??!!!?? ARGH Crooked prints

I don't get it. Parts are printing out of square.



We printed out a paper grid and jogged all the way around and through it. It followed the grid perfectly.



Then print a part..... !@#$% CROOKED. Print the same parts using the same stl's on Cupcake and they come out..... SQUARE!!!! Drooly, but square.

A little frustrated.

We printed at 30mm/s and at 55mm/s and got the same result.

Any help would be kindly appreciated.

Edited 1 time(s). Last edit at 12/19/2011 09:29PM by Bruce.

This is, I think, the strangest problem I've heard of. Congrats.

Since the machine follows the lines correctly when jogging, it would seem to indicate that the mechanics and the g-code interpreter (i.e. the firmware) are fine.

What PC software are you using to generate the G-Code from the STL?
Maybe the software you're using to send the g-code to the reprap is helpfully massaging it somehow?

Some software (skienforge, maybe others) produce DXF's that you could look at and check for squareness. There are also g-code viewers that could be used to the same purpose. Skeinforge at least uses g-code as intermediate steps between "tools" and you could check the intermediate g-code with a viewer to determine which "tool" is causing the problem?

List all programs that touch the g-code as it goes from STL to being sent to the reprap. One of those programs is likely the culprit.

I can't imagine what setting would do this. Perhaps somewhere there's a setting to compensate for non-square mechanics that should be set to zero but isn't? Maybe reset each to the defaults, verify that the problem goes away (crappy print quality, but at least square) and then restore the custom settings one program at a time until the problem recurs?

More info, plz.

---

--
I'm building it with Baling Wire

Try moving the part so there is no chance that it hits the endstop while printing.

What firmware are you using? Can you modify the cupcake firmware so that you are using the same firmware on both?

From your image, it looks like when you increase x, y decreases in a pretty consistent manner. Do you see any similar problems in the z direction?

---

---

Darwin clone, Gen 2 electronics, Arduino Duemilanove w/ AtMega328, 5D Firmware, Pinchwheel extruder
[www.codeerrors.com]

The simplest explanation is that your frame is not square and somehow you didn't notice when following the grid.

Did you use the jigs to set all the frame distances?

---

[www.hydraraptor.blogspot.com]

I'm an idiot....
I'd call it a comedy of errors except it wasn't funny.
I sat down on Saturday and read your responses and this is what I wrote

Thanks for the responses folks

John: I wish congrats was not in order. I have a couple of anxious people waiting for parts. After the first part came off the Mendel... (barring the problem) the prints are so nice that I just don't want to print any more on the Cupcake... There are three of us working on this by the way.
Seems like every time we have a problem it's unique. We scour the forums and scour the wiki, Google our brains out and come up blank. Seems to be our lot in life. Murphy's law prevails. Oh! Here we go, just like right now. I want to take some pictures with my camera so I grab a couple of batteries at home and head out the door. Now I'm sitting at my desk at the shop typing and what the heck, my leg is hot, I mean really hot, like my pocket's on fire. I'm scared to reach in. I start quickly but gingerly pulling things out with my finger tips One set of keys, my Bic butane lighter, one battery, a second set of keys.. ouch are they hot. Then the second battery. It burned my fingers. What's the odds of a key chain shorting out a “AA” battery in your pocket. Apparently pretty good today.

Anyway, I'm the wrong person to be asking about programming, Jeremy does most of that. Unfortunately Him and Mike have a show tonight so he's not available right now. “Spam” If you ever find yourself in Niagara Falls Canada. Be sure to catch the Danny ZZZZ Show , Ask for the Reprap discount.

mccoyn Yes, the prints are well off the end stops.

Nophead: Yes, although it was a while ago I did use the jigs when setting it up, and all three of us sat and watched it zigzag its way up and down and around the grid. All the faces emanating the start point of the print are square.

Then came Sunday...

Growling into my morning coffee: OK... fine... I'll check the frame. Took it to my workshop “10 AM” and plunked it on the table saw “the flattest surface I have”. Grabbed a framing square and stood it up by the left Z bar and took a measurement... OK it's out a tick front to back... 1/8” actually. I checked the spacing between the vertexes and they were all correct. ???? I thought oh great here we go again. And so it went. I sat and scratched my head for a while. What do I adjust? Finally I went to the left front vertex, shortened the upright rod a quarter turn and added a quarter turn to the horizontal and presto, twisted frame. So, I blocked up right corner and applied some pressure on the diagonal to straighten it back out. Took a measurement, 1/16” out. Repeat operation, Perfect.... front to back... Now it was out sideways. And that's how the comedy of errors began to unravel, one thing lead to another, to another, to another. By 8 PM I had adjusted everything thing on the machine that I could possibly adjust. I'm not sure what else was wrong to start with and what came as a result of today's contortions... But finally, it was following the grid perfectly again. I held my breathe and did a partial print in the center of the plate, it looked pretty good. I left it stuck to the plate and did a full print close to home... All square.



So... How many wrongs does it take to create the illusion of right? Not sure, I lost track.

Thanks for posing the question nophead. I wouldn't have even considered it otherwise.

By the way, I saw your Meltdown blog entry. That sucks. I'm using a variation of your steel tube design, but I mounted the heat sink under the carriage and the fan on top so the air blows through the carriage before hitting the heat sink. The carriage stays cool to the touch, and the top of the tube only gets warm to the touch, not too warm to touch... “As long as the fan is running”.



Edited 1 time(s). Last edit at 12/19/2011 09:29PM by Bruce.

Heh, I just did the whole remeasure and adjust every length on my Darwin a few weeks ago because my z-axis was binding up. The nice thing about the Darwin on this front is that everything is a square and you just have to make sure the sides and diagonals that are supposed to be the same length are the same length. The bad thing is there is about 4 times as much stuff to adjust. I ran through the entire operation twice to be sure, and then had to adjust the tension on all my belts and the bed position.

---

---

Darwin clone, Gen 2 electronics, Arduino Duemilanove w/ AtMega328, 5D Firmware, Pinchwheel extruder
[www.codeerrors.com]