Qidi Tech 1 - Replicator 1 clone

[caall99]

Hey, we're all in this together.

I've almost got my 12mm rod upgrade in. I had to strip the machine to the bone to remove the z-axis rods. I then took a bunch of measurements and put it back together. I designed up some new top and bottom rod supports and a platform adapter so I didn't have to drill any holes in the plate. I printed the new parts last night. Hopefully this weekend I will have time to take apart the machine again to install the new hardware.

I'll probably keep whatever acceleration settings I currently have so I can compare results from the 8mm vs. 12mm rods.

Once I determine that the design works I'll post it to Thingiverse.

Maybe with caall99's acceleration finding the upgrade isn't needed, but I already have the parts so I'll go through with it anyway.

Excellent project! Please let me know how you make out! Regardless of being necessary or unnecessary, your modification would further bullet proof this already very promising 3D printer! In all honesty, my print from last night is the highest quality hobby-grade 3D print I have seen so far. I cannot distinguish it from prints made by the Stratasys Fortus 250mc that we have at work. We also have an Objet 300 here, but I doubt I will be able to compete with that.

FYI, what I printed last night was another copy of the aforementioned Kossel Mini BLTouch effector. Even though the Qidi Tech was still shaking from the spurious and noisy gCode that the degraded STL model sliced into, it still turned out beautiful. I can't imagine how nice it would look if I had a clean STL to slice from!

Regarding noise: I installed Astrosyn NEM 17 dampers on my Kossel Mini, and it's COMPLETELY silent. Unfortunately, we are limited with space on the Qidi Tech steppers, and therefore cannot fit Astrosyn dampers on every axis. I am wondering whether or not an upgrade to 1/32 microstepping drivers like the DRV8825s would reduce stepper noise.

I wish Qidi Tech incorporated the frame "lip" on the metal front panel (FF Creator Pro has it), so that the acrylic door would have something to rest against when closed... right now its just kind of hanging there loosely.

[rcleav]

Hey all!

Some updates on my aforementioned problems. I can honestly say that I resolved both.

1. MakerWare != MakerBot Desktop!!! I upgraded to MakerBot Desktop from the Qidi Tech provided MakerWare software and am now slicing 5 times faster, can set minimum layer time, and a variety of other advanced options. This is so much better! I may or may not stay with the MakerBot software, we will see how far I can get my machine tuned.

2. I used ReplicatorG to update my Sailfish firmware settings... and low and behold, my max acceleration was set to 65,536mm/s^2 (yup, you read that right)! Coincidently that is the size of a 16 bit integer, or 2^16. Obviously an incorrect value, that caused my machine to behave as if it didn't have acceleration turned on. I changed that value to 1000mm/s^2, and the X, Y, and Z values from 1000mm/s^2 (default) to 500mm/s^2. MY GOD! Printing is almost buttery smooth! No more "ringing"! I also changed my ABS filament temp from the 230C to 220C, and my bed is now at 85C (I don't use an active cooling fan with ABS). ABS sticks perfectly to the blue sticker and does not warp! Very very happy with last nights result! I may use the spare blue sticker pad with the MagHold spring steel plate (in the mail to me now), because I am having so much success with it.

Things to do:
a. I may reduce the max acceleration parameter to 750mm/s^2, and the Y acceleration parameter to 400mm/s^2, instead of 500mm/s^2. The Y axis has a lot more mass, and should therefore have a smaller acceleration value than the X axis. Still need to decide on values for extruder acceleration. Still trying to figure out why there is a max acceleration parameter and then individual parameters for each axis...
b. I may also reduce my ABS filament temp further, say to 215C. I am working on surface smoothness and bridging next.
c. Install MagHold with BuildTak (option 1) and Blue Sticker (option 2)
d. The wire loom on the right side of the machine is rubbing on the right acrylic window when the machine moves in the Y direction. This is wearing away at the acrylic, and needs to be fixed. Thinking of printing one of those cable chains on thingiverse, just have to decide which one!

OK here are my acceleration settings. It looks your on the right track.
X,Y Max 1000, Z max 150, Extruder max 2000

I'll repost better screen shots later today.

[KludgeGuru]

OK here are my acceleration settings. It looks your on the right track.
X,Y Max 1000, Z max 150, Extruder max 2000

Can't see your images, they are too small and blurry.

Here are my stock EEPROM settings from ReplicatorG:

QidiStockEEPROM.jpg

[jfkansas]

It is best to about halve the X/Y max speed changes. Run them at about 7. 800 or so is good for x and y Max accell. Not sure why that 65000+ number is in there. That should be around 1000 or so.

That picture has way out of whack X/Y offsets and tool head offsets.

Up your JKN K and K2 to .007 and .075 if corners buldge some. Or .006/.065

[caall99]

It is best to about halve the X/Y max speed changes. Run them at about 7. 800 or so is good for x and y Max accell. Not sure why that 65000+ number is in there. That should be around 1000 or so.

That picture has way out of whack X/Y offsets and tool head offsets.

Up your JKN K and K2 to .007 and .075 if corners buldge some. Or .006/.065

I agree the firmware values are total garbage. I will be tweaking mine tonight. jfkansas, can you screenshot your settings for us? What are the offsets supposed to be?

X/Y max speed change is the same as Marlin jerk correct? I hear a higher value is smoother around circles. Is this incorrect information?

[jfkansas]

Not sure on the lower vs higher and how they affect circles. My screen shots would be from a Flashforge, not sure if they are doing something wacky on the Qidi. I will know more in about a week when I get my parts and finally get a machine up and running.

[caall99]

Here is what the Sailfish documentation says about "max speed change": "Know that with the maximum per axis speed change parameters, you tell your bot just how much it has to slow down when changing direction". This leads me to believe that with a high "max speed change" value, the nozzle will slow down MORE when changing directions, thereby INCREASING the print quality around sharp turns and corners. Documentation also says that values closer to your current feed rates are good starting values. In addition Sailfish documentation uses this example for "max speed change": For a Thing-o-Matic (a MakerBot printer) reasonable "max speed change" values are: 30 mm/s for the X and Y axes, 16 mm/s for the Z axis, and 30 mm/s again for the A (extruder) axis.

I am going to try higher values for "max speed change" tonight and see what the effect is. If my interpretation is correct, INCREASING this value will reduce bed shaking and further improve perimeter surface smoothness (i.e. no more ringing). If I set them too high, I may start seeing other defects like blobbing.

I will try the following "max speed change" parameters:
X: 25 mm/s
Y: 25 mm/s
Z: 10 mm/s
E: 20 mm/s

Ultimately, once I have these dialed in, with a suitable acceleration value for each axis, I can really start cranking up feed rates.

I am going to leave my JKN values alone for now, and focus on surface finish of perimeters vs. the top sides of my models.

[jfkansas]

Note that, while high values can increase printervibration and even cause it to jerk about, excessively low valuessignificantly increase print times.

Excerpt from the sailfish manual about max speed changes.

[caall99]

Hmmm, then I am confused...

See Step 1.c here: http://www.makerbot.com/sailfish/tuning

Description reads differently than here (4.3.2): http://www.sailfishfirmware.com/doc/...eleration.html

[caall99]

Hmmm, then I am confused...

See Step 1.c here: http://www.makerbot.com/sailfish/tuning

Description reads differently than here (4.3.2): http://www.sailfishfirmware.com/doc/...eleration.html

I believe Jfkansas is correct. Lower values for "max speed change" result in smoother motion based on my limited testing tonight. Can anyone else confirm or deny this information?