Printable View
Do you level the printer hot or cold. It depends. Heatbed warps when hot. Maybe if you used the zipties provided they are not tie enough on the switch.
Found this post by gmay3
Quote:
If you haven't seen these parts from Clough42, they are highly recommended as improvements on the stock endstops and use screws instead of zipties.
http://www.thingiverse.com/thing:321810
Quote:
Originally Posted by tsteever
Definitely true. Sometimes I get confused between the different people I'm helping in the threads, by PM, and now on IRC. If I didn't bring that up earlier in this thread, it was an oversight on my part.Quote:
Originally Posted by rhonal89
Did not use the zip ties. Saw that in the guide and kinda chuckled. Seemed like everything else was very well thought out and that was an afterthought. I will take a look at those parts and see if I should incorporate them into my machine.
What about the code idea?
For the gcode approach to work, you'd still need to be convinced that whatever problem you're having doesn't manifest itself during a print. For example, you'd need to know that there are no issues with mechanical slop in the x-carriage/extruder/hot end or issues in properly driving the Z motors or the print may suffer even if it starts properly.Quote:
Originally Posted by tsteever
There are some issues with the tentative gcode you listed -
- In your text, you mention manually homing X and Y, yet a G28 X0 Y0 command is shown. Don't bother with the manual homing.
- If you are manually adjusting the nozzle gap for Z, don't include the G28 Z0 command
- The G1 Z15.0 will raise the x-carriage 15mm, presumably for the purpose of priming the extruder. You need to return Z to 0 or the print will attempt to start at that 15mm height
I have a new question related to your original problem - where have you located the clips attaching the glass to the heat bed? I try to locate clips close to the corners, thereby helping to mechanically tie the glass to the corner adjustments. If you only have clips nearer the center of the heat bed, the levelness and Z orientation of the glass could shift slightly as changes in the heat bed temperature alter how flat it is.
Clips in the corners.
Thanks, I am still learning a lot about the code and really (if you can't tell) do not know what I am doing...yet.
I plan, was to manually set the Z axis myself cause every time I autohome the z axis is in a different spot. I am not exaggerating. Right now I have things tore apart for cleaning but my intention is to replace the Z endstop with a printed setup from thingiverse. Before tear down I hit autohome. The bed wend to the corner and the z descended to a level below the glass. Previously i had adjusted it to be a paper height from the glass.
My glass is clipped at the corners and the home is actually off the glass (in case of the nozzle going below glass level). When I hit home and the machine was at zero I moved the extruder over to verify that the nozzle was at "paper height". I then hit home again and this time it descended below the glass level by almost .6mm (estimate). I immediately hit home again without moving anything and the z went up and stopped. It never descended down again. This time the axis was .4 above the glass!
Every time the machine homes, it is at a different location. I cannot get consistent prints at all. My thought is the switch takes too much pressure, has too much play, or is simply faulty. I am going to replace it with a new switch once it arrives.
Going back to the nozzle rising thing. I notice this behavior a lot when the machine homes below the glass. I will hit home again and the Z will rise but not descent again. It is my understanding that homeming on the Z axis is a 3 part move.
1) it depends at the home feedrate until the switch is hit
2) it will then rise up at home federate
3) it will descend again at homefeedrate/2
Where in the code can I change step 2? My thought is that due to the inaccurate switch that when it descends too far in step 1 that the rise is not enough to move the switch to an open state. I'd like to test that theory by changing this rise to a taller number and move the z slightly higher before making the 2nd higher resolution z home descent.
Others can jump in at any time, but before I provide any additional help I want to know that the MAX endstops have been disabled. As I said before, leaving them enabled without actually having the MAX switches on the printer can and sometimes does create homing problems. The symptoms are different, but I simply want to know that any issues with the unconnected/floating max endstop inputs are out of the picture.
I believe I've quoted the last response regarding the max endstops. The code in the quote block has nothing to do with how the max endstop settings are configured, since that code only applies if we have a COREXY printer. Look just above those lines in the configuration.h file. Remove the comment slashes off the first DISABLE_MAX_ENDSTOP line so it looks like the following. I've included the same block of COREXY code for reference:Quote:
Originally Posted by tsteever
Code:#define DISABLE_MAX_ENDSTOPS
//#define DISABLE_MIN_ENDSTOPS
// Disable max endstops for compatibility with endstop checking routine
#if defined(COREXY) && !defined(DISABLE_MAX_ENDSTOPS)
#define DISABLE_MAX_ENDSTOPS
#endif
Okay, I modified the configuration.h file and it now looks like this...
#define DISABLE_MAX_ENDSTOPS
//#define DISABLE_MIN_ENDSTOPS
// Disable max endstops for compatibility with endstop checking routine
#if defined(COREXY) && !defined(DISABLE_MAX_ENDSTOPS)
#define DISABLE_MAX_ENDSTOPS
#endif
Yes, that is the homing sequence, although I've fixed some autocorrect errors and enhanced the words a bit. The retract distance used in step 2 is easy to change - it is defined in the configuration_adv.h file. Look for these lines:Quote:
Originally Posted by tsteever
While you are in that file, consider increasing X and Y retract distances too. That has nothing to do with your troubles, but I changed mine to 10mm some time ago so that the X and Y homing doesn't seem to be just be a bounce.Code://homing hits the endstop, then retracts by this distance, before it tries to slowly bump again:
#define X_HOME_RETRACT_MM 5
#define Y_HOME_RETRACT_MM 5
#define Z_HOME_RETRACT_MM 2
--------------
There does seem to be a problem in how Marlin is reading the state of your Z endstop switch. This likely boils down to one of three things: 1) the switch is bad, 2) the wiring to the switch is bad, 3) there's a problem with the input pin for the Z min endstop switch in the AVR processor on the MEGA2560.
Obtaining a replacement switch was a good idea.
MakerFarm usually ships extra wires in the kits. While you wait on the switch, you could try running new wires between the switch you have and the connection on RAMPS. That would at least confirm there isn't a frayed wire or some other issue involved.
We could also make a firmware change to move what pin is used to sample the Z min endstop and try that while you wait on the switch.
Finally, if you have a multimeter, we could connect it to the switch contacts and monitor what the switch is doing as you fiddle with the lever on it.
Your call on any of these.
BTW - I remember now you saying you didn't use the zip ties to mount the Z endstop switch. Good. BUT!! Don't overtighten the small screws that mount the switch to the endstop bracket. The switches are pretty delicate and can easily be distorted if the hardware is too tight. Distorted switch = imprecise operation.
No Zip ties. I used a dab of hot glue which works great. If I need to take it off I can pop it off and peel the glue off. My new parts did arrive so I am going to print some mounts up.
Working my way through the e calibration and I am noticing some light white smoke/steam coming from the extruder. I have it set at 250 and hit it with a temp gun. It registered at 232 at the tip. I bumped up the temp to 260, still got the steam. I lowered the setting to 230 and I am getting 219.
What is the smoke? Am I too hot?
If your not extrudi g possibly it coild also be water that the plastic has absorbed...
All filament i know of absorbs moisture from the ambient air when left out.
Does it just seem to be coming from around the extruder? Maybe you're just burning off some of the material left on the extruder from the melted shroud. Have you tried wiping down the aluminum block and nozzle while hot with something like a paper towel?
Coming out the nozzle, not around. It really does just look like steam.
I am also having Z homing issues with the 12" that I put together 1-2 weeks ago. I have swapped the Z axis home switch around with one of the others and have everything mounted with screws. It is very frustrating to have to fight with the printer for 20 min to get the Z home position fixed every time I want to print something. I have been very close to throwing the damn thing in the trash because of it. I also have the Rumba board so I am not sure if there is an issue with the Rumba or something else but I would like to get it figured out so I can actually enjoy using the printer instead of fighting with it.
I am beginning to think the issue is with the type of switches sent with the kit. For some reason he included micro switches with roller ends on them instead of standard lever ends. If you are homing off of a flat surface the roller switch would be fine IMO but on the Z axis you are using the tiny bolt as the homing surface so I think there are some inconsistencies related to that because depending on what part of the roller radius the bolt hits on it changes the home position. I have some standard micro switches on order to swap out the roller switch with. I hope that fixes the issue because at this point in time I don't really want to even use the printer because of it.
Quote:
Originally Posted by tsteever
I am finally at a point where I can start to print some parts. No where near the quality of the Afinia or Ultimaker I have used before. Funny too as there should be no reason we can get the same prints out of this machine, on the inside they are all very similar. I think the Afinia has a superior slicer engine. The top layer prints on those look so good! My top layer looks like poop with a stick drug through it. Gotta play with the setting a bit more.Quote:
Originally Posted by Chadd
What helped me was the advice given here in this thread by the awesome forum members. Prints, Sniffle and countless others have kept me from trashing the machine. To start I would change the following in no particular order...
!. Raise height that the Z moves up when it home the second time. See Post #88
2. Change the speed at which this 2nd high-res homing move is done at. I found these settings in the config thread by printbus. There is a line in the code that sets the homing speed and divides it by a number. Increase that number until you are happy.
3. Do everything Printbus and Sniffle have told me to do, they know their stuff!
I am sure there is more but that should help. I can now get a good accurate home position set repeatedly. I will be changing the switch out at some point. I am going to give the Auto Bed Level a try. Can't hurt right?
Now, about that top layer poopiness? Take a look at this pic. It is the top layer od a model printed on an Afina. How can we get this out of a Maker? The part on the left is the Afnia printed at their fine speed, .2 layer height, solid. The part on the right is one of my first prints. Take a look at the top layer on the maker part. See how the nozzle has drug through the top? On my most current prints these lines are even more pronounced. The Afinia part on the left is so nice to look at. Very finished. The parts just print very accurately.
I know it is a 1500 dollar machine with limited user input. But the parts and pieces that make it are so close to the maker. Cannot we get this level pf performance?
http://3dprintboard.com/attachment.p...tid=4981&stc=1
My print quality is good when I am not screwing around with the homing issue. It is hard to say what the print quality issue is without actually seeing the part in hand. I am sure you posted it elsewhere in your thread but what are all your print settings for that part? Looks to me like your nozzle may be dragging. One thing I had to do was to put a shim inbetween the hex hotend and the extruder base because the hot end would wiggle back and forth, especially when retracting to reduce nozzle drool.
Quote:
Originally Posted by tsteever
Print setting for which part? The Afinia is very limited on what you can choose. I need to do some digging to see what the actual settings are. You have the option of layer height, Speed (fine, normal, fast), and fill. The rest is out of your hands. The top layer is a good indication to me of a good print. The Afinia on the Left does an amazing job. The part is very smooth. I have no idea how it does it and I would like to get my hands not he g code to see if I can decode the speeds and temps.
The Maker does okay. Just a bit of work to do to get it to a level I would like to produce at. The nozzle does appear to drag through the top layer a lot. Here is a typical print I am getting. Is there a setting I am missing somewhere that will tell the z to move up when it moves to a new position?
http://3dprintboard.com/attachment.p...tid=4984&stc=1
an interesting thread. newbie come listen.
http://amburadul101.tk/114/o.png
What were your print settings for that print? Looks like you are over extruding.
Quote:
Originally Posted by tsteever
Concerning Pronterface....
If you have been following along you know I had a terrible time with Pronterface on my Mac. No matter how many times I redid the drivers for the FTDI I could not get it to connect consistently. Eventually I followed the steps for reinstalling python via terminal and got it to work sometimes, but with more frequency. I followed the steps listed here towards the bottom...
https://github.com/kliment/Printrun
Once I had that I could get into Pronterface sometimes and sometimes not. I just clicked connect until it worked. I was running the most current version of Pronterface. On a hunch I tried the 12July13 release of Pronterface and I am happy to report I can connect every time! I hope that helps some people.
There's been another person on the Makerfarm IRC also struggling with Z axis problems. That makes three.
One common denominator is that the MakerFarm zip-tie approach to mount the Z endstop has not been used. If screws are used to mount the switches, don't overtighten them since that can and will distort the switch and lead to problems. I personally wouldn't use hot glue to mount the switch since the glue, while hot, could also distort the switch.
Any time Z axis issues are apparent, make sure the coupling between the Z motors and the threaded rods is tight, whatever method is being used. If using the MakerFarm plastic tubing, add zip ties to the tubing and tighten them as much as you can to provide additional clamping force on the threaded rods. If using split-type aluminum couplers, note that they typically will NOT tighten adequately on the undersized threaded rods unless you've added something like foil or foil tape to the threaded rods to increase their diameter to the 5mm the couplers are designed for. Time and time again we've had people who *think* their couplers are tight only to later realize they weren't tight enough.
When Z heights are suspect, always look at the printer LCD and see what the printer thinks the current Z height is. The display will reflect what the printer has been commanded to (via LCD, gcode file or host software). It does not necessarily reflect the actual position if there are movement problems on the printer. This might help determine whether the source of the problem is the command source (host software or gcode), or the printer itself.
When gcode files are suspect, or you want to know what the gcode will be telling the printer to do as far as Z heights, load the gcode in the gcode.ws analyzer. It's not an app you have to download - it's just a URL where you can load your gcode. After loading a file, click on the layer view. The scroll bar on the right side of the GUI allows you to view each layer of the print individually. The scroll bar on the bottom allow you to trace through the movement commands for the selected layer. Note that text data on the left will show what the current print height is for that layer.
I'm wondering whether the Z axis stepper driver current limits have been adjusted for any of these three printers. If not, they could be on the low side, which can cause the Z axis motors to skip. This would manifest itself as the printer not always moving to the commanded Z height, whether during homing or during a print. On RAMPS and RUMBA, the stepper driver current limits are adjusted by small trimpots on the stepper motor driver boards. For RAMBO, this adjustment is done through firmware changes (or maybe LCD? I don't know). Refer to the build guide, electronics documentation, or google for more information on the adjustment process.
Could you be a bit more specific as to print settings? There are so many that I wouldn't know where to start listing. Which ones are you looking for specifically?Quote:
Originally Posted by Chadd
I have been tweaking the Slic3r setting so much I have not clue what is stock! I can attach my slicer config files if that helps.
I worked my way through almost half a spool of filament trying to get the steps right on the extruder. That is not as simple as Youtube makes it look! I first asked it to extrude 100mm after marking at 120mm. I measured and did the math and found I needed to adjust my steps to 848.2 from 841. I redid the test and this time it said I needed to just to 836! That can't be right as it is less than what I started at. Oooof da, my head hurt after all that.
I just did it manually until when I called for 100mm of filament, it extruded 100.05. I called that good and ended up with an step value around 846.2 .
I then moved on to the extruder multiplier by printing the single wall test print and measuring it compared to the gcode which called for .67. I did the math and determined that the spool needed a multiplier of 97%.
That is what I printed the above piece at. Colin has advised me that he uses the included config and tunings posted from his build guide and does not adjust the multiplier. I know his prints are not as good as they could be, but they look nothing like this mess.
After I made the changes you posted out to the code and changed the homing speeds and heights I am getting better Z axis homing. I will measure when i get home.
As far as the current limits go, on the Rumba they are controlled by little pots on the driver. I have checked them and find them to be at .4v per the build guide. I have no idea is this is too high or too low. Since the Z stepper is driving two motors, do You think the V should be higher? I find the advice on setting these to be speculative. Change and listen for sounds. When you don't know what the proper sounds are, how do you know? I know the Z motors are not getting hot, which means I could potentially increase the voltage a bit.
This. You have just located the price gap between a DIY kit and a tested, quality assured, finished product.Quote:
Originally Posted by tsteever
I have the most experience with Ultimaker products, so I will use them as an example. When one purchases an Ultimaker 2, it costs around $2500 USD. The parts come in around $950, and paying someone to build the printer only costs about $50 each. Also, you're right, the parts are very similar to what you get in a makerfarm kit. So, where does your other $1500 go? What are you paying for?
The answer is quality assurance and technical support. I can tell you from experience that an Ultimaker 2 is just as capable of terrible print quality as your Prusa i3v; however, when it arrives at your door, the Ultimaker 2 quietly churns out perfect 50 micron prints. This is the result of lots of testing. Before each printer leaves the factory, it has to run a gauntlet of tests, real torture prints. If the printer makes one mistake at all, no matter how small, it is sent to review where a team figures out what needs to be fixed; furthermore, the Ultimaker 2 has been out for over a year and the company is still regularly updating parts for it as problems present themselves.
The process of releasing a quality product to the world market is extremely complex/difficult, but for our purposes we'll oversimplify it to 3 parts: Design, Testing, Release. First a company designs a printer, then they test it to death, then after they've solved all the problems that they can, the product is released for end users to buy. After that, they continue supporting their customers.
Makers have the choice to enter this process wherever they like:
Design (Free-ish). These are the new printer ideas you find online that only a handful of people have actually made. They can usually be assembled with cheap/spare/printed parts. Be advised, there will be a lot of troubleshooting ahead to get it printing at all, let alone making super high quality prints.
Testing ($ - $$$). These are the RepRap all-stars. Tried and true open source designs that are known for being capable of terrific performance... if one is patient enough to wrestle that from them. Often companies take one of these popular designs and assemble a DIY kit. The value these kits present is simplifying the process of building a printer from scratch (research, part selection, assembly advice). Such is the case of the Makerfarm Prusa i3v. Do not confuse this with a finished product. The company agrees to provide working parts and sometimes build instructions; however, the print quality will be dependent on each individual customer's skill level. In other words, the customer is both quality assurance and technical support.
Release ($$ - ∞). These are the Makerbots/Ultimakers/Afinias/etc. The company has already troubleshot their product and figured out the best practices for its use. If something is malfunctioning, technical support will help you fix it (often one on one). If the problem is too large for you to fix quickly, they will replace the printer. This service is mostly what you have paid them for.
If you're expecting consistent amazing print quality with this printer, it is possible, but you have a long road ahead of you. The good news is, at the end of the road you will have learned a lot. If that doesn't sound like something you would like, then get out now while you're only $600-ish in. Flip that thing on Craigslist and call it a day.
TL;DR: You get what you pay for, and that's a great thing.
Source: I work for Ultimaker's US branch building/testing/fixing printers.
I originally used the Zip ties but changed to bolts after I had issues with Homing.
Quote:
Originally Posted by printbus
I get it, you spend the money for the company support and a pre-tuned printer. I really do understand and fully understood this before getting the printer. I also understood that by getting a "kit" printer I would be doing much of the leg work, with the help of a very supportive community.
I am not placing blame at all, nor complaining. I am just seeking advise and guidance from more knowledgable people than myself and learning as I go. Hopefully this dialogue can help others as well. Now, can you offer any guidance as to how I can get better, more accurate prints from the printer? I have seen many parts produced on an Ultimaker and they are amazing! I hope to get this machine tuned to come marginally close to what that can do. I wish I had the funds to pay for an out of the box solution. I wanted a printer with a large build platform and a reasonable cost in regards to my budget. The ultimaker was just too much money for it's build size. It has quite a premium added to it for the high quality prints.
I made the choice to take on a project and I am up for the challenge. I know this machine can get there is I can just figure out how to overcome the hurdles.
First problem (pronterface) has been overcome
Second Problem, Z-Axis, working on the fine details but it is printing.
Third Problem, Top layer quality. Got some work to do that I suspect is an extruder setup
Forth Problem, accurate reproduction (size) of parts. Suspect extruder setup here too.
Where do you adjust the speed that the Z axis homes at for the final home? Can't seem to find it in the config files.
Never mind, I found it.
Quote:
Originally Posted by Chadd
To be blunt, reducing the homing speed to "fix" what might be wrong is a bad idea. At best it is masking what the problem is. More than likely the underlying problem is still going to surface as z-axis issues during prints. Why would it be that no one has had to slow down the last homing leg before?
After reading this and some searching it appears that a lot of people have issues with the default homing speed. Most say changing it from 4 to 3 takes care of the problem. I am about to upload the firmware to try it out, I am a home built CNC machining guy and have dealt with home building CNC machines.
I know these printers don't have much load on them but in all honestly from a structural standpoint these printers are horrible, I don't think the over all structure of the printer is capable of being repeatable to within the tolerances that we are asking them to be. Slowing down the rate that the machine homes at for sure can't hurt anything.
But I agree the home rate isn't the underlying problem here. Poor design is.
Quote:
Originally Posted by printbus
Who says that? On these printers? I don't recall this ever being discussed in this subforum, except for my INCREASING the homing feedrates. I can't speak to what MakerFarm is providing for firmware these days, but the standard configuration.h file at least used to specify a Z homing feedrate of 50mm/min. That's less than 1mm/sec. So, I'm not sure what the 4 or 3 is referring to.Quote:
Originally Posted by Chadd
Unless I am mistaken this is the home feedrates from the config file.
#define HOMING_FEEDRATE {50*60, 50*60, 4*60, 0}
that is the X, Y, Z, extruder rates, in that order.
There are at least a few threads on the reprap.org forums about this.
Quote:
Originally Posted by printbus
What is the origin of that config file? Is it a stock Marlin configuration.h from the Zalm repository, or the configuration.h that the manufacturer (MakerFarm) provides to be used?Quote:
Originally Posted by Chadd
It is correct that the Z motors on the i3v printers can't be reliably driven to 4mm/sec. I found even 3 to be borderline in my testing.
Yes that was directly from the config file that I downloaded for the Rumba from makerfarm.
I am going to download the rambo firmware from makerfarm and look at it. I have run into a few discrepancies with documentation for the 12" i3v.
"edit" the rambo config file has the exact same feedrates that I posted above.
Quote:
Originally Posted by printbus
The Ramps config is different here are the values from it.
#define HOMING_FEEDRATE {50*60, 50*60, 50, 0}
Maybe this is the issue that people are having? Would that be .883 mm/sec?
I sent an email to Colin to ask him about it.
Quote:
Originally Posted by Chadd
I heard back from Colin regarding the speed setting in the config file. It looks like this was our problem, here is what Colin replied.
Quote:
Hi Chadd,
Thank you for letting me know, I have changed the firmware and uploaded so everyone can download it. The firmware should have been as follows: #define HOMING_FEEDRATE {50*60, 50*60, 50, 0}. Those numbers are the speed at which each axis moves when it homes, if the Z speed is to high the two motors can get out of sync which causes everything to be out of whack.
Thanks
Colin
Okay, so remind me. Is this in the config.h?
Here's my guess at a scenario - at 50mm/min (yes, 0.83mm/sec), the 12-inch printers take just over 6 minutes to do a full-travel home in the Z axis. Um, yeah, that's a freakin' long time. Experimenting, someone might have found that WHEN HOMING, you could appear to drive the Z motors a lot faster than that. Well, somewhat true. The Z motors will move the x-carriage DOWNWARDS nice and fast. Maybe even more than 4mm/sec. The problem is that the motors don't have the torque to lift the x-carriage using M5 rods at that rate, and that someone forgot about the *minor* detail that the X-carriage is lifted upwards a bit for the second leg in the homing process.Quote:
Originally Posted by Chadd
FYI - My take on all sorts of testing the Marlin motion-related parameters such as those in the configuration.h file, including snippets of gcode that are handy for testing the physical movements, can be found in thread http://3dprintboard.com/showthread.p...-MakerFarm-i3v. As incentive, I'll tell you that while the 4mm/sec value is unrealistic for Z, the 50mm/minute feed rate is conservative and the printers can do better than that.