|
Post by theo1001 on Feb 9, 2016 2:16:45 GMT
So I just finished printing a larger job and had the smoothie board lock up again with the heaters on. This is the third time. First two times using USB and this last time printing from the card. I have printed 1.5 roles of PLA so far. This is a major fire hazard in my opinion and anyone running this board needs to be aware that it has a fail mode that allows the heaters to keep going. My two other printers NEVER do this. Smoothie has been called out for their safety programming in the past... It does not look fixed to me. Be aware. Please don't run your Eclips3D with Smoothie unattended. There are a number of threads in groups.google.com/forum/#!forum/3dprintertipstricksreviews that cover these problems and concerns. I choose to ignore this info when I installed my Eclips3d controller but now I have experienced the same problem. I don't have faith in the development team for Smoothie at this point. I would recommend not purchasing another Smoothie until this bug is fixed. Eclipse3D may need to rethink this controller too... Smoothie is a general purpose motion controller and not specific to 3d printing. The development team has a 'fix it yourself' response as makes sense for an open source project but with 3d printing using heaters and flammable plastic it would seem that heater safety would be a foundation concept not an add on afterthought. I guess it's time to try find a new controller for my Eclipse. Safe printing.
|
|
|
Post by ArthurWolf on Feb 9, 2016 14:11:13 GMT
Hello.
I'm Arthur from the Smoothie project.
When you say it "locks up with the heaters on", do you mean it stops printing but keeps the temperature stable, or that it stops printing and keeps the heaters full on, resulting in potential damage ?
The second one is pretty much impossible at this point ( we have a watchdog to prevent this ). Are you using the normal firmware branch, and the latest version ? If this is what is really happening, it is the first occurence in *years*, is very very worrying, and I really insist that you contact me directly at wolf.arthur@gmail.com
What slicing software are you using ? If it is Simplify3D, are you using the Simplify3D-specific firmware branch ?
The information on the 3dprintertiptricksreviews thread about Smoothie is *very largely* false. Some people there have misconceptions about how Smoothie works, and the specifics of some implementations. Please before believing what is written there, go ask for a second opinion from somebody that uses/develops Smoothie.
> « Smoothie is a general purpose motion controller and not specific to 3d printing. »
Smoothie is used mostly by 3D printer users. It's very feature rich in that specific area, all main devs of the firmware are primarily 3D printer users.
It happens that, because we are using a more powerful chip, we also decided to develop the firmware in a modular fashion, allowing it's use for laser cutting and CNC milling also. But those parts, when not enabled, do not interfere in any way with the functionning of the board. Meaning that a Smoothieboard that is configured for 3D printing is essentially exactly the same as a board specific to 3D printing. There is the added benefit that this leads to more contributions and a larger community.
> « The development team has a 'fix it yourself' response »
That is a misrepresentation of our position, which you very likely have heard in the above google group. This is absolutely not our position.
We have worked very hard to fix bugs, and will continue to do so, and we never dismiss them. Just go read the github's issues and pull requests, instead of relying on what people "say" about it. Some persons feel like their opinions on *how* specifically to improve Smoothie are not being listened to enough, get upset, and then will publicly misrepresent our position on dev.
Waiting for your answer.
|
|
|
Post by theo1001 on Feb 9, 2016 16:03:33 GMT
Hi,
I'm a computer programmer... The motion system stops and the heaters stay on, stable. This is a bug. It is very dangerous. It does not happen with my other printers.
I don't want to argue with you Arthur. I don't care WHY at this point I just want a controller that has a watchdog feature that works.
As far as I can tell this is not a Gcode bug. The same file will print fine on a second try. I don't know why it locks up, I'm working on it but when it happens the controller keeps the heaters on. They are not overheating. If I can figure our how to repeat the lockup I will report it to the smoothie team.
thanks
|
|
|
Post by ArthurWolf on Feb 9, 2016 17:25:06 GMT
Hey.
First, glad to hear the temperature stays stable, that's not as bad is if it didn't.
About « I just want a controller that has a watchdog feature that works. », there is a watchdog in Smoothie, and it definitely works. What's likely happening here is the firmware is still running fine, but for whatever reason, it thinks it's done printing when the end of the file was not actually reached.
About « It does not happen with my other printers. », I'm sure it doesn't, but this is the kind of thing no firmware that I know of is completely immune to. Most firmwares ( including Smoothie ) work for most people. I hope we can figure out what is going on here.
So, there are a few things that are not *actually* bugs that could cause this, and that it's completely impossible for the firmware to detect : * File ends prematurely on the SD card ( copy/paste or transfer problem ) * Streaming ends prematurely ( USB corruption, host problem ... )
Can you tell me what slicing and host programs you are using here ? This is very important information for me to determine what's going on.
When it "locks up", are you still able to send commands to the board ? If you have a panel, does it still answer after the "lock up" ?
|
|
|
Post by macgyver03 on Feb 10, 2016 2:11:21 GMT
This problem sounds very similar to one I had on my last printer that ended up being repetier hanging on the serial transmission. I don't know why but I fixed the problem by rolling back to an older version of repetier. I thought it was a long wait being inserted due to the behavior, but I started looking at the serial interface and found that somewhere between the driver and repetier the gcode commands were being lost. It may be something different but considering the heater isn't running away I don't think it's a controller issue (unless smoothie uses scheduling and a servo thread is hung).
|
|
|
Post by theo1001 on Feb 10, 2016 3:26:30 GMT
HI Arthur, The Gcode seems fine and will rerun a second time no problem. I use S3d. I don't think this is the S3D bug. I have locked up both with USB and printing from the memory card. The Smoothie will no longer take commands and I have to power down the printer. I don't know what is causing the lockup and I can't reproduce it.
I am looking into a possible end stop issue that was reported here and presented a similar lock up problem. Possible intermittent short caused by a washer on the end stop mount? Might cause Smoothie to stop moving? I will test a bit to see if this will reproduce the problem before I remove the washers.
|
|
|
Post by ArthurWolf on Feb 10, 2016 8:35:43 GMT
Hey. The problem you are having looks a lot like the S3D bug, except for one thing, so just to be sure, I would ask that you start using the conversion scripts ( smoothieware.org/simplify3d ) on your gcode files before using them, to see if the problem stops occuring. If it does stop, it's a new variation of the bug we've never seen before. If it doesn't, we can keep exploring some other things. Do you have "limit switches" enabled in config ? A short in the endstops would cause Smoothie to reset, which would definitely turn off the temperature control, so I don't expect that is what is happening here. Did you write your config file or did you get it from your vendor ? Would you mind pasting it to pastebin.com and giving the link here ? When the printer locked up while printing from the memory card, was the USB cable plugged in ? Cheers.
|
|
|
Post by 3DprintedLife on Feb 10, 2016 22:59:47 GMT
HI Arthur, The Gcode seems fine and will rerun a second time no problem. I use S3d. I don't think this is the S3D bug. I have locked up both with USB and printing from the memory card. The Smoothie will no longer take commands and I have to power down the printer. I don't know what is causing the lockup and I can't reproduce it. I am looking into a possible end stop issue that was reported here and presented a similar lock up problem. Possible intermittent short caused by a washer on the end stop mount? Might cause Smoothie to stop moving? I will test a bit to see if this will reproduce the problem before I remove the washers. Theo, my bet is that it's the washers on the endstop. I have noticed on occasion that this caused issues for me (usually only when I recently moved the endstop) but one other person has also reported this issue. Try removing the washers or replacing them with plastic washers and see if the problem goes away, I'm guessing it will.
|
|
|
Post by ArthurWolf on Feb 14, 2016 22:24:25 GMT
Hello.
Did you make any progress on this, and could you answer my questions ?
If there really is a bug with Smoothie here, I want to know and fix it ( despite what you might have been told ), and I can't know without your help.
Cheers.
|
|
|
Post by ArthurWolf on Feb 16, 2016 9:28:39 GMT
Hello.
I'm sorry to insist, but nobody else has reported any similar problem, so you are my only line to this possible bug.
Would you mind updating me on your progress, and answering the questions ? It'd help a lot with solving the issue, whatever it is.
Thanks.
|
|
|
Post by theo1001 on Feb 18, 2016 3:33:49 GMT
I have removed the end-stop washers and have test printed for around 15hrs with no problems. I did test the Smoothie by shorting the endstops, the smoothie appeared to reset and the heaters were turned off when it restarted. It did not leave the heaters in the ON state. Arthur, I don't really know how to debug this problem without being able to recreate it. It could be a hardware problem or something else. If it is an S3D bug is is not reproducible, the gcode files that failed once ran fine the next run. I have used S3D for a couple of years on multiple printers with no problems. The fact that Smoothie has a bug with S3D generated gcode is a Smoothie problem as I perceive it. We have to live with the software in our environment and sometimes it is not perfect but we still have to work in the eyes of the users. If you are never going to fix the S3D bug that alone is a good reason to move to a new controller. Attachments:config.txt (14.42 KB)
|
|
|
Post by ArthurWolf on Feb 18, 2016 10:52:53 GMT
Hey
> If it is an S3D bug is is not reproducible
The fact that it's not reproducible means it's not exactly what we call "the S3D bug", but it could have been something related to it. It's quite a complex issue.
> The fact that Smoothie has a bug with S3D generated gcode is a Smoothie problem as I perceive it
Actually, all other firmwares have problems with S3D generated gcode, but handle those more gracefully ( severe slow-downs on high-detail models and/or support generally. Smoothie is less graceful due to having a completely different step generation method ). If you are using S3D with Smoothie, we provide both a firmware that solves the issue completely, and on-line G-code cleaners. So the only problem left on this S3D issue, is the minor inconvenience of having to either use a different firmware build, or clean up your g-code files.
> If you are never going to fix the S3D bug
It is fixed. It has been for quite a while now. And it was *a lot of work* to fix. And this idea that we are not interested in fixing bugs is just non-sense. You just have to go look at Smoothie's github issues section to know that.
> I have removed the end-stop washers and have test printed for around 15hrs with no problems.
This could indicate this was a hardware problem, but I'd still like to know what it was exactly.
If the bug ever occurs again, would you mind, over the serial connection, send the M999 command, and see if the board becomes responsive again after that ? That would mean Smoothie is going into "abort" mode.
If that's what is happening, and Smoothie keeps the heaters on, this could be a problem with your configuration. This is why I keep asking for your configuration file. Would you mind sending it my way in some way ?
Thanks in advance.
|
|
|
Post by theo1001 on Feb 18, 2016 21:12:55 GMT
I did attach my config file to the previous message. I have been printing without the serial cable connected so I can't directly send M999.
The S3D fix is not in the main firmware yet? So all S3D users must reload a forked firmware or post process the gcode file for reliable printing with Smoothie? And the bug is never going to be fixed in the main Smoothie firmware branch?
|
|
|
Post by arthurwolf on Feb 18, 2016 23:01:30 GMT
> I did attach my config file to the previous message.
Sorry, I missed that.
I see you have the kill button enabled. Is it wired to anything ? This could be an issue with kill or another function being triggered by some physical process accidentally.
Also, could you add :
leave_heaters_on_suspend false
to your config file please ?
> I have been printing without the serial cable connected so I can't directly send M999.
Would you mind making sure you can send the command, just in case the problem happens again, so you can try this ?
> The S3D fix is not in the main firmware yet?
It can't be. This has been explained in many places now, but unfortunately it's usually at the end of very long threads.
> [...] the bug is never going to be fixed in the main Smoothie firmware branch?
The S3D fix as it exists now is not intended to be merged with the main firmware. While it does work, merging it would violate Smoothie's coding standards, reduce code quality, and possibly cause problems in the long term.
There is a more sustainable way to fix the issue, but it involves a major refactor of part of the firmware ( which was planned anyway ). While that refactor is planned, it is still a long way away ( other major tasks need to be done before it can be done ).
Let me insist that this is *not* a trivial issue : the S3D problem generates Gcode that no firmware was designed to handle, and Smoothie just happens to be the unlucky one that is least graceful at handling it ( others essentially *do not* handle it and slow down, Smoothie is more capable, goes faster, and chockes ), more because it made different choices for ways of doing things than for any reason related to code quality or good design. Part of what makes Smoothie's step generation/acceleration superior, is also what makes it less able to handle this. There is no graceful or simple solution to the issue.
> So all S3D users must reload a forked firmware or post process the gcode file for reliable printing with Smoothie?
Now that the S3D-fix firmware exists, the impact of this problem on S3D users is extremely minimal : * It only happens to few S3D users ( some of the users with extremely high detail organic models, and some of the users with dual extruders/support ) * If it happens, it requires, one time, taking a firmware.bin file, dropping it on the SD card, and resetting. There, done forever with the problem. Not *that* much trouble. ( Something like Marlin has munerous branches with different features and machine types targeted. Smoothie has one main branch, and temporarily, a separate branch for this issue )
Important to know also, is that S3D has known about this issue for well over a year now. While we made * a lot * of effort ( this was really a massive effort from the community and the dev team, see mailing lists ) to solve the issue on our side, they have completely refused to even acknowledge the issue exists on theirs ( reminder : this issue causes problems with other firmwares too, just not as bad ), and have made no effort to solve the problem ( and even censored users that complained about it ) I'm saying this because ever since the problem was communicated to them, we had been expecting them to fix it shortly ( it's relatively easy to do for them, the community came up with the correction scripts in a few hours : fixing this on their side is close to trivial, and would help many of their users, and better user experience for *all* firmwares. Also, all slicers that are not S3D actively take care of this issue. ), only to see the problem persist month after month.
Let me say this again : we *want* Smoothie to be bug-free, and we want all users, including those few S3D users that run into this issue, to be happy in their 3D printing experience. And we do *a lot* of work to make it so. Some things are just harder to do than others.
|
|
|
Post by AdultReics on Sept 23, 2020 2:51:34 GMT
Goluba <a href=https://maps.google.com.fj/url?q=https://leahgotti.fun/leah-gotti-adult-video-leah-gotti-in-ass-leverage-scene-01-fantasymassage-until-runner-leah-gotti-walks-in-masseuse-spencer-scott-is-furious-that-she-has-no-business-today-after-a-marathon-bl/>thai</a> Patricia <a href=http://whois.hostsir.com/?EntryCode=7615&domain=https://apolonialapiedra.fun/apolonia-lapiedra-adult-video-latina-hottie-apolonia-lapiedra-sopla-bfs-hung-roommate/>you mom</a> Nastasya <a href=https://images.google.as/url?q=https://petajensen.fun/peta-jensen-official-twitter-channel-%f0%9f%87%ba%f0%9f%87%b8-peta-jensen-fuck-my-tits-%f0%9f%87%ba%f0%9f%87%b8-visit-support-usa_pornstars-petajensenxxx-adultbrazil-siliconeros-hugeboobp/>creampies</a> Abigail <a href=http://www.google.com.kw/url?cad=rja&cd=2&ei=BzElVMzgEYjs&esrc=s&frm=1&q=&rct=j&sa=t&source=web&uact=8&url=https://leahgotti.fun/leah-gotti-adult-video-povd-best-fuck-with-leah-gotti/&ved=0CCYQFjAB>tanned</a> Agatha <a href=https://images.google.ws/url?q=https://clairecastel.fun/claire-castel-official-twitter-channel-bonne-soiree-mes-ptits-loups-%f0%9f%98%8a-on-se-voit-lundi-soir-pour-mon-live-a-partir-de-21h-%f0%9f%98%8a%f0%9f%91%84%f0%9f%91%84%f0%9f%91%84%f0%9f%91%84/>wife</a> Delia <a href=https://pornorasskazy.com/forum/away.php?s=https://abelladanger.fun/abella-danger-official-instagram-channel-link-in-bio%f0%9f%94%a5-brazzersofficial/>women</a> Anita <a href=http://www.whatmusic.com/info/productinfo.php?menulevel=home&productid=288&returnurl=https://sashagrey.fun/sasha-grey-official-twitter-channel-incredibly-problematic-further-taking-advantage-of-people-during-a-pandemic-%f0%9f%91%8e/>short hair</a> Agniya <a href=http://catya.co.uk/gallery.php?path=al_pulford%2F&site=https://toriblack.fun/tori-black-adult-video-pov-video-of-tori-black-cumming-on-a-fat-cock/>hotwife</a> Ava <a href=http://9oo9le.me/details.php?site=https://rebecalinares.fun/rebeca-linares-official-twitter-channel-colombia-cumple-202-anos-de-independencia-feliz-dia-a-los-colombianos-que-me-siguen/>dildo</a> Fedora <a href=http://www.easyasinternet.com/Redirect.aspx?destination=https://toriblack.fun/tori-black-reddit-channel-let-me-check-if-youve-done-your-homework/>banho</a> Lorna <a href=http://www.high-pasture-cave.org/index.php?URL=https://lelastar.fun/lela-star-reddit-channel-touching-her-pretty-pussy/>tetona</a>
|
|