Refurbishing my Voron 2.4 300

[claudermilk]

Steve (and Charlie) would approve. 

[Penatr8tor]

Here's the completed electronics bay. Still need to create a Y endstop contactor part and then it's power up time. Probably tonight. I also went ahead and replaced all of the PLA skirts that I had printed on my old Tevo Tornado as well as installing a longer ribbon cable for the Klipperscreen. I have an external usb and ethernet port I can connect if I want to or I might add a USB to the front for my nozzle ADXL345.

Anyways... it's looking pretty clean. 

[Penatr8tor]

Well, It's official!!! I'm part of the "Released the Magic Smoke" club. 

I fried my Octopus. It's been a good board, but its job is now over. So long OG Octopus...

So, after a bit of pondering I have come to the conclusion that it's most likely related to either the probe (TAP) connection or an endstop, although I'm only using 2 wire limit switches but, you wire up stuff and things can get flipped etc. Who really knows at this point. We will figure it out.

I have a couple choices... I can either buy another Octopus board, only $50 and is there any benefit to getting the pro over the V1.1 I just fried or, use the Leviathan board I already bought? Irrespective of my decision, I still need to find the culprit.

Below is a pic of the fried chip on my Octopus board. Does anyone know what this chip controls? That should point directly to my problem. At least that's what I'm thinking at this point. Any input is appreciated.

[Penatr8tor]

I started doing some investigating and I may have found the culprit... 

https://media4.giphy.com/media/v1.Y2lkPTc5MGI3NjExcGtsMGl1eDMzaDloc3l6aDF0NWdhOG52NXIwdjJsMGxrZGRsdnEyZiZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw/QxitttisiHU4diRRcm/giphy.gif

The chip that got fried is an HTC365A

and the schematic for the Octopus V1.1 has the legs going to the fan outputs. The six 4.7K resistors and the cap coincide with the schematic as well, so I think I'm on the right track.

I'm thinking I've got a short in one of the fans, possibly the Nevermore filter I installed. I should have a 24v wall wart squirreled away somewhere and will do a manual fan check on all the fans to see if I have a polarity flip somewhere, although it seems odd that a flipped fan will fry the board but ya never know.

 

[PFarm]

4 hours ago, Penatr8tor said:

use the Leviathan board I already bought

This a good logical solution IMO! Did you change any of the fans? I've had a few fans come in with the polarities reversed something I now always check before installing them.

Well, It's official!!! I'm part of the "Released the Magic Smoke" club. Welcome, been there!

Edited June 22, 2024 by PFarm

[Dirk]

4 hours ago, Penatr8tor said:

The chip that got fried

I am sorry that your octopus got fried. but what is 50 bucks for such a central piece of hardware that already has proven you service for a long time

I have checked the schematics on btts github, and found out that you also did it. It is indeed the chip marked as U4 and that one you have also shown, that it is a chip that apparently regulates voltage to the fans with a 5v control voltage. 

It looks like the most damage is on the top right, which could be the location of FAN2. But what you do is of course best, check em all... And it wouldn't be the first time a board got fried by a fan shortcut. 
(Sorry too late to add anything that makes sense)

 

5 hours ago, Penatr8tor said:

I can either buy another Octopus board, only $50 and is there any benefit to getting the pro over the V1.1 I just fried or, use the Leviathan board I already bought?

Knowing you, I would go for the Leviathan. Which is more expensive, so better. It is quite popular too. So people will be able to provide assistance might you not already figure it out yourself

But I would go for the Pro. The normal octopus gives a visible alert, when there is a short. Pro gives a visible and a zoomer alert. After frying 3 spiders, I also bought a Max EZ, because it cuts off power when there is a short.

The difference between pro and normal is formulated here.

 

[Penatr8tor]

I ended up buying a new Octopus V1.1 just like the old one. Thanks for the link @Dirk. There wasn't really much difference between the two and I already have an SD card with firmware ready to go. I'll manually test fan polarity as well as other plugged in endstops and so on today. 

[Dirk]

2 hours ago, Penatr8tor said:

I'll manually test fan polarity as well as other plugged in endstops and so on today. 

One thing I remembered after pinning down the chip and its function, was that someone on the electricity channel on voron discord told me (after frying my second Spider MCU), that with a short on a board, it can damage any of the other components on the board or any attached to it.

So I am wondering if it is the fan, and if there are any other components involved. For example like the Optotap from your TAP. Which is known to cause problems because it used to be 24v and was later changed to a 5v version.

2 hours ago, Penatr8tor said:

There wasn't really much difference between the two

The normal Octopus seems to be made for the VORON and klipper platform printers, while the pro also aims to run the OS from the board hence the 3 different chip versions on them. I believe Marlin does it. And another one that is mainly used for CNC-machines.

This is also why the pro version has the possibility to set every motor to a different voltage individually by jumper, which could also be used on a Voron, if you were for example doing a AWD conversion and wanted to power the x/y's with 48v and the 4-Z's with 24v.

 

[Penatr8tor]

@Dirk

I was looking right at the board when I flipped the switch and instantly saw the little micro flash. If any thing is got fried as a result... I will address those if they turn up. If it's the opto switch on the TAP, I have an extra as the Mellow TAP uses the same pcb and switch as the printed Voron TAP.

Plus, I remember when I built the Nevermore and it was backwards when I tested it I lazily didn't fix it. 

In the end it will get fixed and it will print and I'll dial it in and it will be a good printer just like the other 5 builds.

[Penatr8tor]

I did some testing last night. I created a 2-wire test lead for the fans with a male 2 pin JST on one end and the other end was attached directly to the 24v power supply in the printer. I disconnected and tested each fan lead and guess what? They all worked.

So why did powering up the Octopus with all fans connected pop the fan controller chip if the polarity and voltage for all the fans is correct?  What other plugged in components can potentially fry the board if wired incorrectly and where do I look next?

While we all ponder this anomaly... I think I might change direction on this refurb mid-stream...

Here's what I'm thinking, and I really need your advice @mvdveer on this one.

I'm thinking, dump the TAP and LDO toolhead PCB and go CAN bus and Beacon H.

What is the best CAN config for...

Octopus 1.1, Steathburner with LGX-Lite, rainbow barf and Beacon-H

or...

Leviathan, Steathburner with LGX-Lite rainbow barf and Beacon-H

or... I might pull the plug on an Orbiter to replace the LGX-Lite

Anyone's input is greatly appreciated.

[mvdveer]

4 hours ago, Penatr8tor said:

Here's what I'm thinking

First decision for me would be what Can board you will be using as the mounting options for these will influence the decision on the extruder. As you know I am a Mellow guy. For some reason have not taken to the BTT can boards (I have more than a few mind you - lying in a drawer somewhere)

The orbitor 2 mounts the SHT36 well, not the SB2040 (SB2209). But with your CAD skills that should not be a problem. I like the SB2040 for the reason of the fan mount in the toolhead. (Eliminates wires) And those tiny JST connectors are not good for my clumsy fat fingers!

I struggled to find a mount for the LGX Lite for the SB2040 or SHT type of mounts which I had installed. So currently running Galileo extruders on most of the bigger Vorons.

You cannot go wrong with an orbitor 2- but then it adds weight to the toolhead. (In my case it does not matter - I am no printman speed demon) There is also a native filament sensor for it that mounts on the orbitor at the toolhead. 

4 hours ago, Penatr8tor said:

I'm thinking, dump the TAP and LDO toolhead PCB and go CAN bus and Beacon H.

ABSOLUTELY! Especially with Beacon Contact - it is GREAT!

I still like the Octopus boards - have not really had any issues with them. Don't own a Leviathan, so cannot comment - I suppose it is a matter of choice or what you have available.

4 hours ago, Penatr8tor said:

What is the best CAN config for...

My choice would be a SHT36 type board if you are going with the Orbitor - the only consideration is the Y-Travel as the Z-Cable chain may interfere. If you using the LGX Lite and can model a mount, then the SB2040 form Mellow.

[Penatr8tor]

Thanks @mvdveer I responded to your reply on the Cartographer thread. I copied it below, no need to further confuse ourselves any more than we already are. 

 

Would this be a good choice for the toolhead CAN PCB?

Fly-SB2040-V2

and pair it with

Fly UTOC-3

Also, what do you use for cabling? Are you buying a premade cable or fabricating one yourself?

I assume that my umbilical will end up with 2 cables, the CAN and the Beacon USB.

I want to try out the Orbiter and it's pretty cost effective at ~$43 shipped from Amazon. So CAN it is, I'm going Orbiter and Mellow SHT36 with a UTOC-3. There are V1 and V2 versions for a number of the boards you mentioned, do you have any idea what the difference/s are? 

 

[TitusADuxass]

13 hours ago, Penatr8tor said:

I'm thinking, dump the TAP and LDO toolhead PCB and go CAN bus and Beacon H.

What is the best CAN config for...

Octopus 1.1, Steathburner with LGX-Lite, rainbow barf and Beacon-H

or...

Leviathan, Steathburner with LGX-Lite rainbow barf and Beacon-H

Once get my VzBot (the troublesome child) up and running to my satisfaction, I'm upgrading my v2.4.

Like you, my plan is to ditch TAP but keep the EBB36 and add a Beacon H.

I'll probably keep the Orbiter 2 extruder, but I'm looking at replacing the stealthburner. I have consistent cooling problems when printing ABS.

For the mainboard - I'm going for a Mellow Fly Super 8 pro - although there's not a lot of info about CAN configuration. Naturally, using the Mellow board means that I'm going have use a Pi.

I haven't, yet, looked at using the Fly SB2040. This would mean keeping the SB and it's cooling issues.

[Penatr8tor]

5 hours ago, TitusADuxass said:

Once get my VzBot (the troublesome child) up and running to my satisfaction, I'm upgrading my v2.4.

Yup, That seems to be the trend, with me it is anyway. Build a printer, rebuild a year or two later. 

I'm looking forward to learning and adding CAN. I really like the LGX-Lite extruders I've been using but also want to try the Orbiter 2.  Mostly though, I want to get rid of the huge wire bundle from the toolhead. 

Hang in there with the Vzbot @TitusADuxass it's a great printer but not without its learning curve and setup woes. Once it's dialed in though... It will print some beautiful prints and print them fast.

Silk and ABS... The silk PLA shows every flaw, especially on flat surfaces. The Astronaut is Overture Diamond Grey ABS, not the best filament but handled flawlessly by the old VZ and both are printed at ~300mm/s with 10K acceleration.

 

[Penatr8tor]

OK so I'm back to working on this printer.

She's been down far too long.

Actually, I have been working on the toolhead for a while now but had issues implementing a Mellow CAN toolhead board which brings us to today...

If you're interested in my recent CAN journey, click here.

Long story short... I had enough extras, leftovers and spare parts to do a complete 180 toolhead and now I'm going to go with an Orbiter extruder (was using an LGX-Lite) and I'm going with the Orbitool CAN board and also implementing the Orbiter filament sensor.

While waiting for the board and sensor to arrive I got to work searching for CAD models or STLs from others to see whether or not I had to design something or use something someone already figured out.

No luck...

Well, no luck finding something that 100% filled my needs. I did find one that was close, so I downloaded the CAD files and got busy.

This is the result:

 

SB Extruder Mount - Rear_Orbiter-V2.0.stl

SB Extruder Mount - Front_Orbiter-V2.0.stl

SB CAN Cable Strain Relief_Orbiter-V2.0.stl

SB Cable Cover_Orbiter-V2.0.stl

SB_Fan Cowl_Orbiter-V2.0.stl

These are all the STL files for the toolhead.

The design should work with TAP or any of the X rail mounts, I'm showing a CNC Beacon mount.

As far as hotends go... I'm using the stock-unaltered Voron designed Rapido SB mount, so you could use any hotend.

Below is a post with the CAD STEP file in it. I hit the size cap in this post.

[Penatr8tor]

Below is the STEP file for the above toolhead.

SB_Orbiter_2.0_Mount_02.step

[Penatr8tor]

OK, Lets build this thing...

 

Orbiter extruder is connected to the front mount and the rear mount is screwed into the back of the front.

I hope that makes sense. 

Connected to the printed mounts is a FYSETC CNC Beacon adapter.

Next up is the Orbitool PCB. It would have been nice if they would have put a chamber thermistor onboard like the SB2040 but it's not that big of a deal to just locate one in the printer. The connections cover all the needed inputs for fans, Neopixel LED, Motor (TMC2209), hotend, x endstop, HE thermistor and an onboard lis2dw accelerometer. There's also no USB on this board, only a CAN connector.

The rear is pretty densely packed. There's a 4 pin header on the rear for the filament sensor. It lines up with the orbiter sensor to accommodate the use of flat bridge connector.

 

Next up is the hotend assembly. Nothing really special going on. Just a stock Voron hotend mount with a Gen 1 Rapido HF. I did add the X endstop, temporarily connected the Orbitool board and connected the motor, heater, thermistor and endstop.

Here's a shot of where the flat sensor bridge goes. I'm still waiting on my filament sensor to arrive, so I printed a stand-in part for fitment.

Next up is mounting the front cowl.

I didn't have to do too much to this to get it to work. There was and interference between the front hub of the orbiter and the rear of the upper fan. There was enough material in the Voron design to permit lowering the 5015 1.5mm into the cowl.

You can see where the duct from the fan is lower than the inside of the cowl. Previously it was flush.

 

A few views with the front cowl mounted.

Wires flow nicely around the extruder.

A really nice design from Dr. Róbert Lőrincz.

Lastly, we button it all up by installing the cable cover.

 

 

[Penatr8tor]

Let's talk about this toolboard, basically, my experience.

 And I have only a little experience with CAN so it's coming from a non-expert.

I would rate the instructions as... Not too bad.

The instructions are in the form of a pdf manual. The manual along with other info, CAD and STL files are on the Orbitool Github. 

Pinout diagram was nicely organized and easy to understand.

 

Orbitool board is connect like most CAN boards. A USB to CAN board and then CAN cable to the toolhead board. The 120ohm resistor is already installed on both the Orbitool and the USB-CAN board, so no farting around with jumpers. In fact, there are zero jumpers on this setup.

 

The manual does a really good job of addressing every connection individually. For exmaple...

When I first saw the programming section... I felt like it was missing something. I'm looking at it saying... this is way less steps than the other CAN board.  Actually... it was a breeze to program. It took no more effort to flash the Orbitool than it does to flash an Octopus and I don't mean opening your trench coat in front of a cephalopod. Or maybe I'm getting better at this stuff, but you just connect the Orbitool the way it's going to be connected to the printer and power up. My board wasn't already in DFU mode so I just did the button combo and rechecked and there I was. Then it was just a matter of running a make menuconfig, setting everything per the examples, save and flash and boom! Done!

I'm currently activating each connection in my printer.cfg and testing.

[claudermilk]

Sounds like this stuff is making progress. I'm still not switching right now--the printer works and I'm keeping it busy lately.

[Penatr8tor]

3 minutes ago, claudermilk said:

Sounds like this stuff is making progress. I'm still not switching right now--the printer works and I'm keeping it busy lately.

Nothing wrong with that.

So far, this device has worked with the least amount of frustration. Probably not any more difficult than a beacon probe IMHO. There are some areas of the instructions that could use a little more detail and maybe some copy-able commands to prevent typos but overall... Thumbs up.

And I have everything working on the bench. 

[Penatr8tor]

This past weekend I got a chance to unhook the electronics from The Rig and install the controller and toolhead on the printer.

Wiring wasn't too bad. Most of what I had connected to the Octopus pretty much transferred over to the Leviathan. A few cuts and crimps and everything went together nicely.

And... I got everything running with enough confidence to print a part. And it turned out really nice actually. This print has literally no input shaping or pressure advance applied, printed in PLA at ~150mm/s and 4K acceleration (no breaking any records). I used a raw Voron Orca slicer profile, dropped a 30x30 cube on the bed and printed 3 walls with no infill and top surface. Printer ran very smooth and quiet on the XY, the Z has always been a bit noisy when it's moving quickly but, that's really only during homing or parking. During a print it's quiet.

And here's the printer... Still need to add panels, some sleeving for the umbilical and a few other bits and she'll be back in service.

[Penatr8tor]

One last mod to finish up the project. Clicky-Clacky Door Mod.

And... The printer is now in my office and printing beautifully.

Printed parts... These were printed on my VzBot.

Material Polymaker ABS Black, 5 perimeters, 5 top, 5 bottom and 30% cubic infill, no supports, no brim.

Bed 110, HE 240, 

Outer walls and top surfaces 200mm/s 5K accel.

All other walls infill etc. 250mm/s 10K accel.

Travel = 800mm/s 20K accel.

Print time = <3hrs

I put all this print info up here because I'm curious to see how fast others are printing ABS. 

And of course, the quality is not bad.