This board is to track any issues with Egret. (Mini Modular Bed System)
Tue, Jan 30
Fri, Jan 26
Wed, Jan 24
To update for clarity:
We have redesigned the 2pc Bed Corners for Gladiola and Foxglove, they have been embeefened so that they provide a tighter fit against the sides of the glass and heater. It is now very difficult to move the bed, and as long as infill remains at least 55% we'll be sitting pretty.
Going through ECO/DCO/MCO now to roll these out
Tue, Jan 23
I just checked the foxglove corners @logan had assembled on a bed plate with a modular bed. I really dont think the bed would shift during a print but upon part removal there may be issues. The corners are really squishy and i was able to rotate the glass completely out of the assembly. Based on this i think the corners need to have a bit more infill density. I could see someone trying to remove a print that is really stuck and having the glass come completely out. This could cause broken glass which could injure a customer, or cause injury without any broken glass. If the glass came out like this then the glass would need reinstalled on the print bed plate which is not difficult but is something that the customer shouldnt have to do. If we can get a slice that prevents the bed from being able to pop out with a rotational force then i think we will be golden.
It does look like we have the same thing going on with the foxglove corners;
These feel incredibly squishy in hand. Both gcodes I found on shared-j look the same at layer 5.
@bigmansas brought in the corners from his foxglove today, super squishy. We tried to find some that felt stiffer but he'll likely be much better off after we get an approved slice in production.
I'm starting an MCO now
Jan 18 2018
Sure, I'll drop in next week with it.
@bigmansas Thanks for the report. Can you bring those in to work for eval? We can get them swapped out for ya
I just purchased the modular bed kit for my Mini at home on Monday (foxglove) and when I went to remove the first print off the bed i gave it a tap with my hand and a putty knife. Instead of the part coming off nicely, like they usually do, if popped the whole bed off and to the left. I thought maybe I hadn't tightened the bed down enough, but it looks like this could be the problem.
Jan 17 2018
All PP-GP0296 stock is pulled and quarantined. Tomorrow I'm going to do some testing to see how many are too squishy to use. Interestingly, not all of them are squishy, even though they appear to be the same slice.
@EricNugent Thank you! That will help a lot!
It should be noted that the slice that is bad has an easy identifier. The bottom layer is concentric. There is no need to purge every single part, just the concentric bottom 2 piece gladiola parts.
@karrad I grabbed a handful of the ones present at lamination (where these are packaged) and every one I felt was too squishy. We'll need to get with production to get all of those pulled.
@logan Glad to hear the approved slice is as expected. Can you check whatever stock we have on the floor, to see what was being printed and delivered?
@karrad I received a batch of the correct print, night and day difference from what I installed yesterday and what we saw at CES. I approved and gave them back to @EricNugent
In response to your other question, no. They either have enough infill or they don't, there's no in between and there's no way to quantify any sort of measurement on that as far as I'm aware. Best answer is; if they can be easily compressed by hand they're bad. The ones I just received were not easily compressed at all, they were rather rigid.
@logan do we have any type of tolerances for this? In order to determine a fallout rate, I think we need acceptable levels. (Preferable to waiting for in the field data)
@benmalouf Looks like we've sold 86, we have 87 in warehouses, and 65 more on the way.
Jan 16 2018
@logan I think there is a mixture of batches. https://code.alephobjects.com/rMINIHIB89a2dbbac14093343ecab46e5f33be37a06342a4 is the commit for 2 piece bed corners for Hibiscus, this is the same file that should be updated in the egret repository. Since the piece is stock on the new machine and an accessory to the previous version I understand how the mix up happened.
@EricNugent @cneff I pulled the .gcode from ~/shared-j/devel/lulzbot/mini/accessories/egret/production_parts/Printed_parts/Flexy_Bed_Corner_for_Mini_2_part_bed_black/
I was not aware of another location for approved slices.
@EricNugent is more involved in the MCO process, @EricNugent would you mind getting that MCO started? Unless one exists already, regardless we need to get the approved slice pushed to devel.
I have not however if the movement isn't too bad then customers may not notice.
Checking the two Egrets we have in Product Planning, there is some slight movement when putting pressure from the side on the glass only. I could see this being noticed during removal, but not affecting prints. @nickp @lansky @Orias Have you heard of any reports in the field of this concerning customers?
So far, AFAIK, it's the Minis we had at CES and the Hibiscus. We should check any other Mini Modular beds we have distributed around the buildings. It would be great if this was a localized issue!
@benmalouf Is this just being seen on Hibiscus? It could be we have an off slice on the alpha, opposed to the bed corners included in the modular kits.
The bed corners on the Hibiscus I have are squishy enough that I highly doubt they have 60% infill. Another Mini modular bed we have over here is totally solid, no sliding at all, and it's got really well filled corners. Like night and day in terms of the infill.
The current gcode cluster has for production use has the infill set at 60% (I just clarified) for PP-GP0296 Gladiola 2 pc bed corners. So long as we are using Ninja Flex filament, there will always be some inconsistencies from print to print, due to how the material will consistently start to clog the nozzle/tool head (resulting in density issues). Density issues are something that cluster QC has been notified of and checks for (Doesn't mean that some don't get through). As far as I know... There have been no issues thus far on the 2 piece beds installed on the cluster machines (installed several months ago).
@logan Do you have a file location for that .gcode you are referencing. I ask because it does not reflect what is currently in production. This is a snap of the .gcode infill I am seeing in :
@logan Lets get an MCO going for these. We will want to have some on hand in case customers report issues in the field. Do we have any idea of the fallout rate?
Looking at the slice, @tutley agreed that the amount of infill is insufficient. From the .gcode I found, I would guess the infill % to be around 15-20%. They should be closer to 50% like Gladiola standard corners.
Below is a snapshot of an infill layer of the current slice in Egret compared to the current slice located in Gladiola.
Jan 15 2018
@kent I thought we had seen the excessive sliding even with the 2-piece bed corners installed and at the time it seemed the reason for this was a low infill density on the corners causing them to not hold as well. I will take a look at the slices available for the 2 piece bed corners for mini and see if I can figure out around what infill density is there.
That's a good question. I don't know. They were pretty tight, but not stripped. I would guess that they were more than the 5in*lbs spec if anything. We will be getting those printers back in MER soon so we could check pretty easily.
@kent were the bed corners tightened to spec?
We brought in an extra set of modular bed corners, and they did seem to help. On my Hibiscus the Glass/PEI slides around a bit, but I supposed I always expected it to. Doesn't seem to impact prints, in any case.
Adding @youngmrcarlson to this ticket as I'm pretty sure he did some of the corner replacement at the show.