Stepper motor on extruder head gets HOT

[Kerry Chin]

Should I be putting some insulation between the alum block and the stepper motor? I don't rememeber it getting this hot after long prints (15+ hrs) but now my filament is jamming after 1 hr or so, so I'm at a loss as to why this is.

I am running the MB extruder upgrade. So frustrating.

Sometimes I can't unload filament because when I hit the UNLOAD command, the motor doesn't even spin. I release pressure on the filament, still nothing. I have to reboot the machine before the stepper responds again.

[Jetguy]

OK, we need to cover some facts here. The botsteps are controlled via
digipots. These digital potentiometers set the current that the
stepper driver is sending the motor. They use what is in theory,
robust communications protocols.
However, if you read the datasheet, there is an application note that
says they default to mid range.

Mightyboard 2 is NOT open source, we don't have schematics to verify
the circuit. A smart design would be that mid range = a value at or
below correct current for the stepper motors used. Simple resistor
changes could make this happen.

Now, theory portion. This is not fact, simply educated thoughts based
on the reports and knowledge of the system.

Theory here is that for whatever reason (firmware, board design, EMI/
RFI, etc..) something is causing the digipots to toggle to invalid
current setting for the motors.

Some folks have reported hot motors- a clear indication of too much
current, others have reported steppers skipping, likely indicating too
low of current settings.

I can't say for sure a fix or even workaround. I would contact MBI
ASAP and document the failures and all steps, firmware versions, exact
mode of failure, and is it repeatable.

If it were me, I would measure the standard ref voltages, cut or
desolder the ref pin clearly marked on the botsteps, and build simple
ref voltage manual pots with 10 K trimmer pots between 5volts and
ground or maybe even 3.3 volts and ground since the refs are below 2
volts. 1.6XX volts? and 0.6 volts for Z?

BTW, rumor has it that bot that had the melted extruder was the
stepper motor, not the heater as I had suspected earlier.

This report appears on the surface to back that up. 1 bot is a random
fluke, 2 bots starts making a person question and 2 bots with
identical errors indicates a possible repeated defect IMO.

On Apr 22, 9:05 am, Damian Gto <damian...@gmail.com> wrote:
> I can not check this, but I think there is a heat issue with rep 2/2x. Most
> of the time it seams that the filament getting soft.
> I am working with a solution here:http://www.thingiverse.com/thing:74333
> but I can not test it yet and I am sure I will change stuff to it.
> In theory is should work very good and it would remove some heat that is on
> the wrong place.
> When I do get my printer I will do more research and monitor all part of
> the bot to see where the heat is leaking.
> I have seen the same things on computers in the past and its most of the
> time bad air flow that cause problem.
> Most of the time its very easy to fix.

> > nothing. I have to reboot the machine before the stepper responds again.- Hide quoted text -
>
> - Show quoted text -

[Eighty]

I think we're starting to get to the meat and potatoes of the elusive "heat problem" with the new generation of bots.  Something is going on here, and we've got to figure it out.

 

My observations have been as follows:

• The stepper clicking is eliminated (or lessened) with aggressive airflow.  So there seems to be some temperature threshold that causes the stepper's performance to degrade
• After printing for several hours, an otherwise ok print will usually start clicking when smaller features are encountered.  Perhaps the gantry just isn't moving enough to provide passive airflow to the stepper, allowing it to cross this magic threshold?

Is it the stepper that is overheating?  Is it wiring?  Connections?  Stepper drivers?  I'm not an expert on all-things-electronic, so I'm out of my league here.

 

I suppose one possible experiment would be to set the digipots a little lower, and see if anything improves.  Anyone have an idea how low you can go before it could cause issues in printing?

 

Just spitballing here.  Maybe as a community, we can figure this out.

[TobyCWoods]

I have done that... a strip of insulation attached to the bottom of the extruder base and against the "cool" block, but I'm starting to suspect heat from the motor. The cool block never gets hot to the touch even after hours of printing. Yet, I just melted my 4th PLA extruder base. This was after a number of cold pauses that lasted hours. The stepper motor is actually in constant work during a cold pause (hence the high freq whine) and I'd bet it's heating up. I am looking into what Wingcommander did with a heat sink on the back of the motor, but I do not think that will solve it given the motor body is not necessarily designed for heat dissipation from the other end of the motor. Moving to an aluminum extruder may keep the parts from melting but the PLA in the extruder will be effected.

[DHeadrick]

When in the cold pause state, do you have access to the menu to "disable steppers"?

[TobyCWoods]

I forgot to ask... Did ya measure the filament???

Always check your filament diameter immediately everytime your extruder stops working... especially during long prints. Too often I have seen diameter variations beyond tolerance and on my bot, 90% of every extruder problem long into a big print started with a filament issue. 5% stringing/oozing  and 5% because I had too much to drink before starting the print.

[Joseph Chiu]

On Mon, Apr 22, 2013 at 8:27 AM, Eighty <adun...@gmail.com> wrote:

• After printing for several hours, an otherwise ok print will usually start clicking when smaller features are encountered.  Perhaps the gantry just isn't moving enough to provide passive airflow to the stepper, allowing it to cross this magic threshold?

One of my hypothesis about this is that, generally speaking, prints tend to have more solid base, while the tops of prints tend to have more towers (islands) and plenty of travel in between them.  

The other condition that is true is that toward the beginning of the print, the platform is closer to the extruder head, and there is more empty air space down below.  I'm wondering if the airflow behavior is different toward the start of print versus later (either because of platform height or the less-solid printing toward the top) -- I haven't given this a detailed thought yet, though.  

It seems to me like any thermal equilibrium would be reached within 30 minutes or so of operation.    I don't think there's any material/construction features in the machine that would have an hours-long thermal lag.   If that was true, you'd expect a longish cool down period would be needed to reset the machine back to the starting state.

Just some of my ideas tossed into the mixing bowl...

[TobyCWoods]

hmmm I'll have to check... but as I understand it you need to maintain the... uhhh what does Dan call it... ? The register? When you start a print all 3 axes steppers all set their min and their max. That has to be maintained during a Cold Pause or you can't unPause. So they are not turned off. dnewman can explain it better.

[DHeadrick]

Yeah, I'm curious as to what is possible and what is not when in pause mode.  I am working on an extruder design and if the head is too close to the right side of the travel, access to the right side of the extruder is severely limited.  I haven't paused a print in the middle yet to see what it does.  I'll have to play with this to see.

[Carl]

I don't believe there is a setting to disable the steppers in pause mode with SailFish... I can't remember if there was on the MBI firmware...

[Carl]

I've measured temps of around 60 deg C on my Rep 2 steppers... all of them - not just the extruder...

[Eighty]

You definitely don't want to disable the steppers during a cold pause.  You might as well cancel the print, because you'll lose the "registration" of the steppers.  In other words, they'll go slack on you.  When you enable them again, they'll click to the nearest convenient "step", which may or may not be the step (or 1/16th step) that you were at before.  Plus, it really increases the chances of moving the gantry when you're changing filaments (or whatever).

 

Joseph,

I agree with you about the steady state timing, plus or minus a few minutes (don't know what it would be).  I've heard varying reports of the actual time elapsed before problems start.  For me, it's usually several hours.  But I've also heard a lot of people say 1 hour (or even 15 minutes).  It could very well be model-dependent.  I could imagine that a difference in airflow (from underneath) would impact the cooling of the stepper and everything else above.  I'm just having a hard time convincing myself that the relatively minor airflow is that crucial to the success of a print.

 

One other theory that I've been mulling is that the stepper drivers are getting overheated, and that there's a loss of current.  I've been meaning to remove the electronics bay cover and try some long printing to see what happens.  The part that doesn't really add up is that I'm able to "fix" the issue by using a fan on top - that doesn't (or shouldn't, at least), help the drivers below. 

[Wingcommander whpthomas]

Welcome to the summer heat soak blues - already been where you guys are at now.

TobyCWoods wrote:

I am looking into what Wingcommander did with a heat sink on the back of the motor, but I do not think that will solve it given the motor body is not necessarily designed for heat dissipation from the other end of the motor.

Yes it is - the back aluminium block is actually slightly larger,  there are steel plates in between, so drawing heat exchange off the back of the motor is as effective as anywhere else.

However if the motor is straining, it will continue to get hotter and hotter.

 

Moving to an aluminum extruder may keep the parts from melting but the PLA in the extruder will be effected.

I have a couple of thought about aluminium extruders - first they are bullet proof - if you stuff up completely - which tends to happen quite a bit when you are beta testing sailfish and my x3g file converter - the extruder will not break, even if the heatsink fans are disabled and you completely overheat your bot, just let it cool down and its ready for round 2.

I think the Aluminium extruder also works as a heat exchange between the front heatsink and the motor - while this is only radiated heat, not direct contact, I still think it having an effect.

Finally I think the filament just runs through the aluminium feed path with less friction, and it just has a lot more grip. We installed a pair on Kobus's 2X a week ago, and the difference was noticeable right form the start.

All that being said - and there may be a bit of confirmation bias in my observations - the aluminium upgrade extrudes filament about as hard as it can be pushed - and since making these upgrades I have not had one air print.

Now with respect to the causes of 'air printing' here is some stuff I posted on the other forum.

1) Firstly, the MBI thermocouples are not super accurate (+/- 3c I think), so the print temperature should be treated as a number that needs to be calibrated. So I would suggest that you use the sailfish gcTemp override to calibrate your printing temperature on your printer rather than using any specific temp recommendation. Start at 200-210 and run a test print like my acceleration test print http://www.thingiverse.com/thing:61022 and keep upping the temperature 5c at a time until clicking stops, do this for black, coloured and translucent/clear filaments.

The easiest way I know to clog up my nozzle and trigger air prints is to raise the printing temp by just 10c.

I think one of the most concerning aspects of these printers is that users develop a mental model that if the temperature is higher, the filament will flow easier, this is true up until a critical temperature above which filament degrades and begins clogging the nozzle. The problems is that inflection point is not the same on every printer - and may also be effected by ambient temperature and PLA formulation - so printing in a cold garage in winter will require different temperature settings to printing in a hot garage in summer.

2) If the extruder is having trouble printing, either because the nozzle is clogging, too much plastic is being put or its pushing against the delrin plunger - or any combination of these, it starts to heat up, and can reach temperatures of 70-80c (depending on ambient) which raises the position where PLA reaches its glass temperature in thermal barrier between the hot nozzle and cooling block, which is running hotter. This causes the filament to melt higher up the hot end causing it to be less efficient at pushing PLA out of the tip. Add to this that the R2 delrin plunger extruder and the R2X extruder designs are prone to not providing sufficient drive torque to push small clogs out the nozzle and these clogs build up and cause print failures.

I have experimentally tested this theory with my Replicator 2, which has a hood and enclosed sides and a ducted extraction fan. If my extruder is printing on a 34c day without the extraction fan running after about 3-4 hours of continuous printing it will start to fail - cleaning out the nozzle and turning on the extraction fan fixes this. The obvious sign is that when I pull out the filament, it is breaking higher up the hot end, and just pulling off leaving a large glob still inside. However when it runs cooler, a much longer piece of filament is pulled out when unloaded with a long string coming up from the tip.

Another contributing factor, is that the molten filament is also squeezing into the gap between the nozzle and the feed tube, causing it to stick even more.

Now the reason why having a large blob of molten filament in the hot end is a problem, is because we will start to exhibit a laminar flow - that is the layers of filament on the sides will be flowing at slower rates than the filament flowing in the centre, slow moving filament that is already heated to temperatures above what is recommended in the MSDS is going to degrade - or cook. As the filament on the walls of the hot end continues to move slower and slower as it continues to cook, and begins to work like plaque in an artery clogging up the works - it is a runaway failure condition that ultimately leads to air printing. Letting the unit cool down and re-loading may work if it causes the blockage to get pushed out. Alternatively it can be forced out using the toothpick cleaning method.

MY SOLUTIONS

So the first thing to do is calibrate print temperatures with a cool printer and clean nozzle and resist the temptation to raise the print temperature higher when the extruder appears to be straining.

The next thing is to reduce friction and keep the cooling block as cool as possible to make the thermal barrier as effective as it can be. In this regard I have added a 40 x 40 x 10mm southbridge heatsink to the back of my extruder stepper, I have replaced the delrin plunger with my Mk8 extruder upgrade, which reduces the strain on the stepper motor - allowing it to operate at a lower (cool to touch) temperature. I have have used two part thermal epoxy to bond the cooling heatsink to the cooling block so it thermal transfer is as efficient as possible (same for the extruder heatsink). Finally, after cleaning out the extruder heat chamber using the toothpick method, I lined the hot end barrel with Inox Mx8 http://www.inoxed.net/mx8.htm which is a high temperature synthetic PTFE (teflon) grease (with a smoke temp well above 260c).

[Carl]

On a side note Henry (not to hijack the thread!)... the mod on your X-axis limit switch looks interesting...

[Wingcommander whpthomas]

On a side note Henry (not to hijack the thread!)... the mod on your X-axis limit switch looks interesting...

That was my first mod - out of the box the limit switch wasn't registering and the x-axis was crashing - its just a zip tie trimmed, crimped and glued in place with plastic super glue - hasn't budged since.

[TobyCWoods]

As usual...  Good stuff from Wingcommander... I think I may go with the heat sink on the stepper motor next even b4 the aluminum extruder arrives (waiting, waiting).

One question though, WC mentions a gap between the nozzle and the tube... ? What gap? Ain't no gap there. If you got a gap, you got a leak. What you do have is an interface between 2 different metals; SS from the tube and (It looks like) brass from the nozzle. Still given that molten plastic is traveling through there... there can be some impact on flow... but no gap.

I would also refrain from too much time in Cold Pause. If you got a long print and want to go to bed and you can't leave it unattended (not a good idea anyways) A long cold pause with plastic extruder parts may not be a good idea.

[Wingcommander whpthomas]

One question though, WC mentions a gap between the nozzle and the tube... ? What gap? Ain't no gap there. If you got a gap, you got a leak.

Maybe I should rephrase that and say that there is a seam - into which plastic can be pushed and to which it can stick. In fact the spare MBI nozzles I have, have a 0.03mm raised edge lining the outside top thread (from thread cutting) and a slight chamfer on the inside edge. Leaving plenty of room for molten plastic to ooz and attach itself, even if it does not leak.

[Dan Newman]

On 22 Apr 2013 , at 9:37 AM, TobyCWoods wrote:

> hmmm I'll have to check... but as I understand it you need to maintain
> the... uhhh what does Dan call it... ? The register? When you start a print
> all 3 axes steppers all set their min and their max. That has to be
> maintained during a Cold Pause or you can't unPause. So they are not turned
> off. dnewman can explain it better.

In cold pause, the extruder stepper motor is disabled. The X, Y, Z steppers
are left enabled. If you disable the X, Y, Z steppers you will then have
a print defect when you restart since the stepper motor rotor will go to some
arbitrary position when re-energized. Not terribly far away from where it was
left but not necessarily very close either: things can be off by several natural
(non micro) steps. Even worse if some outside force was being resisted by the
motors torque.

Dan

[Joey]

On Monday, April 22, 2013 1:49:59 PM UTC-4, Wingcommander whpthomas wrote:

However if the motor is straining, it will continue to get hotter and hotter.

 

Actually, a stepper motor does not care if it is free wheeling or heavily loaded, they get just as hot either way. Unlike a DC motor where the torque is proportional to the current and a higher load means higher currents/temps a stepper is always run at the same current. It can produce a set amount of torque at that current and any load over that causes it to skip steps, which is the clicking you hear when the filament jams.

Even with no load, just holding position, the current is the same.

If you are seeing the steppers getting hotter than normal, something else is going on.

Joey

[TobyCWoods]

OK then... the extruder's stepper is disabled... so we should not be seeing it get hot in a Cold Pause... correct?

[Dan Newman]

On 22 Apr 2013 , at 4:08 PM, TobyCWoods wrote:

> OK then... the extruder's stepper is disabled... so we should not be seeing
> it get hot in a Cold Pause... correct?

In a Sailfish 7.3, that's definitely the case. In earlier Sailfish releases,
I believe that the current to the stepper motor was reduced but not eliminated
entirely. In the MBI firmware, the current is merely reduced; the stepper is
not disabled.

Dan

[John Armbruster]

Just because I can't remember which forum I posted this in, to agree with Toby's anecdote I'll repeat my own anecdote: I placed a thermocouple between the extruder base and the cooling block as close to the filament as possible. This was to test the temperature of the cooling block with the heat sink in place, and the heat sink removed with the fan blowing directly on the extruder/drive gear/filament. In neither case did the cooling block exceed 40C during a two hour print. In fact the temperature stabilized after about ten minutes. There was only a 4C difference with the heat sink in place and with it removed. So I think whatever heat is melting the extruders is not coming from some kind of heat soak through the cooling block, it must come from the stepper motor. 

[Bottleworks]

Hmm, this is interesting and concerning.... Has anyone done any voltage and current measurements yet?

[TobyCWoods]

So... what kind of epoxy did you use. There seems to be a lot out there, some kinda pricy.

[Wingcommander whpthomas]

Here you can see some carbonised PLA in the seam at the top of the nozzle, which I just replaced.

[Wingcommander whpthomas]

Sorry, clicked on the wrong image - this one is in focus.

[Wingcommander whpthomas]

I might just add as a footnote, that the past three colours I printed with the above nozzle were white, clear and yellow - I have not printed black in ages. Despite the gunk, that nozzle was also still quite functional, I just figured I would replace it while I was re-applying thermal insulation to my hot end.

[Eighty]

Why are the nozzles different?  Is the one on the right a different brand?  From the picture, it looks like it has a flatter mating end.

 

And just for the record, are you sure that this grey stuff isn't some PTFE residue?  I recall you mentioning that you swabbed in a bunch of grease with a Q-tip. 

[Wingcommander whpthomas]

Why are the nozzles different?  Is the one on the right a different brand?  From the picture, it looks like it has a flatter mating end.

I chucked it in my lathe and faced it.

And just for the record, are you sure that this grey stuff isn't some PTFE residue?  I recall you mentioning that you swabbed in a bunch of grease with a Q-tip. 

Could be - I was making the point that there is a seam and plastic does gets into that seam. 

This one has done quite a bit of nylon lately too, so like I said, it was performing just fine, but it has done 100s of hours or printing and the replacements are just sitting on my shelf gathering dust - all shiny and new =)

[TobyCWoods]

I was thinking the same about the PFTE. I've never had anything like that in any of my nozzles. My bot runs all the time (which ain't good for my hearing!) and I've had it only a month less time then WC. Yes, I would definitely ssume that there will be some very small migration of molten stuff in the interface or seam at the meeting of the tube and the nozzle and that it certainly could effect flow.

Say you made the nozzle longer so the heated tube and the nozzle are one part, one material. Still have it meet with a cool tube so that remove/replace is still reasonable.