Voron2 Upgrade project

[bdejong11129]

I built my second Voron V2.4 350mm a few years back and recently my son and I finished a Cosplay ODST suit for him and now its time to do some upgrades that I have been planning for some time.  I am going to do all of them to VoronOne first and keep VoronTwo for printing the parts I need and the parts I break. lol.  

The list of upgrades:

1. Convert to CAN Bus to eliminate the multiple wires and tracks. 
   1. BTT EBB SB2209
   2. BTT U2C
2. Build the Armored Turtle AFC
   1. Formbot Kit
   2. Filamatrix modification for filament cutting.
3. Maybe convert to Tap, still on the fence on this one and am hoping for some good conversation on this topic here. 
4. Add beacon for bed mesh. I have lots of questions about this.
5. Replace all belts.
6. Re-Print some of the components since they were printed on an old Prusa without an enclosure and I know I can do better now, plus I am tired of the Red. 
7. Some kind of filament runout sensor.  I have been burning through some almost empty spools but get the timing for a change wrong and waste time and material.  Need the printer to stop for me. lol

First is to do the CAN bus upgrade.  I am going to print a new Stealth burner and just swap it out.  It will contain the filament cutter that I will need for the AFC upgrade. I have several of the parts already printed but am debating Including the dual filament sensors for the AFC.  

This leads me to the first part that I am looking for input on.  I want to maintain the stealth burner so I have on hand the BTT U2C board, and the BTT-SB2209 kit for the CAN conversion.   Are there any good files for routing the Umbilical to the rear of the printer and down to the bottom electronics and the filament tube to the rear?  I don't want to route them through the top since I intend to add the AFC later on.  

Also, any good tips, tricks, links to get can up and running would be appreciated.  I am running Mainsail and Klipper screen on this printer. It also has an Euclid probe on it and am wondering if tap is a good direction to go as well.  I also see various workarounds for the X and Y end stops when going CAN Bus. Lots of questions and options for sure.  

Edited August 8, 2025 by bdejong11129

[bdejong11129]

So the questions will begin now.  
 

I am reading a CAN bus UPDATE tutorial here and in the beginning it is recommending that I change my Pi from the 64bit to the 32bit OS for Linux.  Have others found this to be necessary?

My current Pi version is 6.1.21-v8+  and is aarch64

[TitusADuxass]

3 hours ago, bdejong11129 said:

So the questions will begin now.  
 

I am reading a CAN bus UPDATE tutorial here and in the beginning it is recommending that I change my Pi from the 64bit to the 32bit OS for Linux.  Have others found this to be necessary?

My current Pi version is 6.1.21-v8+  and is aarch64

Mine are or were all 64bit.

[mvdveer]

9 hours ago, bdejong11129 said:

My current Pi version is 6.1.21-v8+  and is aarch64

Mine all running fine on 64 bit

[Penatr8tor]

On 7/28/2025 at 6:54 PM, bdejong11129 said:

Maybe convert to Tap, still on the fence on this one and am hoping for some good conversation on this topic here. 

My advice would be... Don't.

There is a better solution so, don't F around with TAP just get a Beacon Rev-H. I mean if you really want TAP that bad I have 3 CNC TAPs laying around and you can have them all if you want. I don't want them. I'm sure there's at least 3-4 people here that have some old TAPs laying around as well.

Psssst, hey! Come over here, let me show you how long it takes for my printer to probe 32,000 points.

 

[TitusADuxass]

16 hours ago, Penatr8tor said:

I'm sure there's at least 3-4 people here that have some old TAPs laying around as well.

 

Yup.

[bdejong11129]

Ok, so Tap is off for now.  The next question is about beacon and the other eddy current probes.  Somewhere I read that they may not be CAN compatible? Or at least require another cable to be run with the CAN cable. Any validity to this?

[Penatr8tor]

3 hours ago, bdejong11129 said:

Ok, so Tap is off for now.  The next question is about beacon and the other eddy current probes.  Somewhere I read that they may not be CAN compatible? Or at least require another cable to be run with the CAN cable. Any validity to this?

Here's what I know... Beacon was the first, it costs the most and has the least issues. Cartographer is the closest copy of the Beacon, it's cheaper but people have had problems in the past and have thrown their hands up and went to a Beacon. The BTT EDDY is the same formfactor as the old orange Omron inductive probe, so it fits in behind the hotend like the old probe without modification or a special bracket. There's a new EDDY Duo that has the contact feature now. Again, this device has suffered the same growing pains of the Cartographer with lots of failures.

All 3 of the devices will do the same thing or there-abouts... You'll have to decide whether you're willing to pay extra for the reliable one or get one of the cheaper ones and pay with potential failures and frustration. And I will make the disclaimer that I don't know what state the EDDY and Carto are in development is at this point but 6 months ago, it wasn't great.  They could have fixed things.

[PFarm]

Totally agree, Beacon is the best probe out there, yes, it costs more, but it WORKS! I believe @mvdveer was the first to test Cartographer and switched to Beacon on all of his printers, if I'm not mistaken.

Edited July 31, 2025 by PFarm

[bdejong11129]

Don't mind paying a few extra bucks for reliability.  Did that when I tried klicky and ended up with Euclid.  Euclid was just more stable for me but it costs.   Considering my plans for the printer and moving to CAN control using the BTT EBB SB2209 with the BTT U2C, will beacon play well with those boards and controls?  The printer also has the Octopus V1.1 controller board?

I want to be able to settle in on all the work and ensure that I have all the parts before I take it down.  I do have a second V2 that will serve as print it again if and when I break something. lol.  If all goes well the first will be the template for the second update. 

The beacon also will not work with my second voron build bed.  For that one I went with the Mandala Rose Works with the embedded magnets.  So Beacon may just be on one of them. 

Edited July 31, 2025 by bdejong11129

[mvdveer]

3 hours ago, PFarm said:

Totally agree, Beacon is the best probe out there, yes, it costs more, but it WORKS! I believe @mvdveer was the first to test Cartographer and switched to Beacon on all of his printers, if I'm not mistaken.

Correct. Had cartographer as I liked the CANbus option. But as others have said - reliability is key and thus replaced about 6 cartographer probes with beacon probes despite having an extra USB cable in the umbilicus. But this problem is also now solved with the Orbital O2S toolboard having a USB input.

On my recent mercury one , Beacon is plugged into this port and works reliably, thus only one cable again in the umbilicus. (Top = Beacon, Bottom= Motor)

.

As @Penatr8tor, I do not know the state of development of Cartographer now. What I know is, Beacon works, it is reliable and once tuned gives you a perfect first layer ALL the time.

[bdejong11129]

mvdveer, Like to see a mercury one!  My son and I are converting my old Ender 5 Pro to mercury one for him to use.  I may hit you up for tips and tricks when we start that project.  I already have the hardware, parts printed and old hardware stripped.  Just waiting to build it with him so he can learn.

[PFarm]

@mvdveer The Orbitool board does it only mount onto the Orbiter Extruder? Can it be used with Stealthburner?

 

[mvdveer]

5 hours ago, bdejong11129 said:

mvdveer, Like to see a mercury one!  My son and I are converting my old Ender 5 Pro to mercury one for him to use.  I may hit you up for tips and tricks when we start that project.  I already have the hardware, parts printed and old hardware stripped.  Just waiting to build it with him so he can learn.

Anytime! I have done very quick build diary on this forum. Converted a Ender5 Plus to a AWD Mercury One.

[mvdveer]

5 hours ago, PFarm said:

@mvdveer The Orbitool board does it only mount onto the Orbiter Extruder? Can it be used with Stealthburner?

It Is specifically designed for the orbiter extruder, but I am sure you will be able to mount it on a stealthburner. Will look into that tonight and over the weekend. I have another one on hand to test.

Just did a quick search and here you go - looks like it can be done with this mod. Also https://www.printables.com/model/404882-lightweight-orbiter2-mount-for-voron-stealthburner and  https://github.com/sneakytreesnake/StealthOrbiter

[Penatr8tor]

15 hours ago, bdejong11129 said:

For that one I went with the Mandala Rose Works with the embedded magnets.

Yeah, I remember when everyone was worried about having a super accurate build plate. The magnets on those plates are very easily removed. You can pick up a magnet sheet for a couple of bucks and slap it on top.

But what about the holes?

Here's how you do it... You pick up a $10 electric watch demagnetizer from Amazon, remove the ring magnets from the bed, demagnetize them and screw them back in. Then you apply a magnetic sheet over the now nonmagnetic bed, and you still get the benefits of the precision and thermally stable aluminum bed without the magnetic zone issue that screws up scanners. Yeah, you lose a millimeter or two in print height due to the added magnetic sheet thickness, but you gain on the probing side of things.

Not saying you have to do this, but I have a feeling that once you have a Beacon on one printer... you'll want it on the other.

Human nature... 

[bdejong11129]

Thanks for the suggestion on how to convert the build plate of the other printer.  Right now it looks like they are out of stock.  Hopefully they come back soon enough for the printing and assembly of the new hot end. 

[Penatr8tor]

14 hours ago, mvdveer said:

As @Penatr8tor, I do not know the state of development of Cartographer now. What I know is, Beacon works, it is reliable and once tuned gives you a perfect first layer ALL the time.

Yup, turns any printer into a Bambulabs.  I catch myself sending things to the printer and walking away to grab a coffee of something and think... crap. I wonder if that print failed or not. And I check and it's just printing away. Between the Beacon, Nano Polymer and a good filament dryer... I rarely if ever get a print failure.

 

[Penatr8tor]

@mvdveer The Obitool O2S looks like a really nice solution moving forward. My recent Voron upgrade with the regular O2 is working flawlessly and doesn't warrant upgrading to reduce the umbilical by eliminating one cable. But I'm considering converting my VzBot toolhead over to an Orbiter/Orbitool O2S/Filament Sensor with Rapido 2+ UHF setup.

Interesting note on the Orbiter extruder with filament sensor and Orbitool board... I decided to try and print some TPU the other day. So, I press the button on top of the filament sensor, and it heats up and ejects the PLA I had in there. Then I pushed the TPU up thru the bowden tube and thru the sensor and errrrrp the extruder grabs the end, pulls it in a little, heats up the nozzle, purges and loads the TPU like it was PLA. No fuss, no drama. And it printed perfectly. The one thing I thought that was just extra fluff and a nice to have but I don't really need it because I watch my prints etc., component. Turns out to be my most liked and appreciated feature. I'm curious if this sensor setup would work in place of a filament cutter on a multi-filament setup versus having a filament cutter in the toolhead.

Thoughts?

[TitusADuxass]

@Penatr8tor I'm already running an orbiter 2 with a Rapido 2+ UHF on my vz330, I've just got to wait until the O2S comes back in stock.

[Penatr8tor]

Just now, TitusADuxass said:

@Penatr8tor I'm already running an orbiter 2 with a Rapido 2+ UHF on my vz330, I've just got to wait until the O2S comes back in stock.

Yup, I've been watching you from afar... 

Yeah, that looks to be "THE" combo.

I want to upgrade the old VZ not because it doesn't work but mainly because loading and unloading filament is such a pain. Plus, the umbilical is huge, and the heater wires are huge, and they sag and it's just not a well thought out implementation.

My thoughts on the upgrade would be...

Replace the 4, cooling fans for the 5160 drivers with new quiet ones. Currently they sound like a tortured cat when they're running.

Design mounts and implement a horizontal cable chain for the toolhead although, with the O2S and the Beacon connected to it, I'll only have one cable, so I might not need this. 

Implement new toolhead and design a better cooling duct and CPAP hose mount.

[mvdveer]

9 hours ago, Penatr8tor said:

I'm curious if this sensor setup would work in place of a filament cutter on a multi-filament setup versus having a filament cutter in the toolhead.

Nope, you still need a filament cutter. The sensor will work for sensing when the filament reaches the toolhead, but in order to get reliable filament changes, you need a filament cutter. You can do tip forming, but that is a pain and not as reliable as a cutter.

[bdejong11129]

Long update but had a few days to really hit this project and made some good progress. 

Here is the original Voron that I built wayyyyy back in the day.  I think it was like version 1.8 or something.  I printed the parts on a Prusa MK3 using my sun room as a heated enclosure.  The parts were good enough and held up well over the years.  Printed in good old red and black.  However, there are several wire repairs that were done over the years and the belts, well, lets say they are ready to retire. 

 

 

I took to ChatGPT to help me with deciding on a color scheme and then used that data in Fusion to color up the model.  This gave me an idea of where I wanted to use what colors. 

 

Once that was done I used all of the printers to get all the parts printed before tear down and assembly.  Hardest part was printing the teal parts on a teal bed, with initial layers you cant even see them until they are tall enough to cast a shadow. 

 

So, Thursday night I took the old girl down and took it apart.  I stripped it down to the frame so I could rebuild it all with the new parts.  I didn't break the main frame down since that was all good to go.  It was a larger job than expected when it came to removing all of the cable chains and the wires from them.  Its pretty surprising how many of them are eliminated when going with CAN Bus and sensorless homing. 

 

At the end of the night on Thursday the printer was tore down and I had rebuilt all the Z-Drive assemblies. 

 

I even had time to put all the skirts on as well as the Display mount. 

 

That was it for Thursday night and I need to start another post due to file size.... lol

 

 

 

[bdejong11129]

Friday....  This was a good and veryy long day.  It started as just do some gantry work and call it a day and it ended up with a fully assembled printer with Can working at @4:30 AM Saturday. lol

I started out by rebuilding the gantry and getting all the new belts installed.  This went pretty smoothly other than having to reprint one part for the XY joint that had a weird layer shift that resulted in a misalignment of the pulley.  Good thing the little Zero was ready and willing to print one off parts on the fly. 

Once that was done I built the new Stealth burner and added the necessary parts for the box turtle.  Wire mess was still visible, lol

 

I spent about an hour or so removing the unused wires from the bundle and getting on with the management of the CAN bus cable and the two A/B motor wires.  I was thinking about calling it a night but then thought, why not go and read up on programming the U2C and the tool head board.  Well, one milestone led to another, and another, and another.  I was able to get a functional printer, with all of the CAN boards up and running and sensorless homing started.  It was sensorless homing that ended the night for me as it was proving more difficult on the 2.4 than it was on the zero.  Figured it was a good idea to call it a night and hit it up in the morning. 

 

 

 

 

Edited September 6, 2025 by bdejong11129

[bdejong11129]

So, Friday night, leading into Saturday was pretty productive.  The whole CAN bus thing was easier than expected.  I mostly followed the instructions on BTT git hub with a little help from Esotericals site.  Only real issue I had was trying to follow all the If this then that and then is this do something different.  I figured it all out and it was up and running early Saturday morning.  I adjusted the sample config file and merged it with my printer.cfg fairly easy and got euclid and auto-z working as expected.  

One thing that was still an issue was the whole sensorless homing thing.  I had odd behaviors where individually they would repeatedly home perfect but refused to work when I asked to home all.  I ended up using this macro with some tweaks for my setup that seemed to resolve all the issues.  I do have to use a fairly long delay but I really don't care since most of what's printed on this big printer are hours long so a few additional seconds to ensure a reliable home sequence is alright by me.  Most of what I changed had to do with the probe since its really only for QGL but it assumes you use it for Z as well. 

Homing macro

With sensorless homing now working and reliable I went to testing the bed, fans, lights, etc.  I redid my hot end PID just for safety measures and also found a bad crimp on the extruder fan.  I don't like dupont and I really don't like the tiny super micro fit stuff on the tool head board.  I do like not having any cable chains on the X and Y axis. 

 

Most of today was fine tuning the homing, Auto-Z, QGL and print settings that have changed.  What I need to finish up is configuring the NeoPixels in the tool head, the filament cutting, filament sensor, Can Bus cable management, new door hinges, Z-Drive covers and the bottom plastic supports.  Then I will tidy up the wiring.  So at least a few days to a week now that I am tired and ready to call it a day. 

So far I am pleased with how it has turned out.  Figure that once I get the remaining details done and the enclosure back on it will be time to get it a serial number.