Hot end jamming fix

[ohcolinho]

Hello all, longtime lurker first time poster here

So far I've been loving the Rostock Max, its a well designed kit and great features. However, since the beginning I've been having extruder woes, not limited to the hot end jamming, retraction inconsistency, and requiring more force than what seems normal (compared to other 1.75mm hot ends i've had) to extrude through the hot end.
After finally disassembling the hot end to unclog a hotend jam, I think I've figured out the reason for the jamming!

Looking at the Hot end 3D CAD models, the total length of the small 4mm OD PTFE tubing should be 45.59mm long in order to ensure that the filament is fed directly into the hot end nozzle
Measuring the ptfe tube that was in the hot end that I received, I found it to be 40mm long. Thus there was a 1.5mm gap creating volume for the ABS to squeeze into and sit around, rather than directly extruding the nozzle.

Here is a diagram showing the gap, as well as what the proper length of the PTFE tubing.
[img]http://colin-ho.com/wp-content/uploads/ ... oblem1.png[/img]

Heres the picture of the length of the PTFE tubing I got with my extruder
[img]http://colin-ho.com/wp-content/uploads/ ... .14.16.jpg[/img]

Lesson learned: don't assume the hot end was assembled perfectly, take it apart and inspect the PTFE tube length and make sure there aren't any gaps!

Another problem I found was that the PTFE tubing was that a bit of it was crimped down on the end facing the nozzle, thus creating a corresponding dent in the PEEK piece.
[img]http://colin-ho.com/wp-content/uploads/ ... .16.57.jpg[/img]

Corresponding indentation in PEEK
[img]http://colin-ho.com/wp-content/uploads/ ... .17.52.jpg[/img]

I suspect that this may have enabled the ABS to initiate its creep into the gap, and the eventual pressure of the filament shoved more molten ABS into the gap further widening it.

I Hope this helps for others with hot end issues!

[626Pilot]

Nice find.

When I install Windows, it comes with Internet Explorer. Some people like IE, but to me, its only purpose is to download Chrome or Firefox. I look at the stock hotend the same way: use it to print some calibration pieces, then use it to print mounting hardware for another hotend like the Budaschnozzle. Mine has jammed so many times printing PLA. I've used one fan, even TWO fans blowing directly across the PEEK section - it still jams. With ABS, I have much better luck, but the damn thing STILL jams once or twice a day, probably because the method of using two nested PTFE tubes to keep the filament straight has too much to go wrong with it. Eventually I have to unscrew the nozzle and clean everything out. What a mess.

In my honest opinion, SeeMeCNC should just ditch this hotend. The guy at http://hotends.com is selling a better J-head hotend for about 55 bucks, vs. the $40 the SeeMe hotend is listed for. It has less internal friction, and the PEEK is milled so that the thing will actually cool well enough for PLA with a single fan. It's a proven design that has already gone through a bunch of iterations. On top of that, the EZStruder is designed to mount directly to this type of hotend. Sure, they're trying to sell to more than just Rostock owners, but I've begun to wonder if SeeMeCNC is considering dumping the old hotend in favor of making their own J-heads. I doubt it would cost much more to produce, if anything, because it's designed to require the least machining possible. It's not like they don't have the know-how.

[Michi86]

Got the same problem
i fixed it (because i have no PTFE tube at home) by miss out the rubber from the bowden tube holder (is this the correct name?), so that the bowden tube holder pushes in the too short piece of PTFE tube into the nozzle

[ohcolinho]

Update: The cool people at SeeMeCNC is shipping me a new PEEK adapter to replace the damaged one! I'm impressed by their service

[JohnStack]

I tried taking off the Peek on the stock hot end, it was _really_ jammed I guess.

SeeMeCNC is working on a new hot end so there's that.

Tricklaser too!

I have used the HotEnds.com version. Peek Fan MANDATORY. Don't run it without, no matter what temp.

I have not had time to install the eBay version here: http://www.ebay.com/itm/new-1pcs-RepRap ... 19dc6515ef

And a while back, I ordered what all the long timers run: the E3D from http://e3d-online.com

[RichW]

I think the teflon liner acts like heat shrink when it's heated too much. One of my problems has been with the inner liner shrinking in diameter. The first one I replaced, had to have the hot end drilled out to 2 mm. It was hardened as well as shrunk, I never thought to check the length. It would seem that ABS is just too much for the SMCNC hot end.
I expect to change to an all-metal hot end soon, either a custom piece or the E3D looks good to me.

[PartDaddy]

Starting with a stock SeeMeCNC Hot End and resistors for the heating elements (not cartridges). . .


First: Oly & I designed this hot end for ABS and PLA. It can not handle high temperatures (ie. over 250C). You begin to take a change melting the PEEK.



Second: review ohcolinho's post. I'll be reviewing hot end assemblies in production. The 2mm ID x 4mm OD teflon tube should be cut square and seat into the nozzle. However, it should not compress much, if any, when the nozzle is tightened. On the top end, it will touch the push-to-connect fitting. It should not be short.



Third: What we found is that a cooling fan is needed on the RAMBo board. The direct drive EZStruder requires more power to push the filament. What many people think is a "plug" is nothing more than the stepper driver overheating. The chip has thermal protection and shuts off in an over-temp situation. It shuts off for only a fraction of a second, then turns back on. What you notice is a "skip" in the extruder. Plugs do happen, however, many problems we're seeing involve an overheated stepper driver chip. A fan, like our little 25mm fan, pointed at the driver chip solve the problem.

My notes on plugs: PLA is more sensitive to plugging rather than ABS. This is why we always recommend starting with ABS. The speed, layer, temp, retract, etc. all effect printing. With the SeeMeCNC hot end, too much heat can cause problems when printing with PLA because heat creaps up the hot end. A fan, say 40mm in size pointed at the top half of the hot end solves a lot of problems. I've been printing with a 0.35" dia nozzle, 50mm/s, our translucent orange PLA, bed 60C, nozzle 175C with a 40mm fan blowing on it. John has been printing about the same, but with our little 25mm fan inside the fan shroud available on our seemecnc github.

We've spent many hours testing a new nozzle design which gives even better results when printing with PLA. It will be available soon. It is backward compatible with all hot ends shipped with the Rostock MAX.

[JohnStack]

Hey PartDaddy, What are the official tube lengths inside the nozzle?

Recommended fan size to be that small or something larger to cool the entire board?

[edited]

[RichW]

Thanks to Part Daddy for the skinny on the EZstruder and Rambo situation. I have been having pretty good results with the EZstruder on small jobs, but the printing of a new magnetic effector plate has been giving me trouble. As long as a large area is being printed, with a steady flow of ABS thru the hot end, everything is Jake. When the small details are all that's left, the flow stops. I will try it again with a 30mm fan on the Rambo.

[626Pilot]

I'm glad to hear SeeMe is working on a new hot end. Will be interesting to see what it looks like.

[Octavio]

Hello Everybody,

I finished assembling my Rostock MAX just recently and I have not been able to get a good piece out of it.
I'm testing it with the hallow cube and after millions of different problems (blown thermistor, ABS not sticking to the heated bed, heated bed not reaching the 90˚ temp) now I'm stuck with pieces that start nicely and end up bad.
It seems the flow of the melted ABS starts failing in the middle of the piece.

Here is a photo of two bad pieces:
https://www.dropbox.com/s/iyle8lzsnm89b ... owCube.jpg

Does anybody know what is the problem or even better, the solution?

Many thanks in advance!

Octavio

[JohnStack]

Stop printing and trying different combos. What are all of your settings? What software?

[lordbinky]

I suggest making a new thread for you're issue Octavio, it is likely to get more attention that way. John is right in that we'll need alot more information about your settings to give any better than general advice. Also while troubleshooting, turn off/disable/zero any extra features in your slicer such as retraction, Z-Lift, etc. We need as few variables in play to narrow down the problem.

Now for that general suggestion, IF your filament settings were correct which is a big assumption at this point (there's lots of setting tweaks that go on for each filament color/brand/type/etc), that problem looks similar to what happened when my hotend could not keep it's temperature up. If you suspect this, check the temperature graph in Repetier Host or watch the LCD display when watching your print.

[Octavio]

Hello John and Lordbinky.

Thank you for your help.
I've started a new thread and added information to it (is this enough?)
You can find it here:
http://forum.seemecnc.com/viewtopic.php?f=62&t=2492

Thanks' again,

Octavio

[dsnettleton]

Starting with a stock SeeMeCNC Hot End and resistors for the heating elements (not cartridges). . .

Third: What we found is that a cooling fan is needed on the RAMBo board. The direct drive EZStruder requires more power to push the filament. What many people think is a "plug" is nothing more than the stepper driver overheating. The chip has thermal protection and shuts off in an over-temp situation. It shuts off for only a fraction of a second, then turns back on. What you notice is a "skip" in the extruder. Plugs do happen, however, many problems we're seeing involve an overheated stepper driver chip. A fan, like our little 25mm fan, pointed at the driver chip solve the problem.

I didn't know about the thermal chip in the board. Is there an easy way to tell whether the EZStruder is skipping from the over-temp rather than from a plug, or is the behavior more or less the same? Also, would it help to decrease the motor current in the firmware to try and keep temps down?

[Apollyon25]

I recently had many of the same problems with the hot end jamming and/or the extruder (EZstruder) slipping.
There were multiple causes of slippage:

1) print head too close to the bed
As noted elsewhere on this forum, if there isn't room for the filament to come out, the internal pressures in the nozzle will cause the EZstruder to slip (magnetic loss of holding torque)

2) incorrect nozzle temps
Again, noted elsewhere on the forum, too hot, and it slips, too cold and it slips - try using the calibration print (tall print with 180, 215, 185, 210 etc listed on the front of it) - <link>

3) trying to print too fast
unless you have everything dialed in perfectly, don't mess with increasing speeds...

4) the NEMA17 (400step/rev) having insufficient torque to push
Slipped from day one… all the current I could give it, minimal current, micro stepping or not…

5) the current supplied to the motor being too low
6) the driver entering current limit
7) the driver entering thermal shutdown
The driver, (pololu using allegro chip) was set by default to only supply <1A to the motor. This is to prevent consumers blowing the chip.
To fix this, I had to replace the top resistor in the Rdivider (bottom being the 10k trimpot) to 4k7 to allow the set point Iref to able to be set to max (2.4V).
This allows full current into the motor. This improved matters but there was still incidents where the Ezstruder would slip.
Also, do not do this without a heatsink on the chip AND a fan - there is too much internal dissipation at full current.

8) a weird Slic3r bug which would try to extrude 10mm of filament every so often.
Swapped for KISSlicer - does not have this feature.

All result in the extruder slipping.
The only solution that I found that actually worked was to replace it with a 1:5.18 planetary driven motor based extruder.

I hope this helps someone. If anyone requires further information, send me a PM.

[Eaglezsoar]

It's good to see that you did resolve that issue. A lot of extruders are based on that 5:1 geared stepper and they work great.

[626Pilot]

There is also the filament, which will often absorb moisture from the air and swell. I have a bucket with some molecular sieve dessicant in it where I keep my most troublesome filament. Some people do the same thing, but they put a light bulb in the bucket to help the process along. It's also possible to heat the filament in an oven. You have to use very low temperatures. Even if the filament doesn't melt it can still stick together, and if the spool is made of plastic, it can melt at a much lower temperature than the filament. Maybe someone else can chime in with the best temperature for that.

[mhackney]

Great troubleshooting list.

And, 1, 2 & 3 also are frequently the culprits for not getting the first layer to stick. Never take anything for granted. I've been producing 50+ parts a day for the last few days. Everything is running smoothly, printer nicely calibrated, the world was a happy place. Then yesterday, I loaded some new pink PLA. It simply would not stick to my paper (I've been printing with 100% success on typing paper). Perplexed, I checked the Z=0. Sure enough, it was about .07mm too far above the build surface. A quick update to the EEPROM settings and the pink started sticking like glue.

[Apollyon25]

Further to my list...

I was getting jams again so have ordered, as a backup to the J-Head mk5b, an E3D mk5 3mmm bowden hotend, but the day after that, my samples of some custom PTFE tubing arrived...
I got 5.0mmOD with 3.0mmID and some 5.0mmOD with 3.2mmID (all tolerances assumed to be +/-0.1mm).
The 3.0mm ID as expected was too close to the OD of the filament 2.9 worst case, 2.775 average.
Finding PTFE tubing here in NZ has, til now, been fruitless.

Comparing actual Dupont Teflon PTFE to PFA (which supposedly has 'similar' friction properties) and the differences are poles apart!
Even my bent, ribbed filament slides MUCH MUCH easier through PTFE than PFA.

One 3hr print (at 100mm/s) later, no jamming!
So (touch wood), this may be solved finally.

[Eaglezsoar]

Excellent troubleshooting skills! We all hope that all of your problems are gone and you can get down to some serious printing.

[Apollyon25]

Touch wood... but I think that was it.

The printer has been running solidly for several days straight on a bunch of different prints - torsos, terminator head, component drawers, bits for itself etc etc and not a single nozzle jam-type issue...

I had to re-add some part cooling fans (PLA)...

I guess adding a second extruder is now in the works... and adding ABS.

[Eaglezsoar]

Hey PartDaddy, What are the official tube lengths inside the nozzle?

Recommended fan size to be that small or something larger to cool the entire board?

[edited]

The 4mm tube length on the newest hotends is 45.1mm

[teoman]

A quick question about replacing the ptfe inside the hotend.

Can you remove the nozzle and use needle-nose pliers to pull the old ptfe out and then insert the new one or does the PEEK section need to be disassembled?


Thank you in advance.

[Mac The Knife]

A quick question about replacing the ptfe inside the hotend.

Can you remove the nozzle and use needle-nose pliers to pull the old ptfe out and then insert the new one or does the PEEK section need to be disassembled?


Thank you in advance.

Yes you can.