Folger Tech Kossel Delta 2020 Full 3D Printer Kit w/Auto-Level

Quote
thevisad
This is the same exact thing I have an issue with. It is the servo motor causing the issue, drawing too much power and crashing the ramps machine.

In my case it's not a matter of drawing too much power, but the arm hitting the end effector and not able to complete the motion. I mounted my servo so it stores alongside the hotend rather than sticking out the back. I had to adjust the X and Y in Z_PROBE_OFFSET accordingly.



Edited 1 time(s). Last edit at 04/10/2015 06:12PM by Deguello.

Just ordered! Now the wait begins.

Finally we have lift-off - atleast in a postal sense. Placed my order on the 23rd March through ebay. To my surprise, Shipping notice issued on the 8th April and would you believe it today it's on the move..........

Just thought that I would share my ordering timescale incase it's helpful to someone

Quote
rotten_carrot
Finally we have lift-off - atleast in a postal sense. Placed my order on the 23rd March through ebay. To my surprise, Shipping notice issued on the 8th April and would you believe it today it's on the move..........

Just thought that I would share my ordering timescale incase it's helpful to someone

Ordered mine on the 18th March, got my shipping notice on the 8th of April and mine was on the move yesterday, should have it Monday morning,hope everything is in the box.....

Quote
3D52

Quote
rotten_carrot
Finally we have lift-off - atleast in a postal sense. Placed my order on the 23rd March through ebay. To my surprise, Shipping notice issued on the 8th April and would you believe it today it's on the move..........

Just thought that I would share my ordering timescale incase it's helpful to someone

Ordered mine on the 18th March, got my shipping notice on the 8th of April and mine was on the move yesterday, should have it Monday morning,hope everything is in the box.....

Good stuff! It would certainly sound like Folger were genuine when they said that the most recent delay was caused due to waiting for parts to come in. They must have had orders all ready to go aside from whatever they were waiting on. I certainly hope that yours has everything in the box since, having been put together at around the same time, I'm guessing that means mine is more likely too (or atleast that makes sense in my mind ). You'll definitely be receiving yours before me as I'm in the UK - though I'm hoping that any missing parts wont be too big of a problem. I ordered a 2004 LCD display at the same time as my delta (before realising that they don't combine postage - atleast not to Europe) and it turned up in about five days which I thought was pretty good going

Thought id give some feed back on my build.

The printer arrived all parts included except for the extra 10mm screws as mentioned though didnt pose a problem. There are difference between the pdf manual and the online video ie; the direction stepper motors are wired, the probe arm, etc

The extruder need a bit more voltage as mentioned in earlier posts, deffinately calibrate your stepper drivers mine were all about .65mv by default(no heating problems for me). Also the extruder was slipping which was a result of the gear not being tight enough on the shaft, eluded me as not very noticable at first. Also honed the extruder injected housing to give plenty of room for gear as it was rubbing.

I didnt have alot of success with the auto calibration, to many variations in the results i think there is to much slack in the server arm. Also switched to a version of marlin that supports the m666 m665 command [github.com] with better success , though still not accurate enough. ( for me)

Added springs between bed and frame to give more granular calibration and followed this guide: [www.youtube.com]

After doing manual calibration not only did I gain valuable knowledge but have had excellent results. dosnt take long and far greater accuracy than auto leveling.(for me)

Glue stick has been giving good result for first layer adhesion.

I should mention that one of the rails was running rough, followed and earlier suggestion and cleaned with wd40, then cleaned and applied white lithium grease to rails so far no problems.

After a long wait due to delays, I was knocking out prints a couple days after arrival. all in all pretty happy.

Edited 1 time(s). Last edit at 04/13/2015 09:02AM by vorpal.

Im currently stuck at the end of the build process.

for some reason none of my stepper drivers are anywhere close to the proper voltage. the highest i can get them is about 0.003v .

the power supply is geting 12.3 v in and im reluctant to keep going untill i can resolve this .
[www.youtube.com] - Video of me checking them all right away.


Also i didnt get a ceramic screwdriver to adjust the pot so ive been doing it with something less then ideal.


Anyone had these sorts of voltage problems?

make sure your multimeter is set to millivolts and the power supply is on

I cant get anywhere close to X – 0.350V closest I can get is like 0.003v

Quote
Deguello

Quote
thevisad
This is the same exact thing I have an issue with. It is the servo motor causing the issue, drawing too much power and crashing the ramps machine.

In my case it's not a matter of drawing too much power, but the arm hitting the end effector and not able to complete the motion. I mounted my servo so it stores alongside the hotend rather than sticking out the back. I had to adjust the X and Y in Z_PROBE_OFFSET accordingly.

servo1.jpg

Quote
Deguello
. Kicked off a new print. Watched the bed heat up (a bit quicker), then it ran autolevel, then started heating the hotend. About that time pronterface lost communication.

After watching it crash every time the hotend started heating, I decided I must have fried either the Arduino or the RAMPS. I had the new set from Folger, so I swapped first the Arduino, then the RAMPS. Same problem every time./quote]

Actually, this is the same exact problem. I have sorted it out already, but glad your issue is different

Quote
vorpal
Thought id give some feed back on my build.

The printer arrived all parts included except for the extra 10mm screws as mentioned though didnt pose a problem. There are difference between the pdf manual and the online video ie; the direction stepper motors are wired, the probe arm, etc

The extruder need a bit more voltage as mentioned in earlier posts, deffinately calibrate your stepper drivers mine were all about .65mv by default(no heating problems for me). Also the extruder was slipping which was a result of the gear not being tight enough on the shaft, eluded me as not very noticable at first. Also honed the extruder injected housing to give plenty of room for gear as it was rubbing.

I didnt have alot of success with the auto calibration, to many variations in the results i think there is to much slack in the server arm. Also switched to a version of marlin that supports the m666 m665 command [github.com] with better success , though still not accurate enough. ( for me)

Added springs between bed and frame to give more granular calibration and followed this guide: [www.youtube.com]

After doing manual calibration not only did I gain valuable knowledge but have had excellent results. dosnt take long and far greater accuracy than auto leveling.(for me)

Glue stick has been giving good result for first layer adhesion.

I should mention that one of the rails was running rough, followed and earlier suggestion and cleaned with wd40, then cleaned and applied white lithium grease to rails so far no problems.

After a long wait due to delays, I was knocking out prints a couple days after arrival. all in all pretty happy.

With the current version of the Marlin, there is a huge flaw in the auto leveling system. I have rewritten part of it to resolve part of this issue, but need to go back through it again. The auto leveling only levels a spot on the bed every 3 spaces, this causes the system to skip over large portions of the bed. This is not an issue in itself, as you cans imply extrapolate the level between the points. Except, that is not what Marlin does, it copies the location for coord x to coord y. This causes the bed to look like steps instead of a slope.

Some upgrades done to the system this weekend: dropbox (https://www.dropbox.com/sh/omyysmqwmxj0tu8/AABlddOcWMMlBA8e5cPoPXtea?dl=0)

1: Added fans on each of the stepper motors. Pending the creation of a new motor block with fan exhausts out the sides, LED (for visual temp monitoring) and Thermistor mounts and monitoring software written into Marlin. Added in a POT to slow.stop the fans during startup phase. This will be switched to being managed by the RAMPS via the monitoring of motor temperature.
2: Found the bed would no longer heat past 84 celcius with the fans located where they were. Added a 3/8 plywood and a single layer of motorcycle wrap (http://www.jpcycles.com/product/161-452) sandwiched in between the board and the heatbed. Bed now properly raises to 106, is insulated and the RAMPS is protected properly. The underside of the board at the RAMPS is now running 31-33 degrees celcius after 1 hour of printing, instead of the 80+ it was getting before..
3: Added a 8 inch Hobby Lobby mirror to the heat bed, with silicon heat sink compound spread through the locations. This sits perfectly between the screw posts on the hotbed.

Prox Sensor/Fotek Sensor Update:
The new prox sensors came in and they are just not going to work at all, it was a great idea that petered out when I couldn't get a reliable reading. The 5mm Fotek sensor I bought was a bust as well, it won't detect the bed under the mirror and only detects the bed when I am right ontop of it. I ordered a 15 mm and some pressure resistors to play around with.

Edited 1 time(s). Last edit at 04/13/2015 11:02AM by thevisad.

Quote
vexon
I cant get anywhere close to X – 0.350V closest I can get is like 0.003v

You could be at the point where your POT needs to be dialed up or a bad board, do you have a second one to test with?

Quote
vexon
I cant get anywhere close to X – 0.350V closest I can get is like 0.003v

If it's just one driver that won't adjust you can swap the one from the E1 position. If it's all of them you either have a bad RAMPS or the power connector isn't wired correctly. Make sure you have plus, minus, plus, minus (left to right) on the power connector and measure 12 volts on each of the red wires.

no i dont, i honestly feel like redoing all of the underbelly and putting the board in a seperate case.

Quote
vexon
no i dont, i honestly feel like redoing all of the underbelly and putting the board in a seperate case.

That is a great idea, look at my upgrades above. You can do exactly the same thing with a silicone pad (ALA [www.amazon.com]) that it is sandwiched (BOARD/SILICON/HOTBED/THERMALPASTE/GLASS). This stabilizes the heated surface, provides an even temprature across the bed surface, prevents air from quickly lowering the bed and prevents the RAMPS/motors from being cooked by the hotbed. The board I used was too thick, it was what I had on hand. I lost an overall of 11.2 mm total with the new bed/glass setup but it is a rock solid heating solution now. I used motorcycle pipe tape (can get to 1500 Celsius without failure) and you can put your hand directly on the wrap with a bike running. I dropped a total of 70ish degrees on the underside of the hotbed).

Quote
vexon
no i dont, i honestly feel like redoing all of the underbelly and putting the board in a seperate case.

Here is a nice video of the testing of the RAMPS setup [www.youtube.com]

Quote
vorpal
I didnt have alot of success with the auto calibration, to many variations in the results i think there is to much slack in the server arm. Also switched to a version of marlin that supports the m666 m665 command [github.com] with better success , though still not accurate enough. ( for me)

I tried to convert to the RC Marlin as well. Got it almsot all the way there, switches all worked, and it would home fine. When I run the G30 A D250.0 command to autocalibrate it would probe and probe and probe and after about 2 minutes it times out saying it is aborted because my Z endstops were too far apart (listing values of -17.xx, -17.xx, and 25.xx). I went to start a test print and it drove the effector down into the bed on the 2nd probe point, as if it thought the bed were tilted at something like a 45 degree angle, luckily no damage as the arm slipped on the servo gear, and I had put cardboard over my bed to protect it.

This is where I stopped because it was 4am and I needed to sleep, did you run into this issue? Would you care to share your config.h so I can take a shot and see what happens because I am sure I missed something somewhere as I am still new to arduinos and programming.

Quote
jculver09

Quote
vorpal
I didnt have alot of success with the auto calibration, to many variations in the results i think there is to much slack in the server arm. Also switched to a version of marlin that supports the m666 m665 command [github.com] with better success , though still not accurate enough. ( for me)

I tried to convert to the RC Marlin as well. Got it almsot all the way there, switches all worked, and it would home fine. When I run the G30 A D250.0 command to autocalibrate it would probe and probe and probe and after about 2 minutes it times out saying it is aborted because my Z endstops were too far apart (listing values of -17.xx, -17.xx, and 25.xx). I went to start a test print and it drove the effector down into the bed on the 2nd probe point, as if it thought the bed were tilted at something like a 45 degree angle, luckily no damage as the arm slipped on the servo gear, and I had put cardboard over my bed to protect it.

This is where I stopped because it was 4am and I needed to sleep, did you run into this issue? Would you care to share your config.h so I can take a shot and see what happens because I am sure I missed something somewhere as I am still new to arduinos and programming.

I have attached the configuration file I use. You will need to change values to match your setup. I found it important to use m502 to load new values after a new firmware upload, then m500 to save to EEPROM. you can check current settings using m501 to make sure values are as expected.

Make sure you endstops are all the same height from the lower frame, after build mine were out 5-8mm after balancing this made for less software adjustments using m665 x-- y-- z--

Also use m500 after using m665 or m209, to commit changes, a good way to test then commit to firmware once happy. I got it down to about 5-7 iterations with the m30 A. But best accuracy was achieved manually leveling the bed as mentioned earlier.

I went down this path as the supplied firmware didnt work for me as expected. Also make sure you havnt got you z min enstop inverted or wired backwards and always re home after applying any changes using m500.

some commands to note:

m209 zxxx - Set max z height
m500 - Save to EEPROM
m502 - Load default values from firmware
m501 - Show current values
m114 - Show current position
m119 - Show endstop state

G30 (without any parameters) - Probe bed a report current values
G30 Ax.xx - Full autocalibration (x.xx = optional value to specify required precision - if this is not specified the value in configuration.h will be used)
G30 E - Autocalibrate endstops (and then stop)
G30 R - Autocalibrate Delta Radius & Endstops (and then stop)
G30 D - Autocalibrate Diagional rod length (and then stop)
G30 Ix - Autocalibrate Tower position for Tower x (x= 1 for X tower, 2 for Y tower, 3 for Z tower) - This adjusts the A,B,C M666 values
G30 Tx - Autocalibrate Tower radius for Tower x (x= 1 for X tower, 2 for Y tower, 3 for Z tower) - This adjusts the I,J,K M666 values

hope this helps

Edited 2 time(s). Last edit at 04/14/2015 09:04AM by vorpal.

Quote
jculver09

Quote
vorpal
I didnt have alot of success with the auto calibration, to many variations in the results i think there is to much slack in the server arm. Also switched to a version of marlin that supports the m666 m665 command [github.com] with better success , though still not accurate enough. ( for me)

I tried to convert to the RC Marlin as well. Got it almsot all the way there, switches all worked, and it would home fine. When I run the G30 A D250.0 command to autocalibrate it would probe and probe and probe and after about 2 minutes it times out saying it is aborted because my Z endstops were too far apart (listing values of -17.xx, -17.xx, and 25.xx). I went to start a test print and it drove the effector down into the bed on the 2nd probe point, as if it thought the bed were tilted at something like a 45 degree angle, luckily no damage as the arm slipped on the servo gear, and I had put cardboard over my bed to protect it.

This is where I stopped because it was 4am and I needed to sleep, did you run into this issue? Would you care to share your config.h so I can take a shot and see what happens because I am sure I missed something somewhere as I am still new to arduinos and programming.

That seems really weird and makes me think things are wired up wrong. I would check your wiring and by that I do not mean just your wiring runs, but the pins at the plugs. Possibly one of them got inserted wrong.

Quote
thevisad

Quote
jculver09

Quote
vorpal
I didnt have alot of success with the auto calibration, to many variations in the results i think there is to much slack in the server arm. Also switched to a version of marlin that supports the m666 m665 command [github.com] with better success , though still not accurate enough. ( for me)

I tried to convert to the RC Marlin as well. Got it almsot all the way there, switches all worked, and it would home fine. When I run the G30 A D250.0 command to autocalibrate it would probe and probe and probe and after about 2 minutes it times out saying it is aborted because my Z endstops were too far apart (listing values of -17.xx, -17.xx, and 25.xx). I went to start a test print and it drove the effector down into the bed on the 2nd probe point, as if it thought the bed were tilted at something like a 45 degree angle, luckily no damage as the arm slipped on the servo gear, and I had put cardboard over my bed to protect it.

This is where I stopped because it was 4am and I needed to sleep, did you run into this issue? Would you care to share your config.h so I can take a shot and see what happens because I am sure I missed something somewhere as I am still new to arduinos and programming.

That seems really weird and makes me think things are wired up wrong. I would check your wiring and by that I do not mean just your wiring runs, but the pins at the plugs. Possibly one of them got inserted wrong.

It is definitely the firmware, because I can re-upload the original marlin provided and it works just fine. I was just tinkering with this newer version because I was not happy with autocalibration results either. I know my bed is not perfectly flat and I can see in the first layer where it is incorrectly interpolating and leaving some spots too high, or making some so low the nozzle scuffs the surface of my blue tape. This is really only apparent when I set my layer height to .1mm which I rarely use because of how slow the prints are.

Quote
jculver09

Quote
thevisad

Quote
jculver09

Quote
vorpal
I didnt have alot of success with the auto calibration, to many variations in the results i think there is to much slack in the server arm. Also switched to a version of marlin that supports the m666 m665 command [github.com] with better success , though still not accurate enough. ( for me)

I tried to convert to the RC Marlin as well. Got it almsot all the way there, switches all worked, and it would home fine. When I run the G30 A D250.0 command to autocalibrate it would probe and probe and probe and after about 2 minutes it times out saying it is aborted because my Z endstops were too far apart (listing values of -17.xx, -17.xx, and 25.xx). I went to start a test print and it drove the effector down into the bed on the 2nd probe point, as if it thought the bed were tilted at something like a 45 degree angle, luckily no damage as the arm slipped on the servo gear, and I had put cardboard over my bed to protect it.

This is where I stopped because it was 4am and I needed to sleep, did you run into this issue? Would you care to share your config.h so I can take a shot and see what happens because I am sure I missed something somewhere as I am still new to arduinos and programming.

That seems really weird and makes me think things are wired up wrong. I would check your wiring and by that I do not mean just your wiring runs, but the pins at the plugs. Possibly one of them got inserted wrong.

It is definitely the firmware, because I can re-upload the original marlin provided and it works just fine. I was just tinkering with this newer version because I was not happy with autocalibration results either. I know my bed is not perfectly flat and I can see in the first layer where it is incorrectly interpolating and leaving some spots too high, or making some so low the nozzle scuffs the surface of my blue tape. This is really only apparent when I set my layer height to .1mm which I rarely use because of how slow the prints are.

There are only a few different lines in this firmware vs the original marlin (I did a bit level compare). The issue I ran into with the auto level was understanding that I needed to level the bed from 253.9. So if you make changes to the bed height at all, reset the firmware to 253.9, then reupload it and start the z leveling again.

Do yourself a favor... go to Michaels or Hobby Lobby (or other similar craft store) and buy either an 8 inch mirror or order a borosolicate plate. The plate warps as you heat it, so it will be different each time. As soon as I added the mirror my issues that were similar to yours. It was printing perfectly fine on one side of the kapton tape and cutting a strip through the other side.

Next, get a better switch, if your's is doing what mine is, then it is causing an issue. In the guide, we level these until the button clicks the mechanical switch. However, during the leveling the switch isn't clicking due to the fact it's tripping the switch prior to the physical "click" occurs. This is causing an issue with the bed height as well.

So, in this design, with these switches and this aluminum plate here are the issues against us. 1) Auto level has major flaw in leveling system causing it to look at the bed as "stepped" instead of slopped, 2) Servo is overloading the RAMPS controller, 3) Switch causing inconsistent "clicks" creating a leveling situation, 4) a4988 drivers that are incorrectly stepping our motors half of the time. (below)

Also, found that there are known issues with the A4988 drivers they are using that are causing skipped steps. I am investigating this right now and a solution to this problem that should cause ultra smooth prints. Add this to the Simplify3d Upgrade I did yesterday and I am hoping the print quality will go through the roof.

Edited 2 time(s). Last edit at 04/14/2015 09:38PM by thevisad.

Quote
thevisad

Also, found that there are known issues with the A4988 drivers they are using that are causing skipped steps. I am investigating this right now and a solution to this problem that should cause ultra smooth prints. Add this to the Simplify3d Upgrade I did yesterday and I am hoping the print quality will go through the roof.

Care to speak more about your problem and possible solution to this? I personally have never noticed skipped steps since I run my motor currenst high with heat sinks and fans.

Quote
mmcginnis9272

Quote
thevisad

Also, found that there are known issues with the A4988 drivers they are using that are causing skipped steps. I am investigating this right now and a solution to this problem that should cause ultra smooth prints. Add this to the Simplify3d Upgrade I did yesterday and I am hoping the print quality will go through the roof.

Care to speak more about your problem and possible solution to this? I personally have never noticed skipped steps since I run my motor currenst high with heat sinks and fans.

The symptoms, videos and fixes are located here. The A4988 drivers were designed for a different motor, with high amp motors there is one issue, with low amp another.

http://hydraraptor.blogspot.com/2012/04/stepstuck.html

https://www.youtube.com/watch?v=cJDn1L-EUtY

Hey guys,

So i accidently Juiced my setup ( i ruined the ramps board through missadventure )

does this look like a decent price to replace the parts?

[www.amazon.ca]

I understand that i might not need to replace everything but i want the extra parts for future use in case they do work.

edit: cant communicate with arduino either : <


Edit 2 : folger sells it for 50 ish (american so the exchange will do me harm )

Edited 2 time(s). Last edit at 04/15/2015 08:30PM by vexon.

Quote
vexon
Hey guys,

So i accidently Juiced my setup ( i ruined the ramps board through missadventure )

does this look like a decent price to replace the parts?

[www.amazon.ca]

I understand that i might not need to replace everything but i want the extra parts for future use in case they do work.

edit: cant communicate with arduino either : <


Edit 2 : folger sells it for 50 ish (american so the exchange will do me harm )

If you can deal with waiting 2-3 weeks to save 30ish then buy these. I paid 25.16 for the last set from these guys and have ordered dozens of things from them.

Ramps 1.4
Mega 2560
a4988 Drivers

Quote
thevisad

Quote
vexon
Hey guys,

So i accidently Juiced my setup ( i ruined the ramps board through missadventure )

does this look like a decent price to replace the parts?

[www.amazon.ca]

I understand that i might not need to replace everything but i want the extra parts for future use in case they do work.

edit: cant communicate with arduino either : <


Edit 2 : folger sells it for 50 ish (american so the exchange will do me harm )

If you can deal with waiting 2-3 weeks to save 30ish then buy these. I paid 25.16 for the last set from these guys and have ordered dozens of things from them.

Ramps 1.4
Mega 2560
a4988 Drivers

Wow thats nice and cheap, I am a bit conflicted though with my desire to get rolling and my desire to save oodles of money.

hmmmmmm

My extruder failed last night, I couldn't print anything unless I could push the filament through. I printed a new extruder while manually feeding the filament. I was able to get a good enough part to replace the original, but the gear is wrong and is pinching the filament too much.

I tested the ohms on the stepper motors I have in my kit. The XYZ steppers are 2.85 OHMS and the exruder is 2.2 ohms. If you are using true Pololu drivers, then you will need to use the fix for missed steps. If you are using the chinese versions, then you need to make sure your resistor in this location is a 0 ohm or jump it. Looks like the ones I have on both the red and green sticks are 0 ohm already and have the fix applied.

I finally got to the point were i turned on the power, adjusted the Vref on the drivers and installed the firmware. I went through the configuration manual and tested all the settings against the actual as built printer. All the functions seem to work as intended. I want to finish securing the wires now that I am certain that everything is working. Then on to printing.

Got my kit finally delivered, assembled, configured (re-assembled, re- reconfigured)...

My X axis endstop doesnt seem to work right... - because of this, the carrage ended up sliding off the linear rail. (was able to get this back on without losing any ballbearings)

When homing, all the carriges rise up to the top and bump the switch -which activate the stop but the X axis seems to keep 'grinding' away after the switch is pressed and the light comes on. It grinds like this for a few seconds, causing alot of vibration and belt skipping - then it seems to finally register and stops... -like theres a delay.

I believe the connections are all good and the motor seems to perform like it should otherwise... I haven't tried swapping the end switches yet.