Re-Arm, any experiences?

I still have to confirm this but i have the feeling that the tilt correction that is done by the G32 endstop correction is influencing the G31 grid compensation.
Meaning my the quadrant that is too high is the same for both G32 and G31, just a bit better for G31.

That's possible, my rearm powered machine is a corexy I haven't tried this board in a delta yet.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

In the process of setting up my Rearm to be transplanted into my delta, I've been distinctly unimpressed with Smoothie's delta support. For example, you can't put the Z height offset for the probe vs the nozzle into firmware anywhere. And the documentation has a sneering tone to it. It also doesn't support three point levelling on a delta.

Not really the best introduction to new firmware

AMEN brother, panucatt hasnt even answered my support ticket since two weeks.

The z offset eludes me still (see my earlier post) but it should theoretically work with the probe z heigtht + your probe offset from earlier firmware.
3 point levelling i think is done via G32 which does endstop correction, so that's kind of the same functionality.

Im doing some more tests today to figure out how g31 and g32 play together.

I have to say there is a distinct difference in tone from Smoothieware compared to Duet. The Duet guys can't do enough to help, but then they are selling all the hardware, there are no cloned DuetWifi boards (yet) and are commercially funded by doing so where as the Smoothieware guys sell some Smoothieboards but a lot of their users are running MKS Sbase. I don't know the commercial relationship between Panucatt and Smoothieware but there must be one.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

DJD,

Where is the RRF documentation for DC42's fork? I've looked at his github, but there's nothing there, not even a pointer to where the actual documentation is. I'm half considering porting it to the Re-arm. I've had a look at the code, but it certainly doesn't look like it has a proper hardware abstraction layer. I noticed a few #ifdef's to do with the RADDS, not really built for portability sadly.


I've got to say, the software ecosystem for 32 bit boards is pretty... limited. You've either got Smoothie with it's lack of proper delta support, or RRF with it's lack of normal LCD support. Neither of them support babystepping, and neither seem to be nearly as developed as the two main 8 bit forks. Maybe I might port Repetier instead.

The documentatin for RRF DC42 branch is the Duet Documentation [www.reprap.org]
1.18beta3 Does support babystepping via gcode or now via paneldue and web interface.

I know nothing about porting software.

Edit yes I meant - [duet3d.com]

Edited 2 time(s). Last edit at 03/24/2017 04:04PM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
nebbian
Where is the RRF documentation for DC42's fork? I've looked at his github, but there's nothing there, not even a pointer to where the actual documentation is.

The master documentation for Duet and RRF is now at [duet3d.com].

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Im testing the calibration at the moment and its working a bit better but I have the feeling the calibration is not all that reliable. Sometimes it actually goes haywire and crashes the probe and other times it works fine.

I tested G32 and G32+G31 and the latter is working better. But G32 alone is not sufficient. It consistently has one quarter of the bed to high and even with G31 you can still see that its not optimal.

and one other thing. I havent found any release notes for master or edge firmware. Whats teh difference there?

No the documentation for each release is lacking. I know the grid levelling isn't in master which is why I compiled edge.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

well thats a bingo. Loaded the firmware to the sd and immediately lost connection to the board and havent regained it

I have tried to change the Mega to Re-ARM today and I have got a problem (not a big surprise )

When I turned the power on, the power led on RAMPS and on Re-ARM turn on for a few miliseconds and then go off. I cannot connect to Re-ARM in this case using USB connection. I have a graphical discount LCD and I also have the adapter for Re-ARM. I have observed that if the adapter is unplugged, Re-ARM and RAMPS can turn on. If the adapter is plugged except the 5 Volts pin, Re-ARM and RAMPS can still turn on and I can connect to the board using USB. If I plug the 5 volts pin then turn on the unit, it does behave exactly like I have described in the beginning of the paragraph. These symptoms make me think there might be a problem with the supplied cable. Any other ideas?

Thanks....

Did your ramps board work normally before, i.e. would work without 5v usb power?

Does your ramps board have the D1 diode, and is it the right way around?

Are you sure you have the 5v extra supply wire connected to the right pin on re-arm?

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

" I have a graphical discount LCD" which model? a real one from reprap discount or the cheap clone from geeetech

The latter has the cables around the wrong way... see [dustsreprap.blogspot.co.nz] for details.

DJD:

My RAMPS did work normally before. I didn't need to plug any power source to the Mega.

The board has the D1 diode and according to the images I can find on Internet, its direction is also right.

Yes, I tried both of the 5 volt options on the back side of the re-ARM, didn't make a difference.

Dust:

I've bought it from Aliexpress . Here is the link : [www.aliexpress.com]
I don't think it is the geetech one, it doesn't have a white pcb.

I did another test. I normally have a voltage adapter which I use to convert 12 volts to 5 volts to be used to power the servo ports (for BLTouch). I unplugged it from RAMPS and plugged it directly to the wire on the re-ARM adapter cable, which is supposed to carry 5 volts to the LCD. It doesn't power on the LCD. This is definitely a cable problem. Maybe I should try your solution but I am afraid of using 7 different wires, as it could be a very unstable solution and could dislocate easily.

That is a 2004 LCD ie a 20x4 character lcd. it is not a graphics lcd

You cannot connect these directly to any form of smoothie board. (including re-arm)
You need to use a interface [github.com] which converts parallel type lcd's to SPI

Edited 1 time(s). Last edit at 03/26/2017 03:58AM by Dust.

I am very very sorry, wrong link. That was a past acquisition and I mixed it up.

Here is the real link: [www.aliexpress.com]

And a very quick update:

I had a spare LCD cable. I cut the same cable that the adapter cable has splitted and put a Dupont connector on it, replaced the adaptor with that new cable and everything works now.

The problem was with the cable.

Now, let's figure out how that BL-Touch thing is going to work. It seems to be a little bit more complicated compared to Marlin, as Snoothieware has some problems accepting the Z min probe also as the servo probe. I assume G29 won't be working now I can issue M280 S3.0 and M280 S7.0 commands to deploy and retract the pins. People seem to use GCode to do the bed leveling.

And just another quick question: My stepper drivers are these : [www.aliexpress.com]

Apparently they support 1/128 microstepping. Shall I go on and set them at that level? Can the Re-ARM cope with the required amount of pulses? I was using 1/32 microstepping on Mega so they are set at 1/32 right now.

I have z min endstop and probe configured as the same pin it doesn't seem to mind despite the docs saying you shouldn't.

As for step pulses that's defined in config as 100khz, so you can calculate how fast you can go for any given microstepping. I don't know if it can go faster.

Edited 2 time(s). Last edit at 03/26/2017 05:00AM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

DrMaestro,

From what I've read about the Re-arm, 1/64 should be a good starting point to balance speed and noise/accuracy.

I figured out how to start printing. I have to use this gCode:

M280 S3.0 ; BLTouch Pin down
G32 ; Bed mapping
G1 X123 Y0 F3600 ; Go to the first point of the 3 point leveling strategy
G30 Z1.4; Set Z offset
M280 S7.0 ; BLTouch Pin up

My 3 point leveling strategy has homing X and Y set to true, so I don't need to home the X and Y axes first. There is no Z homing, you just send the head to the first point of the 3 point mapping coordinates and you have to set a Z offset (as BLTouch has a Z offset). Interestingly, in Marlin my Z offset was -1.40, so I assume here the logic is inverted.

That works quite well.

My actual problem is some intermittant freezes. The board seems to freeze at a random interval, approximately 20 minutes after being turned on. I don't know if this behaviour will happen when I am printing (I am testing it now).

I am still at 1/32 stepping, just to be sure that everything works.

Update:

The prints didn't seem to freeze for now (the print job took 2 hours).

I changed to 1/64 microstepping. The motors are very very silent now. I can compare it to TMC 2100s (but my LV8729s are very cool while TMCs got so hot to touch despite my 80 mm fan cooling RAMPS).

Print quality wise, I am disappointed a little bit. I don't have any worse quality prints at all but I was expecting to see some improvements on surface quality. It seems to be exactly similar to what I was getting on RAMPS, including som odd behaviour that I menitoned on [forums.reprap.org] I think it is mostly related to the slicing software's behaviour (which regulates nozzle pressure). Maybe the results will be different on a delta printer.

So is it worth it to upgrade? In my opinion it is if you want a really silent printer but don't really expect better quality prints on a cartesian printer. However those random freezes and some artefacts (horizontal lines appearing on the LCD) on the LCD are some concerns. If it ever freezes during a print, I'll be going back to Mega.

Edited 1 time(s). Last edit at 03/26/2017 11:16AM by drmaestro.

Quote
drmaestro
However those random freezes and some artefacts (horizontal lines appearing on the LCD) on the LCD are some concerns. If it ever freezes during a print, I'll be going back to Mega.

Nice work on getting it up and running Mine's still sitting on the bench as I work through how to get it fully functional.

Regarding those horizontal lines, are you printing using the external SD card? I noticed those lines when I was playing around with reading files off that card.

Quote
nebbian

Quote
drmaestro
However those random freezes and some artefacts (horizontal lines appearing on the LCD) on the LCD are some concerns. If it ever freezes during a print, I'll be going back to Mega.

Nice work on getting it up and running Mine's still sitting on the bench as I work through how to get it fully functional.

Regarding those horizontal lines, are you printing using the external SD card? I noticed those lines when I was playing around with reading files off that card.

Exactly. Those lines only appear if I print using the LCD's SD card or if I turn the encoder while the SD card is in or printing. It is quite interesting to see that Smoothieware's LCD support is worse than Marlin and they actively discourage printing from the SD (which also works perfectly on Marlin).

By the way, I had a layer shifting issue yesterday. I haven't seen how it happened (I wasn't supervising it) so I don't know if the stepper driver overheated and did a thermal shut down or if the torque was not enough. The motor was cold to the touch, so I supposed it could be the second one but then I am at 1/64 microstepping so the torque could be the issue.

Edited 1 time(s). Last edit at 03/27/2017 06:41AM by drmaestro.

Its hard not to conclude here (even though my re-arm is working and print quality is good) that given some of the struggles people seem to be having:
1) if you are on a tight budget go with decent Ramps/Mega and possibly Repetier firmware for improved performance over marlin, Using some of the newer stepper driver chips and potentially 0.9 deg motors will improve your print quality.
2) If you are not on a tight budget get a Duet board, where none of these issues are present, print quality is amazing and the support and community involvement is second to none, right now users get to vote on new features being added to the firmware.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

I have the lines as well, but they dont seem to influence the print. I did some prints from the sd card now and it appears reliable.

Quote
DjDemonD
2) If you are not on a tight budget get a Duet board, where none of these issues are present, print quality is amazing and the support and community involvement is second to none, right now users get to vote on new features being added to the firmware.

About print quality on Duet: I don't have one and cannot buy one to compare it, because apparently every seller accepts only PayPal, which we can't use in my country since June 2016. Would be nice to have a seller who accepts credit cards. I understand from DC42's comments elsewhere that it doesn't use approximation for calculating movements, so it seems to be more precise compared to RAMPS but doesn't Smoothie also use a more exact movement calculation algorithm (I have no idea and would like to know it)? If it does, and those two algorithms are not very different, I am still skeptical about improved print quality, as I don't see a definitive improvement on re-ARM.

Edited 1 time(s). Last edit at 03/27/2017 04:20PM by drmaestro.

Quote
drmaestro
About print quality on Duet: I don't have one and cannot buy one to compare it, because apparently every seller accepts only PayPal, which we can't use in my country since June 2016.

I suggest you create a thread about this here [www.duet3d.com]. Payment by wire transfer may be possible.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Hi that's a shame about duet board. Where are you based? I'll take your card payment if you want me to buy one and send it on to you.

I have noticed improved print quality with rearm. But duet is a level above especially for deltas. Perhaps I see it more as the duetwifi is running my kossel Xl which was a no expense spared build where my corexy is only middling in terms of precision.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
dc42

Quote
drmaestro
About print quality on Duet: I don't have one and cannot buy one to compare it, because apparently every seller accepts only PayPal, which we can't use in my country since June 2016.

I suggest you create a thread about this here [www.duet3d.com]. Payment by wire transfer may be possible.

I'll do that, however wire transfer is generally expensive (the last time I checked, it required around 30 USD as a transfer fee).

Quote
DjDemonD
Hi that's a shame about duet board. Where are you based? I'll take your card payment if you want me to buy one and send it on to you.

I have noticed improved print quality with rearm. But duet is a level above especially for deltas. Perhaps I see it more as the duetwifi is running my kossel Xl which was a no expense spared build where my corexy is only middling in terms of precision.

I live in Istanbul (Turkey). How do you accept card payments? Do you own an online shop?