QU-BD MBE Extruder Failure Analysis

General discussion of 3D printers

QU-BD MBE Extruder Failure Analysis

Postby rudi » Sat Jul 06, 2013 6:28 am

Hi all,

As many others, my start in 3D printing began with an MBE Extruder, which I bought from QU-BD some months ago. At first, I did not get it to work reliably. As many, I nearly gave up on it. However, because other parts of my printer were misbehaving as well, I had ample reason to keep reassembling and reconfiguring until I finally got it to work reliably, and I'd like to share some insights.

At this time, my 3D printing setup consists of an Ord Bot Hadron, a plain vanilla QU-BD MBE Extruder, a Prusa Heated Bed, glass plate, a switching 29 amps power supply and a Sanguinololu driver board, plus Repetier Host and firmware. I am printing with ABS. My setup works reliably.

The QU-BD MBE Extruder originated in 2012 as a Kickstarter project and consists of just a handful of parts. It is either bought as a kit or completely pre-assembled, both from the QU-BD webstore. Some parts are optional and some parts can be upgraded. The design is said to be working 'out of the box', but a lot of owners have modified their MBE Extruder in order to get more reliable results. That said, many owners of an MBE Extruder have given up on it completely because of the lack of results. However, with some care and patience, the MBE Extruder can give great results, right out of the box.

mbe-extruder.png
Schematic of MBE Extruder

Root Causes
The expected results of an extruder are easily summed up: Extrude the precise amount of Filament that is required at any one given time. There are three Root Causes for the extruder not living up to these expectations (refer to the drawings for part descriptions):
  • Too much friction in the Filament path
  • Feed Gear has not enough Torque to drive the Filament
  • Feed Gear has not enough grip on the Filament

Of course, all three Root Causes can happen at once as well, which it often does.

Too much friction in the Filament path
  • Filament too wide
    Make sure the filament you use does not exceed 1.75mm, use a calipers to check.
  • Filament not clean enough
    Although I've not experienced any problems with this myself, I have seen reports of people suspecting either filament that has some kind of grease on it.
  • Gap between Filament Barrel and Brass Extruder Nozzle
    A gap between the barrel and the nozzle is killing: a build-up of ABS between this gap allows the ABS to get out of the Aluminium Heater Block and form a sticky substance that adds a lot of friction tot the filament feed path.
  • Impurities in Filament Barrel or Brass Extruder Nozzle
    Although I only experienced this once while experimenting with PLA, a number of people report about the build-up of impurities in barrel and nozzle. This adds to the friction.
  • Misalignment of the Filament Feed Path
    There are a number of possible misalignments: the place of the drive gear, the tube of the barrel over the nozzle and the angle of the chassis over the feed block.
  • Extrusion Temperature too low
    The temperature is measured with the thermistor. The exact place of the thermistor determines if the measured temperature is the temperature of the brass nozzle, the block or the nearby air. The nozzle should be about 225 degrees Centigrade.
  • Filament Adjustment Screw has rough edges
    It pays of to polish the adjustment screw to a smooth sinhy state. Rough edges, for instance when the gear has scratched the screw, add to the friction.

Feed Gear has not enough Torque to drive the Filament
  • The Current from the Driver Board to the Stepper Motor is not enough
    Adjust the current from the driver board
  • Wiring from Driver Board to Stepper Motor too thin
    Although I've not experienced this myself, too thin wire will result in a voltage drop proportional to the current drawn. That will mean stalls when friction gets too much.
  • Filament Adjustment Screw is too tight
    The screw has to be just soo tight that the filament gear has enough grip.
  • Driver Board Thermal Shutdown because of too high Current
    This happens real quick when either the friction is too high, or the current to the stepper motor is too high.
  • Drive Gear Screw too loose so it slips on the axis
    This happens to me all the time: the screw in the drive gear gets loose. I suspect it happens when the stepper motor heats up and cools down afterwards.
  • Stepper Motor Axis has no flat area for the Gear Screw
    This is a design flaw: there should be a flat area on the stepper motor to prevent the gear from slipping.
  • Acceleration of Filament Feed too high (reduce to 1 mm/s2)
    This is a software setting: if the acceleration is too high, the current is too high and the stepper motor will heat up quickly, resulting in less grip on the filament.
  • Filament Feed speed too high (reduce to 3 mm/s)
    If the speed is too high, the stepper motor draws too much current. That said, 60 mm/s is easily obtainable, but start much slower.

Feed Gear has not enough grip on the Filament
  • Drive Gear is not aligned properly with the Filament Path
  • Drive Gear Teeth are not sharp enough (worn out or filled with ABS)
    Once the gear has scraped of some ABS, it will fill the teeth and reduce grip.
  • The Filament Adjustment Screw is too loose
    The filament has to be pushed against the gear for the teeth to sink in the ABS. This adjustment is an ongoing process.
  • The Drive Gear teeth sink in the Filament
    This is a big problem, that happens over time: if the friction is just a bit too high, the stepper motor heats up. If it gets so hot you can barely touch it, the gear is hot enough to melt the ABS all by itself. Grip reduces fast this way. Another common problem is that the upper end of the stainless filament barrel is becoming too hot. That is easily solved by applying heat plaster between the filament barrel and the chassis block, and between the chassis block and the heat sink.

Some advise
In order to get the Filament fed reliably, it is important to get the Filament Path aligned and as friction-free as possible. At 225 degrees Celcius, you must be able to feed the Filament manually between two fingers. During the assembly of the extruder, pay special attention to the possible gap between the Filament Barrel and the Brass Extruder Nozzle. Even a small gap can let melted Filament through, which sticks to the Filament and thus adds friction. Also, make sure the Drive Gear aligns properly with the Filament Path. If it doesn't, the Filament has to bend some to get through the Filament Barrel, which in turn adds friction.

barrel-nozzle.png
MBE Hot End Detail

Once the Filament Feed path is friction-free and the temperature is set properly, two other problems arise: not enough torque or not enough grip. It is very important to adjust the current from the Driver Board properly: just enough current to have no stall, but no too much current to engage Thermal Shutdown of the driver board. This adjustment has effects on the short term (stall) and in the long run (thermal shutdown, or heated-up Stepper Motor). Because the temperature of the Stepper Motor itself increases proportionally with the current over time, adjustments are preferably done in cold state and then observed for at least half an hour. What I do is turn the current half-way, have the extruder extrude some ABS and while it is working, I turn the current way back until you see and hear the Stepper Motor stalling (it makes a loud thumping noise, about twice a second). Then I turn the current up some and leave it that way.

The grip on the filament and torque of the Drive Gear are getting more important as the friction in the Feed Path increases. If the Feed Path is near friction-free, grip and torque aren't all that important anymore. Therefore, if the extruder isn't working properly, always start with root cause 1. Always assume there is more than one issue.

Best regards,

Rudi
Last edited by rudi on Sun Jul 07, 2013 9:42 am, edited 2 times in total.
Ord Bot Hadron + vanilla QU-BD MBE Extruder
rudi
 
Posts: 4
Joined: Fri Apr 12, 2013 7:08 am

Re: QU-BD MBE Extruder Failure Analysis

Postby cvoinescu » Sat Jul 06, 2013 10:12 am

Thank you for this analysis.

I'd like to point out that the choice of material is essential in the case of the QU-BD extruder. It can be made to work with ABS, but PLA is much more difficult.

PLA is glassy and hard, and melts at a fairly definite temperature into a relatively low viscosity liquid. ABS is flexible and soft, and softens and melts gradually over a broad temperature range; even completely molten it's still quite viscous. Also, PLA melts at a lower temperature than ABS.

This makes ABS better for the QU-BD in three ways.

Firstly, ABS deals much better with the board melt zone and high heat flow between the hot and the cold ends, but PLA tends to jam. One hypothesis is that heat creeps up during idle moments (I've seen retraction blamed too) and too much PLA melts, and then, when the filament is driven down, the low-viscosity molten PLA flows up between the barrel and the solid filament and freezes there, jamming the extruder. I have no evidence that that's actually the case, but it seems plausible. It's also consistent with which modifications appear to work.

Secondly, PLA, being much stiffer, is very sensitive to an exact adjustment of the pressure screw. ABS yields more and the teeth of the filament drive gear can sink into it quite a lot, giving a wide enough range of adjustment for the system to work. Also, the QU-BD requires the filament to be round and of constant diameter. With ABS, it can tolerate slight variations, but with PLA, where a few tens of microns make the difference between no grip at all and a jam, the filament needs to be perfect. Needless to say, that's usually not the case, which is why the QU-BD drive is unreliable on real-world PLA filament.

Thirdly, the area of the drive gear can become hot enough in normal operation (heat from the stepper motor and from the hot end transmitted to the cold end) to soften PLA. Even a small difference can compromise the grip of the non-self-adjustable drive gear and pressure screw system.
cvoinescu
 
Posts: 501
Joined: Thu Aug 09, 2012 9:12 am
Location: Camberley, Surrey, UK

Re: QU-BD MBE Extruder Failure Analysis

Postby rudi » Sun Jul 07, 2013 9:08 am

One thing that helped me enormously getting reliable results is to make sure that the upper end of the filament barrel is cooled properly. The black anodized aluminum chassis block is meant to cool the barrel, so the barrel has to make perfect contact with the chassis block, which means heat sink plaster. Also, the cooling profile in turn is meant to cool the chassis block, so also a good contact is required. Once the upper end of the barrel is cooled properly, a whole lot of feed-related problems are solved. Use the provided heat sink plaster between the barrel and the chassis block, and between the chassis block and the cooling profile. You could extend that treatment to the stepper motor, although I recon that wasn't meant in the design. You will notice the cooling profile getting a lot hotter this way, actually dissipating the heat from the upper end of the barrel, resulting in no heat traveling up the filament.
Ord Bot Hadron + vanilla QU-BD MBE Extruder
rudi
 
Posts: 4
Joined: Fri Apr 12, 2013 7:08 am

Re: QU-BD MBE Extruder Failure Analysis

Postby rudi » Sun Jul 07, 2013 10:40 am

Come to think about it, I wonder if the stepper motor actually would benefit from a better thermal connection with the chassis block. It could also backfire: instead of helping to cool the stepper motor down, better thermal contact with Th chassis block could just as well transfer more heat to the stepper motor. It might even be a better idea to thermally isolate the stepper motor from the chassis block.. Something worth to investigate: what generates more heat: the stepper motor or the chassis block.
Ord Bot Hadron + vanilla QU-BD MBE Extruder
rudi
 
Posts: 4
Joined: Fri Apr 12, 2013 7:08 am

Re: QU-BD MBE Extruder Failure Analysis

Postby kbob » Sun Jul 07, 2013 4:23 pm

rudi wrote:Come to think about it, I wonder if the stepper motor actually would benefit from a better thermal connection with the chassis block. It could also backfire: instead of helping to cool the stepper motor down, better thermal contact with Th chassis block could just as well transfer more heat to the stepper motor. It might even be a better idea to thermally isolate the stepper motor from the chassis block.. Something worth to investigate: what generates more heat: the stepper motor or the chassis block.


Actually, which is at a higher temperature. You can find out. Just touch both pieces with your finger and see which feels warmer. (I'm betting the motor is warmer.)
Bob
"If you didn't code it, it will never own you." (-:
kbob
 
Posts: 151
Joined: Mon Jul 09, 2012 6:39 am
Location: Eugene, OR, US

Re: QU-BD MBE Extruder Failure Analysis

Postby Winder » Mon Jul 08, 2013 5:51 pm

rudi wrote:Come to think about it, I wonder if the stepper motor actually would benefit from a better thermal connection with the chassis block. It could also backfire: instead of helping to cool the stepper motor down, better thermal contact with Th chassis block could just as well transfer more heat to the stepper motor. It might even be a better idea to thermally isolate the stepper motor from the chassis block.. Something worth to investigate: what generates more heat: the stepper motor or the chassis block.


I put a thermal barrier between the chassis block and the stepper motor (a small sheet of PLA from a bad print). My stepper motor tends to run a bit warm warm so this helped a lot.
Winder
 
Posts: 16
Joined: Fri Sep 10, 2010 11:56 am

Re: QU-BD MBE Extruder Failure Analysis

Postby bloomingtonmike » Mon Jul 15, 2013 3:23 pm

Since installing PTEF teflon tubing in my barrel I have had ZERO issues with the MBE.
bloomingtonmike
 
Posts: 113
Joined: Thu Dec 20, 2012 3:04 am

Re: QU-BD MBE Extruder Failure Analysis

Postby jamesarm97 » Tue Jul 23, 2013 1:18 pm

I wish I searched for qu-bd extruder problems before I purchased two of them and my friend bought 6. I quickly found out that they don't print!!. At first I thought maybe it was just me so my friend was building a new 3d printer and had it ready to test and used a qu-bd also. He had the exact same problems. We had already installed the tensioner during assembly and never even tried the pinch wheel, using the MK7 instead. We have tried shortening the assembly, added thermal grease to the aluminum block and the part threaded into it. I have even replaced my nozzle with an old Mk7 I had. Nothing seems to work. At the rate I am replacing parts I should have just bought the makerbot MK7/8. I ended up buying a new stainless tube and nozzle from makerbot to see if that would help. The only item that will be left original is the aluminum blocks, heater and stepper. So frustrating. It looks like the filament (PLA) is expanding right at the hot / cold junction and creating a wedge. We did notice that the tube hole is larger than the other tubes we measured so maybe that is giving the PLA room to expand.

photo 1-21.JPG

photo 2-19.JPG

photo 3-11.JPG
jamesarm97
 
Posts: 2
Joined: Sun Jul 21, 2013 11:30 pm

Re: QU-BD MBE Extruder Failure Analysis

Postby orcinus » Thu Jul 25, 2013 1:23 pm

That "plug" is actually normal.
You'll see some form of it in nearly all extruders out there.

What's potentially troubling is its length - it should be as short as possible.
If it creeps up too far, the friction will increase and extrusion will stop.
orcinus
 
Posts: 720
Joined: Tue Feb 14, 2012 4:03 am

Re: QU-BD MBE Extruder Failure Analysis

Postby jamesarm97 » Thu Jul 25, 2013 3:19 pm

Yes, I just found out today after receiving genuine makerbot tube and nozzle that the lower hole below the thermal break is drilled at a larger diameter. We have 6 qu-bd hot ends and the first two we tried jam so badly we can't print anything if a few test cubes. We were trying to find the root and possible fix of the problems but I am about to install the makerbot parts and the only original part from qu-bd will be the fan, heater cart and aluminum blocks. I think I will try the teflon take on the steel tube where it screws into the heated aluminum block as suggested somewhere. We were using thermal grease on the top part of the tube going into the cooling block and where the fan heatsink touch the aluminum block. That did not work. Moving the tube further into the cooling aluminum block did not help either. So frustrating.
jamesarm97
 
Posts: 2
Joined: Sun Jul 21, 2013 11:30 pm

Next

Return to 3D Printer General Discussion

Who is online

Users browsing this forum: No registered users and 12 guests

cron