So how fast were you able to go on print speed? How noisy is printing? I like the simplicity of the concept just weighing the pros & cons as you push the limits.

 

I am just about to start to think about wiring and I was re reading your setup. I do have 2 external drivers to run the 2 steppers on Z... I am right in thinking that I can just take the 4 pin out EN, STEP, DIR and GND from the Z axis port and split them to my 2 External Stepper Drivers to run each Z Motor. There is nothing I have missed?

I was reading up on using the Z_DUAL_STEPPER_DRIVERS function in Marlin and although that seemed simple enough it does reduce the option of an easy change to Dual extruders in the future.

 

What are your

#define DEFAULT_MAX_FEEDRATE

and

#define DEFAULT_MAX_ACCELERATION

for your X and Y axis with he leadscrews

 

Hello evilc66,

I have started my build. It is very close to your design (thank you!).
So far I have finished the structure (except for the 45deg rails that I am still waiting for), it is super stiff ! Maybe we could use that same chassis to build a (small) CNC machine too ? Note that I don't need one or want one, it is just an idea that crossed my mind because that chassis is so stiff.

Since it's my very first 3d printer, I am taking it very slow to set up the electronics, because I don't want to fry anything.

I use a rumba board with marlin firmware. As you advised me, I would like to use an external driver for the Y axis, to get all the power from the NEMA23 motor. However, it seems it is not possible (or at least not straightforward) on rumba to wire an external driver because drivers are integrated to the board, and there is no direct access to GND, STEP, DIR and EN outputs like on the smoothieboard for example. Maybe I'll just use 4 small pins to get in the GND, STEP, DIR and EN outputs of the Y driver connectors, but I am not very comfortable with that. Do you think it's a good idea ? I really don't want to fry that board...

Finally, I am planing on using an e3d Titan extruder with an e3d V6 hotend. I believe you use the same setup. Could you point out a link to the CAD of the Titan mount to the C-beam gantry (if it's a thingiverse CAD file or similar ? It seems you have a prety nice mount with an integrated duct fan, and I would like to have the same one, if possible.

Thank you !
Jacques D

 

Jacques,

I think you have some miss information the rumba board like most RAMPS based boards are very easy to get the GND,STEP,DIR and EN straight from the driver port.

You could use one of these to just plugin MKS CD 57/86 Stepper Motor Driver Current Expansion Board For 3D Printer | eBay

Or just plug into the 4 pins direct. Here is a good tutorial. Using a Single-Axis TB6560 Stepper Driver with GRBL/RAMPS

Antony

 

Amazing ! I was planning on plugging the 4 pins directly in the driver ports (although I wasn't comfortable doing it), but the MKS board seems way cleaner, I just bought it !

Thank you antonyking !

 

This is what I was intending to do but the MKS board I got already had a pin out for external stepper. There is another guy selling the similar pinout driver boards that was cheaper but I can't find it now.

 

Thank you, evilc66. I will probably explore this solution as well, since I have a BLTouch sensor too.

Jacques D

 

FYI here is my Titan bracket with BLTouch sensor on Thingiverse E3D Titan Extruder Bracket with 3DTouch or BLTouch by ant0ny

It bolts by STD 23 x 23mm pattern. Note this design is for 3mm brass heat inserts.

I will have a write up on my build about the setup... I am just printing a endstop and assemble it today so by tomorrow I'll have pics.

I came across my C-Beam Gantry plate adapter, so I poped it up on Thingi maybe you can use it.

OpenBuilds C-Beam Gantry Plate to Titan Extruder by ant0ny

 

Thank you antonyking for the C-Bean gantry plate ! Downloaded that part, and the same mount as evilc66.

I sent that to a friend who owns a Prusa. He will try to print me that in the next few days.

Tonight I got all the three axis running. Yay ! Baby steps but I'm getting there !
50mm/s so far for X Y. I don't know if I'll go higher.

Thank you guys !

 

evilc 66 how noisy is your Y axis? mine is very noisy and it could be just the fact I have such a large area but it is driving me crazy. I need to update my build but I have printed out parts to run the Y axis by belt to see the difference in speed and noise.

 

hmm ok I'll do another print today and see if it gets better thanks.

 

I think the noise on my printer could come from plays between the bearings and acme screws at both ends of the linear actuator. I have no axial play, but all the screws have radial play at at least one of the end bearings.
jamesdjadam used plumber's tape on the inside and outside of the bearings on his build, and said it worked well : C-Beam XL - TinyG - Aluminium Plate - Dewalt 611
Maybe I'll do the same thing.

 

My motors don't run properly. When I make quick changes in direction (using the LCD controller), sometimes the motors block. It vibrates, makes a lot of noise, but doesn't move.
For example, if I command a movement from +160mm to +200mm back and forth multiple times, at some point the motor will stall. It happens mostly on the X axis, but it did happen on the Y axis (which runs with an external driver).

They run smoothly in both direction as long as I don't command quick changes in direction.
Note that I am still in the testing phase, so they move almost no loads. I still haven't mounted the extruder/hot end assembly, or the glass/heatbed assembly.
The X axis has very little friction. The Y axis has a little more, but nothing that worries me.

For the X axis, I tried to increase the current up to the limit (1A = 1.6V for Vref) on the stepstick A4988. It didn't solve the problem.

Acceleration is at 2000 mm/s/s. Jerk is set at 20mm/s. Are these numbers too high ?

When I lower acceleration, the problem disappears (so I guess the problem comes from here). What are your settings ?

Thank you !

 

Hello,

I tried to command my printer with my computer through Repetier and Pronterface, without any success. It says connecting for a couple seconds, then fail... COM and bauderates (115200) match the setting of my firmware.
Note that the software Arduino 1.8.1 has absolutely no problem connecting to the printer to upload firmware changes.
Additionaly, with Pronterface, when I click on the reset button, the printer does reset... I just cannot connect to control anything else...

The software Arduino is OFF when I try to connect with Pronterface or Repetier (I read somewhere that there could be port access limitations to only one software at a time).
I started both software with admin rights but it did not help.
I tried changing the COM port (COM 3 to COM 1) but it did not solve the problem.
I did not try yet to lower the bauderate.

I don't understand. Do you have any idea on what is going on ?
I am using a rumba board, with Marlin 1.1.0 RC8, on a Windows 10 OS. USB 2 ports.

Thank you,
Jacques

 

Thank you for your replies, I appreciate !

I tried using Arduino 1.8.1 with a 250000 bauderate, but it didn't work, that's why I'm using 115200 at the moment.
Maybe I should try again though because when I first tried 250000, I was only starting to play with Marlin code/Arduino interface. I did not understand anything from what I was doing (it's not much better right now but at least I'm a bit more confident). Maybe I could make it work now, I should try.

When I press the reset button on Pronterface, I see the LCD screen of the printer resetting, and I can hear the motors being powered off and on. So it does look/sound like the printer is actually resetting... However I cannot do anything else : can't get the motors to turn, or send gcode commands (which is what I am trying to achieve).

 

OK I managed to command the printer. I changed the bauderate to 250000, restarted the computer, the printer, and it worked after that. So thank you for the tip !

Now I'm trying to make the autolevel sensor works. It is supposed to be a BLTouch 3DTouch, but it turns out it's a fake one : a TLTouch from Trianglelab. And I didn't read good things about that clone...
When I launch the self test, the servo gets sometime crazy (most of the time would be more accurate), going up and down super quickly. I'll try to play with the little hex screw, but I'm not confident. I did clean the little rod, even with acetone.
Of course, when I try to home the Z axis, it just does not work. I guess due to these flickering of the probe.

I'm a bit disappointed at the website I bought this sensor on (3despana.com). It does say on the description that it is a BLTouch, and never mention it's a copy from trianglelab.
I thought this website was legit. Now I don't really trust the other components I bought from this website, in particular the SSR that runs my 220V heatbed...

Jacques

 

Hello,

I finally got the clone BLTouch to work !

First, I took it appart and clean it completely and thoroughly, including the internals (with stitching thread, just like it's recommanded on the ANTCLABS website, in case of problems with the sensor).
I played with the little hex screw, and at the end I got the pin to act correctly (no flickering). Good !

Then I had a hard time to get the sensor to provide the right info.
It turns out, the clone BLTouch (reminder : a clone from trianglelab sometimes refered as TL-Touch, which is not the same clone as the one from Geeetech) does not have the same logic as the original BLTouch.
So in Marlin, if you use a TL-touch from trianglelab, you actually have to leave commented the #define BLTouch (disabled) ! It took me a long time to try/understand that !
It worked well from there !

From what I understood (but do not quote me on that), the clone TL-touch has an opposite logic to the original BLTouch : it says triggered when the other says open. Now the solution would only be to switch the Z_MIN_ENDSTOP_INVERTING from true to false, or the other way around. But it was not that simple.
The main "problem" is that BLTouch provides the triggered signal for a short time (5 ms) when the rod is pushed, while the clone TL-touch remains open (inverted logic) as long as the rod is pushed.
I believe the fact that the original BLTouch only sends the triggered signal for only 5ms led the Marlin devs to create the #define BLTouch option. I don't understand why, but with #define BLTouch enabled, switching the Z_MIN_ENDSTOP_INVERTING and Z_PROBE_ENDSTOP_INVERTING from true to false (and vice versa) was uneffective. And because the logic of the clone TL-touch is inverted (open state when rod is pushed in), homing wouldn't work.
When commenting #define BLTouch, everything went back to normal : Z_MIN_ENDSTOP_INVERTING and Z_PROBE_ENDSTOP_INVERTING worked again and I got the sensor to behave properly. It took me a long time to figure this out, but now it works !

Evilc66, I hope you don't mind if I use the thread from your build to ask questions and give status about my build. My build is similar to yours, so I feel that it is not necessary to open a new build.
Maybe I will open one when/if i get the printer to work, though.

Jacques

 

Hello,

I finally got my first print : a calibration cube !
Here are pictures of the sixth iteration. Result is okay, but there are two things that bother me :
- on the first picture, the edge on the right is not perfectly straight. If you look closely, it is bent outward at the bottom. Note that it is the only edge that looks like that. I'm guessing a temperature issue here. Bed is around 60ºC and nozzle to 190ºC. I never turn on the duct fan. I'm not sure, but the first layer may be printed at higher temperature (I need to check that setting. I'm pretty sure I changed it, but maybe I didn't save it). If yes, it could explain this.
- on the second picture, the edge on the left is not perfect. Again it is the only edge like that. The 3 other vertical edges look good to me. I have no idea what could be the problem...

Any idea on how to improve the quality of my print ?

Printer settings:
0.4mm nozzle at 190ºC,
Layer height of 0.15mm (first layer of 0.3mm),
bed at 60ºC,
Speed at 20mm/s for the outer perimeter (50mm/s for the rest),
Jerk 15mm/s
Acceleration 1500 mm/s2,
Duct fan off,
Marlin 1.1.0 RC8
Repetier with Slic3r software

Thank you,
Jacques

 

Hello,

Thank you for your help !

I figured out last night what was the problem : it's the heatbed. Now I don't know exactly why it's a problem but I have 2 leads.

On the picture attached, the cube on the right was printed at 220ºC and heatbed at 60ºC. The cube on the left was printed at 200º without bed.
Appart from the major layer problem that appeared toward the end of the print (bug that I still don't understand), all the cubes printed with heatbed ON were looking like the one on the right, no matter the nozzle temperature (from 180ºC to 220ºC).

The cube on the left, which was printed without heatbed, has a better finish, except for two lines (that I also don't understand). It is not perfect, but far better IMO.

Note that the lighting is pretty "tough" for the cubes, and even the cube here on the right doesn't look so bad in real life.

My guess is that these lines, that we see on the right cube, come from the on/off cycle of the (powerful) bed. I measured approximately the duration of a cycle : bed is powered for 5s (rarely up to 10s), and turned off about 35s (sometimes 40s, rarely up to 60s).

It could be that the bed "breathes", i.e. every time heatbed is turned on/off, it goes slightly up or down, messing with the layer height. My bed design is different from yours ; it is far less elegant, and I have measured high temperature differences between the borders (20mm all around the bed are not heated) and the center of the bed. I'm going to redesign it next week.
Or it could be a drop in voltage, due to the suddent power on of the bed (although the bed is powered with 220V and controlled with a SSR). This voltage drop would mess with stepper motors speed and/or accuracy.

I don't know, just hypothesis here. The fact is that it is the heatbed that messes with the print quality.

I'll have to wait for the end of next week to keep testing and working on the print.

Thank you for your help,
Jacques