Eddie Current Probe Comparison - a PERSONAL viewpoint

[Nurgelrot]

The new Mellow Fly SB2040V3 up on AlliExpress. That was discussed above. ~35-40 USD in the config with the scanning z probe.

Edited April 4 by Nurgelrot

[mvdveer]

13 hours ago, Nurgelrot said:

The new Mellow Fly SB2040V3 up on AlliExpress.

The SET3 is already sold out (One with 2209 drivers and probe)

[Dirk]

12 hours ago, mvdveer said:

already sold out

I think that's just for Australia  

According to my AliExpress all 10 they had yesterday are still there.

I am still doubting cause they do not have the sweeping movement while doing a bed mesh which I like a lot.

And I have some bad experience with getting this when they are barely out...

 

 

[Nurgelrot]

Its the same chipset as utilized by the Duet3d Scanning Z probe (in fact you can plug the duet coil board into this tool board. It does the sweeping movment there in RRF. So at least for RRF it works as you'd expect. No idea why Mellow did that video of it probing points like a normal zprobe. sorta missed the point...

 

[Dirk]

The only reason I bought my my X5 4.4i back in 2009, even though it was financial crisis in Europe, because of the great sound of the V8.

Such things matter to men, you know...

The sweeping sound must have the same effect. 

The LDC chip is the Eddy current chip which is on Beacon, Cartographer/IDM. So I guess that must be the Eddy current chip.

The way things are implemented, are in the firmware I guess.

I will ask the lady from the helpdesk and if she can confirm it sweeps I'll get one

[claudermilk]

4 hours ago, Dirk said:

The only reason I bought my my X5 4.4i back in 2009, even though it was financial crisis in Europe, because of the great sound of the V8.

Such things matter to men, you know...

 Truth. That's why I bought a Mustang GT and not the more logical, efficient V6 model. Of course the fact that both my wife and I were giggling like 5 year olds that had just set off M80s after the test drives might have had something to do with it. I know I would have been saying "I coulda had a V8" if I didn't. 

[mvdveer]

11 hours ago, Dirk said:

the great sound of the V8.

6 hours ago, claudermilk said:

Mustang GT and not the more logical, efficient V6 model.

Unfortunately the Porsche 911 does not come in a V8, but I'm happy 

[claudermilk]

I'm not jealous. Really.  The boxer 6 is every bit as satisfying as a V8.

[Dirk]

As promised, my findings about the sound of the engine...

The lady from Mellow said that Klipper does not support sweeping so... It was a translation from Chinese to English, so I thought that some information was lost. But I was wrong.

Today Big Tree Tech launched their Eddy Current sensor. The Eddy. I found it here. It is not shaped as the Beacon or the Cartographer. Not even like the Mellow one. It looks like the Omron inductive probe. And meant to replace it.

When I dug deeper I found out where the sound of the engine comes from: Klipper has apparently a built in module for Eddy current sensors. And both Mellow and BTT use it. While Beacon has its own code, as well as Cartographer.

So the sweeping comes from the code of Beacon and Cartographer. And the points analyzing from Klipper.

Hence the expression of BTT in their marketing: 20 sec high speed leveling, instead of the 100 points per 2 seconds of Beacon  

 

[ChicagoKeri]

On 4/4/2024 at 6:38 AM, Nurgelrot said:

The new Mellow Fly SB2040V3 up on AlliExpress. That was discussed above. ~35-40 USD in the config with the scanning z probe.

 

Confirmed arrival of "Set 6" in Chicago already.  This is the one with the scanning coil and the RepRap Firmware support.

The part arrived before any support is available but that's OK...

[hypercore]

The new mellow sht36v3 does have a CAN output on the board itself, so there should be no more need for splicing. Mellow support also said that it will depend on klipper devs or a third party to add support for the probe.

[Nurgelrot]

12 hours ago, ChicagoKeri said:

Confirmed arrival of "Set 6" in Chicago already.  This is the one with the scanning coil and the RepRap Firmware support.

The part arrived before any support is available but that's OK...

yeah The Teamgloomy Devs in England don't have boards to test with yet. (we all know Mellow wont make docs anytime soon) I've got a Trident working today no real issues. The new SB daughterboard took a bit to figure out since the fans moved. Normal pin out for HEF/PCF is now flipped. no big deal just somthing to be aware of. Reach out if you have trouble with the the pins but is io.0=3pin middle of board io.1=2pin next to it and io.2 is the 3pin HV-endstop -the [V S G] order is reversed from the others on the HV-endstop. 

Aslo found they run a bit warm so i used one of the many 3007 fan shrouds out there for the SB and hooked it up to fan2 told it to come on if it gets over 45C Just did a 7 hour print on the 350 Trident and the board stayed ~ 50C

Edited April 15 by Nurgelrot

[chrisalbertson]

Did you all see that BTT now offers an eddy current probe?  They say with temp compensation.  There are two versions (1) I2C interface and (2) USB interface and they say CAN bus interface is "coming soon".    The price is just under $17. for the I2C version which is what most people with Voron printers would want.

With a $17 price, if this works as well as the specs claim, BTT will own is market.   

 

https://biqu.equipment/products/bigtreetech-eddy?variant=41018205405282

 

When I first read about eddy curent sensors and saw how they worked, I figured I coud make one for a few bucks.  They are not complex.   But now there is no reason to design an open source version

Edited April 20 by chrisalbertson

[mvdveer]

Just a quick update and an embarrassing story. 

This happened on the Micron 180.

As you know I commented that I had a IDM probe fail - Coil shorted/disconnected and then errors on abnormal temperature readings. (BTW - this probe's coil is destroyed)

Replaced this with a Cartographer probe - after some 20hours printing (which by the way is irrelevant in this instance) - Coil shorted/disconnected

Well onto the USB beacon - replaced the cartographer probe - re-started the same print, and not long after  - Coil shorted/disconnected -- What The......

Restarted the print and observed - failed at EXACTLY the same part of the print's first layer - in the back corner. (X165 Y 5)

The light went on - I read a user posting not so long ago (sorry cannot remember who, but THANK YOU) about having to define an exclusion zone due to nozzle brush interfering with the probe. It turns out the print failed whilst in the area of the nozzle brush. The probe, protruding to the rear of the printer was moving over the nozzle brush, and most likely touching it - that seemed to short the probe and caused the disconnection.  Let's see if this actually was it.

Back to the slicer and moved the parts forward on the plate - resliced, uploaded and........ Successful print.

So BEWARE of EDDIE current probes and NOZZLE Brushes. Made me decide to change all the brass nozzle brushes to silicone. (Add that to the To-Do-List)

 

[Dirk]

32 minutes ago, mvdveer said:

NOZZLE Brushes

Auch...

So conclusion is... the Micron is an Eddy Current probe destroyer?

Damn.. 

Actually, this was one of the things I was warned about:

 - do not use brass nozzle wipers (I removed the brass nozzle brush when doing my cable cleaning CAN move).

 - do not leave the z-endstop (SexBolt) when you switch to an eddy (same story with the nozzle brush)

A much more frequent eddy current probe destroying problem, is warping prints during printing that crash into the coil and destroy it.

What I also wanted to say now we are talking about the brush.. I did not have the need to reinstall a silicone one, because I do not have a problem with a bit of dried plastic on my nozzle anymore... With TAP it was making a difference.. but Carto does not have this problem. 

Maybe I will install a silicone wiper when Carto comes with auto-z

 

[mvdveer]

52 minutes ago, Dirk said:

What I also wanted to say now we are talking about the brush.. I did not have the need to reinstall a silicone one, because I do not have a problem with a bit of dried plastic on my nozzle anymore... With TAP it was making a difference.. but Carto does not have this problem. 

 

Very good point, the only reason for the nozzle brush was probe accuracy. Dank u well!

[chrisalbertson]

17 hours ago, Dirk said:

Actually, this was one of the things I was warned about:

 - do not use brass nozzle wipers (I removed the brass nozzle brush when doing my cable cleaning CAN move).

 - do not leave the z-endstop (SexBolt) when you switch to an eddy (same story with the nozzle brush)

The instructions that come with the BTT Eddy sensor explicitly say to use the physical contact z-stop.

If the sensor has the ability to be turned off and on, I can't see a problem with any of this.

I'm planning on using the BTT sensor on a Prusa i3 clone rebuild.   If ever there was a printer then needs bed leveling it is that printer.  The bed on bowel-shape at any temperature. I plan to toss out the old 8-bit Marlin-based system and the 12-volt  and use Klipper and 24 volts.

 

[Dirk]

17 minutes ago, chrisalbertson said:

BTT Eddy sensor

It is special in many aspects. It also does not need a metal free area above the probe.

I do not see a very good reason to keep the z endstop, when you have a sensor fully capable of detecting the bed. But I am sure there is a logical explanation.

Neither Beacon nor Cartographer work together with a z endstop. In fact, they do not require it. If Eddy requires a physical endstop, it means it can not function as a z endstop, which I find strange. 

By the way, I did not say you have to remove the z-endstop or the metal brush. Please keep it. As you saw in @mvdveers experience, it can be okay... (for up to 20 hours).

Let us know of your experience with Eddy!

[mvdveer]

4 hours ago, Dirk said:

it can be okay... (for up to 20 hours).

Just to clarify - I don't think it is the amount of hours per se, but the amount of time the probe actually moves over the brush before it gets damaged. In small prints, where the toolhead does not move close to Y0, this would not occur. Moral the story as you said - don't use a nozzle brush with Eddie current probes

[xyleth]

On 4/21/2024 at 1:29 AM, mvdveer said:

The light went on - I read a user posting not so long ago (sorry cannot remember who, but THANK YOU) about having to define an exclusion zone due to nozzle brush interfering with the probe. It turns out the print failed whilst in the area of the nozzle brush.

That was me!  Glad it was of use

[Buurman]

On 3/25/2024 at 11:13 AM, mvdveer said:

Eddie Probe Comparison:

Feature	Beacon Probe	Cartographer	IDM
Packaging	Plastic Box. USB Cable and Plugs included	Cardboard Box, Connectors and short cable included	Plastic bag, connector included no cable
Connection	USB	USB / CAN	USB /CAN
Temperature Limits	No temperature warnings	No temperature Warnings	Warnings - coil temperature exceed limits
Performance	Very Reliable and consistent (4months +)	Very reliable (3 weeks)	Inconsistent readings (2weeks, then changed)
Installation	USB from probe to Raspi	Can H and L spliced into toolhead, 5V and GND from probe input	Can H and L spliced into toolhead, 5V and GND from probe input
Cost	1.AUD$149.00 incl postage 2.AUD$219.00	AUD$55.00	AUD$64.95
Supplier	1.Beacon 3D (USA) 2.PhaserFPV (Aus)	Orvil3D (Australia)	Budi Store (Aliexpress)

·      Beacon Probe:

          

  Plastic box with connectors, probe and USB Cable

·      Cartographer:

          

  Cardboard box with connectors and pre-crimped cable

·      IDM:

            

  Plastic Bag with one connector and crimps. No cable.

Here are the probes in order: IDM, Cartographer and Beacon:

   

 Each probe has identifying marks on the coil, but the Beacon Probe has the name written on it as well.

Whilst I had the Trident 300mm in for some repairs and maintenance, I decided to try out the IDM Eddie scanner (Canbus) that I got from Aliexpress.  First impression was not good. The probe was delivered in a plastic bag, no cable, no instructions.  Went ahead and installed this on both the Micron 180 and the Trident 300. 

Had trouble with the alignment of the probe PCB mounting holes on both printers with the IDM probes. They were out of alignment by about 1mm and had to enlarge the mounting holes on both probes. The probe on the Micron was mounted to a printed piece which could have accounted for this, but the one on the Trident was mounted on a CNC carriage from Funsor.  Also test fitted the Beacon probe on the Micron and mounting holes aligned.

Installation after that, was simple – mounted the probe and crimped a cable – one end has 4 wires from a JST connector for the probe, the other a JST connector for the toolhead board and spliced the CanH and CanL into the existing can cables on the toolhead.

Had to update the firmware on the IDM probes. Very simple process, as I followed the instructions on the cartographer github page.   Calibrated the probes and set the Z-Offset as per the instructions.

The Beacon probe (Rev D) has been installed on the VZBot and has been operating for a few months now – USB from board back to Raspi. Since then Beacon released the Rev H which is larger, but flatter than the Rev D. The overall dimensions on protruding behind the printer is similar though. Just less height.

             

 The IDM probe on the Micron has been nothing but trouble. Seems to be losing the Z-Offset and routinely comes up with “temperature out of range for coil – shutting down”., which I suspect is the cause for the inconsistent Z-Offset.  The IDM probe on the Trident has given the occasional “mcu lost communication with probe” error. I put this down to a bad splice on the CanH and CanL connections.  Or maybe it is the Canbus bridge on the Octopus board? (Had this issue on another Trident – may switch to a UTOC 1 on this whilst it is out for service)

Since installing the beacon probe (months now – Rev D version) – have not had to touch it. Works flawlessly.

Since installing the IDM on the micron – ready to chuck it out. (Must qualify and say this was the cheapest probe I could get on Aliexpress.) Did this and replaced with a Cartographer probe (CanBus) - No issues as yet - works flawlessly

Since installing the IDM on the Trident – some communication issues with the probe – will need to give it a bit more time. Did that and all issues settled. Been running for 3 weeks now and no problems. Had to take the printer out of commission for some maintenance and upgrades. Think I will change to a Cartographer Probe.

Picked up some Cartographer probes from a local supplier.Very cheap indeed - AUD50 for Canbus and AUD55 for CanBus/USB.

The Cartographer probe performs very well, probably on par with the Beacon so far.  But it is early days.  Will have to see what time and the high chamber temperatures does in the long run.

What I love about the cartographer is that it is Canbus - no bulky USB running to the toolhead. Still just the 4 wires in the umbilical. 

Which is best? - you decide

Love it, thanks Buddy! I will start looking for a Cartograph (really great name.. not) and looking forward to trying it out!!! 

Great tests!!

[Sparkss]

On 4/20/2024 at 7:29 PM, mvdveer said:

Just a quick update and an embarrassing story. 

This happened on the Micron 180.

As you know I commented that I had a IDM probe fail - Coil shorted/disconnected and then errors on abnormal temperature readings. (BTW - this probe's coil is destroyed)

Replaced this with a Cartographer probe - after some 20hours printing (which by the way is irrelevant in this instance) - Coil shorted/disconnected

Well onto the USB beacon - replaced the cartographer probe - re-started the same print, and not long after  - Coil shorted/disconnected -- What The......

Restarted the print and observed - failed at EXACTLY the same part of the print's first layer - in the back corner. (X165 Y 5)

The light went on - I read a user posting not so long ago (sorry cannot remember who, but THANK YOU) about having to define an exclusion zone due to nozzle brush interfering with the probe. It turns out the print failed whilst in the area of the nozzle brush. The probe, protruding to the rear of the printer was moving over the nozzle brush, and most likely touching it - that seemed to short the probe and caused the disconnection.  Let's see if this actually was it.

Back to the slicer and moved the parts forward on the plate - resliced, uploaded and........ Successful print.

So BEWARE of EDDIE current probes and NOZZLE Brushes. Made me decide to change all the brass nozzle brushes to silicone. (Add that to the To-Do-List)

I had a similar occurrence with an IDM probe, but what would happen on a 2.4 is it would QGL and bed mesh then would fail/klipper would shutdown. Just prior to printing I would run the nozzle over the brush to give it a clean start, and after cleaning it would move forward to the bed/print start position (moving across the brush) and shorting out the probe. Thankfully I caught it on about the 3rd failure, before it eventually fried the coil. I clipped the "pointy bits" of the wire sticking through the probe (just beyond the solder, did not touch the solder) and then used 2 layers of kapton tape over the solder "nubs", that solved it for me. I also added a z-hop to the post-brush macro, just to play it safe.

I also got advice from discord to lower the brush to below the bed and drop the nozzle down to clean it, then back up again to return to the print position.

[chrisalbertson]

On 4/21/2024 at 11:18 AM, Dirk said:

It is special in many aspects. It also does not need a metal free area above the probe.

I do not see a very good reason to keep the z endstop, when you have a sensor fully capable of detecting the bed. But I am sure there is a logical explanation.

Neither Beacon nor Cartographer work together with a z endstop. In fact, they do not require it. If Eddy requires a physical endstop, it means it can not function as a z endstop, which I find strange. 

By the way, I did not say you have to remove the z-endstop or the metal brush. Please keep it. As you saw in @mvdveers experience, it can be okay... (for up to 20 hours).

Let us know of your experience with Eddy!

BTT does say to keep the physical end stop.  In fact they supply a design for adding thiseddy cureent snsor to the Vorn "Tap" sensor  But also note that BTT found a way to make this probe so that it can retail for $17.    It might be built differently from the $80 probes.  My guess is that the sensor measures relative but not absolute distance.

BTT put up a GitHub for their sensor.  The user manual has a 19-April date and has been there now for 19 hours.  I just read it.  There is also some CAD date there.

One key bit of information is they are using a ldc1612 chip for the coil sensor and an RP2040 as the MCU.   Yes, the BTT sensor looks  like a Klipper MCU

Here is the datasheet for the idc1612.  It looks like it is very straightforward to design an eddy current sensor using this chip https://www.ti.com/lit/ds/symlink/ldc1612.pdf

As I had guessed there seems to be a way to power this sensor down using one of the pins on the chip.

 

 

 

[Dirk]

1 hour ago, chrisalbertson said:

ldc1612 chip for the coil sensor and an RP2040 as the MCU

retail for $17

Thank you very much for this information.

This is the annoying and sneaky tactics of BTT to keep information around their products limited. Everyone has just heard the hype and nobody knows what is what or which by the cryptic naming, missing documentation, confusing and nothing-saying acronyms.

You are kind of comparing apples with pears.

Beacon,Cartographer as well as Mellow use the LDC1612 chip for Eddy-current sensing. 

The similarity stops there.

Like you guessed, BTT manufactures Eddy with a single layer PCB.

- And this is what you are getting for 17$ (plus in my case 5$ shipping) and a cable for i2c. This they call Eddy-COIL. Eddy coil uses i2c as a communication protocol with the main MCU to process the signals. So in COIL, there IS NO MCU.

- The same quality PCB and packaging is used for Eddy-USB, which has additionally a RP2040 MCU and a USB cable, which is not certified for motion. The BIQU.equipment price shows me 25$ +5$ 

- Eddy-CAN not there yet.

Both products are cheaply made and will most probably cause troubles with their cables (check out the price of a motion certified USB cable) and the accuracy with an inferior protocol.

2 hours ago, chrisalbertson said:

the $80 probes

Suggesting you did not have a look at the prices of the other products we have been talking about, so I will post you the link for cartographer here which costs 35$ and see above where @ChicagoKeri mentions purchasing not only an Eddy-Current probe for 35-40$, but also a CAN toolhead PCB of a decent quality. 

The Cartographer probe which I own, has a 4 layered PCB, a superior on board MCU (STM32) with an accelerator on board. It came in my case with a CAN cable, but they provide higher quality USB cable that works perfectly in an umbilical.

So when you look through the hype it is not a difficult choice.

[Buurman]

For me, the only reason to look at something else than Beacon is just the cable, if the beacon came on CAN, no brainer... 

I been using nonstop the beacon for 6 months or so, printed out many full beds with parts, its insane accurate, in combination with the KAMP meshing its just perfect.

So thats why the Cartographer (really cant get over that name) sounds interesting, reliability on can, have to test it to believe it...  (thanks @Dirk!)

The only downsound of the Beacon, if I can mention it, is that when you manually change the offset on display during tuning, etc, and if you want to commit it to the default setup, a simple "save" wont do it, you have to use a seperate beacon command to commit the offset, since your normal offset setting is no longer in use.

Z_OFFSET_APPLY_PROBE