opulo-inc / docs Goto Github PK
View Code? Open in Web Editor NEWDocumentation for the assembly, maintenance, and usage of Opulo designs.
License: GNU General Public License v3.0
Documentation for the assembly, maintenance, and usage of Opulo designs.
License: GNU General Public License v3.0
Document how to mount tray feeders on the staging plate and define where they are in software
Users of the FTP will need to get the kicad files, import them as a new board into OpenPnP, and then set the location of the board roughly.
There should be only one guide for the assembly Belt Tension Arm (FDM-0037) which should be linked from every use of this part.
Additionally, the whole assembly guide should start with the requirement to build all these simple sub-assemblies (e.g. mounting of the pulley on the motor). Assembly of the third Belt Tension Arm separately is much more annoying than building all three at one time. Because the screws should not be searched again, tools should be used in the same way at one time... etc.
The current docs refer to Index_Mobo_Rev03
which no longer exists, it should be updated to Opulo_Lumen_REV3
or Opulo_Lumen_REV4
.
In Pump Mount no9 it says F31 F33, for me this is G31 and G33, as F31 and F32 do not exist on my staging plate.
is that a typo ?
the photos show button head screws, and the text is correct, they are socket head screws.
@sphawes is getting gabe photos
Docs say:
Import Board
1 Download the FTP board design. It’s in the LumenPnP_PCBs.zip artifact on the latest release.
There is no LumenPnP_PCBs.zip file at the latest release page.
2 Go to File > Save Job As and save your FTP job.
Before I load it? Do you mean to say to download the file, then load the board files? This needs heavy clarification.
We changed the names of the FDM printed parts to be more consistent. We also got rid of the part numbers, in favor of internal part numbers that won't be referenced on the docs site.
This is an issue to update all the assembly steps to use the right part names.
Was talking with Stephen and noticed this photo is of an older model. the newer one is much more recessed. Could we update this photo to match what the customers are getting in the box currently?
Should there be an O-ring on the screw end of the CP40 nozzle holder where it attaches to the NEMA11 hollow shaft stepper motor (like there is on the Pneumatic Coupler mounted on the other end of the stepper).
Got a note that our docs don't account for the fact that sometimes the vacuum sensors can have different addresses, so the default testing gcode command doesn't work for them. Reference see this discussion
Hubspot Ticket: 1034656629
This page https://docs.opulo.io/openpnp/ftp/installing-the-feeders/ has you install the strip feeder using mounting holes C11 and E11, which is nearly impossible given the proximity to the vacuum hose connections and cables. I flipped it trying holes C15 and E15.
Slightly different documentation required for step 8 here, and for bottom camera mount steps which will need to include the foot.
In general, how to make sure you get consistent results from the vision pipeline to make your fiducials detect reliably.
Looking at a picture it looks like pcb-04 mounts on A and B above the camera on the build plate, however the top bolt interferes wit the pcb. looks like there is a notch out of the pcb for that bolt, is this correct?
The old image shows LH on top and RH on the bottom. This has been swapped in manufacturing to prevent having to twist hoses more than necessary. Now it has RH on the top and LH on the bottom. New photos have been added to google photos shared folder for product photos. Thank you!
Fine-tuning the board's location with fiducial dots, or manually defining it with the pad location tool
Make it so that you can click on the images in the docs so that they grow. Currently we have images that are getting down-scaled which is bad!
needs commands for installing mkdocs and mkdocs-material
This section gives a variety of options for gcodes you need to run to test the vacuum sensors. None of them are correct for my printer, and i only found this out by going to the discord and searching historic messages.
I'd recommend opening this section with a description of the options for gcodes:
Only once the user has identified enough of that for them to be able to talk to the i2c sensor should they attempt to test vacuum
In the "Wiring" section steps 15-18 the motor cables are shown as labeled LH and RH, mine were labeled LM and RM.
In the "Wiring" section step 34 the strain relief is attached with hardware that was already inserted, my strain relief had no hardware already inserted and there was not sufficient additional hardware in the kit.
In the OpenPnP section it should be mentioned to remove the dust covers from the cameras.
Just received my Lumen v3 kit, and went to the docs.opulo.io, clicked on the "Docs" link and was immediately taken to the "Lumen PnP Kit Assembly" page, which is where I've gone each time I've loaded the docs site. I had also almost purchased a bunch of filament to begin printing the parts needed for assembly.
However, having now opened the box and seen that most of the parts already seem to be printed, and looked around and saw the "Lumen PnP Semi Assembly" in the top menu, I realized I've been looking in the wrong place the entire time.
I would recommend renaming "Kit Assembly" and "Semi Assembly" to "Version 2" and "Version 3" since I don't see "semi assembly" anywhere on my packaging. A landing page that allowed you to go to the relevant documentation based on your version, rather than being sent straight to "Kit Assembly" would be helpful as well.
Now that we're not using Hugo we have some flexibility with naming the site's markdown pages for easier navigation when writing them. I'm gonna move stuff around!
Documentation on how to affix the FTP PCB to the staging plate
Please let me know if you can use a MobO board on OpenPNP, I find MOBO boards much cleaner and easier to make.
LumenPNP build V3.0.2:
Most V3 instructions were clear and photos useful. A couple of more important points are noted with ***.
*** Assembling the frame, step 17 and further: These hammerhead t-nuts may not be intuitive--typically, the bolt can't be screwed far into them, or they can't rotate to engage the extrusion. As shipped, the t-nuts are screwed too far on to rotate for some of the parts, and so must be loosened.
Related, in step 18, it may make sense to tighten the "bottom" bolt first (whose t-nut can't be seen); it is then obvious whether this engaged properly in the extrusion. Then later the top bolt can be tightened; in this case, the t-nut can be seen to rotate (or not if it was too tight).
In Wiring, step 7: Suggest noting that M5 bolt is going to be a tight fit through the cable chain. (This may be reassuring.)
In Wiring, step 10 (and beyond): Recommend being clear on the orientation of connectors. You can probably use wire color (e.g., "with the black wire on the left"). In other cases, there are orientation tabs etc that can be called out.
General note: I suggest adding a brief note early on describing how to safely unplug these electrical connectors (vs yanking out by wires), in case the user has made an incorrect connection. Also, how to release the tubing by pushing down on the blue ring. Neither may be obvious to some users.
In Wiring, step 13: Clarify this is the unlabeled cable, since RM is also a 6 position/4-wire connector. (Might also suggest the language "position" vs. "hole".)
*** In Wiring, step 15: Cable is actually labeled LM, not LH (which is a tube).
*** In Wiring, step 17: Cable is actually labeled RM, not RH (which is a tube).
In Wiring, step 24: In my LumenPnP, the pneumatic tubing was very tight to the board. No reason not to have some more slack, which would reduce the chance of putting force on the stepper driver heatsinks as the board shield is removed and installed.
In Wiring, step 25: Note that the cable "clamp" (3D printed cable clamp that fits into the extrusion) may need to be shifted.
*** In Wiring, step 30: XM was far too tight--I had to cut and re-do the zip ties on the strain relief print to get enough slack to route XM properly. (This is not a documentation problem, but wiring harness problem.)
In Wiring, step 34: Need to call out hardware sizes. (BTW, I was missing a M5-10 bolt, only had one left.)
Additionally, it may be handy to provide a printed sheet with the various hardware types noted--in particular bolt lengths.
The newest update to OpenPnP has either a bug or a new way of confirming the measurement when calculating the mm/pixels. The old one showed "measure" and, once clicked, would change to "confirm". This is no longer the case. The button does not change the text but keeps a blue glow around the button until it is clicked again, which saves the data at that point.
We had some additional content before we moved the docs to their own repo. Lets bring it back!
There can be occurrences where LumenPnP users think they have two of the same cameras, rather than distinct Top and Bottom camera. However, it is rare that LumenPnP users with cameras from Opulo received the wrong hardware. It is also rare for ELP to program (or label) these camera models with the wrong Device ID.
What seems to be the issue more often than not is actually that OpenPnP get the camera's Device IDs mixed up due to the modules having identical UUIDs.
While OpenPnP thinking the top and bottom camera have the same name is confusing, it usually can be ignored after setup. It would be a good idea to add a disclaimer to the docs that alerts customers on how to handle this situation themselves.
On this page: https://docs.opulo.io/openpnp/calibration/camera-fisheye-cal/#bottom-camera-calibration
starting at step 6 for bottom camera calibration it talks about holding the cal target under the top camera and continues on with the top camera after that.
in number 6 my measures would be:
DL 14 inches.
LM 14 inches (which may cause shortage on the other end)
The LumenPnP v3 kit will ship semi-assembled, so it will need new documentation on how to assemble it. This should be a lot shorter and simpler, but will need new CAD screenshots and photos from real life.
Nozzle holder No 14 it says M3x10 in the description in contrast to the parts listing of the chapter ( there it says M3x8)
Common symptoms of problems and what to do about them when setting up your first board.
There's a few extensions we can enable to get some functionality back in mkdocs that we lost in the transition from hugo. Click to zoom images, being the main one.
On the Mounting the Motherboard page of the docs, we could be better about step 4 which is about installing the stepper motor drivers
We write Note that the whale icon on the motor drivers will be right-side up and the two upward-facing pins will be facing to the left.
but we don't have a photo of what this looks like up close. Users can really only see the UT logo in the silkscreen of the PCB for reference as is. I suggest that we add a photo showing the whale and also a section about installing the heat sinks.
Many community members have had issues with the USB C receptacle tearing off of the motherboard. We should add a note for folks to check and reenforce the receptacle's feet while they're soldering all the through-hole parts.
This has been a quite common stumbling block of users including myself. I first had trouble with vision recognizing the homing fiducial and couldn't sort out how to resolve it until Bobbe asked in Discord if I had tried modifying the pipeline. I had no clue what the vision pipeline was or how to get to it. I've seen this same exact thing happen to each user in discord as they complete the lumen calibration.
I completely agree that it shouldn't be a goal of the lumen docs to teach users details about openpnp, that's the job of the openpnp docs which are quite good. But the situation that occurs with everyone is that once you've completed the lumen calibration procedure and it's time to migrate over to openpnp and it's docs, you simply don't even have a clue as to all the things you don't know. So if the lumen docs simply introduced the users to some of these topics with links to openpnp docs to learn more would help quite a bit.
Here's a list of some of the common stumbling blocks I see that users will have better luck skipping over with a list & links to openpnp:
1- vision pipeline: https://github.com/openpnp/openpnp/wiki/CvPipeline
they even have a nice quick little video which had I found early on would have saved me from going deep into the wrong rabbit hole.. ;) https://www.youtube.com/watch?v=iUEP0bILAU0
2- fiducials: https://github.com/openpnp/openpnp/wiki/Fiducials
Might be good to suggest they test out finding fiducials on their PCB if present. Might even be good to include in the homing calibration procedure as something to investigate if having issues as this is when I first had problems that needed to be solved with modifying a vision pipeline.
3- feeder setup: https://github.com/openpnp/openpnp/wiki/Setup-and-Calibration:-Feeders
CurlyTaleGames had a good recommendation to suggest users setup their first static feeder: https://github.com/openpnp/openpnp/wiki/ReferenceStripFeeder
4- Bottom vision: https://github.com/openpnp/openpnp/wiki/Bottom-Vision
advanced explanation of stages: http://openpnp.github.io/openpnp/master/org/openpnp/vision/pipeline/stages/package-summary.html
That's all I can think of at the moment. I think just pointing these things out similar to how the lumen docs point out where to go for your first openpnp job would be of great help. :)
Its either F32 F34 or G31, G33 as G is the odd numbers and F is the even numbers. Just a nit pick.
Can you provide the resistance value of the resistor in the FTP board and the color of the lamp beads? Thanks a lot
In Pump mount parts list, it says 4 M3 hex nuts.
But only two are used on the pump
Since OpenPnP came out with 2022-06-10, there's a new Vision tab. This ticket is for reviewing the OpenPnP setup guide and making sure it still works with the new tab.
Mounting the calibration PCB is missing
Add instructions for adding a toolhead to v2 LumenPnP kits (before the semi-assembled v3)
We discussed that it would be nice for the Docs project to pull the latest BOM and photos so that we can link to it in the docs and can use the photos. The idea would be to replace the current hover-text snippet we use to show part numbers (which are now gone) and instead have the hover snippet show the relevant photo of the printed part, as well as be a link to the BOM hosted on the docs site so that people can see what the part is, download the .stl, etc.
If you accidentially (or on purpose) send a "home all axis" command from the OpenPnP interface or console, this can seriously damage your nozzle or other parts of the machine.
When the nozzle is already in a low position, it will get draged along to the origin point in the front left corner of the machine, possibly hitting everything on the way there, like the nozzle magazine, camera reference, component trays,...
This is a common issue, since you want us to use the G28 command in your documentation to do an initial motor test (https://docs.opulo.io/docs/testing/motor-check/)
I think the marlin firmware needs to be aware that the Z-axis has to be moved into to a neutral position before it starts to home the X and Y axis. Unfortunately, trying to home Z first does not help, as that would lower the left Z carriage all the way. So there needs to be a way to move Z into a neutral position (both Z carriages in the middle position) before X and Y should be moved.
There are some options in the marlin config files that deal with homing. But i'm in no way proficient enough in marlin configs to understand all these options.
Adding a warning in the documentation might be a good idea too, to make sure new users are aware to look out for the Z-axis position, before they are tasked to send a G28 command to test the motors.
Initial paragraphs in this section call out default Z-Heights as 61mm for V2 and 31.5mm for V3.
In section 5, "Set the Camera's Z-Axis locations:", Step 1 says for V2 set the values to 31.5mm and step 2 says for V3 set the value to 61mm.
We need guidance on how to tune your bottom camera's vision to detect your nozzle tips!
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.