Prusa Mendel's Firmware update and Calibration

[Crenn]

Hi all,

Some of you might know that on Saturday, the space's Prusa Mendel had a firmware update and currently it's being tweaked to ensure good print quality and possibly a faster printing speed by Rob Gannon with the assistance of other members as well. At this time, the printer is mostly working with a few last remaining tweaks left relating to the extruder speeds and the temperature readings. I'm planning to try to get the printer behaving well before the 3D printer night tomorrow and need some opinions and advise before I make any more changes.

The attached photo is of the calibration cubes printed on Saturday; the progress can be seen from left to right, with the last cube printed at 150mm/s with some interesting results. However, there is a problem printing at this speed, mainly as it was completing the last block, the block was wobbling until I started to blow on the object which cooled the plastic down enough to stabilise it. This could be solved with a fan, but there is also another issue which relates to the filament retraction.

During printing, it was noted that between layers there would be a pause between layers causing small blobs in the test prints (and is visible in all of the objects, although it's been diminishing as I've tried to reduce the effect). The pause has been pin pointed to the retraction process which currently isn't working correctly. The first action was to increase the amount of retraction, from 2mm to 3mm, which seemed to have no effect. Later the problem was found to be related to the max speed of the extruder which was set to 5mm/s, with Slic3r wanted to retract at 35mm/s. I bumped up the retraction speed to 10mm/s (the middle cube) and then to 25mm/s (the right cube) with some degree of success. The question I have is, the problem is still present, so what should I do? The retraction length is still 3mm, one thing I can think of is to retract either 2mm or even 1mm. The other possibility is to allow the firmware to do retractions at 35mm/s (it's limited to 25mm/s presently) although it should be noted that when attempting to extrude at 800mm/min (13.3mm/s) it was grinding the filament, although it was able to extrude at 750mm/min, although is set to extrude at 300mm/min currently. Is it safe to allow a higher retraction speed even though it will grind filament at a third of that speed?

The calibration cubes will be back at the space tomorrow, I brought them home to get a good picture of them.

Kind Regards,
Crenn

[Darren Freeman]

On Sun, 2013-05-05 at 04:46 -0700, Crenn wrote:
> Some of you might know that on Saturday, the space's Prusa Mendel had
> a firmware update and currently it's being tweaked to ensure good
> print quality and possibly a faster printing speed by Rob Gannon with
> the assistance of other members as well. At this time,

I would like to thank all of you for doing this.

I have been comparing the print quality between the two, and sadly the
slow Cupcake has become my favourite while the Prusa was having print
quality issues.

I know I'm not alone in appreciating the effort you guys have put in!

Have fun,
Darren

[Michael Sullivan]

You are not going to extrude that quick with the gearbox ratio and the makerbot filament drive wheel and in fact think that the maximum speed of the space printers should be kept down due to all the inexperienced users on them. 

The reason Frankencake prints so well is because it can only print slowly

Mick

--
You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

[Rob Gannon]

I realised while driving home from the space yesterday that the temperature sensor setting in the configuration.h is wrong, it is set to 7 but should be set to 1. Without the hotend temperature being regulated correctly the extruder performance can’t be correctly measured. We need to fix that problem and then re-test the extruder performance, and workout the maximum reliable extruder speed. Using that as a benchmark we can calculate expected maximum printing speeds.

 

The blobbing on layer change issue may be resolved by increasing the retraction distance, not just speed. But looking at the photos of the models it appears that the general filament flow rate may need to be adjusted. The flow rate settings for high speed printing may need to be a little different from the slow speed settings, and of course,  only once the flow rate has been correctly calibrated can the retraction be correctly calibrated.

 

The other issue that is apparent from the photos is the cooling settings in Slic3r need attention. High speed printing on small layers will cause problems that need to be compensated for, but again anything to do with the hot end temperature needs be dealt with once we have corrected the temperature sensor settings, and adjusted the PID values.

 

Trystan, what time are you planning on getting to the space, I’d also like to get these issues sorted early?

 

Rob.

--

You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/Or66qXFfppYJ.

[Trystan -Crenn- Jones]

I finish work at 2:30PM today so should be at the space probably around 3:00PM. I can change the temperature sensor setting easily enough and rerun the PID auto-tune to get the correct values for that.

I agree that the flow rate needs to be adjusted as there is visible gaps in the external fills, whether this is a result of this particular filament, is unknown though. The increase in retraction distance I believe was part of the issue due to the slower retraction speed, but increasing it further will possibly help the issue due to the higher retraction speed. But this is something you'll need to judge as I'm not experienced enough.

What time do you expect to be able to get to the space today?

-Trystan

[Trystan -Crenn- Jones]

I'm inclined to agree that in normal prints, the speeds should be kept down to ensure reliability of the printers or at least until the new speeds can be demonstrated to be robust and reliable at.

The Frankencake prints well, but it does do it slower than the Prusa works at normally. If we could get similar printing speeds/times from both, that could be useful. But I'm not in a position to make that call, only give my opinion.

-Trystan

[Stuart Young]

There's lots of small problems with the space printers that cause bigger issues all round.

 1. The motor and hot end heat up the gearbox, and as it is all metal, this transfers heat to the hobbed filament drive. Due to the increased temperature, the filament is chewed instead of being pushed (the plastic deflects more easily, rather than staying resiliant and pushing the rest of the filament). I have measured the temp at the hobbed filament drive in the past and it was ~70 Deg C (which is also not good for the motor). Might be worth putting a thermocouple in the gap in the hobbed filament drive shaft (don't extrude while it's there) and measuring that temp after about 10 minutes of operation.

 2. The idler tension arm on the extruder reduces over time as the idler arm bends/deflects due to heat. This eventually reduces idler tension in long prints, which affects filament grip. The usual way people resolve this is by adjusting the tension mid-print, which is fine while the extruder is warm, but once it cools down again, it puts more tension on the filament, crushing it leading to a high likelihood of it being chewed. Some of the heat is due to the fact that the idler bearing presses close to the hobbed filament drive, but the most significant part of this is due to absorption of radiated heat from the heated bed and the hot end [*]. If we're going to stay with this extruder design, then it would be good to get someone to make the extruder tension arm out of aluminium (perhaps on the CNC Mill?) so it doesn't have the capacity to bend.

 3. The gearbox doesn't seem to like going fast (seems to cause the motor to skip steps due to back pressure), which means fast filament feeds (which are needed for fast print speeds) are pretty much out of the question. I personally think we should switch the extruder to a guided Wades design, with a decent filament drive cut specifically for the task (such as an Arcol Hyena).

Regarding heat:

 Put a 40mm fan on the back of the motor (like is on the Frankencake), running constantly. This will help remove heat from the motor, gearbox and the hobbed filament drive, which should improve things. Having airflow also pass over the top X carriage plate (which is aluminium) and near the extruder idler arm will reduce the heat in those areas too, reducing or removing the problems they suffer entirely. Also, it seems as though heat is flowing up the (stainless steel?) barrel above the heat block, so this may need to be cooled as well. As the barrel is smooth, it has very little surface area to dissipate heat (unlike other designs I've seen), but any amount of cooling should help. We just need to deflect the air so that it doesn't flow over the heat block.

Regarding flow speed:

 The smaller the nozzle, you need higher pressure or higher temp you need to use to extrude. The Prusa has a 0.4mm nozzle, whereas the Frankencake has a 0.5mm nozzle. As the hot end and extruder is similar (pressure), you will find that any temps you use on the Frankencake that get decent results will be lower than those you will need to get decent results on the Prusa. That said, too high a temp will lead to the plastic dripping out of the nozzle like a tap, which is not going to help with retraction issues.

[*] - This isn't as much of a problem with the Frankencake, as the heated bed under the hot end is 1/4 the size. This, combined with the fan on the back of the Frankencake's extruder motor has basically stopped this becoming an issue.

Stuart Young (aka Cefiar)

[John Bosua]

My 2 cents worth . As an experienced 3d print person, printing fast always creates bad models!! if something is important print it slowly. 

The spaces prusa ( and most others) that use the threaded rods for frames will always have movement in the frames and the faster you go the more movement hence inaccurate and bad looking models.

50mm is as fast as I would recommend on the prusa.

speed is you enemy when 3d printing if you want good looking and accurate models on a prusa.

If you look at Rob Gannon's printer design you will see that it is very rigid and will print beautifully at faster speeds.

John Bosua

--
You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/Or66qXFfppYJ.

For more options, visit https://groups.google.com/groups/opt_out.
 
 

<IMG_4710-gg.jpg>

[Rob Gannon]

Stuart,

 

I agree with what you are saying about heat and the need for a cooling fan for the extruder. However, the testing we did on Saturday showed that the extruder is capable of running at least twice as fast as we need it to for the printing speeds we were testing at before there is any issue with its ability to feed the filament. Also, during our testing the filament was being chewed well before the motor power and gearbox limits were reached, and that was without the temperature being correctly set.

 

I think that getting the firmware and slicing software configuration well-tuned and adding a fan to the extruder should be enough to get the printer working well at a range of speeds without needing to re-design the extruder. The only other change I would make is to add a cooling fan for the extruded filament as well as one for the extruder.

[Rob Gannon]

I agree that in a lot of cases high printing speeds are not appropriate, and I wouldn’t suggest that everyone tries to print everything at 150mm/s. The original Slic3r configuration set to 40mm/s is still there and we have added another optional configuration that we were using for testing at 150mm/s. I would expect that a range of speeds should be used depending on the requirements of each actual print job.

 

I often use high speeds, thick layers and low fill to print a quick rough draft of a design, in say 20 minutes while I’m tweaking it, and then re-print at a lower speed, higher fill and lower layers for my final print that then takes 4 hours for the same model. As John says “if something is important print it slowly”. If you just want to print a cat to see a 3d printer in action then maybe trying faster speeds might be acceptable.

 

The purpose of the testing we are doing is to find what the maximums are that the printer is capable of and the set those maximums in the printers firmware so that if someone tries to go beyond them the firmware will ignore the instructions that it can’t handle and only operate in a workable range, not to do all printing at high speed from now on.

 

From: connected-commu...@googlegroups.com [mailto:connected-commu...@googlegroups.com] On Behalf Of John Bosua
Sent: Monday, 6 May 2013 9:42 AM
To: connected-commu...@googlegroups.com
Subject: Re: [CCHS] Prusa Mendel's Firmware update and Calibration

 

My 2 cents worth . As an experienced 3d print person, printing fast always creates bad models!! if something is important print it slowly. 

[Michael Sullivan]

I think the max speed should not be set as fast as the printer can print, it should be as I said before a speed that is a little more failsafe when people with little knowledge are printing and  can not be watched all the time.

Beginners will struggle at high speeds and we should always be thinking the the longevity and lifespan of the space printer rather than fast printing.

People find ways to break them now!

Mick