The Voron kit build thread

@Ken226 Thank you for discovering this issue and sharing your insights. I just got done building and installing my gantry a couple days ago and I was wondering about its ability to hold squareness. It seemed like it did not take much to move the gantry out of square. I was actually looking at how I could add some rigidity to the corners of the gantry with some machined aluminum pieces. Obtaining mounting points though that are clear of all the mechanisms might be challenging.
 
7milesup said:
@Ken226 Thank you for discovering this issue and sharing your insights. I just got done building and installing my gantry a couple days ago and I was wondering about its ability to hold squareness. It seemed like it did not take much to move the gantry out of square. I was actually looking at how I could add some rigidity to the corners of the gantry with some machined aluminum pieces. Obtaining mounting points though that are clear of all the mechanisms might be challenging.
Click to expand...


And the gantry frame itself only has two corners, in the back. It's open in the front.

But, the squareness thing wasn't really related to those corners anyway, it was in the X axis cross rail. Since each end of the X rail is mounted on a linear bearing, each end of the X rail can move fore/aft a little, independent of the other end.

As you tighten the belts on one side, it pulls the X cross rail into a skew. Tighten the opposite belt, and it starts to correct that skew.

And theoretically, in a perfect world, if you tightened them to the exact same tension you should have an X cross rail that is perfectly perpendicular to the Y axis.

But in my case, evenly matched tension didn't quite = a perpendicular x rail. So, I decided that a squared up X rail is better, regardless of how evenly matched the belts tension. It seems to have really paid-off.

It should have been obvious to me from the beginning of this project. But, I didnt see it until I had the printer laying on its back, looking in through the top. I was like, "holy sh**, that thing looks like it melted In the sun!".
 
Last edited:
Interesting story on the 2.4 Gantry. I have read that it is flexible to allow the quad gantry levelling (tramming) system to work. This is the first I've heard about the XY issues, but it makes sense.

The Trident has a rigid gantry. De-racking XY is still a balance of belt tension and mounting X on the Y rail sliders.
 
Last edited:
I did a custom bottom plate out of 0.060" aluminum, then buffed to a near mirror finish. The acrylic panel is fastened to the underside of the aluminum with VHB tape.
PXL_20221216_223752747.jpg
 
Last edited:
The din line in neopixels uses 5 volts right? What happens if you accidentally connect the din terminal to a 3v pin? What about if you connect it accidentally to a strait 5v pin. Pretty sure I did both at different points. Those tiny pins are hard to see, and the various wire diagrams out there for these things vary, and some suck.

Trying to diagnose my stealthburner leds. They are misbehaving after the PCB upgrade.

The logo works fine, but the nozzle diodes wont come on. When I first connected it they came on, but in the wrong colors, flickered a little, then a day later the went off and won't come back on.
 
NeoPixel Problems

I was about to post the below article when the posting came up about neopixel issues.

As I recall 3V data to Neopixels might work but be unreliable, Adafruit notes confirm that. I would not expect 5V supply to DIN to cause harm to the neopixel, though if there was no supply voltage at the time it might be damaging. Those are small wires and through a difficult mechanical path, I would review for damage or shorts. The source of the data might be damaged if it tried to drive while the DIN was connected to 5V.


Drying Silica Gel Desiccant

I did an experiment yesterday with a small batch of used silica gel desiccant packs. Generally I reactivate them in a filament dryer at 65C for several hours. I decided to test the often mentioned recipe of using a microwave at 50% power for a few minutes. A lot of moisture was driven off. The packs got warm but not terribly hot. After that they did not seem to work, as measured by one of the inexpensive RH meters. So I tossed them into the filament dryer overnight and tested them again. They seem to work poorly now. I suspect the microwave overheats and damages the small structures in the silica. Even though the recipe is written right on the bulk package, it is damaging to the product. Some sources say that 200C will damage the silica. In a microwave the temperatures are basically uncontrolled. Most recipes indicate 60 to 120C for drying. Some sources mention that the indicator in the silica is damaged at about 150C.

So 60C for 6-7 hours, or 120C for 1-2 hours is probably the best recipe that I have found, and skip the microwave.

After that I toss them into a tennis ball can and put an RH meter in. It drops to 10% after they cool down. The damaged silica slowly drops to 15% and doesn't go lower. Tennis ball cans seal pretty well and are easy to open. Zip lock bags are pretty permeable to moisture, only the very heavy ones, or the metalized mylar ones are better. Thick screw top bottles with good seals should also work. The digital RH meters are very handy to tell when the silica needs reactivating, and the tennis ball cans have a nice wide mouth so the meter can fit easily. The labels generally peel off easily so the meter can easily be seen without opening the cans.
 
AlanB said:
NeoPixel Problems

The source of the data might be damaged if it tried to drive while the DIN was connected to 5V
Click to expand...

Can you expand a little more on that?

When you say "the source of the data might be damaged". The source of the data being my LED config file? Or, the neipixel signal out pin, pb0, pin on the octopus board?

The stealthburner has 3 diodes. Led 1 (the logo) is working fine, diodes 2 and 3 (the nozzle), aren't working. Is there a mode of failure in the octopus board that could cause diodes 2 and 3 to have trouble, while 1 is still working?
 
I haven't looked at that wiring but I expect they are all driven serially from the one Octopus output. Neopixels behave like shift registers, so the data feeds through from the first to the last. What order are they chained in?

Essentially each Dout is the driver for the next Din. So if there was a temporary short to 5V between two Neopixels that might damage the Dout of the previous LED. It might continue to work but not pass the data on to the LEDs after.
 
Silica gel should be heated to about 105C, give or take. A toaster oven set to 220F or even 250F is fine. Leave the gel in the toaster oven for a half hour. The color changes much faster than that so you need to leave the gel in the oven for the full time to get all the moisture out. A low temperature bake of 60C is not effective.

As you found out it's possible to ruin silica gel in a microwave. Basically you can't adequately control the temperature and a lot of it gets overheated and destroys the structure of the gel which is responsible for adsorption.

I just put mine in a foil lined tin, spread it out thin, set the timer for about 30 minutes and walk away. Come back after the ding of the bell and leave it there until it cools off. Then pour into a glass mason jar with a rubber seal. It stays fresh indefinitely, until you need it. This has always worked well for me.

For a standard reel size bag, the RH goes to about 10% within 5 minutes, with a spool of filament in it at atmospheric pressure.
 
You must log in or register to reply here.
Back
Top