1. Got a question or need help troubleshooting? Post to the troubleshooting forum or Search the forums!

Bed Can't Get Above 90Deg Before Error

Discussion in 'Troubleshooting' started by JeffreyB, Jun 23, 2017.

  1. Kilrah

    Kilrah Well-Known Member

    Joined:
    Apr 18, 2017
    Messages:
    498
    Likes Received:
    332
    No luck, looks like it's Octoprint deciding to disconnect. Serial log also seems to support this:

    Code:
    2017-06-30 16:05:22,442 - Send: M105
    2017-06-30 16:05:22,453 - Recv: ok T:190.6 /190.0 B:50.0 /0.0 @:76 B@:0
    2017-06-30 16:05:27,444 - Send: M105
    2017-06-30 16:05:27,459 - Recv: ok T:192.1 /190.0 B:-40.0 /0.0 @:47 B@:0
    2017-06-30 16:05:32,447 - Send: M105
    2017-06-30 16:05:32,466 - Recv: ok T:192.4 /190.0 B:-40.0 /0.0 @:39 B@:0
    2017-06-30 16:05:32,527 - Changing monitoring state from 'Paused' to 'Offline'
    2017-06-30 16:05:32,533 - Connection closed, closing down monitor
    
    No error from printer, Octoprint just decides to disconnect when bed temp is out. Haven't found where that's decided yet.
     
    #21 Kilrah, Jul 1, 2017
    Last edited: Jul 1, 2017
  2. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
  3. Kilrah

    Kilrah Well-Known Member

    Joined:
    Apr 18, 2017
    Messages:
    498
    Likes Received:
    332
    It's not an "issue", it's an intended safety feature. On most printers the temp sensors are permanently connected and having one get disconnected or feeding an invalid value is indeed a cause for concern, expecially at the low end since that will cause the temperature regulator to dump full power to try to get it to raise temp to the setpoint and cause a "thermal runaway".

    BUT on a printer like the R2 where the bed is intentionally easily removable it's a bit of a pain since it prevents doing it at some times where it would be convenient. Plus since the heater power is limited and isn't likely to casue dangerous situations anyway I'd rather live without the safety feature and with the bed removal possibility, hence why I'm looking into how to get rid of it. This has obviously not been considered by Robo.
     
  4. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    Then it is a Firmware issue, OctoPrint doesn't do this, it only reports. Marlin controls the RAMPS USB port as well as communications. Robo has their own fork of Marlin here: https://github.com/Robo3D/Marlin/tree/R2

    Robo is doing a lot of work stripping things that don't actually get compiled into the firmware depending on Configuration.h settings. They also made changes that are very specific (in my opinion totally unnecessary) for the R2/C2. There is also some code in the KIVY based interface used for the touchscreen, although I haven't looked at that source in so long that my memory might be faulty here. It is my opinion that Robo is trying very hard to make modifying the firmware as difficult as possible for casual users of Marlin, by rearranging code and stripping large chunks of code from the source.

    Not all changes are unnecessary, some were made because of the custom not 100% compatible RAMPS host controller used in the C2/R2.

    One more thing, this is just more of an annoyance than anything else. I am pretty sure this Marlin source is based on Marlin 1.1.0 RC6 but Robo calls it version 1.1.8. Marlin is currently at 1.1.3 so you could think that Robo's version is more advanced or has more features than base Marlin, when in fact it has less features. It would sure be nice if they gave it it's own separate numbering scheme or at the very least added a 4th digit to the Marlin code. If this is truly the 8th revision then something like 1.1.0.8 would make more sense, that way it would be easier to tell what version of Marlin was used as the base.

    The other funny thing is, I don't believe that Robo intends to add anything back to Marlin proper, so forking the build was really not necessary, a cloned repository might have been better. It's a GITHUB thing.
     
    #24 WheresWaldo, Jul 1, 2017
    Last edited: Jul 1, 2017
  5. Mater

    Mater New Member

    Joined:
    Dec 26, 2016
    Messages:
    11
    Likes Received:
    1
    I got a new pogo pin board yesterday from Robo and installed it. The print bed heats up normally, but now I have a Hotend not heating up properly.
     
  6. Kilrah

    Kilrah Well-Known Member

    Joined:
    Apr 18, 2017
    Messages:
    498
    Likes Received:
    332
    Well I had mentioned 2 observations that meant it could only be on the Pi side...

    Anyway I found that it's the RoboLCD plugin that triggers disconnection, and not actually related to the safety features. If someone else wants to disable it it's in \RoboLCD\lcd\pconsole.py:97, comment out the code as shown below:

    Code:
                    #disconnect if the bed reports a negative number two times in a row
                    #if float(self.temperature['bed']) <= 0:
                    #    self.t_counter -= 1
                    #    roboprinter.printer_instance._logger.info(str(self.t_counter))
                    #    if self.t_counter == 0:
                    #        roboprinter.printer_instance._logger.info("Shutting down")
                    #        roboprinter.printer_instance._printer.disconnect()
                    #        self.t_counter = 2
    
    Note that as per Robo's warning you don't want to remove the bed while powered (most likely due to the pogo pin problem) so after pausing also manually set the bed to Cooldown, and set it back to where you want after putting it back. The reason they implemented the disconnect is probably because of that too, if you remove it while powered at least when you put it back it will be unpowered.
    Also, if you leave the heating on it actually still freezes the controller board even with the change, so...

    Now I can pause, remove the bed for inspection / convenient removal of the string caused by pausing, and do things like changing filament without risking to damage the part.

     
    #26 Kilrah, Jul 2, 2017
    Last edited: Jul 2, 2017
    WheresWaldo and mark tomlinson like this.
  7. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    @Kilrah, when I experimented with the Robo flavored OctoPrint in the past, I found a lot of issues with Robotheme. So many that I removed it completely and went back to the original OctoPrint user interface. The Robotheme was based on work by Voxel8 and I don't believe Voxel8 ever updated their UI for OctoPrint 1.3.x compatibility. RoboOS works perfectly fine without the skinning of OctoPrint that Robo includes (IMHO it works better without the Robo skinning).
     
    #27 WheresWaldo, Jul 2, 2017
    Last edited: Jul 3, 2017
    spanishcop likes this.
  8. Kilrah

    Kilrah Well-Known Member

    Joined:
    Apr 18, 2017
    Messages:
    498
    Likes Received:
    332
    I actually like the RoboLCD interface, and apart ftom this it hasn't casued me any issues.
    Much less of a fan of the Robo Octoprint theme though especially since it's a bit broken, but after disabling it I still kinda prefer it to the default Octoprint theme...

    I didn't try the "standard" Octoprint touch plugin, should probably give that a try once.

    On the original issue, even with 5min delay for 2°C I will get disconnects >85°C when coming from cold. Switched to 10min now, but it seems it's hopeless to want to start print with >85°. It can be increased in steps once the print runs, but it's kinda pointless since it's for the first layers that it matters...
     
  9. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    @Kilrah I meant to say robotheme not the LCD interface, Although the LCD interface has issues also, not as many are apparent as the theme has. I have corrected my post above.
     
    Kilrah likes this.
  10. JeffreyB

    JeffreyB Member

    Joined:
    Mar 25, 2016
    Messages:
    71
    Likes Received:
    29
    Well I was finally able to get to 110 degs. I did a modified version (I left my working pogo pins intact to help with current) of Kilrah's bypassing of the pogo pins wiring and that didn't totally fix the problem, but got me closer. I then took the R2 version of the firmware and changed the "MAX_BED_POWER" to 255 which put the under-powered bed heater into bang, bang. I also changed the timeout on "WATCH_BED_TEMP_PERIOD" from 100 seconds to 200 seconds. I've never received the connection error since and the bed heats up much faster.

    I will say that Robo support is just terrible. When I first got my R2 I called with several questions. The guy said they had not been trained on them yet. OK, I got that, I was a kickstarter backer so it's early in the game for them, I'll give them that. But he told me tier 2 support would get back to me. They never did and it's been over a month. I figured out everything on my own. Then I called and told them about my warped pulley rods. They sent me to a webpage to order replacements under warranty. Cool, that was easy enough. I did that and have been patiently waiting on a shipping notice. That was on June 22nd. I go back today to check the status and it says "Your order has been cancelled", no reason why, no email notification, nothing. I'm still waiting to see why they cancelled my warranty replacement of the warped pulley rods. I'm not holding my breath... The pains of a kickstarter backer I guess, although I only saved $200, so it wasn't that great of a deal.
     
  11. Kilrah

    Kilrah Well-Known Member

    Joined:
    Apr 18, 2017
    Messages:
    498
    Likes Received:
    332
    Cool, thanks for the tip... indeed it seems that with PID control it never actually heats with 100% duty cycle for some reason.

    I just added one thing to my mod post, everyone who's done the mod should pay attention:

    IMPORTANT: Look at the last image and be sure to isolate the original connections. Otherwise if you insert the bed roughly without being careful with alignment it is possible to have one of the front magnetic holders already connected, and the matching original connection touch the top thermistor pogo pin. Since the heater switch is on the low side that would send 24V into the thermistor input and fry the Atmel on the printer control board.

    20170711_195313.jpg
     
    #31 Kilrah, Jul 11, 2017
    Last edited: Jul 11, 2017
    JeffreyB likes this.
  12. Kilrah

    Kilrah Well-Known Member

    Joined:
    Apr 18, 2017
    Messages:
    498
    Likes Received:
    332
    Regarding support I sent a mail with lots of details about all my issues with it, and I just received an order for a free pogo pin board and that's it, no text, communication, or comment about the other problems. Bleh.

    It's one of those stupid situations where it's the manufacturer's duty to fix these issues and one should insist until that happens, but on the other hand it would take so frustratingly long and you end up fixing things yourself and at your expense just because you can do so in so much less time than it takes to just get an answer from them.

    Hopefully they get better by the time more "standard" customers without the background and skills get theirs, otherwise it's going to become a nightmare. In theory even just replacing the pogo pin board would be expected to be beyond the capabilities of the kind of customer they target with their marketing and such a streamlined machine.
     
  13. Geof

    Geof Volunteer Moderator
    Staff Member

    Joined:
    Nov 9, 2015
    Messages:
    6,729
    Likes Received:
    2,330
    :D
     
  14. GATed73

    GATed73 New Member

    Joined:
    Jun 2, 2017
    Messages:
    15
    Likes Received:
    10

    Same problem just sprung up on me this morning. I've been doing only PLA prints, with the bed at about 60C.

    Actually, my message says Heater_ID: bed
     
  15. JeffreyB

    JeffreyB Member

    Joined:
    Mar 25, 2016
    Messages:
    71
    Likes Received:
    29
    These pogo pin issues are starting to pile up. I'm guessing Robo needs to get ahead of this, hopefully they're working on a redesign before they start shipping 1000's of these. Otherwise they are going to have a big mess on their hands when normal "consumers" start buying these as plug-in-play printers. Face it, for most us Kickstarters "this ain't our first rodeo". Lol...

    Jeff
     
    spanishcop, Kilrah and mark tomlinson like this.
  16. Seamus

    Seamus Member

    Joined:
    Jun 13, 2017
    Messages:
    41
    Likes Received:
    49
    According to the email I received a few days ago, they are, in fact, redesigning the pogo-pin interface and/or board.

    Unfortunately, I know this in the context that they were telling me that the printbed assembly isn't available as a unit *because* of this redesign. I had been attempting to get my hands on a second printbed assembly so I could modify one to bypass the pogo-pin adapter board with a connector harness, without making any permanent modifications to my printer and voiding the warranty (since I've already had to get warranty service on it for the pogo-pin issue, and can foresee needing to do so again in the future).

    My first failure came at a very inopportune time - I was running a print job of parts that will hopefully result in my getting a better job than the one I have now. I was able to improvise a solution that lasted exactly as long as it took to get a replacement pogo board (it literally failed again the day the replacement arrived), but I'd like to have a fallback plan to avoid finding myself without a functioning printer, should a similar situation come up again. Even a returned/refurbished printbed would work for my purposes, but I haven't heard anything from them following their initial "no".
     
    spanishcop likes this.
  17. Kilrah

    Kilrah Well-Known Member

    Joined:
    Apr 18, 2017
    Messages:
    498
    Likes Received:
    332
    Just in case, my mod above is reliable and pretty much reversible.
    Just don't forget the isolation stuff.
     
    JeffreyB likes this.
  18. Seamus

    Seamus Member

    Joined:
    Jun 13, 2017
    Messages:
    41
    Likes Received:
    49
    The direction I was thinking is also reversible, in the same sense - requiring some soldering to the pads and such - but since I don't know just where the line has been drawn as to what voids the warranty, I was trying to play it safe.

    My approach would solder to the four pads on the heater board, and connect to the latching black 0.1" header plug that normally goes into the pogo adapter board, via a set of powerpole or similar high-current connectors, trading the "connectorless" interface (not an issue for me) for reliability. Since it would use the existing connector, the modification would be non-intrusive to the printer itself, but the heater board would technically be modified by soldering the connector to the pads.

    It may just be a case of me being a bit too nit-picky about this, but considering that I've already had three support issues, involving two warranty replacements, I'd rather keep the warranty intact until things stabilize a bit more.

    EDIT to add: For what it's worth, I do like your solution, using the magnetic mounts.
     
  19. OutsourcedGuru

    OutsourcedGuru Active Member

    Joined:
    Jun 3, 2017
    Messages:
    752
    Likes Received:
    141
    Too bad one can't print a mixed-media bed with part conductor and part plastic. In a way, it might work a bit like those embedded inserts from another thread here. Only you might pause the printout of the bed mid-way, lay down the copper pathways (as cut from a laser CNC perhaps) and then finish the print over it. Then you'd use that soldering iron technique to melt in the inserts (which would attach to those pins).
     
  20. jwmueller

    jwmueller Active Member

    Joined:
    Jun 5, 2017
    Messages:
    169
    Likes Received:
    51
    My third print on this machine was the stock file on the USB for the extruder clips, and that file keeps failing with the same error message as the first post. Letting it run the box and cap now, and that had a lower temp setting and has been fine.

    On another note, I was having trouble with the back right corner of the bed staying connected to the magnet. So I think the pogo pins might have actually been part of the issue.
     

Share This Page