Greetings,
We are a group of undergraduate students who are currently working on the winding of a pressure vessel. We have successfully assembled the machine and calibrated it to account for any offsets. During our initial tests using the trace function to ensure proper alignment, everything appeared to be in order. However, we encountered difficulties when attempting to initiate the 4-axis hemi end winding process.
To kickstart this process, the machine's executor homes it, aligning itself with the mandrel's starting position and patiently awaiting our manual input. Here is where the unusual part of the sequence begins.
The G1 command is executed, moving the carriage to the end of the mandrel. Subsequently, the G2 command should follow, but it seems to become stuck either before or after the DH (Directional Heading) moves towards the mandrel's holders. This delay can last for approximately a minute, even in fast mode. Following this, the machine moves away from the mandrel and towards the right end, where it encounters a similar issue with the G3 command.
It's worth noting that the occurrence of these issues is entirely random, and they can happen after varying numbers of passes.
For a visual demonstration of the problem we've described, you can watch two videos by following the link below.
In this video, the G2 command does not get executed properly which is obvious by the DH not moving foward towards the mandrel holders at the ends.
http://www.youtube.com/watch?v=ltH0eNW6hME
In this video, a similar problem is encountered. This time at the right end with the G3 command, although this time the DH moves towards the mandrel holder but gets stuck.
http://www.youtube.com/watch?v=HOlsJ_A7YKU&
Things we have tested to try to eliminate the problem without any success:
- Different winding patterens angles and all the other parameters you can change in the designer.
- Different filament width and thickness
- Running the designer software on different PCs although they are all running windows 11.
- Running the software as an administrator and trying with windows xp, 7 compatibility modes.
- Switched the DHL and DHR motor.
- Used another USB-hub with dedicated powersupply.
- Turned off all power saving mode options on the laptops.
Is there a fix to this issue?
Comments
We have recently seen this error with Win11 PCs, most notably the G2 error where the DHL refuses to advance after the first pass to the far end and the carriage seems to sit indefinately while the DHL remains in the retracted position. Once this happens, waiting on the machine at this point is futile.
In your specific case after watching your videos, it appears that you need to check and set your Z-axis and Y-Axis offsets in the Designer program. Open the Designer program and go to menu items File --> Settings --> 4-Axis. The Y-Axis Offset on the standard 4X-23 Winder is about 4.5 inches. This is the distance from the DH concave roller (the last roller that the filament touches) to the mandrel centerline when the DH is fully retracted. I can see that your Y-Axis Offset is set to 6.89 inches (line #14 of the G-Code execution with the Y parameter). "Home" all the Axes by poressing the "Go Home" button on the Designer Setting 4-Axis page and make sure you re-measure the Y-Axis offset and input to the designer.
Likewise, do this with the Z-Axis offset as described here: https://youtu.be/MxvUU7Ig1yI?si=k97RjEdK8kNDKV7S&t=38
It also appears from your video that the DH Rotary head does not fully rotate from positive angles to negative angles as the Executor is running the winder, but yet as I watch the DHR is indicating that it should be moving. this may be because the set screw on the DHR gear is slipping on the DHR motor shaft - so give that a look.
Also, I see that your Phi and Delta angles for this layer are set to 180 and 35 degrees respectively. Since you are just getting this machine up and running, I would suggest setting at 200 and 50 as the best first guess. 50 gives the DHL and DHR more time to complete their motions in coordination with the mandrel movement. Until you know otherwise, stick with 200 and 50 degrees for Phi and Delta.
Also, try running at slow or medium speeds (you are running on fast speed) after you've first set up a new machine. As you gain more experisnce with the Winder, you can start to expand its operating envelope, including faster speeds.
Thanks and let me know results.
Turner
This is the youtube video, I cannot post the whole link because of an error
/watch?v=_tYrjkHA5UA
This is the link to the images on imgur
https://imgur.com/a/YBftJcK
Thank you for your response!
We are aware of the issue with the DHR stepper not performing as expected. In our efforts to address this problem, we ran the program with the head's rotating belt disconnected, hoping that relieving the stepper of its load might resolve the issue. However, this adjustment did not yield any improvement in the winding process, and we continued to encounter the same problem.
It appears that the stepper experiences skipping when subjected to a load, preventing it from returning to the horizontal position during the homing process.
We attempted to execute the program on a Windows 10 laptop with the machine set to its default offsets, but unfortunately, we encountered the same issue. The default offsets caused the DHL to attempt to move beyond its designated path, prompting us to halt the operation and reevaluate our offset settings.
Upon measuring, we discovered that the offsets were as follows: 6.9 inches for the Y-offset and 2.7 inches for the Z-offset, as depicted in the images below. Despite implementing these measurements, our subsequent attempts were still unsuccessful.
In our testing, we employed 200 and 50 degrees for the Phi and Delta angles. To provide you with a better understanding of our situation, we have included a video of our testing process. Initially, we perform a trace operation to ensure the distances to the mandrel are accurate before proceeding to build and execute the program, where we continue to encounter the same issue.
Attached Images:
The first image illustrates the Y-offset while the DHL (Delivery Head Linear) is in its homed position.
The second image displays the Z-offset while the carriage has been homed.
thanks for the picture of the Y-Axis Offset. I think the winder frame may be assembled incorrectly for the model 4X-23 - or your model is 4-axis model 4X-34 which we have discontinued. The difference between the 4X-23 and the 4X-34 is the size of the carriage and mandrel stepper motors. The 4X-23 has 2.3" width carriage and mandrel stepper motors and the 4X-34 has 3.4". The maximum Y-Offset, once assembled of the 4X-23 is no more than 4.5" and the maximum Y-Offset, once assembled of the 4X-34 is about 8.5". The 4X-34 model has been discontinued.
Here are the assembly plans for the 4X-23: https://www.dropbox.com/s/iwbswgem6xlm3f1/XWinder_Assembly_Diagrams...
What is the width of your carriage and stepper motors ?
Turner Hunt
Do you have an earlier version of x-winder software that we could possibly try?
yes - what version number are you after ? or givce me approx date the build was released
Greetings again Turner Hunt
We have not yet received any communication from you, and we remain interested in the possibility of testing an earlier iteration of this software. Specifically, we are interested in obtaining the three most recent versions, but if that is not feasible, we would appreciate the opportunity to access a version prior to v4.19.
Thank you in advance.
Greetings Turner Hunt ,
I am also a part of this project, and my group and I have decided to test the three latest versions of this software, if that is possible.
Thank you for your prompt response.
We have measured the steppers for both the carriage and the mandrel, and they both measure 2.3 inches, indicating that we indeed have the 4X-23 model. This purchase was recently made by our university, further confirming its model. The carriage itself is 5 inches wide. (Linked images for reference.)
As for the frame construction, we utilized three pieces of 2-meter-long bars measuring 30mm x 60mm due to their availability in our region (Europe). We followed the assembly instructions meticulously, while also ensuring that the top and bottom of the left and right crossbars align flush with the front and back of the long bars. (Image linked for visual clarity.)
Upon closer inspection and comparison, we may have identified a potential explanation for the approximate 2-inch Y-axis offset discrepancy. This pertains to the choice of extrusions used in our frame. Assuming we had employed long bars with cross-section measurements of 25mm x 75mm, it would have allowed the carriage's front to sit approximately 30mm (1.18 inches) closer to the mandrel, thereby reducing the offset. The combined width of the two 30mm x 60mm rear long bars is 120mm, whereas two 25mm x 75mm extrusions would measure 150mm in total width. Assuming these extrusions are flush, it would bring the Delivery Head (DH) closer to the mandrel's central axis.
Simultaneously, a similar effect would have been observed in the placement of the mandrel itself, with its central axis situated approximately 37.5mm (1.5 inches) from the outer edge of the frame, thereby reducing the Y-offset. In our case, the distance from the central axis to the outer edge is 15mm (0.6 inches). By calculating the differences, the fact that we used the second recommended type of extrusions could account for approximately 2.06 inches of our offset.
[Images and sketches related to these observations and calculations can be found in the linked materials.]
We have the capability to modify the frame by adjusting the forward long bars on which the mandrel is positioned, bringing them closer to the DH to ensure that the machine's default Y-offset is not exceeded. However, it remains unclear whether this adjustment would resolve the issue of the machine becoming stuck at the hemi ends of the mandrel. This uncertainty arises from our previous attempts to run the machine, following your instructions, without the mandrel and with default offset settings (also 200 phi and 50 delta at all three speeds). Regrettably, we still encountered the exact same issue, as illustrated in the previously linked videos.
We appreciate your continued support and guidance in resolving this matter.
https://imgur.com/a/7tvRXa5