Getting better prints out of the DP5 Part 1
Apr 18, 2017 0:12:00 GMT
kioskfan, ralf, and 1 more like this
Post by 3dprintingmeathead on Apr 18, 2017 0:12:00 GMT
This series is going to be random and unorganized, don't try to find rhyme or reason for the parts. There is none.
The DP5 has a few strong points, but like any sub $300 printer, it has it's weaknesses. Without buying upgrade parts, or by using parts of minimal cost, I'll go over a few things I learned that got me better prints. I will also be omitting basic maintenance things like belt tensioners and lubing the bearings, etc.
1)Print at a 45 degree angle. This is very beneficial for speed, accuracy, and artifacts. This goes for all printers out there. If you send 1 axis in motion, you will be limited to the speed of that axis. But if you send 2 axies(is that how you spell it?) in motion, you get the "hypotenuse effect"- yup, I just coined that phrase. But the math doesn't lie here. If the x moves 5mm, and the y moves 5mm, you will deposit a 7mm line of filament. If you do that in 1 second, you are printing at 7mm/second, but the motors are only moving at 5mm/second!
Why is that important? Well, it all comes down to acceleration and jerk, which translate into artifacts. By printing faster without moving the motors faster, you avoid issues like ringing or ghosting.
There's another benefit here though. My board had some interference that caused some obnoxious zebra stripes in my prints. This phenomenon only occurs when only 1 axis is in motion. So these marks are virtually non-existent when both axies are in motion. Win freaking win!
2)Part cooling fan. The one that comes with the printer blows...or, doesn't blow, that's the problem. This is due to poor design of the duct along with a super quiet fan. Here's the formula you should be looking at:The volume of air a fan moves is directly proportional to the noise it produces. Turbulence is loud, not the motors. Sure, there's some really advanced fan blades out there that reduce the sound, but not in these sizes.I recommend a fan like this. In fact I just installed this one and with a properly designed duct, I wonder if it's too strong.
Which is the perfect segway into the next point:proper duct design. This is where us DP5 users are in a pickle. Out cooling fan is mounted to our leveling sensor. And the current fan duct is even worse than the fan itself by a magnitude of 7.58(approximately).
At that point, there are mainly 3 options to get an awesome cooling fan. 1) Cut the fan out, leaving the sensor in place. 2)redesign a fan and sensor mount, utilizing the bolt pattern on the existing carriage. 3)Start fresh and print an I3 carriage with the standard bolt pattern and use whatever mount that tickles your fancy.
I should note that there are some adapter mounts made by dp5 users that keep the same duct, but use a larger fan. This will give a boost in performance, but you will see a greater gain with a more aerodynamic duct.
3)Level your bed. This is more theory than backed by science, but screw it. Yes, we are blessed with auto tramming. But the calculations needed to move all 3 axies when printing a layer can be a lot for our 8 bit board to handle. Also, there's wear and tear to consider. The acrylic plate doesn't help too much either. I'd say every 3 days or so, level it.
4) Calibrate your extruder. The extruder is really the only thing that needs calibration. The x and y are set. Well, there's debate on whether 80.5 steps/mm or 80 is the right one, I used 80, but that's just what I found worked. DO NOT!!!! print a calibration cube and change your steps to adjust for how far you are off! IT DOESN"T WORK THAT WAY! Sorry to be rude here, but the pulley determines how many steps are required per mm. Period. There are no outside factors, aside from loose belts, that change this. The calibration cube method does nothing to take into account issues like: Over or under extrusion, loose belts, x axis out of square, filament out of dimension, etc. These issues are the ones you should be fixing, not compensating for.
But the extruder is a different story. Instead of grooves for the filament to ride in, it has teeth that bite into the filament. How deep it bites in is determined by factors like spring tension and filament stiffness. So calibrate it.
5) play with your acceleration and jerk settings. Lower is better, higher is quicker. Do you want to print 2 ok parts, or one excellent part? the choice is yours.
6)Slow down. My benchys are done at about 45mm/s tops. 35 is better.
7)Cool off. A lot of people on reddit have prints that look like they sat in the sun too long. Then they say "it's at 200c, I shouldn't go lower". Well, is it at 200c? How do you know? The thermistor says it is, but is the thermistor accurate? What if it's off by 5 degrees? These are cheap mass produced thermistors, which are just a resistor that changes it resistance with temperature. The board sends out a signal and measures the voltage, compares it to the voltage it sent, and calculates the temperature from there. If the sensor isn't at spec, the readings will be off. So if it looks hot and melted, cool it off. I printed PLA at 180c before, and it came out pretty crisp. Was it at 180c? It doesn't matter, it's just a reference for the next time I print high detail where strength doesn't really matter.
That's enough for 1 day, I gotta eat and hope the bruins can dig themselves out of this 2 goal hole. More to come.
The DP5 has a few strong points, but like any sub $300 printer, it has it's weaknesses. Without buying upgrade parts, or by using parts of minimal cost, I'll go over a few things I learned that got me better prints. I will also be omitting basic maintenance things like belt tensioners and lubing the bearings, etc.
1)Print at a 45 degree angle. This is very beneficial for speed, accuracy, and artifacts. This goes for all printers out there. If you send 1 axis in motion, you will be limited to the speed of that axis. But if you send 2 axies(is that how you spell it?) in motion, you get the "hypotenuse effect"- yup, I just coined that phrase. But the math doesn't lie here. If the x moves 5mm, and the y moves 5mm, you will deposit a 7mm line of filament. If you do that in 1 second, you are printing at 7mm/second, but the motors are only moving at 5mm/second!
Why is that important? Well, it all comes down to acceleration and jerk, which translate into artifacts. By printing faster without moving the motors faster, you avoid issues like ringing or ghosting.
There's another benefit here though. My board had some interference that caused some obnoxious zebra stripes in my prints. This phenomenon only occurs when only 1 axis is in motion. So these marks are virtually non-existent when both axies are in motion. Win freaking win!
2)Part cooling fan. The one that comes with the printer blows...or, doesn't blow, that's the problem. This is due to poor design of the duct along with a super quiet fan. Here's the formula you should be looking at:The volume of air a fan moves is directly proportional to the noise it produces. Turbulence is loud, not the motors. Sure, there's some really advanced fan blades out there that reduce the sound, but not in these sizes.I recommend a fan like this. In fact I just installed this one and with a properly designed duct, I wonder if it's too strong.
Which is the perfect segway into the next point:proper duct design. This is where us DP5 users are in a pickle. Out cooling fan is mounted to our leveling sensor. And the current fan duct is even worse than the fan itself by a magnitude of 7.58(approximately).
At that point, there are mainly 3 options to get an awesome cooling fan. 1) Cut the fan out, leaving the sensor in place. 2)redesign a fan and sensor mount, utilizing the bolt pattern on the existing carriage. 3)Start fresh and print an I3 carriage with the standard bolt pattern and use whatever mount that tickles your fancy.
I should note that there are some adapter mounts made by dp5 users that keep the same duct, but use a larger fan. This will give a boost in performance, but you will see a greater gain with a more aerodynamic duct.
3)Level your bed. This is more theory than backed by science, but screw it. Yes, we are blessed with auto tramming. But the calculations needed to move all 3 axies when printing a layer can be a lot for our 8 bit board to handle. Also, there's wear and tear to consider. The acrylic plate doesn't help too much either. I'd say every 3 days or so, level it.
4) Calibrate your extruder. The extruder is really the only thing that needs calibration. The x and y are set. Well, there's debate on whether 80.5 steps/mm or 80 is the right one, I used 80, but that's just what I found worked. DO NOT!!!! print a calibration cube and change your steps to adjust for how far you are off! IT DOESN"T WORK THAT WAY! Sorry to be rude here, but the pulley determines how many steps are required per mm. Period. There are no outside factors, aside from loose belts, that change this. The calibration cube method does nothing to take into account issues like: Over or under extrusion, loose belts, x axis out of square, filament out of dimension, etc. These issues are the ones you should be fixing, not compensating for.
But the extruder is a different story. Instead of grooves for the filament to ride in, it has teeth that bite into the filament. How deep it bites in is determined by factors like spring tension and filament stiffness. So calibrate it.
5) play with your acceleration and jerk settings. Lower is better, higher is quicker. Do you want to print 2 ok parts, or one excellent part? the choice is yours.
6)Slow down. My benchys are done at about 45mm/s tops. 35 is better.
7)Cool off. A lot of people on reddit have prints that look like they sat in the sun too long. Then they say "it's at 200c, I shouldn't go lower". Well, is it at 200c? How do you know? The thermistor says it is, but is the thermistor accurate? What if it's off by 5 degrees? These are cheap mass produced thermistors, which are just a resistor that changes it resistance with temperature. The board sends out a signal and measures the voltage, compares it to the voltage it sent, and calculates the temperature from there. If the sensor isn't at spec, the readings will be off. So if it looks hot and melted, cool it off. I printed PLA at 180c before, and it came out pretty crisp. Was it at 180c? It doesn't matter, it's just a reference for the next time I print high detail where strength doesn't really matter.
That's enough for 1 day, I gotta eat and hope the bruins can dig themselves out of this 2 goal hole. More to come.
