Mhackney's Rostock Max

[aehM_Key]

Had no time to run the machine till than. I'll probably go for the full 255 current and tweak the acceleration. Will see, if the stepper is not getting too hot.

[mhackney]

I'm running 1125 on acceleration and 222 on current. However, I have now been plagued by the filament buckling problem and I'm trying to address that. There is some dynamic in this Bowden setup that we need to get to the bottom of and understand!

It does not take an unreasonable amount of force to push filament through the hot end by hand through the Bowden tube. That is good as it demonstrates that there is not a lot of extra friction introduced.

In order for the extruder to buckle and eject filament between the rollers and Bowden connector, there must be good grip on the filament, otherwise it would simply slip. This is also a good thing (the grip that is).

So what is going on between these observations that is leading to the buckling? Something must be preventing the filament from extruding and the extruder tries to chug along. The only place where the pressure can be released is that little gap between the rollers and Bowden connector or the filament could slip. I've been conducting some controlled experiments to try to get to the bottom of this. I ordered a few things to fabricate an instrument for measuring the pressure required to push the filament so I can collect some data.

Interestingly, in my case, these problems did not occur when I had "0" retraction and ran the extruder at 60mm.s. It wasn't until I enabled retraction and increased the extruder to 100 mm/s to improve print quality that it reared its ugly head. Also interestingly, I have a part that fails at the exact same place on the 2nd layer infill right at the time a LOT of short extrude-retract cycles happen - leading me to think that maybe hydraulic pressure is built up in the hot nozzle and the only relief is back at that Bowden/roller junction. This part did not fail with the previous settings but had some globbing that I was trying to address. Unfortunately, now that I realize this, I can't go back and quickly reset to 60mm/s and no retract to test since I have my extruder disconnected while I work on the problem.

[aehM_Key]

Do you have a link to this "complicated" part? Maybe I'll try it.
You could install this extruder-update part, which prevents buckling.

[mhackney]

The part is the precision block in The Essential Calibration Set

I saw the printed extruder cap but I don't think it will work on my machine - the gap is already so small, .5mm or less. When I get ejection, the ribbon coming out is flat! I am working on an alternative - a delrin insert that will ride very closely on the rollers. Should be done today.

[aehM_Key]

Any special settings needed to reproduce the problem? Infill settings?

[mhackney]

This was happening at layer 2 so infill at that point was solid. The rest of the part I had at 50%. 5mm retract 1mm Z lift, 5mm extra length on restart (this could be part of the culprit).

What I'd like to do is hold the Slic3r configuration fixed and adjust extruder stepper speed and acceleration.

[aehM_Key]

5mm extra length on restart (this could be part of the culprit

Hm...I remember vaguely, that this is not recomended.

[mhackney]

Perhaps not, but with 5mm retract I was getting starving too. Just another variable!

[mhackney]

Well, I decided to put the extruder back on my machine and do some controlled tests before making any mods.

Firstly, I left my retract and restart settings, accel and velocity settings alone and tested the precision block.
Result: filament ejection at the same place it usually occurs. This is my "bad" baseline.

Then I reconfigured Slic3r and removed all retract related settings (exempt the 1mm Z lift) and reconfigured Repetier firmware back to 60 mm/s feedrate and 1125 mm/sec^2 acceleration. I did leave the Flow Multiply slider at 75% in Repetier host - I had been testing this before and wanted to make sure I had a control.
Result: this part printed to completion and had, as expected, some blobbing and minor infill issues (from the low flow setting)

I am now printing the same but with 100% flow.

[mhackney]

The 100% flow sample is complete and other than being better infilled, is about the same as the previous. I am now going to test my hydraulic pressure theory by extruding into air and handwriting some extrusion/retraction gcode.

[Polygonhell]

In order for the extruder to buckle and eject filament between the rollers and Bowden connector, there must be good grip on the filament, otherwise it would simply slip. This is also a good thing (the grip that is).

So what is going on between these observations that is leading to the buckling? Something must be preventing the filament from extruding and the extruder tries to chug along. The only place where the pressure can be released is that little gap between the rollers and Bowden connector or the filament could slip. I've been conducting some controlled experiments to try to get to the bottom of this. I ordered a few things to fabricate an instrument for measuring the pressure required to push the filament so I can collect some data.

FWIW the buckling filament problem isn't related to the Bowden tube, there are people in the H1.1 forum reporting the same issue using the extruder directly.
However it is made worse by the somewhat extreme retract/returns that the Bowden tube requires.
IMO it's an oversight in the design, any gap between the rollers and something that constrains the motion of the filament will allow the filament to buckle and get ejected.
If you look at other extruder designs, they usually closely fil the idler and drive component with a tube, and for example people have a similar issue trying to use a Wades extruder designed for 3mm with 1.75mm filament because the guide at the bittom is wide enough for the filament to buckle inside of.

FWIW it's a complete none issue with PLA because it's stiff enough to not buckle, but ABS and even worse Nylon it'sa nightmare IME.
I have a partly completed direct drive extruder I keep meaning to finish and try, but my motivation waned after I swapped back to printing PLA.

[mhackney]

Thanks Polygonhell. I have an H-1 for a year and not experienced that. My son's H-1.1 I believe did experience it a couple of times. When I look at mine, the gap between the rollers and bottom of the adapter is really small! I can't understand the physics that would cause the filament to eject like that!

I have designed a Delrin insert that will close the gap that is the same concept as your cap but instead of fitting around the outside, I am boring a 3/16" x 1/2" deep hole in the aluminum part. I'll post photos when it's done.

I have 2 different extruders and hot ends arriving this week to try out. You probably sense that I have a scientific nature and like to experiment using controls and measurements! I've long wanted to experiment with extruders and hot ends so I finally bit the bullet.

[Av8r RC]

Mhackney, just curious. Which hot ends did you purchase?

About a year ago I bought the Sumpod 3D printer, and it was plagued with hot end problems. I neve could get the stock one to work. I ended up trying my hand at designing and making my own. I did have some success with it (way more than I ever got with the original one), but the prints weren't that great. Instead of trying design after design, I decided to buy one. I went with the Budascnozzle, and it proved to be a great hot end. My only complaint with it is the aluminum on aluminum construction at the nozzle. When it's brand new you can change out the nozzles easily. But once you get a few heat cycles thru it, it's virtually impossible to swap out the nozzle without damaging the whole assy.

If this is one of the ones you got. I would recommend putting some anti-seize compound on the threads (haven't tried it yet). Or better yet make a new threaded body out of brass, I plan to make one of these at work if we can ever get a lathe open at the end of a day.

[mhackney]

I bought a JHead hot end (here tomorrow) and I'll print an extruder for that. Not sure which. But I also wanted to try it with the Steve's setup and take some force measurements. I also bought a QU-BD MDE with the 12 V cartridge heater. Looks like a nice simple little unit. I considered the buddashnozzle and might still get one at some point.

cheers,
Michael

[Polygonhell]

I bought a JHead hot end (here tomorrow) and I'll print an extruder for that. Not sure which. But I also wanted to try it with the Steve's setup and take some force measurements. I also bought a QU-BD MDE with the 12 V cartridge heater. Looks like a nice simple little unit. I considered the buddashnozzle and might still get one at some point.

cheers,
Michael

I have Budda and a JHead, on other machines, both are good hotends, but they are very different.
The Budda has a very short nozzle, and requires very little pressure to exrude, the JHead is a lot closer to the Steve's with the current nozzle.
I've heard bad things about the QU-BD (I know they use a different design in the printer they sell), but don't have one, and have no first hand knowledge.
I would really like to pick up one of the all stainless hotends that seem to be coming into vogue, notably Prusa's.

[gabrielk]

i have buckling problems yesterday also, I tried to use Polygonhell cap first, but when i read other posts about that kiond of problems only in the begining, i started to think something is wrong with my extruder. so I disconnected bowden tube from the extruder. And i measured with dial calipers filament before extruder and after it. Difference was .3mm!!! so it was to tight and after few retracts (9mm) it was weakened a lot and try to crack. that was the first issue. Second one is to very very precisely aseembly of rollers and tube connector so they are in line.

[mhackney]

Yes, too tight and it will distort your filament. I've been through all of that and have my extruder a) not distorting the filament and b) not slipping. Your point about retracts perhaps weakening the filament is an interesting observation. Perhaps the friction of this rapid motion several times heats/distorts the filament enough to initiate the problem.

[aehM_Key]

@mhackney: Have you made comparisons in print-quality with and without the 25mm-fan?

[mhackney]

The 25mm fan is to cool the barrel - a requirement for printing PLA and perhaps a nicety for ABS. I have not attempted to print PLA with this machine yet. I do however think it prevents "melt creep" of the ABS since I have had no clogs in the teflon tube since installing it. So, does that affect print quality? Only in a gross way - it prevents premature print abortion! Other than that, no, there does not appear to be any improvement in surface quality, etc of the print.

cheers,
Michael

[geneb]

Found this - should work well!

http://www.jameco.com/webapp/wcs/stores ... 2100363_-1

The big deal with the fan on the upper section of the hot end is due to heat creep. The upper section of the hot end will begin to heat up and cause PLA to jam HARD. Can do all kinds of evil things when that happens.

g.

[mhackney]

That has a lower volume than the one I have. It most likely will be enough though. I am about to purchase 50 units of another model to resell. These are higher volume and just a hair quieter.

[geneb]

Jameco sells the model # you listed, but it's the three wire version. If memory serves, third (signal) wire is a PWM output, correct?

g.

[Av8r RC]

Mhackney, do you or anyone else for that matter happen to have some dimensions of the onyx bed. I just had a mill open up this afternoon and I can fab myself an aluminum plate for it. But I don't have my printer here with me.

I know the dia is 12 3/16. Anyone know the bolt circle and the distance to the flat at the top?

Thanks

[aehM_Key]

Thought the fan would probably influence the retraction and with this the blobbing. I'm not planning to print with PLA.

[mhackney]

@geneb - yes, I got the originals at Jameco. The 3 wire version is for PWM, however, you just clip the white wire and hook the red and black. If you are connecting to RAMBo on the extruder 0 fan port, you will get speed control. I currently have mine direct wired.

@Av8r RC - I actually make my plates 12 1/8" diameter. I'll machine a few this weekend and offer them on my site on Monday. I have enough material to fabricate 6 (had enough for 10 but 4 are already spoken for). The bolt hole pattern looks to be on a 285mm diameter. I just looked at the ROSTOCK dxf files but they do not have the Onyx holes on them. The SeeMeCNC site says: Six 3mm diameter mounting holes spaced 300mm center-to-center so there you go!

You will need to cut out for the electronics pad area too.