What's new
VORON Design

Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members!

Question Huvud with Stealth Burner and CW2? (CANbus)

brendanm720

Well-known member
Hello!

Does Huvud work okay with SB and CW2, or should I plan on changing over to a different board (probably a Mellow SB2040) once I upgrade?

Ideally, I'd like to use the Neopixels (my current printer has a nozzle light and it's pretty nice) and have the ability to add a toolhead filament sensor at some point (though I'm not sure I need that anymore, since ERCF no longer appears to require one).

I've been watching Eddie's videos, and saw his series on Huvud. I dug around in discord and noticed quite a few folks were using Huvud (and loved it to little tiny bits) but none of them seemed to state which toolhead and extruder they were using with it.

Some backstory:
I joined the PIF queue thinking it would take about a year for the providers to pick me up, but I ended up getting my order in something like 7 months, which was prior to Stealthburner and CW2 being released. (This was not long after Stephan spun up his new printers and started chewing through the queue with extreme prejudice) I've got some beautiful Afterburner and CW1 parts, and plan to assemble the printer using those, and then print the new toolhead and extruder parts and upgrade after I break it in and get it tuned.

Thanks!

-Brendan
 
The Huvud is a 42x42mm board with the exact form factor of the back of a NEMA17 motor. This made for a very convenient mount onto the back of a CW1 or LGX extruder.

Stealthburner uses a different form factor motor, a round "pancake" NEMA14. The EBB36 or FLYSHT36 (guys, seriously, marketing department) are both boards designed to fit on the back of this motor. The SB2040 from Mellow is, as of 11/22, the only SB-specific form factor canbus board -- but it's got its own issues that may cause you to want to wait for another option.

If you are going StealthBurner, a Huvud (or EBB42 / FLYSHT42) is probably the worst option for integration due to form factor. An EBB36/FLYSHT36 will be the second-best and most universal if you go with a different toolhead or extruder. An SB2040 or similar will be the best integration with StealthBurner & CW2 specifically.
 
The Huvud is a 42x42mm board with the exact form factor of the back of a NEMA17 motor. This made for a very convenient mount onto the back of a CW1 or LGX extruder.

Stealthburner uses a different form factor motor, a round "pancake" NEMA14. The EBB36 or FLYSHT36 (guys, seriously, marketing department) are both boards designed to fit on the back of this motor. The SB2040 from Mellow is, as of 11/22, the only SB-specific form factor canbus board -- but it's got its own issues that may cause you to want to wait for another option.

If you are going StealthBurner, a Huvud (or EBB42 / FLYSHT42) is probably the worst option for integration due to form factor. An EBB36/FLYSHT36 will be the second-best and most universal if you go with a different toolhead or extruder. An SB2040 or similar will be the best integration with StealthBurner & CW2 specifically.
What's the problem with SB2040? On the way to install one...
 
The Huvud is a 42x42mm board with the exact form factor of the back of a NEMA17 motor. This made for a very convenient mount onto the back of a CW1 or LGX extruder.

I figured as much. It'll work fine out of the gate with the AB and CW1 I have parts for.

Stealthburner uses a different form factor motor, a round "pancake" NEMA14. The EBB36 or FLYSHT36 (guys, seriously, marketing department) are both boards designed to fit on the back of this motor. The SB2040 from Mellow is, as of 11/22, the only SB-specific form factor canbus board -- but it's got its own issues that may cause you to want to wait for another option.

I was aware that the CW2 changed over to the round NEMA14 motor, I just wasn't sure if Huvud would work at all with that SB/CW2 combo. So, to sum up -- No, it won't work cleanly and I'll have to do a bunch of farting around to make it work, which will most likely be bad. (Agreed on the names on those Mellow boards. I snickered when I saw that when they came out.)

I like the fact that the SB2040 fits so nicely, but I'm not set on using that board. In fact, I don't even have all of the parts I need yet to begin assembly, so another can board option may come out before I am able to get this going.

I must admit I'm curious as to the issue with the SB2040 also... Can you share?

If you are going StealthBurner, a Huvud (or EBB42 / FLYSHT42) is probably the worst option for integration due to form factor. An EBB36/FLYSHT36 will be the second-best and most universal if you go with a different toolhead or extruder. An SB2040 or similar will be the best integration with StealthBurner & CW2 specifically.

Thanks for the Info!
 
I figured as much. It'll work fine out of the gate with the AB and CW1 I have parts for.



I was aware that the CW2 changed over to the round NEMA14 motor, I just wasn't sure if Huvud would work at all with that SB/CW2 combo. So, to sum up -- No, it won't work cleanly and I'll have to do a bunch of farting around to make it work, which will most likely be bad. (Agreed on the names on those Mellow boards. I snickered when I saw that when they came out.)

I like the fact that the SB2040 fits so nicely, but I'm not set on using that board. In fact, I don't even have all of the parts I need yet to begin assembly, so another can board option may come out before I am able to get this going.

I must admit I'm curious as to the issue with the SB2040 also... Can you share?



Thanks for the Info!

The SB2040: Yenda has a google doc going here:

 
Hello!

Does Huvud work okay with SB and CW2, or should I plan on changing over to a different board (probably a Mellow SB2040) once I upgrade?

Ideally, I'd like to use the Neopixels (my current printer has a nozzle light and it's pretty nice) and have the ability to add a toolhead filament sensor at some point (though I'm not sure I need that anymore, since ERCF no longer appears to require one).

I've been watching Eddie's videos, and saw his series on Huvud. I dug around in discord and noticed quite a few folks were using Huvud (and loved it to little tiny bits) but none of them seemed to state which toolhead and extruder they were using with it.

Some backstory:
I joined the PIF queue thinking it would take about a year for the providers to pick me up, but I ended up getting my order in something like 7 months, which was prior to Stealthburner and CW2 being released. (This was not long after Stephan spun up his new printers and started chewing through the queue with extreme prejudice) I've got some beautiful Afterburner and CW1 parts, and plan to assemble the printer using those, and then print the new toolhead and extruder parts and upgrade after I break it in and get it tuned.

Thanks!

-Brendan
As far as uses, I am sure before much longer there will be if there isn't already a ERCF that is connected via can bus. The ercf repo does have user mods for the SB/CW2 that have filament detection. Just not filament motion or filament 3 axis width/cross sectional measuring (yet).

Choosing Can Bus for your printer right now, you are signing up for a bit of beta action, not only in inaccurate advice, directions and user mods, but also in continuous education. Give it a year and most of the guides will be a lot more accurate for all the hardware combinations, but also klipper and the macros will become more polished for the hardware too.

I certainly wouldn't tell anyone not to use it, but here is some of that bad advice I was talking about (LOL):
  1. Set your self up with canboot (Thank you Arksine) on as many can bus devices as you can for your deployment. Flashing new canbus settings / updating klipper with out disassembling things = win.
  2. Don't use cheap can bus components, stay far away from any MCP2515 chipsets aka waveshare rs485 can hats IMHO.
  3. Using a flex rated insulated and shielded twisted pair for the CAN H/L is smart.
  4. Planning ahead for different use cases of your tool head is also smart, so if the bare minimum spec for wire is xx gauge you should probably assume that a hotend and or led combination is coming down the pipe (or even as an upgrade path to ultra high flow) that will take more power than the bare minimum spec can provide so don't cut yourself short on power.
  5. Every printer set up should be different for it's own specific use case. If your stepper on your extruder is driven harder using more power, then it will get hotter and you might need to consider some active cooling for the stepper, MCU (can bus pcb) and stepper driver on the tool head MCU.
  6. There is cross over uses for can bus from the automotive for more than just printers, even quad racers and drones are implementing can bus to save weight so ironically you might find the answers you need to your specific problem in a forum other than for printers and be prepared to dig in if an issue comes up.
  7. All the software is evolving, so back up your configs, and installs just incase the "upgrade" button nukes your set up (something everyone should do anyway, but more important for bleeding edge adoption of technology).
 
As far as uses, I am sure before much longer there will be if there isn't already a ERCF that is connected via can bus. The ercf repo does have user mods for the SB/CW2 that have filament detection. Just not filament motion or filament 3 axis width/cross sectional measuring (yet).

That's *very* interesting. I knew there was a way to get ERCF to work without the filament sensor, and I knew someone was working on the SB/CW2 filament sensor. I was kind of kicking around the idea of one of the BTT filament sensors outside of the printer for motion and runout detection (since the toolhead presence sensor is mainly there for ERCF to know if the filament made it or not).

Choosing Can Bus for your printer right now, you are signing up for a bit of beta action, not only in inaccurate advice, directions and user mods, but also in continuous education. Give it a year and most of the guides will be a lot more accurate for all the hardware combinations, but also klipper and the macros will become more polished for the hardware too.

Yeah, I was thinking it would be good to reduce the amount of wiring being sent over to the toolhead. If it's CAN, I can be somewhat reasonably set up for future upgrades/enhacements/et cetera, by just swapping in a new board if the existing board doesn't do what the new hardware needs.

I certainly wouldn't tell anyone not to use it, but here is some of that bad advice I was talking about (LOL):

  • Set your self up with canboot (Thank you Arksine) on as many can bus devices as you can for your deployment. Flashing new canbus settings / updating klipper with out disassembling things = win.
I was definitely going to do that. It looks much more convenient, and I'm all for not moving the printer.

  • Don't use cheap can bus components, stay far away from any MCP2515 chipsets aka waveshare rs485 can hats IMHO.
Noted.

  • Using a flex rated insulated and shielded twisted pair for the CAN H/L is smart.
I was planning on using the same kind of PTFE wires that we call for in the stock Voron BOM in an umbillical. Do you have a favorite shielded TP cable for the CAN H/L?

  • Planning ahead for different use cases of your tool head is also smart, so if the bare minimum spec for wire is xx gauge you should probably assume that a hotend and or led combination is coming down the pipe (or even as an upgrade path to ultra high flow) that will take more power than the bare minimum spec can provide so don't cut yourself short on power.
Heh. Yeah. I tend to go overboard with power wires -- I was thinking 18 or even 16 AWG (18 should be good for 10A, and 16 should be good for 15A, if my back of the napkin math is correct). I honestly don't expect to use more than 5-6A on the toolhead, but better to have it and not need it than to need it and have to rip everything out and start over.

  • Every printer set up should be different for it's own specific use case. If your stepper on your extruder is driven harder using more power, then it will get hotter and you might need to consider some active cooling for the stepper, MCU (can bus pcb) and stepper driver on the tool head MCU.
I had wondered about that. I know some of the little pancake steppers can get pretty hot. I guess I'll cross that bridge when I get to it.

  • There is cross over uses for can bus from the automotive for more than just printers, even quad racers and drones are implementing can bus to save weight so ironically you might find the answers you need to your specific problem in a forum other than for printers and be prepared to dig in if an issue comes up.
Interesting. I hadn't thought of drones and quad racers being CAN Resources. Good to Know.

  1. All the software is evolving, so back up your configs, and installs just incase the "upgrade" button nukes your set up (something everyone should do anyway, but more important for bleeding edge adoption of technology).
Definitely. I do that even on non-experimental systems... like my Home Assistant instance and my UNRAID Server.
 
With regards to power to the toolhead I have run 6 wires to my EBB 42 of the same type that I had on the original spec umbilical. I have doubled up both 24v and ground to allow for a bit more current headroom. Data is obviously fine.

My EBB 42 seems to be running well from my Octopus Pro can bridge but I have yet printed anything in anger. Make sure that you follow the Arksine instructions carefully - particularly in respect to adding the 'hotplug' instruction in the config.txt otherwise Klipper gets sulky on the restart command.
 
The SB2040: Yenda has a google doc going here:

If I'd have read this before moving to canbus I doubt I would have bought a SB2040 (I think this doc was created after I had already bought mine :) ). Anyway, I did get one and it has worked perfectly since installing it. Maybe I've been lucky buy I've had no issues with it at all (except having to fit. a slightly smaller heatsink to clear the CW2 cover). I've also got the CW2 mod that includes a filament sensor just above the extruder gears, and this uses on of the endstop connections on the SB2040.
 
I ended up going with the EBB36, figuring on it being the most versatile for both configurations (CW1 while I print parts for CW2, and then CW2 going forward).

Thanks, Everyone for all of your input.
 
With regards to power to the toolhead I have run 6 wires to my EBB 42 of the same type that I had on the original spec umbilical. I have doubled up both 24v and ground to allow for a bit more current headroom. Data is obviously fine.

My EBB 42 seems to be running well from my Octopus Pro can bridge but I have yet printed anything in anger. Make sure that you follow the Arksine instructions carefully - particularly in respect to adding the 'hotplug' instruction in the config.txt otherwise Klipper gets sulky on the restart command.
Forgive me - I’m new - but what is the arksine instructions?? I’ve never heard of this and could use all the help I can get!
 
Forgive me - I’m new - but what is the arksine instructions?? I’ve never heard of this and could use all the help I can get!

Arksine has developed a utility and boot loader for the MCU's we use (can bus pcbs) on our tool heads. The purpose for this boot loader is so we can flash our tool head MCU's over the can bus instead of having to either pull the tool head out or fiddle with connecting a usb cable to the tool head while it's on the printer and hitting the tiny buttons on the MCU to get it into DFU mode to flash klipper on it. With Arksine's boot loader, you can flash the MCU right were it is over the can bus using your SSH connection on your RPI.

Depending on which tool head MCU you use, you might have to flash it directly the first time using STM32 programmer over USB, there are some quirks with some boards.

Here is a good primer:

 
I''m running a Huvud (original spec) with my CW1. I've got an ERCF sensor waiting to get plugged into the 3x2 debug header (I need more dupont crimps). When/if I move to SB/CW2 I will need to replace this Huvud with a SHT36 (I bought a bunch of LL Huvuds, SHT42s and SHT36s just in case the chip shortage never really resolves itself). The main reason for replacing the Huvud with a SHT36 is that the Huvud does not make the 5V available other than through the fan headers (it only supplies 3.3, and only via the debug header). So for things like the LEDs in the SB which will need 5V, and eventually the Tap optical sensor (which can't go on until after I switch to CW2) I will need 5V available.
 
Top