3DR "Simple" Delta Printer

Hi CAD321,
Great contribution documenting all the parts in the spreadsheet format. Wish I had that when I built mine earlier this year.
I bought a Printrbot Simple kit without the wood just before they stopped selling them for $249, came out with the all metal PB Simple and raised their prices by $100 for a kit with the metal.

You might consider adding a little more detail to your list. For example, the carbon fiber rods from TriDprinting are 180mm.
Misumi will give you the option of shipping everything as one package, so you don't pay as much.
McMaster Carr on the other hand will ship everything separately if the availability dates are different and so I ran up a big shipping bill for a bunch of small parts that were all dated differently.

The Printrboard is a very common board to use with RepRap printers of the cartesian design.....the ones that are not Delta design.
With the Delta design you will need to be able to upload revised firmware several times. With my MAC, and already having a Printrbot Simple, I was trying to connect two Printrboards and my MAC did not like that.
If you have a PC with Windows, you will need to download some software and get the right drivers to connect the USB to a COM port....but you will figure it out, maybe you already know this.

I really like the Ubis hot end from Printrbot, so if the Anubis is a Ubis clone, you will probably like it.

With the 3DR Delta, you have fixed dimensions, with the printed plastic parts in the horizontal dimension.
One option you might consider is building a Mini Kossel using the QU-BD parts plus the Mini Kossel plastic parts and aluminum extrusions for the horizontal and vertical parts.
But that would mean not using the bearings on smooth rods and instead using linear rails and carriages.
The Mini Kossel design is scalable. So the extra cost of the linear rails gives you the option of ordering longer horizontal extrusions.

Well, I found the 3DR Simple was less expensive to build than the Mini Kossel, but part of that was that I already had two 3D printers and printed all my own parts.
Also, I got the Printrbot parts cheap before they raised their prices.

I like your plan because you get the parts to make a Delta cheap, print your own plastic parts and buy a few other items.
You might be able to do it for less than $500. I estimated it cost me $650 to build my 3DR Simple.
Take Care

CAD321,

One other thing about the 3DR Simple is the choice of Spectra line versus belts.
I built mine with Spectra line and have the belts and pulleys to upgrade but have not done that yet.
Some people have complained about skipped steps, loose spectra line that can not get tight, and the plastic pulley spinning on the motor shaft.
I have to agree that with the spectra line there are skipped steps.
I was very lucky to be careful when winding my line onto the plastic pulleys and followed the directions exactly, not something I do well.
I did have some difficulty getting the spectra line tight enough.
So, anyway, you will want to plan on upgrading to belts. Otherwise, skipped steps and loose spectra line.

Quote
CAD321
I've ordered myslef a QU-BD one up and given its similar build volume to the printrbot simple and low price point, am considering converting it to a 3DR simple. However I am on a budget (as I'm sure many are) and was curious to know what the price would be to upgrade. After looking around I was unable to find a full BOM so I decided to make one in Excel. I thought I would share it here in case others find it useful. It includes everything John has posted in his tutorials such as printed parts, electronics, pricing, where to buy etc... The only thing not included are the parts for the extruder as there are so many different options out there. If you notice an issue with the BOM or find something is missing, please feel free to mention it so that myself and others have an up to date BOM.

Nice work on the spreadsheet!

I don't know what everyone else's experience has been, but my Misumi smooth rods have started to rust. I've since added some oil to the rods to clean them up and prevent additional rusting, but you may want to look to see if they have rust-resistant material available or source them from a different vendor. Granted, I am in the high-humidity environment of South Florida, but I have smooth rods on my RepRapPro Mendel as well and they've been in use for over a year now without a spot of rust on them. The only other thing I can think of is that the smooth rods were a very tight fit in my frame and I took off some kind of finish or scuffed them up during assembly.

As an unrelated side note, I've put up a blog post with some pictures and videos from my build for anyone who is interested: [www.jiggywatts.com]

Edited 3 time(s). Last edit at 05/05/2014 11:33PM by PierreB.

Quote
RTurnock
CAD321,

One other thing about the 3DR Simple is the choice of Spectra line versus belts.

I have changed the spreadsheet to allow for the user to choose between either belts or spectra line. The areas highlighted in yellow indacate a customizable option.

Quote
RTurnock
You might consider adding a little more detail to your list. For example, the carbon fiber rods from TriDprinting are 180mm

I dont actually have the carbon fiber rods from tridprininting in the spreadsheet. The rods themsleves that are listed are from McMaster Carr where as the rod ends are from tridprinting. I did this to allow for flexablility when determining rod lengths.

Quote
CAD321

Quote
RTurnock
You might consider adding a little more detail to your list. For example, the carbon fiber rods from TriDprinting are 180mm

I dont actually have the carbon fiber rods from tridprininting in the spreadsheet. The rods themsleves that are listed are from McMaster Carr where as the rod ends are from tridprinting. I did this to allow for flexablility when determining rod lengths.

Great, thanks for updating the spreadsheet.
My mistake trying to eyeball the spreadsheet and not aligning the item with the source properly.
Thanks for showing that option.
I have ordered carbon fiber rods from TriDprinting that are 0.227 inches OD and you list the McMaster Carr at 0.156 inches OD. Just an observation. I don't know if that is OK or not.
TriDprinting has 12 inch CF rods that I bought, added Traxxas rod ends and have 340mm diagonal rods.

Thanks for doing all the work to document the 3DR Simple.
Great for others to follow.

MakerFaire, here we come. For those of you going to MakerFaire this weekend, I'll have my 3DR "Simple" Delta printer there. You can find me in booth 403. I'll be showing my printer, as well as a circuit board that a friend of mine and I designed for automatic bed leveling with Force-sensitive resistors.

Hi all,

I finally got time to work on my 3DR simple after two months. It's working now. Not perfectly, but making shapes. What are the recommended retraction settings for this machine?

Quote
umdpru
Hi all,

I finally got time to work on my 3DR simple after two months. It's working now. Not perfectly, but making shapes. What are the recommended retraction settings for this machine?

I have slic3r configured to retract 3mm at 30mm/s.

Thanks, Pierre. I seem to recall that bowden fed extruders needed a longer retraction distance?

@umdpru

In my limited experience seems like the extruder retraction distance and speed depend on the hot end characteristics. I followed Jay Couture's experience with the E3D hot end to remove a jumper to get 1/8 stepping on my extruder motor, doubled the steps/unit, double the speed and increased the distance retracted. All of this to make sure the heat creeping up the E3D hot end did not make the PLA expand and clog the hot end. The Bowden feed results in needing a high torque motor on the extruder. either a geared motor or a Greg's Wade extruder. Having a weak spring on the extruder that does not push the bearing hard enough against the filament so the gear can bite into it will result in problems at the hot end. Retraction distance is a minor problem unless you have a sensitive hot end like the E3D.

Thanks RTurnock. I have a J-Head Mk IV from Makerfarm on my machine. I'm using a geared Wade's extruder.

Does anyone have a problem with the upper end of the machine oscillating with print head direction changes? Mine can oscillate quite heavily during print head changes. I'm thinking of manufacturing a base and adding triangular supports for the arms to stop them from oscillating. It's introducing print irregularities.

Edited 1 time(s). Last edit at 05/23/2014 09:01AM by umdpru.

@umdpru,

Hard for me to imagine oscillating at top of 3DR Simple as designed by JohnSL.
three vertical 2020 extrusions, two smooth rods on each side of each extrusion tightly fit into plastic.
Mine is rigid.

Is your's printed in PLA or ABS? I wonder if I didn't use enough infill. I'll make a video once I have it printing something again.

In any event, I'm going to move on to calibrating the printer.

I'm following this guide here:

[minow.blogspot.com]

How are we 3DR users supposed to calibrate the end-stops without adjustment screws? I'm going to have to learn about M666, I guess.

@umdpru

Mine are PLA. Yes, the amount of infill makes a big difference whether ABS or PLA. I bought a set of ABS parts for mini kossel that cracked and broke because they had been printed at 20% or less.
You need to print with at least 30%.

Yes, you need to use M666 after tightening down your endstops. Yes, follow the minnow.blogspot instructions.

Thanks. I printed mine at 50% infill in ABS. I may take the pieces apart and glue them together. ABS isn't anywhere near as rigid as PLA is so that oculd be some of what I'm seeing. The assembled pieces are flexing somewhat and I do have my filament reel mounted on top further exacerbating the problem.

I have found this calibration "tutorial" that I'm trying to get through now.

[reprapandme.blogspot.com]

The major steps I have done so far are to set MANUAL_Z_HOME_POS so that I can barely get a piece of paper between the nozzle and the print surface.

Then I have moved onto the X tower and used M666 commands to offset the end-stop to get the nozzle height calibrated.

Then I move on to the Y tower and get the endstop offsets done. Then I move to the z tower and M666 has not effect there at all. The nozzle hits the print surface and end-stop offsets aren't changing the height of the nozzle at Z0. Then, I go back to X0 Y0 Z0 and now the MANUAL_Z_HOME_POS is too long (the nozzle hits the print surface).

Is this an iterative process that needs to be gone through a few times to nail down? Am I missing something here?

EDIT: I double checked the physical build measurements of the machine and everything appears to be spot on.

Edited 1 time(s). Last edit at 05/23/2014 04:17PM by umdpru.

Yes, iterative to the point of exhaustion. But it must be done. You have to get the manual leveling done at least to the point where it is close enough for the auto probe leveling to work OR like me, no auto leveling yet, although I have had it working with FSRs then pulled back to fix other problems.
I use mini POSTIT notes that are 0.1mm thick and easy to hang onto with the little bit of glue.
HInt, I found that I had to set the Manual Z home Pos to LESS than what I measured. I have no use for over 200mm of height to print so it was easy for me to give up a few mm.
Then you can set the M666 values as negative. Like this: M666 X-0.50 Y-0.30 Z-1.40 OR even higher values depending on how much to decrease your Manual Z Home Pos.
Here is the exhausting process I have gone through, although others might not agree with the exact sequence, this is the general idea.
Measure, adjust with M666 X-1.5 (or your value) then M500 to save in EEPROM and M501 to confirm.
Your firmware must have the two EEPROM #define lines NOT commented out.

Iteration sequence:
1. measure and set M666 X value
2. measure and set M666 Y value
3. measure and set M66 Z value
4. go back and adjust X value
5. adjust Y value
6 Adjust Z value
Repeat until they are all 0.1mm

Move the nozzle tip to 0,0,0 the center.
Measure the center gap......change the firmware the value of delta radius by either changing the smooth rod offset or by directly adding or subtracting from delta radius. compile and upload the revised firmware.

Repeat the Iteration sequence. Repeat adjusting the center gap.

Then work on getting the plastic to stick to the surface. If the nozzle is too close then the plastic can not come out, if too far, there is no "squish" and the plastic won't stick.

Yes, Jay Couture's tutorial is excellent.
>>>>>>>
I have found this calibration "tutorial" that I'm trying to get through now.

[reprapandme.blogspot.com]
>>>>>>>>>>>

That is basically what I have done with the exception that I first set the MANUAL_Z_HOME_POS to the thickness of a piece of paper (call it the paper gauge here forward). that left me with MANUAL_Z_HOME_POS = 124.2. Then I moved to X and then to Y and got the paper gauge set.

Then at Z, the M666 command to offset the end stop has ZERO effect on the height of the nozzle. I mean, I set it to Z-5 and Z5 and confirmed it with M501 then went to Z0 and the nozzle goes to the same height, regardless. Is that strange, or what?

the minnow blog threw me for a loop when it discussed changing screw adjustors for the endstops so I moved on thinking it was outdated. I realize now that M666 is in effect adjusting the same thing just in firmware. I need to figure out why the Z tower is giving me a problem with M666.

Edited 1 time(s). Last edit at 05/23/2014 04:38PM by umdpru.

Ok. Here's an update. I'm magically not having the issues with the Z tower. In fact, the arbitrary Z-5 worked out to get it right where it needed to be.

Now I'm going to move on to adjusting smooth rod offset because the nozzle is "below" the build plate by about .4mm at the center.

Maybe you could attach a copy of your configuration.h file to help me understand your printer better.

Hmmmmmm, the MANUAL_Z_HOME_POS = 124.2 seems way too short. My 3DR Simple has over 200mm height.
I would like to see your configuration.h file.

My printer is short. It was a mish-mash of John's design and Rich Raps. The extrusions are 500mm and the effector arms are 200mm. I think I could probably shorten those up and get better height. That MANUAL_Z_HOME_POS was determined from measurements and M114 to verify. Here's a copy of my Config.h. I went through a lot of pains to get Marlin to work. Keep in mind, that I'm using Ramps 1.4 not printrboard, too.

Thanks for taking a look and thanks for the help.

I'm getting closer to where I need to be. the first layer on prints is still too close to the build surface. I've iterated twice on the smooth rod offset and then tower calibrations. How much should I be messing with smooth rod offset? I decreased it twice in .5mm increments then calibrated the towers. The tower offsets changed slightly but not by much. I **AM** a bit concerned that the Z-tower offset is so much more than the other two.

HEre's the EEPROM echo

Quote

SENDING:M501
echotored settings retrieved
echoteps per unit:
echo: M92 X55.50 Y55.50 Z55.50 E610.00
echo:Maximum feedrates (mm/s):
echo: M203 X500.00 Y500.00 Z500.00 E29.00
echo:Maximum Acceleration (mm/s2):
echo: M201 X500 Y500 Z500 E380
echo:Acceleration: S=acceleration, T=retract acceleration
echo: M204 S380.00 T380.00
echo:Advanced variables: S=Min feedrate (mm/s), T=Min travel feedrate (mm/s), B=minimum segment time (ms), X=maximum XY jerk (mm/s), Z=maximum Z jerk (mm/s), E=maximum E jerk (mm/s)
echo: M205 S0.00 T0.00 B20000 X20.00 Z20.00 E15.00
echo:Home offset (mm):
echo: M206 X0.00 Y0.00 Z0.00
echo:Endstop adjustement (mm):
echo: M666 X-1.10 Y-0.80 Z-4.85
echoID settings:
echo: M301 P61.25 I5.34 D175.60

Edited 1 time(s). Last edit at 05/23/2014 05:35PM by umdpru.

You could lower your DELTA_SEGMENTS_PER_SECOND 160 down to 100 and still be OK. This can increase the speed of the processor to calculate results.

Well, this has become frustrating. For some reason, the M666 commands aren't having any effect on the Z tower offsets. At. All. I go from Z10 to Z-10 and there is absolutely no change in the offset on the tower.

Is there some setting in config.h that I am overlooking?

You have 500mm X 240mm frame, right? the horizontal extrusions are 240mm?
According to Jay Couture's Kossel Calculator:
"DELTA_RADIUS:", 104.555, "mm"
"DELTA_SMOOTH_ROD_OFFSET:", 151.555, "mm"
"DELTA_DIAGONAL_ROD:", 202.699, "mm"
"DELTA vertical length:", 173.653, "mm"
"Delta_rod_angle:", 58.9483, "mm when homed"
"Build plate radius:", 84.2773, "mm"
"Max Build Height:", 153.027, "mm

You wrote your diagonal rods are 200mm, that fits the above design.
Your max build height should be closer to 150mm but it is 124mm, hmmmm, I wonder why?

One of your problems is going to be that the Delta rod angle when homed is 58.95 degrees.
All of the calculations in Marlin assume the angle is exactly 60 degrees.
In addition to adjusting the M666 values and the Delta Radius, you are probably going to have to adjust the delta diagonal rod length to get the calculations in Marlin so that the angle is closer to 60 Degrees.

Here is the process your are using, right?
M666 Z-10.0 // after the Z carriage hits the endstop it will go down 10mm, right? if not you need to use M119 to test your Z endstop. While holding the endstop switch, and issue command M119. let go of the switch and use M119. You should get "Triggered" and "Open" respectively.
M500 saves this value to EEPROM
M501 asks the Mega2560 to list all the values saved in EEPROM

In configuration.h you have uncommented EEPROM both of the lines, so that is correct.

Ah, there it is. You have:
// Horizontal offset from middle of printer to smooth rod center.
#define DELTA_SMOOTH_ROD_OFFSET 132.54 // mm (original 133.54)
As a result, your Delta Radius is currently 132.54 - 22.87 - 16.5 = 93.17

Your DELTA_SMOOTH_ROD_OFFSET should be 151.555 or use 151.0.
So that your Delta Radius is closer to 104.555.
your delta radius is 10mm too low.
problem solved.

I have a 3DR simple by JohnSL not a mini Kossel. The firmware was originally for a mini-kossel because the firmware used by JohnSL RichRap on his 3DR didn't support M666 as he used hall effects for the end stops which contained adjustments. Mine doesn't have any manual endstop adjustments so M66^ is the only way.

The original smooth rod offsets are taken right from JohnSL's 3D CAD model. Here is the blog post [trains.socha.com].


The process I am using to get the offsets correct first is as follows:

1. Set MANUAL_Z_HOME_POS such that a paper fits in the gap with little drag.
2. move carriage to a point right in front of X tower at "Z=0"
3. Jog nozzle manually until paper fits in gap (~.1mm gap).
4. Issue M114 to get currect Z height.
5. Use Z height to alter endstop offset such that Z=0 is at the print surface (.1mm gap) using M666.
6. M500 to save.
7. M501 to confirm.
8. Home
9. Repeat tower z=0 to confirm gap.

I then move on to the y tower and repeat the same process.

Lastly, I go to the Z-tower but now the M666 endstop offsets aren't having ANY effect on the actual offset distance. The nozzle still goes to the same point everytime regardless of what I enter into "M666 Z ". I can enter M666 Z10 or M666Z-10 and the nozzle will still go to the same exact height as if the endstop offset hasn't been accepted even though M501 confirmed the new Z offset.

I have triple checked the physcial build measurements at the Z-tower to see if something was wonky and it's all the same on all three towers. The physical part of the build is correct.

I just found something interesting. I have commented out the min end stop pins in pins.h to disable them (used -1) and I just confirmed that fact.

M119 with teh nozzle 50mm down shows all max pins untriggered, however, the Z-min pin reports as triggered.

That's strange, right?

EDIT: I just tried compiling with "DISABLE_MIN_END_STOPS" uncommented to disable minimum end-stops and I got a compile error referencing auto bed leveling. That clued me in that I had auto bed leveling enabled and I'm not using it. I commented out "define ENABLE_AUTO_BED_LEVELING" and now I get a compile error referencing "accurate_bed_leveling_points". Commenting out "enalbe accurate_bed_leveling" still gives the same compile error.

I'm going to go get the original Marlin from ErikZalms Git and upload that. I think I might have shot myself in the foot by trying to force a mini kossel firmware onto my 3DR.

EDIT2: It appears I was smarter than I gave myself credit for. the firmware I am working with did come from Erik's Git.

why can't I compile with auto bed leveling commented out?

Edited 2 time(s). Last edit at 05/24/2014 12:29PM by umdpru.

My 3DR Simple firmware has auto bed leveling commented out.
//#define ENABLE_AUTO_BED_LEVELING // Delete the comment to enable (remove // at the start of the line)

Here is my configuration.h for my 3DR Simple. But I am currently having a problem with it starting a print and then stopping....not sure if I have a wiring problem or a firmware problem.
Configuration.pdf