T-Shirt Forums banner

Behavioral Problem - Image elongated CIS on Workforce 1100

[DIY DTG] 
9K views 59 replies 8 participants last post by  colorfinger 
#1 ·
OK Team....

Well, I loaded up my CIS on my Workforce 1100 and tried to print a design... Now, the image is elongated. No banding, it's just a circle that prints as an oval... Anyone have any ideas how this can be happening? It happens in Windows and Linux under Gutenprint.

Thanks,

Bob ?;O)
 
#8 ·
Re: Behavioral Problem...

Is this on your diy dtg machine? Please forgive me if I don't know what I'm talking about - but I have always wondered if the revolution ratio between the drive rod and larger conveyor rollers would cause distortion of the printed design (assuming you are using the conveyor-belt style). I imagine you would need conveyor rollers that are the same diameter and overall dimensions as the drive rod.
 
#10 ·
Re: Behavioral Problem...

Is this on your diy dtg machine? Please forgive me if I don't know what I'm talking about - but I have always wondered if the revolution ratio between the drive rod and larger conveyor rollers would cause distortion of the printed design (assuming you are using the conveyor-belt style). I imagine you would need conveyor rollers that are the same diameter and overall dimensions as the drive rod.
if the roller is bigger than it should be then the belt goes faster and you have white horizontal lines and longer prints if it is smaller then the lines print on each other and you have darker horizontal lines

colorfinger can you post a picture of the print?
 
#11 ·
Re: Behavioral Problem...

You should have seen my spindle... Perfect, and it mounted the rod well... The poor motor did not like it at all... Has to travel four times as far to turn.... I'm scraping it as a bad idea. I think it is more logical to place a 5/16 rod on the tension end and have the belt drive it instead of the 1" roller. I think this is the ticket here... This should preclude any ratio issues... Well, I'm off to Home Depot to find a rod.

Bob ?;O)
 
#12 ·
Re: Behavioral Problem...

You know, if the image is elongated then the conveyor is moving too fast! I spent the whole weekend looking at it the wrong way. It definitely is not a gear ratio problem. The speed of the belt is constant. I am wondering if the encoder is not at an exact right angle to the wheel if this could be causing the problem.

Bob ?;O)
 
#28 ·
Re: Behavioral Problem...

Here's the print... This logo should be a circle....

Bob ?;O)
Bob, Are you absolutly sure that it's not the artwork? I saw this happen when an artist had his monitor ratio wrong, and making what he thought was a circle an oval. I could be wrong, but if X and Y were not synced.. I would think that you'd have banding and stairstep edges. I would try some kind of calibrated grid pattern.

Ian
 
#18 · (Edited)
Re: Behavioral Problem...

bob,

Im sure its the roller setup...bigger rollers on your conveyor will make your travel distance longer ie the elongated shape..

it has to be a 1to 1 ratio... your set up has the conveyor geared up..

i cant say that at 100% but im pretty sure.. check out the 1270-1290 conveyor build thread..looks like hes pulling right of the pf shaft...some of his rollers are bigger, but there lined on top... here is the vid...

[media]http://www.youtube.com/watch?v=yCRQsEfGX0A[/media]
 
#19 ·
Re: Behavioral Problem...

hmmmm
when the roller is bigger or smaller you have banding, i guess your drive system is ok cause i cant see any lines.
even in full quality mode you can see horizontal lines
how thick is the roller?
have you added any other gears?
what are the print settings?
have you tried printing through epson driver?
can you post a closer pic?
 
#20 ·
Re: Behavioral Problem...

A closer picture will reveal nothing. It is a solid print. First, I've spoken with several engineers where I work and they confirm that the belt's rate does not change. If the rod moves the belt one inch then it would move one inch under the print head. Logically if it were due to the fact that the roller is larger than the rod then it would be moving slower and not faster... The rod's rate of movement is fixed by the sensor. It will not slow down

That's what I was thinking but the elongated image means that it is moving faster not slower. The larger gear moves at a slower rate. You would have to run another gear to take advantage of this change... No new gear exists.

No, I really think that it might be because the encoder is not at a 45 degree angle and the rod is moving a bit farther to get it's tick count. I will test this out when I have time to put it back together... This is the simplest change else I will replace the first conveyor roller with a 5/16" rod to test you theory.

Thanks for the input guys,

Bob ?;O)
 
#21 · (Edited)
Re: Behavioral Problem...

Bob,

Hope you get it going, i really believe it has something to do with the small 5/16 drive rod and the bigger 1 rollers..

on any motorized vehicle with sprockets or gears...like a motor cycle,atv, etc for performance they change there sprockets...and i gaurantee you it makes a differrence..i did this on my 2004 suzuki gxxr 750 and it makes a world of difference... a smaller sprocket on the back wheel makes the wheel spin faster, regardless of the power source..

i understand your encoder theory, but it thinks its moving a 5/16 rod..your transferring that same movement to a 1 or bigger roller?

good luck:)
 
#22 ·
Re: Behavioral Problem...

Jeff, Perhaps I'm wrong... Let's look at it this way... If you are driving a bicycle in low gear it takes less energy to peddle. This is matching a small sprocket to a small sprocket. But if you increase the gear (A larger sprocket) it takes more energy to push... If you peddle at the same rate as you did with the smaller sprocket you will go faster.

I won't have any cash to pour into my money hole till this weekend. Sorry we'll have to wait till then to find the answer. I'll have to purchase a 5/16" rod and figure out how to eliminate the friction on it in place of one of the conveyor rollers... The drive rod will have to drive this new rod instead of a conveyor roller.

Bob ?;O)
 
#23 ·
Re: Behavioral Problem...

Jeff, Perhaps I'm wrong... Let's look at it this way... If you are driving a bicycle in low gear it takes less energy to peddle. This is matching a small sprocket to a small sprocket. But if you increase the gear (A larger sprocket) it takes more energy to push... If you peddle at the same rate as you did with the smaller sprocket you will go faster.

I won't have any cash to pour into my money hole till this weekend. Sorry we'll have to wait till then to find the answer. I'll have to purchase a 5/16" rod and figure out how to eliminate the friction on it in place of one of the conveyor rollers... The drive rod will have to drive this new rod instead of a conveyor roller.

Bob ?;O)

bob..

heres what im referencing, when speaking about gear ratios on a bike..here is a great article on the subject...should help you ..i hope..
Motorcycle Sprocket Selection and Gearing Fundamentals
 
#27 ·
Re: Behavioral Problem...

If you altered the stock 12mm axis you will definitely need to add gears or replace the encoder for another one.
In my design i used a 60mm diameter conveyor drum (5x12mm=60mm) so i added 2 gears to achieve a 5/1 reduction in order to have the same linear speed. You also have to add the belt thickness in order to calculate the final thickness of your conveyor roller. The thicker the belt the faster the platen will advance. In our situation when we say "faster" we mean bigger steps

Edit
PF adjustment test will be your best friend
 
#29 ·
Re: Behavioral Problem...

Simon, You have me stumped. More technical than I am grasping at the moment. Perhaps if you could explain the math it would make sense to me. I'm a very visual person and I'm no engineer by any stretch... I would think that a 1-1 ratio will do the trick.

Z, No I'm absolutely sure the art is correct. PF adjustment? Just how is that done?

I split the frame drilling a hole for the bearing roller to support the rod. I glued it back tut it will take 24 hours to be fully cured. I will resume tomorrow night.

Thanks everyone...

Bob ?;O)
 
#30 ·
Re: Behavioral Problem...

Its just school geometry. The math type to get the perimeter of a circle is 2π*r (r is half the diameter) so
2*3,14*6mm=37,68mm. So when you rotate a full rev (360deg)a 12mm shaft it will travel a linear distance of 37.68mm. The PF encoder controls the angular rotation of the pf shaft.
Do the math and you will see what needs to be done regarding your conveyor drum.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top