Category: Configuration

Have you ever looked at your 3d print and wondered what those little patterns are in the sides of your print? It’s possible that they are artifacts from resonance in your X and Y axes. Resonance, meaning that your axis is vibrating, which means that your nozzle, which rides on the axis, is also vibrating, which means that you are transferring those vibrations to your print.

There are a few things you can adjust.

Most printers have off center holes in the bearings that the axes ride. This is done intentionally, so that you can adjust the tension. Make sure that your bearings are tight. Make sure that your belts are tight too. Try wiggling the motors, the print head, the bed, and the vertical supports. 

What does this have to do with resonance compensation? I’m a believer that you can’t solve a mechanical problem with software. If something is loose, don’t try to apply compensation, fix the problem.

Even after you have a mechanically solid machine, however, you will still have some resonance remaining in the system. I believe that this is what resonance compensation is for, not to mask underlying mechanical problems.

A couple of days ago, I wrote about calibrating your extruder with Klipper. Today, I would like to talk about how to calibrate your extruder e-steps with Marlin firmware. Here is how I do it:

  • Move printhead to center of bed and about 150mm off the bed
  • Mark a spot (I use tape) on the filament about 125mm down from the bottom of the extruder
  • Use calipers to measure actual value of tape from extruder. We’ll call this “A” and say we have 126mm for this example
  • M503 – read current values and make a note of the current E value. We’ll call this value “B” and say we have a current E value of 400 for this example
  • M83 – enable relative mode on extruder
  • G1 E100 F100 – feed 100mm of filament at 100mm per minute
  • Let it finish the extrusion and measure the distance between the bottom of the extruder and the tape that we marked. We’ll call this value “C” and say that we have a current value of 24 for this example. If everything were already calibrated correctly it would be 126 – 100, or 26mm.
  • Calculate our new E-Steps value. 100 * B(A-C) is the formula. Substituting our values from this example we get 100 * 400/(126-24) = 392.16.
  • Update our esteps on our machine with M92 E392.16
  • Save our new value with M500

After I rebuild my hotend, I always calibrate my extruder esteps. Every extruder, nozzle, Bowden tube, heatbreak, and filament has a slightly different flow rate, so this is something that I do pretty frequently. The process is pretty simple and straightforward.

  • Move the print head at least 100mm above the bed
  • Using calipers, mark off (I use tape) a spot around 125mm on the filament
  • Measure the exact length of the filament
  • Command the extruder to extrude 100mm
  • Measure the amount that is left
  • Calculate the actual that was actually extruded
  • Update rotation_distance for the extruder

Ideally, you would have a drive system for your filament that is:

  • lightweight
  • responsive
  • durable
  • does not obstruct or adversely affect movement of the printhead

Unfortunately, you have to choose. You can’t have all of these qualities. Each setup has pros and cons and every 3d printer needs to decide what is best for them or accept what their printer comes with.

A Bowden system mounts the extruder off to the side of the printer frame, allowing the printhead to be more lightweight, which allows higher speeds and acceleration. The tradeoff is that you have a tube that connects the extruder to the printhead, which makes accurately controlling the amount of extruded material more difficult (think trying to turn off a lightswitch with the end of a broom handle).

A direct drive system is just the opposite. You get more responsiveness with your extruded filament, but you are adding weight to your printhead.

If frequency were the only indicator, you would guess that printing different versions of the Venom Symbiote was my favorite thing to print. I’ve printed green ones, black ones, and in this case, I printed a transparent one.

I had to completely rebuild my hotend after this one, over a very simple mistake.

I had purchased a new hotend (complete) to replace one that I was having problems with. I put it on my printer and started printing and you can see the results for yourself.

What I forgot to do is “hot tightening.” When assembling the hotend together (at room temperature) you screw in the nozzle and the heat break until they touch. My heat sink also has some setscrews that need to be tightened. At room temperature, everything is tight. However, when things heat up and expand, little gaps in between each of the components will occur. The solution to this is to heat up your hotend to around 20 degrees beyond what you intend to print at, and then finish tightening the components.

The topic of first layer adhesion comes up often in 3d printing. Sometimes, the footprint of the part itself is not enough to keep the print stuck to the bed. I’ll go through some of the bed adhesion options below.

Priming Line: One of the main reasons for making an extrusion before you start on the part is to make sure that there is filament in the nozzle. A priming does this and nothing else. I don’t use this option much, but using a 50mm priming line accomplishes the purpose of filling up the nozzle.

Skirt: Your printer will create a quick circle all the way around the perimeter of the part. I use this option frequently. It accomplishes a couple of things for you.

  • A skirt fills the nozzle with filament
  • A skirt goes all the way around the perimeter of the part, confirming that your part will fit on the bed
  • By going all the way around the part, you also ensure that your bed level is accurate, just in case you have doubts about it

Brim: I use this option for tall, skinny parts. A brim is actually attached to the part, whereas a priming line or skirt are not. A brim provides more of a base for your part that is pretty easy to remove when you are done printing.

Raft: Every option is a compromise between robustness and printability. A raft leans heavily towards the robust side of the scale, but the cost is a longer print time and the fact that you will have to remove the raft from the print later. A raft creates a base for your print to build off of, rather than printing directly on the bed.

Someone on a forum had a question about why horizontal lines might be showing up on their 3d prints. Someone else suggested that maybe their infill was showing through and that increasing the number of walls from 2 to 3 thicknesses might fix it. That’s a good place to start and I thought that was the solution to the problem.

Well, it didn’t fix it. It turns out that they had to go through their whole 3d printer. Their z rod had just a little bit of play in it. Once they adjusted that, the horizontal lines went away.

Every once in a while, I am asked by someone about what their acceleration should be set at. My answer, as in so many cases, is “it varies.”

If you go too low, you won’t necessarily hurt anything, but your prints will take forever. If you set acceleration too high your belts will jerk abruptly each time the axis changes direction and you risk skipping teeth on the belts and getting layer shift.

I’ve run tuning test towers, and I know that my printer starts to have problems around 2,500 or 3,000 mm/s². It skips teeth and my layers end up all over the place.

Unless you want to do a tuning tower for each print (not practical) you are better off doing a little educated guessing based on what you are printing. 

For small, intricate, parts I set my acceleration low. Somewhere around 500 mm/s². For larger parts that don’t have a lot of detail, I might set at 2,000 or 2,500.

Maybe this is an unpopular opinion, but I don’t think it matters a whole lot for small parts. When you do a 3-point turn in your car, your acceleration doesn’t matter much. You are changing direction too frequently for it to even matter. I don’t have scientific data to back this up, but I would guess that it takes a Ferrari about the same amount of time to do a 3-point turn as my Dodge Dakota. If someone has a Ferrari they want to let me borrow to do that test, I’m game. The same is true on your 3d printer when you print small parts, so your acceleration doesn’t matter a whole lot.

I saw someone recently who was asking in a forum if they should incorporate BLTouch into their 3d printer. Their logic was that they didn’t know how to bed level.

I know it’s tempting, but don’t do it. In theory, it’s a great idea. Just get the probe to automatically do the bed leveling for you. But, in my experience, it’s never worked like that. The best option has always been to mechanically get everything as close as you can to perfect. Then use the software to fine tune that. The more the software has to compensate for, the more error you are going to have in your bed level, and in your prints.

Just to be clear, I’m not opposed to BLTouch, or any other automatic bed level devices. What I’m opposed to is the idea that you can somehow skip the effort of mechanically setting your bed level. 

I switched filaments. I’ve been meaning to change my nozzle, so while I was in there I did that too. When I printed a test print to make sure everything was good my first layer had holes in it.

Here is what I had to do to fix it.

  • my new nozzle doesn’t necessarily have the same height as my old one, so I releveled the bed and reestablished my mesh bed leveling.
  • my new nozzle probably doesn’t have the same heating characteristics as the one that I replaced, so I tuned PID.
  • my new filament may not feed in quite the same way as the old one, so I updated my e-steps.