthy for responding to my whinge.

I agree that a more modular design would be good way to go.
i.e. Build X. Test X. Move on to Y knowing all new problems are in part Y.

I still feel the current design is too hard. Maybe a third world surviver would do a better job than a first world wannabe building this machine. But I am not sure.

However

Surely it would be possible to have long bits of plastic with a hole at each end to put the rods onto so that it would be impossible to have the wrong measurement? Sort of like a permanent story stick.

Eg

A------------------B
|
|
|
|
|
|
C------------------D



Pole A-C would have a distance between the holes of 40cm
Pole C-B would have a hole distance of X (Pythagoras 40*40*2 square root)

This way an idiot could put it together. (that would be me - Grin)
and know that every measurement was exactly correct.

Sure it would cost a little more. Since you would need more parts.
But save some serious cursing. and would not move simply because a
screw was not tight enough. (Arrhr @#[email protected]#[email protected]#$)

Note these bits of plastic do not replace the steel rods that keep the machine a cube under pressure. Just help in the build process.

Ian (from bits from bytes) What do you think?
Sounds like an extra kit you can sell as an add on.

regards

Stephen

p.s.
I never heard of a story stick. Thanks I will be trying that. Seems a lot easier than the tape measure and more accurate
Sorry, only registered users may post in this forum.

Click here to login