Extruder voltage increase

[david.e...@gmail.com]

I started using a cubex with non proprietary filament but at higher speeds it racks. Apparently, increasing extruder voltage help a lot but the owner is concerned, especially as we almost had it die on us (sd card issues). Could somebody go over the risks so that I can get it working better?

[Duplicat]

The print is racking ?  or the printer is racking (shaking)?
I'm currently modding a CubeX, and I noticed that without the case on,  the frame isn't stable enough to print. (mostly just on the Y-axis). The forward and back motion causes the bed-arm bearings to bounce the entire bed (potential problems with Z-height layers).
As beautifully engineered and machined as these machines are,, it appears they are still reliant on the casing for additional frame/structural support.

[Joe Kuter]

The voltage adjustment isn't the actual extruder voltage.  It is a  reference voltage that determines the amplitude of the outputs to the stepper, which are bipolar and vary due to microstepping. Too complicated to explain here. 

After bumping up my reference voltage from 1.1 to 1.5 I have noticed no adverse effects.  The pulses to the stepper peak at about 15 volts which is well within the limits of the motor and driver.  Didn't do any before readings so I don't know if there is a 1:1 correspondence.  Stepper output torque is directly proportional to the driving voltage.

The major risk is in your level of knowledge, expertise and dexterity.  You must get meter probe on to the test point without shorting adjacent connections on the board.  Also, the adjustment pot is very sensitive. Do your adjustment in 1/16 turn increments.

Hope this helps in your dialogue with "The Owner".

[Hugues]

Hi,

I push my two voltage at 1.5v, doing some 60h print without issue.
It's easy to do, take your time !

[david.e...@gmail.com]

On Tuesday, January 27, 2015 at 1:30:36 PM UTC-8, david.e...@gmail.com wrote:
> I started using a cubex with non proprietary filament but at higher speeds it racks. Apparently, increasing extruder voltage help a lot but the owner is concerned, especially as we almost had it die on us (sd card issues). Could somebody go over the risks so that I can get it working better?

Sorry, I kind of let this die. So their should be no problems, no chance of failure? (assuming we take precautions and are careful).

[david.e...@gmail.com]

By the way, racking is probably the wrong word. The gear cannot push the filament through because of the friction from the nonmoving idler wheel. It makes a loud "CLUNK CLUNCK CLUNK" noise.

[Jetguy]

Covered IN DEPTH in another thread.

Here is the driver data sheet http://www.allegromicro.com/~/media/Files/Datasheets/A3979-Datasheet.ashx

Per the data sheet, the formula is 

The maximum value of current limiting is set by the selection of RS and the voltage at the VREF input with a transconductance function approximated by: ITRIPmax = VREF/8RS

RS is the resistor values. By looking at the board, we see the giant current sensing resistors near the chip are marked R100 http://www.mouser.com/ProductDetail/IRC-TT-Electronics/LR2512LF-01-R100-F/?qs=8k2Nf0ft%252bwguafT9kteOCg%3D%3Dbasically we know they are precision 0.1 Ohm resistors.

So using the equation again, our target is max allowable current of 1.68 Amps per the data sheet for the motor.

We know we have 0.1 Ohm current sensing resistors. 8 times RS is 0.8 

Basically, whatever Vref is voltage wise, times 0.8 (8 times 0.1 Ohms) = output current.

1.6V vref times 0.8 = 1.28A

This is why I wanted you to do the math and understand what is logical here.

1.6V is a safe V ref voltage, but it's ALSO not the max Vref voltage that equated to maximum rated torque for the motor.

In order to reach rated torque, YOU MUST feed the motor with rated 1.68A of current.

In other words, given the motor spec, given the data sheet, and given the resistor populated on the board, in order to get maximum torque out of the motor, you can set the Vref as high at 2.1V

2.1V times the RS factor of 0.8= 1.68A

A quick check if 1.68A is max, and 1.28A is a setpoint, then that only equates to 76% of maximum rated torque.

Again, for you, I would go somewhere between 1.7V to 2.1 volt with the understanding, the closer you go to 2.1, the hotter both the driver and the motor will get. The driver has a flan blowing right on it- the motor does not.

[Joe Kuter]

Jetguy..........Great technical breakdown!

The discussion in this thread has gone down the voltage adjustment path without exploring other factors.

Increasing the motor torque can fix minor issues, but there could be other factors at play.

A clicking sound means that the filament is captured properly by the extruder wheels and that the motor is unable to advance it into the hot end. This can be a push or pull problem, or a combination.

Here are a couple of quick experiments:

On the pull side, execute the cartridge change function for the extruder. After pulling the filament loose, make sure that it is not binding in the feed tube. Also make sure there isn't excessive drag from the feed spool. Pull the filament all the way out and examine the end. The bite marks will show how far it is passing into the hot end beyond the wheels. Also, the the tip should have evidence of melting.

On the push side, Is it trying to push into a pillow, or a brick wall? Go to the PJ-CON menu and set the hot end temperature to the max, 260. Wait a minute after it reaches 260 and increase the RPM up to

.2. If filament starts to extrude without clicking, bump up the RPM in .2 increments until it starts to click. Use this information to determine your slicer settings.

If nothing comes out at any speed, there is possibly a jam in the hot end. The internal PTFE tube has a 2mm inside diameter which could bind with thicker filaments. I have used bulk filament from several suppliers with no binding problems.

[James Kincaid]

I see this is an old thread.  I found it when doing a search about something else.  I just wanted to comment that the problem may not be the stepper motor at all, but rather the wattage of the extruder.  My thinking is as you increase the speed the flow rate of the filament coming into the extruder is increased.  This would mean that the extruder needs to be able to add more heat in order to keep up with the increased rate of filament cooling it down.  What I see happening is the extruder is not supplying enough heat quick enough,  thus the filament as it's coming out of the extruder is more viscous, thereby causing symptoms similar to a jammed up extruder.

[Jetguy]

No, sorry I cannot support that conclusion.
#1 this is an all metal contact design extruder and aluminum of all highly heat conductive materials.
#2 the thermistor is in direct contact with the internal aluminum wall inside the heater core. You'd be hard pressed to get a more accurate system.
#3 there is NO way volume is every high enough on this extruder to exceed flow rate and cool the bore.

Being honest, the massive gantry in this CubeX and in theory the Cube Pro is the weakest point. That mass is what limits the feedrate of everything else.
However, there have been massive discussions in other groups about specific internal geometry of all metal extruders and to my knowledge, the CubeX does not have a stepped bore required to control the melt cap during retraction.
That leads to what we call a retraction based cold end jam. Hot molten filament is pulled up during a retract and cools and sticks to the upper cold end bore. That prevents a feed later
The stepped bore design in the thermal zone of other extruders shaves off this melt cap and prevents the cold end jam.

Again, there is NO way, and I can say that without ever a doubt, this is not a  feedrate based temperature problem. If anything, the stock system uses too high of temps anyway.

[James Kincaid]

Interesting comment, but my intuition is causing me to feel like I may be correct here.  One way to find out would be to do some actual tests.  For that, if the original poster is seeing this, two questions I'd have would be, what type of filament were you using and at what speed where you printing at?

I would love to be wrong here because I like to learn new things.  So let's find out.  :)

[James Kincaid]

Well, I guess I'll need to set up the test on my own parameters.  The orriginal poster replied to me privately, but only to say that they agreed with Jetguy.  When I asked what the parameters were that caused the filament injection failure characteristic they did not reply back, strangely.

One rudimentary test that I did was to determine roughly how much force a stock CubeX filament drive system could apply to a 1.75 mm filament at 10 rpm and it felt to me like around 2 kg.  I need to get a digital force meter of some type to see what the exact number is.

As for simulating the failure, I think what I'll do is create some custom Gcode that steadily ramps up speed on a print of a simple cylinder so that I can find out at what feed in millimeters per second the failure occurs using 1.75 mm filament.

I have numerous things I'm working on so it may be a while before I report back but I do have it on my to do list so I will report back eventually.

[Giovanni]

Hi,
I've increased the voltage up to 1.8V.
When testing with PJCON everything was OK, no clicking even with high RPM ( 8-10RPM), but while printing, after a few layers the extruder started to click exactly like if the voltage was low and maybe even more causing bad print quality and sometimes aleatory abort printing with "Flow Fail Errors".
I discovered that the driver was very hot.
The driver is this chip on the print jet board :


I placed a cooler on this chip and the problem is solved, now I run 1,8V and I forgot what a clicking extruder is.


The cooler is glued with a thermal epoxyde paste.

Hope this may help others.

Giovanni

p.s. I've started to work on a new version of Cubeitmod

[Jetguy]

The problem is, higher current ALSO heats the stepper motor and the surrounding aluminum filament drive section.
When printing PLA, this will cause a feed jam from softened PLA inside the feeder.

FIX the PROBLEM!!!
By that, you changed to non-standard settings- that caused problems at the driver- you hacked that, now you are going to find the next problem further down the chain.

Your problem is hotend related, it shouldn't take that kind of force to extrude/

[Giovanni]

Hi Jetguy,

the main problem is not hotend related, the main problem is that Cubify has intentionally set a low voltage to produce "Flow Failure Errors" if you print with bulk filament, I had a lot of problems with a PVA filament that was very irregular, the diameter is from 1.7mm up to 1.85mm, this filament requires a lot of torque to pass into the drive section but not into the hotend.

It's true that the motor heats a little now, but not enough to cause a problem, the drive section is connected to the hot end sink and both are in aluminium, this helps to cool the part.

Since this mod I've printed large parts in PLA, more than 10 hours of printing time with a single extruder and had no problems.

[Peter Gregory]

I've increased my voltages to 1.6v and have pretty good torque for 3rd party filament (Matterhackers or WinBo).
However, I notice it still has problems with older filament.
Once I take it out of the shrink wrap, it works great.
After a couple of weeks in the air, I start getting hard blobs and clumpy extrusions.
I tried keeping my filament in a bag with desiccant, but it still seems to clump and blob after a while.
Maybe I need to buy vacuum bags?
Anyway, I was wondering if older filament and high torque would damage the hot end.
Maybe the clumpy parts would expand the extruder hole or even break the ceramic casing around the heating element due to high pressure flexing.
Any thoughts?

[Jetguy]

The software doesn't set voltage- you honestly don't know what you are talking about, you failed to do the math.

I have done the math, the research and know the system like the back of my hand.

If you have flow errors IT"S A DIFFERENT PROBLEM.

How else can you explain how 90% of ALL consumer printers using the Replicator design use a SHORT single stack motor, similar drive gear pushing 1.75mm filament and DO NOT skip steps, yet the CubeX has a giant honking long body NEMA17 easily higher torque- and yet skips steps all the time??

Exactly, it's crap. By IT, I mean the CUBEX hot end system. Yes, i very much like the filament drive, yes, I like the encoders so that we could detect errors.

However, it shits the bed. There is no polite way of saying this and previous attempts seem to not clue you in.

[Hugues]

There are too much functionality on this extruder without filament diameter liberty. I unmount it lot of time, sometimes you find dust on the encoder weel, sometime the wheel is bend... Those are causing error and sometimes (mostly) there are no visible reason, hotend is not jam.
I think the solution will be to use the stocked extruder as a slave (without motor) for information feedback and a classic direct extruder in bowden for filament pushing.

[Giovanni]

Le samedi 28 novembre 2015 23:28:00 UTC+1, Jetguy a écrit :

The software doesn't set voltage- you honestly don't know what you are talking about, you failed to do the math.

English is not my mothertongue, certainly I made mistakes but after reading my message I didn't understand how do you can think I wrote that the software sets the voltage ??? I never said that !

 

I have done the math, the research and know the system like the back of my hand.

If you have flow errors IT"S A DIFFERENT PROBLEM.

How else can you explain how 90% of ALL consumer printers using the Replicator design use a SHORT single stack motor, similar drive gear pushing 1.75mm filament and DO NOT skip steps, yet the CubeX has a giant honking long body NEMA17 easily higher torque- and yet skips steps all the time??

Exactly, it's crap. By IT, I mean the CUBEX hot end system. Yes, i very much like the filament drive, yes, I like the encoders so that we could detect errors.

However, it shits the bed. There is no polite way of saying this and previous attempts seem to not clue you in.

You better redo your maths, because you didn't understand where the problem is.

The main difference between the Cubex extruder and others brands extruders is that the distance between the wheel on the motor and the friction roller is fixed, other systems have the friction roller pushed with a spring.

If that distance is fixed and you try to pass a larger filament the wheels will press on the filament very hard and the toque required to pass the filament increases a lot, this will not append with other brands because the spring will keep quite the same force on the wheels while the diameter of the filament changes.

This "wonderful" design coupled with a very low voltage on the motor will make the motor loosing steps while trying to pass bulk filament, because Cubyfy filament is smaller than bulk filaments,  1.7mm for Cubify and 1.75 +/- 0.05 for the best bulk filaments, but a lot are 1.75mm +/- 0.1.

And all this coupled with the "fabulous" encoders is just a system to avoid the usage of bulk filament and there's absolutely nothing to improve the print quality.

To see that's not a hotend problem you just can remove the hotend and try to pass a bulk filament with a diameter of 1.8mm, with the Cubify's recommended voltage of 1.15V, your extruder will click and you'll have a FlowFailure.

[Giovanni]

Hi Peter,

even with 1.6V I suggest you to put a sink on the driver chip to prevent lost steps.

If you still have original Cubex extruders the weak point is the PTFE tube between the sink and the hot point, I suggest you to use another hot end, I'm using a full metal CustomOne from Cubexupgrade and pretty happy with that, because it helps a lot on improving the print quality.

PLA and PVA filament are very sensitive to humidity, when not used you have to place them it into a closed bag with desiccant.

[Jetguy]

"Cubify has intentionally set a low voltage to produce "Flow Failure Errors""

 

None of that makes any sense.

Cubify is not the name of the company if that was your intention- that is the software name. 3D systems is the company and the model is CUBEX.

Why would any company intentionally set the current low to cause a fault to users?

Flow errors simply indicate a jammed hotend because the filament is not moving at the rate expected.

[Jetguy]

Fixed pinch or no fixed pinch, the argument is silly.

Cooking a motor to reach maximum rated torque is not required. If you have to push the driver to the point heatsinks are required on top of a fan already blowing right at the board it's just wrong.

[Giovanni]

I hope for you that you're trolling, if not I can't do much for you and in both cases I'll not waste more time.

"Il n'y a pas de pire sourd que celui qui ne veut pas entendre"

[Jetguy]

No, I'm trying to prove beyond a shadow of a doubt your answer is wrong/incorrect/ basically bad information provided to other users and owners of this printer.

I HAVE PROOF
You have speculation.

[Jetguy]

Again, it's such an amazingly simple piece of information I have provided
#1 I provided calculated and ACCEPTABLE current ranges for the known stock motors
#2 I have explain WHY using values outside of that is bad
#3 i have debunked your fixed pinch torque requirements theory which is beyond bogus and you have nothing to back it up.
#4 if you have to add a heatsink- you have violated rules #1-3 and run the risk of heatinng the filament drive components to soften PLA before it ever reaches the hotend creating an even worse jam.

[Peter Gregory]

Running at 2.2v (where it was from the factory) would work fine with new filament (just opened package).
I have removed as much friction from the delivery as possible (no feed tubes, spools suspended overhead directly into the feed holes).
After about two weeks of use (I would store PLA in zip-lock bags with descant bags when not printing) the filament jams would start.
I live in Tennessee and the summers are very humid.
I would notice "clumpier" extrusions. The filament would stream at an angle and would appear thin.
After running a long fill (continuous extruding) the suck / prime would not start well, leaving gaps in prints.
Perimeters and infills would not be consistent.
Either back pressure was a problem, wear in the hot end was a problem, overheating and filament melting too far up the tube was the problem.
However, increasing the stepper torque helps with the problem.
Running at 2.6v gives noticeable torque improvement without getting the drive motor hotter than usual (I'm not sure about the motor driver chip, I haven't checked it while printing).
The suck / prime appears to work better with older filament.
Maybe I am masking the problems with a bad hot end, but it keeps me printing. :)

[Bog Ki Min]

i had also this "filament flow fail" error. but now i have solved finally because of your Arguments!

thank you everyone.

now i share my tip.

filament flow fail error have relationship with extruder filament drive gears power.

we knows there are two wheel.

and Cubex have Concave drive gear. Concave drive gear ist generally not bed choice. (see this link. http://reprap.org/wiki/Drive-gear)

but another wheel has problem. it is fixed. and we need more torque and Grip Power(Friction) of  drive gears.

for more torque, we can increase Extruder voltage (max. 2.2V).

but then for Grip Power?

there are minimum 3 solution.

1. using 1.8mm Filament.

2. Gearbox redesign (Spring structure) it is good, but expensiv.

3.  heat shrink tubing (my try)

and i think heat shrink tubing method is Affordable and simple.

you need only heat shrink tube and hairdrier.

If the size of the wheels slightly increased, grip force is stronger.