I have been printing heavily with an I3 that I built about a year ago. It works well but I always have issues with feed problems on longer prints (5 hours plus). I kept improving cooling, the extruder, etc etc and had about given up getting long prints consistently.

Every now and then my bed would slip as well causing the layers to misalign. It didn’t happen often so I just started again and usually succeeded the 2nd time. Recently I decided to try to knock that out and I ran across an article talking about the acceleration and jerk settings (in Marlin) and how they affected bed slippage so I decided to try it.

Originally I had the following:

#define DEFAULT_MAX_ACCELERATION {9000,9000,100,10000}
#define DEFAULT_XYJERK 20.0 // (mm/sec)

Based on what I was reading I changed em to:

Originally I had the following:

#define DEFAULT_MAX_ACCELERATION {5000,5000,100,8000}
#define DEFAULT_XYJERK 10.0 // (mm/sec)

I started noticing that not only did the bed not slip but I almost never get feed issues now. I am finishing my 4th print over 10 hours in a week and so far no slips and no feed issues.

Does it make sense to you guys with more experience that changing those settings would have a great positive affect on feed?

I get the bed slippage thing cause I have seen that many times in steppers over the years when you try to accel too fast but I guess I’m not getting why it help the feed issues.

I’m very happy. I’d just like to understand what is at work.



1 Like

5000 is very high acceleration for an i3.

You do not mention at what printing speeds you have skipping steps.

I never had skipping steps.

I have mine to:

define DEFAULT_MAX_ACCELERATION {3000,3000,100,10000}


Also look at your Jerk values:

#define DEFAULT_XYJERK 20 // (mm/sec)
#define DEFAULT_ZJERK 0.4 // (mm/sec)
#define DEFAULT_EJERK 5.0 // (mm/sec)



Skipping steps might also be that the stepper motors do not supply the correct max current. Check that they are setup (the pot) correctly.

1 Like

Here is acceleration setting for Slic3r of Original i3 MK2.

Perimeters: 800

Infill: 2000

Bridge: 1000

First layer: 1000

Default: 1000

1 Like

Can you describe what your feed issues were? Did it stop feeding entirely, heat creep, etc?

My guess is that the higher acceleration was either causing a retraction issue or your extruder stepper was unable to match the acceleration. In the first case, the retraction may have been too aggressive, causing the filament to retract while still too soft - this would result in complete loss of feed, as the melted filament would get pulled up into the hotend. This would cause restricted feed in the form of a clog. In the second option, the stepper couldn’t keep up and would slip - this would result in grinding filament, and skipped layers.

Without knowing what exactly your feed failures looked like - were there clogs and complete loss of filament feed? Skipped layers?

Hope this helps a little!

1 Like

So I’m the third George to answer, lol. Anyways, stepper driver current and acceleration are related. Higher current almost equals higher speeds. For the fast retraction or quicker movements, need more current. Plain and simple. The nema 17 are the right balance of power in my opinion. I also have a cheaper i3. GT pro b to be specific. I run my driver’s at 0.9 amps for X and Y, while I run 1.0-1.1 amps for the extruder and the dual steppers for the z axis. I share my firmware and Slic3r settings on my YouTube channel page. Google mk8 autolevel and will see some of my stuff. Links will take you to store and whatnot. Blue w for wordpress, has all my firmwares and Slic3r configs Even ones i have experimented with. During and acceleration or de-acceleration, the stepper motor will take more current. If these not enough current or torque, motor will skip a step. With higher current comes higher noise. So if you sleep next to it then you may have problems. Turn up your current on the driver’s. With a DMM set to 2 volts, turn on power to printer, take positive lead and place it on the small potentiometer, metal disc with a cross, while you place the negative lead on gnd pin of the driver. Adjust the potentiometer clockwise to increase current and CCW to decrease the current. Has no stop on most so may go all the way around. The voltage be read by the DMM will be a reference to the current that is being used by the stepper motor. So in this case the volts is actually the amperage that is being fed to the stepper motor. Good luck and have fun. Hopefully it’s not able to print great every time. Also forgot to say that your slic3r settings will also affect this. One can directly control the speed with g-code if absolutely necessary. And I have been testing my higher speeds, 60-80 mm a sec, and I can turn up the feed rate to about 300% for a short bit or if drivers are cooled well enough, can get some nice speeds. Driver’s tend to burn out when running any faster. At 330% my X axis failed. A spot on the potentiometer was burned, small adjustment and was back in business.

1 Like

On long prints it would just stop feeding completely. Usually I would have to pull the filament and reprime and could not just push some through to get it restarted.

OK you described what I kind of thinking. Something about the too high accel rates caused issues on the retractions.

Eventually the filament would completely stop feeding, the extruder would slip and if left too long it would usually cut the filament at the extruder gear.

I just finished another 11.5 hour PLA print yesterday. All perfect.

Thanks for the response.

OK good to know. Thanks.

Yes they are.

My wife knows all too well cause she has to hold the multimeter where I can see it while I fiddle with the trim pots and probes :slight_smile:

We have checked em 4 or 5 times just to make they are not changing.

Wow OK. That confirms something else I was going to ask, What settings others have that work well. Right now it is running very well so I will probably leave it but if I have any more issues I will definitely try going lower with those settings. My retract acceleration matched yours by default.

I really appreciate all the answers. I definitely stumbled into the settings I have now and you guys help me feel better that they are more where they should have been all along. Crazy that you can suffer a year of iffy prints cause the defaults you find are wrong and you haven’t learned to tweak a setting yet. People helping in this hobby keep other people sane I think.

My current settings match yours almost exactly. I’ll go get your firmware and see what else I might have off. Super nice of you to share that.

Thanks to all for your help!


No problem. If you need tech support also sell that on online store. If anything runs backwards don’t worry as you just have to invert the logic on that axis via Arduino ide or c++ under the tab of configuration.h Can make basic settings there, pretty much self explanatory. However, in my version of firmware I also use autolevel, the mk8 auto level bracket I made it for. And there is another I’m testing and trying with the filament sensor. In my firmware I enabled the thermal runaway protection, and have tuned the pid loops to the setups I have, both for Sanguinololu board and the gt2560 board. This firmware can run multiple boards as any standard firmwares but I added in a line for the gt2560 rev a itself. Make sure whT board you have. Rev a gt2560 is slightly different the the original gt2560 (legacy board). RAMPs 1.4 is easiest to change for. So just a heads up. georgeroblesjr.wordpress.com is my WordPress blog. There are three posts.

Also have parts in stock for some stuff, some filament, heaters cartridge, mk9 upgrade for the extruder and prebuilt extruder as well. Have plastic parts but is better for ninjaflex, flexible filaments. Look into filament cleaner/ oiler. Will also help. Over time the PTFE tube can swell and burn so good to keep some handy. 4mm od with 2 mm id.