Beacon Probe - cloned already?

[Paul Trautner]

Got everything working but I have a weird issue with QGL when I run the QGL its showing 0.0058mm but if I look at the z height measurements its much greater which shows up in the bedmesh when I do it. Anyone know what happening?

[cjkennedy72]

@Paul Trautner If you have the IDM version, have you performed the step to calibrate the IDM? I have one of the USB versions, wish I had seen the CAN versions. But, the sensor trips at 2mm from the bed. When you run 'IDM_CALIBRATE' it goes through the normal Z off-set calibration process and once you have your Z offset configured it makes the mesh more accurate. 

The trigger distance of 2mm is why your Z measurement is 1.9xxxxx. The difference between that number and z=2 should be pretty close to the Z_OFFSET after you run IDM_CALIBRATE. If you already ran IDM_CALIBRATE, then check the saved Z_OFFSET and see if that number plus the probed "z=" value doesn't get you close to an even "z=2".

I have not put my 2nd IDM on my Voron 2.4 yet, but I have it on the VZTrident I am working on and that was what I found to be true. As for the 0.0058, that is the standard deviation between the 4 probed points after the final set of adjustments (the line right below that) is applied.

The output looks good to me, and I would be happy with that low variance.

[Paul Trautner]

6 minutes ago, cjkennedy72 said:

@Paul Trautner If you have the IDM version, have you performed the step to calibrate the IDM? I have one of the USB versions, wish I had seen the CAN versions. But, the sensor trips at 2mm from the bed. When you run 'IDM_CALIBRATE' it goes through the normal Z off-set calibration process and once you have your Z offset configured it makes the mesh more accurate. 

The trigger distance of 2mm is why your Z measurement is 1.9xxxxx. The difference between that number and z=2 should be pretty close to the Z_OFFSET after you run IDM_CALIBRATE. If you already ran IDM_CALIBRATE, then check the saved Z_OFFSET and see if that number plus the probed "z=" value doesn't get you close to an even "z=2".

I have not put my 2nd IDM on my Voron 2.4 yet, but I have it on the VZTrident I am working on and that was what I found to be true. As for the 0.0058, that is the standard deviation between the 4 probed points after the final set of adjustments (the line right below that) is applied.

The output looks good to me, and I would be happy with that low variance.

When you run backlash estimate what type of values are you seeing? I think the harmonic drives are causing issue and the idm probe shows around 0.06-0.07mm of backlash which I confirmed with a dial gage. Not sure why there is so much backlash when I can't feel any backlash in the system. I tore the system apart to try the original belt reduction which is printing now.

 

[cjkennedy72]

The VZbot is down right now... I am changing the Z-Axis motion system from the TronXY to VZBot-330 conversion, which maintains the linear rods on Z. I am replacing it with a mod I found that uses a version of the Trident Z motion system. So, once I either get the probe mounted on my V2.4 or finish rebuilding the Z motion system on the soon-to-be VZTrident, I will let you know what my backlash value is.

With that said, I was doing some research a while back and found that some amount of backlash is inherent in these types of systems. The point is to get everything dialed in to minimize the backlash, or compensate for it in some way. Input shaping is a big part of overcoming this on the horizontal plane.

I think 5-7 100ths of a millimeter backlash is pretty nominal, based on my reading.

[cjkennedy72]

BTW, the build is looking good so far! I look forward to seeing it finished.

[Dirk]

1 hour ago, Paul Trautner said:

This truly looks amazing. Not the process of IDM scanning, but also the printer in just pure black. 

15 hours ago, Paul Trautner said:

Anyone know what happening?

Unfortunately I have no idea what is going on either. Maybe @mvdveer can add something sensible once he is back from new years festivities, since he has several beacons running on his printers.

1 hour ago, Paul Trautner said:

the harmonic drives are causing issue

I do not see why they would. As far as I understand, they are just z-drives to Klipper. So it should not interfere with your probing or bed leveling method.

Edited December 30, 2023 by Dirk

[mvdveer]

Found a good link with comprehensive instructions to installing these devices on GitHub for anyone interested.

[Dirk]

9 hours ago, mvdveer said:

instructions to installing these devices on GitHub

These are the exact documents I received in a zip file from my aliexpress seller.

Nicely translated into English

[Paul Trautner]

Switched to the original belt drive for the z axis and I have backlash numbers that seem reasonable but after I do qgl it's shows 0.007 but when I do a bedmesh there is a pretty big tilt.

[Dirk]

Hmm.. that looks like a big stripe in your bed...

Anything to do with interference from something metallic on or above the bed at that location? From the left to the right side of your bed?

I know these sensors are sensitive to metal above the sensor... But this....

[Paul Trautner]

7 hours ago, Dirk said:

Hmm.. that looks like a big stripe in your bed...

Anything to do with interference from something metallic on or above the bed at that location? From the left to the right side of your bed?

I know these sensors are sensitive to metal above the sensor... But this....

It's real measurements, crappy fystec bed. Working on trying to lapping it out

[cjkennedy72]

So.. The latest Klipper update breaks these probes. To fix mine, I was able to parse the klippy.log file, figure out the offending Python code, and was lucky that Beacon3D had already updated their code. I copied in the new lines and the probe started working again.

Due to this little episode, I have ordered an authentic Beacon-H from them. I will replace the other one eventually. But I wanted to reward the creator and maintainer since who knows how long before that gets into the IDM git repos.

[WINEDS]

On 12/23/2023 at 9:51 AM, Paul Trautner said:

Picture shows I decided to wire it to the stealthburner with a SB2040 mellow board. Anyone using the SB2040 pro board with the 2240 driver? I am getting a uart error when trying to move the extruder and thier docs only show the 2209 driver

I have the Mellow Fly 2040 as well which does not have a Can output.  Looks like you did a splice on the Can cable?  Can you show this more please?  Also isn't the 120ohm terminator supposed to be the on the last device in the chain?  Did you terminate at the IDM PCB or on the Mellow Fly 2040?

Edited January 17 by WINEDS

[Bender_Rodríguez]

On 12/22/2023 at 10:51 PM, Paul Trautner said:

Picture shows I decided to wire it to the stealthburner with a SB2040 mellow board. Anyone using the SB2040 pro board with the 2240 driver? I am getting a uart error when trying to move the extruder and thier docs only show the 2209 driver

Where do I find STL for this stealth burner front?

[Dirk]

On 1/17/2024 at 10:30 PM, cjkennedy72 said:

The latest Klipper update breaks these probes

No idea what was the update, but I just checked the IDM gitee website. They have "pushed a new commit - 2 days ago" with an update you can view. So I guess these IDM people are also keeping their products up to date

From the Cartography3D discord I know, they have had no problems.

I wish I had them installed and working so I could share, but not there yet.

[cjkennedy72]

@Dirk, I am not a software engineer, but I am a cloud consultant and platform engineer with enough Python knowledge to navigate and survive. Basically, the update changed how the MCU TRSYNC values are being passed to sub-modules like idm.py or beacon.py. In the log, it shows the offending line in the respective code. I had to replace the

for

statement in this section.

This is the section that changed... The 'for' statement had to be expanded. I didn't keep the original code, but if I remember correctly, it stopped at 'for' and didn't include the 'try' section.

 

    def home_start(self, print_time, sample_time, sample_count, rest_time,
                   triggered=True):
        if self.idm.model is None:
            raise self.idm.printer.command_error("No IDM model loaded")

        self.is_homing = True
        self.idm._apply_threshold()
        self.idm._sample_async()
        clock = self._mcu.print_time_to_clock(print_time)
        rest_ticks = self._mcu.print_time_to_clock(print_time+rest_time) - clock
        self._rest_ticks = rest_ticks
        reactor = self._mcu.get_printer().get_reactor()
        self._trigger_completion = reactor.completion()
        expire_timeout = TRSYNC_TIMEOUT
        if len(self._trsyncs) == 1:
            expire_timeout = TRSYNC_SINGLE_MCU_TIMEOUT
        for i, trsync in enumerate(self._trsyncs):
            try:
                trsync.start(print_time, self._trigger_completion, expire_timeout)
            except TypeError:
                offset = float(i) / len(self._trsyncs)
                trsync.start(
                    print_time, offset, self._trigger_completion, expire_timeout
                )
        etrsync = self._trsyncs[0]
        ffi_main, ffi_lib = chelper.get_ffi()
        ffi_lib.trdispatch_start(self._trdispatch, etrsync.REASON_HOST_REQUEST)
        self.idm.idm_home_cmd.send([
            etrsync.get_oid(),
            etrsync.REASON_ENDSTOP_HIT,
            0,
        ])
        return self._trigger_completion

[Dirk]

Python... Isn't that so.e sort of snake we here in Europe only see in the zoo?

 

The second link shows the commit difference (red removed, green added). ( So glad code in Chinese stays code) see attachment.

But they seem to have applied the same fix. 

[cjkennedy72]

I am not surprised and will even wager that they saw the fix from Beacon and copied it into their code... The commit diff was how I found it in Beacon3d's git repo. Good to see it was handled quickly by them as well... Still looking forward to the Beacon Rev-H that will be here Monday

 

Edited January 19 by cjkennedy72

[Paul Trautner]

On 1/17/2024 at 5:32 PM, WINEDS said:

I have the Mellow Fly 2040 as well which does not have a Can output.  Looks like you did a splice on the Can cable?  Can you show this more please?  Also isn't the 120ohm terminator supposed to be the on the last device in the chain?  Did you terminate at the IDM PCB or on the Mellow Fly 2040?

I spliced into the wire just before the connection for the SB2040 that should make the SB2040 the last device which has the termination resistor. Let me know if you need pictures

[Paul Trautner]

On 1/18/2024 at 1:13 PM, Bender_Rodríguez said:

Where do I find STL for this stealth burner front?

I got it off cults3d but there are a few on printable is you search light weight stealth burner

[WINEDS]

7 hours ago, Paul Trautner said:

I spliced into the wire just before the connection for the SB2040 that should make the SB2040 the last device which has the termination resistor. Let me know if you need pictures

Thanks very much!  Yes photos would be handy please.  I also am tossing up repurposing the PT100 connector as a CAN output as its unused on my SB2040 and has nothing connected to any of the connector pins (i did not order the MAX chip version).  But this would require terminating at the IDM probe end... 

[Paul Trautner]

22 hours ago, WINEDS said:

Thanks very much!  Yes photos would be handy please.  I also am tossing up repurposing the PT100 connector as a CAN output as its unused on my SB2040 and has nothing connected to any of the connector pins (i did not order the MAX chip version).  But this would require terminating at the IDM probe end... 

Here is how I did it, I added a connector on the idm scanner wires to make it easier to service the hotend because the wires run down the main body with the hotend wires 

[Paul Trautner]

Got everything back together and pretty happy with the flatness of the build plate now. I did one scan at 60C and another at 100C and the shape does not change which I can't say for my other printer. Overall I am pretty happy with the scanner and love that its operating by canbus and has a built in accelerometer that is placed closer to the nozzle. I think the beacon might be a better product but I was not willing to run extra wires to the tool head. I will start parting parts on the printer soon so I will report how the first layer comes out.

60C

 

100C

 

repeat 100C

[WINEDS]

6 hours ago, Paul Trautner said:

Here is how I did it, I added a connector on the idm scanner wires to make it easier to service the hotend because the wires run down the main body with the hotend wires 

Thank you very much Paul!  I think I'm going to attempt to repurpose the PT100 connector as a canbus out though.  If that doesn't work I'll follow your work.  Great work also on lapping the bed out.  

 

Edited January 21 by WINEDS

[WINEDS]

What type of screws to mount the IDM probe?  I was going to use button head M3 but just realised they may protrude too much?