Beacon Probe - cloned already?

[Paul Trautner]

9 hours ago, WINEDS said:

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?

I used M3 button head

[mvdveer]

14 hours ago, WINEDS said:

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?

I used flathead screw (FHSC) to connect the beacon probe.

[WINEDS]

Thanks @Paul Trautner and @mvdveer for the heads up on the mounting screws.  A pictorial progress so far..

Probe as received had poor solder joints on sensor mount.  Interestingly the Aliexpress photos show the same!

So I put it in the bench vice and touched up the bad joints

I also ordered the companion CNC X carriage with the IDM probe 

I recently installed SB2040 Canbus board which has no Can output for the IDM probe.  So I added one using the unused PT100 connector (No max chip on my SB2040).  I also had to remove a small SMD capacitor on the component side (c41).

Next I had to knock up a quick and dirty part for the CNC X carriage to trigger the X axis gantry endstop

Now at the point where I can home and QGL using the IDM probe.  Also Bed Mesh but still more config file tweaking to go..  Upgraded toolhead with CNC X Carriage and IDM probe :

 

Edited January 28 by WINEDS

[Bender_Rodríguez]

Is the spacer for x endstop needed? Have you got the STL used?

[WINEDS]

On 1/29/2024 at 12:43 AM, Bender_Rodríguez said:

Is the spacer for x endstop needed? Have you got the STL used?

Yes.  Here is the one I used.  

XY_endstop_adapter3.stl

[WINEDS]

So in the process of updating Klipper I flashed my IDM probe with the IDM github firmware and now its not showing UUID.  Looks like I need to put it into DFU mode and flash using a "USB Cable terminated to work w/ Catographer"?  What is a USB Cable terminated to work w/ Catographer?

Edit :  Looks like I got lucky.  Hot plugged/unplugged 4 pin IDM connector multiple times and blue led starting flashing.  After that UUID was found.  So flashed with latest Can and Klipper firmware from IDM gitee site.

 

Edited January 31 by WINEDS

[Bender_Rodríguez]

Nice man, glad it's working for you! I've just had the right crimp connectors for the new beacon revision delivered so hopefully I can start on my new cable and get printing

dont know why but I've got a generic USB version and can version idm probe

[WINEDS]

Thanks but it's not all good.  Gitee firmware seems unstable and shows high idm probe CPU usage which makes Klipper crash during QGL or bed mesh.  Not sure what firmware it came loaded with but didn't have these issues with older Klipper version.  

[mvdveer]

8 minutes ago, WINEDS said:

Thanks but it's not all good.  Gitee firmware seems unstable and shows high idm probe CPU usage which makes Klipper crash during QGL or bed mesh.  Not sure what firmware it came loaded with but didn't have these issues with older Klipper version.  

Have you applied the fix that was posted previously. Sorry I can't help as I only have the beacon probes.

[WINEDS]

39 minutes ago, mvdveer said:

Have you applied the fix that was posted previously. Sorry I can't help as I only have the beacon probes.

Um which one was that?  Do you mean for bed_mesh.py?

Edit :  ok from @cjkennedy72. ?   I think I went down that rabbit hole today but I'll check his post.  Thanks.

Edited January 31 by WINEDS

[cjkennedy72]

@WINEDS The fix is incorporated into the currently available idm.py, older versions could be problematic. There was a Klipper update that changed the info sent to the probe when queried or right before a probe operation, Beacon3d fixed their code, and the IDM clones followed soon after.

The CPU usage on the IDM probe seems to sit around 40% as I have observed. that was with the shipped firmware and the latest as well. I just updated the one I still have a couple days ago. It is using an STM32F042G chip, so I am a little surprised by the load. The authentic Beacon3d probe I have sits near zero most of the time. But I am unsure what chip it has. I will look later.

I hope you guys have better luck... I bought two of the clones and as of this past weekend, one of the two has suffered a hardware fault. I am now getting an error about the coil frequency being too high, which I found in the code for both the clone and beacon, but I cannot tell what the real issue is. I suspect poor solder-joints between the board and coil, but chose to order another authentic Beacon to replace it... which happened to show up about an hour ago.

[Dirk]

Since the start of this thread, I have purchased an IDM and a Cartographer3D probe. I have also been hanging around on the Cartographer3D discourse, had a few chats with the guy thats behind it (A British 'lecturer' living and lecturing in ShenZhen). 

The way these kinds of sensors work (eddy something) are all the same. However the IDM and Cartographer ones, use the generic STM chip. You know, the one from our MCU's that drive our printers (and on all these CAN tool/head boards). Same family. So also compatible with CAN and USB and whatever other peripherals are supported for these chips.

The single difference with the Canadian Beacon probe, is that beacon uses a different chip. NOT STM. That is why it does NOT support CAN (as I understood). The reason Beacon uses this other chip, is the ability of these chips to work well with higher temperatures.

Cartographer sells chips, that can both be used with CAN or USB. The only problem is a USB cable that is certified for folding/bending in a drag-chain.

The STM chips, like on our MCU's (and CAN boards) are certified up to somewhere 70 degrees centigrade. 

As you have installed and know, when you print ABS, most beds are at 100-110 degrees and the first layer is very close to the  eddy probe. And you also do not use any ventilation, because you do not want to fuck up the first layer of ABS. So these Eddy probes heat up to 70-80-90 degrees. Luckily the chip sits a few millimeters more to the back and a bit higher, but I think they get as warm. So this is maybe a problem with the IDM/Carto STM chips Eddy probes. But I have seen several examples of prints where people have a temperature around 70-80 degrees measured at their cartographers and afterwards the sensor works perfectly well still...

When Klipper updated and changed something structural in how it communicates with all these peripheral boards (apparently they not only changed the CAN communication, but also the communication over USB), all kinds of peripheral accesoires broke with the update.

This is a dumb method of upgrading from Klipper. They apparently do not keep the backwards compatibility. If it had been commercial, I can assure you, they would go out of business with a few more practices like this.

Anyway, what I can say, is that IDM apparently updated their software (No idea about how they did with their other code). There were just 2 lines of changed code. As you can see in an earlier post I placed.

Same happened on the Cartographer discord. The guy that maintains everything updated the source code on their repository within a few hours. They also provide (like the chinese IDM) compiled binaries from the repository. (I guess for people that need precompiled binaries).

The change in the code was really very small. So I doubt that that causes problems for overheating of excess use of your CPU. But anyway. If you do not trust the Chinese IDM, (do know that the rest of your printer is Chinese, including the Beacon Probe if you do use it), you should buy a Beacon Probe. 

But hey... I am no expert. Just writing down what I observe. I hope it helps someone.

 

 

 

 

 

 

 

[Paul Trautner]

Been doing more testing and have found 1 negative while printing Nylon at 290C. I found the nozzle heats up the probe and causes a shift in the measurement which causes issues when I print back to back and the probe does not cool off before quad gantry level. I am printing polymaker so the bed is at 30C so the heat can only come from the extruder. 

[WINEDS]

On 1/31/2024 at 10:33 PM, cjkennedy72 said:

@WINEDS The fix is incorporated into the currently available idm.py, older versions could be problematic. There was a Klipper update that changed the info sent to the probe when queried or right before a probe operation, Beacon3d fixed their code, and the IDM clones followed soon after.

The CPU usage on the IDM probe seems to sit around 40% as I have observed. that was with the shipped firmware and the latest as well. I just updated the one I still have a couple days ago. It is using an STM32F042G chip, so I am a little surprised by the load. The authentic Beacon3d probe I have sits near zero most of the time. But I am unsure what chip it has. I will look later.

I hope you guys have better luck... I bought two of the clones and as of this past weekend, one of the two has suffered a hardware fault. I am now getting an error about the coil frequency being too high, which I found in the code for both the clone and beacon, but I cannot tell what the real issue is. I suspect poor solder-joints between the board and coil, but chose to order another authentic Beacon to replace it... which happened to show up about an hour ago.

Much better day today.  Spent all morning trying to get the IDM probe to respond to firmware updates.  UUID disappeared probably due to flashing firmware from the translated IDM github.  Tried hot plugging/unplugging til the cows came home but nothing I could do to get it into bootloader mode.  I then made a USB cable and tried to get it into DFU without success.  So I ordered a STLink from Amazon (should, come tomorrow).  Nothing much to do but keep fiddling.  Stopped Klipper using systemctl.  And tried hot plugging/unplugging again and I finally got a blinking blue led to indicate bootloader mode.  Maybe stopping kilpper was the secret sauce? 

At this point I decided to try switching to Cartographer firmware and software.  Flashed the latest FW for the V3 and it seems to be working.  Calibrated and started a print and Bed Mesh crashing klipper was back (I had it yesterday too).  Apparently this is related to a fix required for Bed_mesh.py.  As my RPI klipper was dirty I removed it and recloned from github.  This solved the Bed Mesh crash.  I still have had one instance where QGL failed with stupid high values for one corner but it been fine since then.  So onward and upward we go.  

I must say I am missing the setup I had with Euclid and AutoZ.  Now it back to tuning first later before every print.  @Paul Trautner apparently there is a feature called "models".  Maybe that will help. 

@cjkennedy72 Yes the fix seems to be in cartographer.py so looks like I am good to go in that respect.  My CPU load seems to be mostly 0.43 now although occasionally goes to zero for a while (not sure when though).  It would be good to know what CPU Beacon are using.  

Sorry to hear about you clone probe woes.  If you haven't been there yet the Cartographer discord is very active and Richard seems to be there a lot.  Might be worth pinging him about your coil frequency issue.

On 1/31/2024 at 10:54 PM, Dirk said:

Since the start of this thread, I have purchased an IDM and a Cartographer3D probe. I have also been hanging around on the Cartographer3D discourse, had a few chats with the guy thats behind it (A British 'lecturer' living and lecturing in ShenZhen). 

The way these kinds of sensors work (eddy something) are all the same. However the IDM and Cartographer ones, use the generic STM chip. You know, the one from our MCU's that drive our printers (and on all these CAN tool/head boards). Same family. So also compatible with CAN and USB and whatever other peripherals are supported for these chips.

The single difference with the Canadian Beacon probe, is that beacon uses a different chip. NOT STM. That is why it does NOT support CAN (as I understood). The reason Beacon uses this other chip, is the ability of these chips to work well with higher temperatures.

Cartographer sells chips, that can both be used with CAN or USB. The only problem is a USB cable that is certified for folding/bending in a drag-chain.

The STM chips, like on our MCU's (and CAN boards) are certified up to somewhere 70 degrees centigrade. 

As you have installed and know, when you print ABS, most beds are at 100-110 degrees and the first layer is very close to the  eddy probe. And you also do not use any ventilation, because you do not want to fuck up the first layer of ABS. So these Eddy probes heat up to 70-80-90 degrees. Luckily the chip sits a few millimeters more to the back and a bit higher, but I think they get as warm. So this is maybe a problem with the IDM/Carto STM chips Eddy probes. But I have seen several examples of prints where people have a temperature around 70-80 degrees measured at their cartographers and afterwards the sensor works perfectly well still...

When Klipper updated and changed something structural in how it communicates with all these peripheral boards (apparently they not only changed the CAN communication, but also the communication over USB), all kinds of peripheral accesoires broke with the update.

This is a dumb method of upgrading from Klipper. They apparently do not keep the backwards compatibility. If it had been commercial, I can assure you, they would go out of business with a few more practices like this.

Anyway, what I can say, is that IDM apparently updated their software (No idea about how they did with their other code). There were just 2 lines of changed code. As you can see in an earlier post I placed.

Same happened on the Cartographer discord. The guy that maintains everything updated the source code on their repository within a few hours. They also provide (like the chinese IDM) compiled binaries from the repository. (I guess for people that need precompiled binaries).

The change in the code was really very small. So I doubt that that causes problems for overheating of excess use of your CPU. But anyway. If you do not trust the Chinese IDM, (do know that the rest of your printer is Chinese, including the Beacon Probe if you do use it), you should buy a Beacon Probe. 

But hey... I am no expert. Just writing down what I observe. I hope it helps someone.

@Dirk Valid point about the robustness of the STM32 on the IDM probe.  Is there a way to monitor its temperature in Mainsail (not Coli temperature)?  I am going to try some longer ABS prints later on to do a bit of stress test.  I hope Beacon comes out with a Canbus version.  I imagine they could do so by adding a Canbus transceiver as per the IDM/Cartographer. 

I think your comments about Klipper updates are valid.  Not sure what their testing regime is like for major updates like we have just encountered.  I am going to try and preserve for a while with the clone probe.  I do want a Canbus probe though (reluctant to run USB to the Toolhead) so hopefully Beacon will release one soon.

 

 

   

Edited February 14 by WINEDS

[netzwerg]

[WINEDS]

What CNC X carriage is that?

[Paul Trautner]

So I changed over my build to the Chaotic Lab CNC tap and I did a 50X50 Mesh so I could compare the results to the idm scanner. The results are almost identical but I was able to do the 50X50 mesh with the scanner in less than 5 minutes and the tap probe took 1.5 hours.

 

[Dirk]

3 hours ago, Paul Trautner said:

The results are almost identical

Thank you for sharing this information. It is very valuable.

[Bender_Rodríguez]

On 2/3/2024 at 9:53 PM, WINEDS said:

What CNC X carriage is that?

https://a.aliexpress.com/_ExjkWJR

[netzwerg]

12 hours ago, Bender_Rodríguez said:

https://a.aliexpress.com/_ExjkWJR

No, this one: Super Leichter Wagen für Voron Drucker - Etsy.de

[Dirk]

I am wondering why there is a need to buy a cnc carriage that is as expensive as the probe itself. 

I have printed this  mod, which replaces the Omron probe in the Stealthburner (and Mini Stealth) and perfectly fits these probes.

If you want to per se buy a cnc carriage, you should do a combi deal, like the one with cartographer3d. 

 

[netzwerg]

1 hour ago, Dirk said:

I am wondering why there is a need to buy a cnc carriage that is as expensive as the probe itself.

There is no need, it's just hobby.

[H4kunaMatata]

I wouldn't say clone but alternative and it seems to be working just as good.

I only know about the cartographer probe, its first version was running RP2040 witch doesn't do well with heat (60c), the current version is running STM32 which deals with heat better(85c).

Also, RP2040 does CAN via software while STM32 does CAN via hardware. Hardware level is always better than software.

I've used Beacon before and I'm joining the Cartographer club. This is the only Beacon alternative probe that I'm aware of without costing a kidney.

Edited February 12 by H4kunaMatata

[mvdveer]

15 hours ago, H4kunaMatata said:

running RP2040 which doesn't do well with heat (60c),

That worries me, if it is so,  with the LDO_Nighthawk-SB board - a RP2040 chip.  Although the spec sheets for the RP2040 rates it at -40C to +85C. assume they would have tested it at these temperature extremes.

[Dirk]

@mvdveer, the rp2040 chip was in the first version of cartographer. Cartographer is now at version 3, with an STM chip. No idea why @H4kunaMatatamentions it, since it's not relevant.

However, the stm chip c11an not withstand much higher temperatures. It is also around 85 degrees max.

The location of the chip is a bigger problem. The Eddy probes are close to the bed ( heated 110 degrees for pla) and directly behind the nozzle (200-300 degrees) see also experience of @Paul Trautner above.

the beacon chips you have can handle up to 120 degrees.

The sb2040 is also the chip used in BTT 2part can board as well. And since these boards are further from the nozzle/bed suffer less from this problem.

I remember a post you made on one of your v0s, 70 degree chamber temps with an EBB36 board, you installed a fan for it. Also not an option at nozzle level, but possible at the toolhead (sbb2040 and nite hawk)