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!

NiteHawk 36 communication loss

Onewayne

Member
hey everyone,

I decided to give the nhk a go, mostly for the ability to run beacon without another cable but at this point it def wasnt worth it. This is a brand new build and the nhk worked as expected through all preliminary tests and configurations ie input shaper, PA, flow , max vol, all the usual stuff to get up and running. It wasnt until the first few longer prints that this problem came to light. First time around the 3 hour mark , Lost communication with MCU "nhk" yadda yadda yadda. Restart print, fail, same error around 2 hour mark, running some shorter prints of 30-45 minutes succeeded longer prints continue to fail.

Nitehawk is rev 3
All prints done with no panels, open air so its not heat, nhk temp reading high 50s,
also last few fails seem to be happening quicker into print time than earlier on.

Corrective action takin.

1, checked 24v connections, all connectors perfect, nhk is wired directly to uhp-200 24v
2. resecured toolboard cable, new zip ties and hot glue to insure zero movement.
3. Obtained 2 more usb cables, the really nice ones, not the junk it came with. Both resulted in failure along with another cable that I use to diagnose things since it always works.

It is def not the cable nor the connections at this point.

Everything running thru a new Pi 4, board is an octo pro

Not really knowing what to do I assumed maybe its a power issue, im running one uhp 200 24v for the board , toolhead and z motors plus cpap cooling
toolhead consists of a dragon ace volcano with 70watt cart and extruder motor LDO nema 14 at .7amp and 12v delta hotend fan
X and Y are powered separately on 48v supply

Last night I ran the same print with no filament turning hotend and fan to zero along with one of the new cables , it was successfully completed. Although when tried again for real it failed within 40 min.

I assume its the power but not sure what the exact cause is, Is the 70w cart to much or is the 200w supply not enough? Is the cpap making the hotend work to hard to keep temp and pulling to much thru the nhk?
I have no idea what or why so any help appreciated, my plan was to move the nhk outside the chamber to cut wiring down once i paneled and insulated but def leaning back towards dummy toolhead board and custom umbilical back to the board which i will most likely do as soon as the xol board arrives but i would like to use this nhk on my other trident and need to figure this out.

thanks for the help.
 
Last edited:
I don't think anyone ever says this but I think it's the Pi4. It can't run fast enough to keep up IMO. I just don't know if its the bus speeds or processing power.
I had the same issues and I switched to a Pi5 and have not seen it come back, not even once.
 
Possibly interference - I have the same issue with USB connected EBB36 - just waiting for the double shielded CAT8 cable to arrive - will the use 6 pairs for power and 2 for CAN data and see if it still disconnects. Shielding will be grounded.
 
Possibly interference - I have the same issue with USB connected EBB36 - just waiting for the double shielded CAT8 cable to arrive - will the use 6 pairs for power and 2 for CAN data and see if it still disconnects. Shielding will be grounded.
Possible and I did think about that although I would think more people would have similar issues and the cable supplied seems to be well made . I tossed it back in the box , for 40 bucks it’s not worth chasing and now that I’m paneled up and insulated it won’t last 10 min so it’s back to xol pcb and custom umbilical . I think it’s funny that after all the upgrades and chasing minimal wiring with canbus and usb I’ve gone full circle back to direct connections , sometimes more is more :) thanks for the reply
 
I don't think anyone ever says this but I think it's the Pi4. It can't run fast enough to keep up IMO. I just don't know if its the bus speeds or processing power.
I had the same issues and I switched to a Pi5 and have not seen it come back, not even once.
I agree and would swap it out if I was going to keep the nitehawk but it won’t last in the heat now that I’m finally insulated but I would have liked to figure it out , I did push it running full throttle at 25k accel and 50mm3/s and the failure was still random and didn’t show any clues it was speed related although maybe it’s just to much data no matter what. I may have just received a bad board or a ghost in the machine.
 
I don't know what chamber temps you are chasing but I run approx 50 to 55c with 85c on the nitehawk mcu on a nitehawk-SB with no problems. I have had no issues in 500 hours of run time.

The Nitehawk-36 first release did have issues and, im pretty sure, there was a recall due to this. If yours is not the latest rev-c release it might be worth a RMA and replacement as they spent some time rectifying the problems.

Here is the link to their announcement:


Definitely worth having a good one in the draw :)
 
I don't know what chamber temps you are chasing but I run approx 50 to 55c with 85c on the nitehawk mcu on a nitehawk-SB with no problems. I have had no issues in 500 hours of run time.

The Nitehawk-36 first release did have issues and, im pretty sure, there was a recall due to this. If yours is not the latest rev-c release it might be worth a RMA and replacement as they spent some time rectifying the problems.

Here is the link to their announcement:


Definitely worth having a good one in the draw :)
Hey Brian, ya that’s the issue with running insulated , my average is 70c so the nitehawk will hit triple digits more than likely , I can easily hit 80 and above running the bed hotter. I would love to use the nitehawk for the ease of install. I was toying with going back to canbus since my other machine runs 65-70 and it hasn’t had any issues but decided to go back to direct with the xol pcb .
 
Top