Seb's Formbot 2.4 300mm Adventure

[Penatr8tor]

People are going to tell me I'm crazy but...

I use gun oil.

Zero grease.

There's just not enough load on the rails to need grease plus... These linear rails are of the recirculating ball type and what happens with grease when there's a very light load is that the balls don't always recirculate.

So, if you want to use grease... by all means use grease. I use the same oil I use in my Colt 1911 Mark IV Gold Cup Match.

p.s. I do wipe and reapply a couple drops to the groove every so often.

[sebster]

Gun oil is an interesting choice

I haven't worked with linear rails before so I went with the official recommendation. Hopefully the rails are ok as they are. I guess I'll find out.

Squaring the gantry was tricky. It's a barely noticeable distance off on the front. One side touches about 1mm before the other but only at the front. I think that is because the X/Y joint isn't perfectly flush in one corner. I'm pretty sure everything is square though and I assume the gantry won't ever print at the very edge.

Assembling the TAP took a lot longer than anticipated. The manual isn't in sync with the STLs which confused me at first.

How much resistance should the TAP provide before the magnets release?

It feels like mine provides almost none and that is after changing the formbot screws for some that feel a lot more magnetic as long as they are not part of the assembly. The magnets are aligned but it still seems like there isn't much holding force. If that turns out to be an issue I might go for quadruple magnet. I'll leave it as is and see what people say. Did I miss the screws intended for TAP or does formbot not provide magnetic ones? Is quad magnet recommended? My TAP sled barely snaps back down... edit: Looked at Nero's video and it looks somewhat similar to mine... until he sets the magnets. Hm. Three magnets will create shear forces but it would be a compromise and the rails should be able to handle it I assume?! editv2: My new hypothesis is that the TAP rail is slightly sticky so the force of the magnets gets partially lost in that. I made sure they are greased evenly. Maybe oil would be better for the TAP rail?

I'm starting to see why some people recommend Tridents over 2.4s. That gantry has a lot of variables to get right.

The belts are in and TAP is installed. Next stop: Stealthburner.

QC is keeping an eye on proceedings

I plan on using a piano tuner to get the belt tension equal so I don't care about the length of the belts. This worked great for the banded Z on my ender.

Edited March 16 by sebster

[mvdveer]

4 hours ago, sebster said:

QC is keeping an eye on proceedings

Hi QC!

[Dirk]

On 3/15/2024 at 9:26 PM, Penatr8tor said:

I use gun oil.

I would have loved to try that out... Any product with 'for guns' would be considered illegal around here (and the rest of the world except in the US especially around the areas where you live )

I used this. Good thing is... I now also have parts to smear my espresso machine parts after cleaning it

I doubt it matters so much... as long as it makes the rails "slippery when wet"

[sebster]

Is this enough overlap between the gears?

The gear from the kit has the exact dimension as specified in the manual. There is nothing to adjust the gear on the stepper as far as I can tell. The motor sits flush against the plastic. The label on the motor says: CSE14HRA1L410A-01

[mvdveer]

Looks pretty good. Should be some (minimal) play/backlash between the gears (about 0.5mm as a guess). I just adjust be touch.

[sebster]

Thanks for the feedback, @mvdveer

 

I'm surprised the design doesn't make it possible for the gears to overlap 100%. I tried to make sure the motor is mounted with a little bit of play. The number of parts where I hope rather than know it's correct is starting to reach a level that feels problematic It might be time to let go of the aspiration to get everything right first time and hope I "learn" without causing damage  

Formbot just upgraded their 2.4 kit to come with Canbus out of the box

Up next is the canbus wiring. Those pins will be a challenge...

 

Edited March 17 by sebster

[mvdveer]

2 hours ago, sebster said:

Up next is the canbus wiring. Those pins will be a challenge...

They are tiny, but even with my fumbling hands - managed it. Patience is a virtue and in this case it is true. 

[sebster]

Stealthburner is done, I hope. Those 1.25 crimps are comical but they seem to be connected.

Is this the correct wiring for the extruder motor? I'm not sure how to read the spec sheet and this is my best guess in consultation with GPT4

The CAN cable goes directly to the BTT Pi, right? I want to make sure because the cable of BTT's own CAN kit is the wrong connector size for the Pi.

My brain is on fire. Doing this for the first time means there is so much to learn and figure out. I hereby decide to act like an and adult and look forward to continuing tomorrow

My ender is running on 1500/3000 acc and 50-100mms with pretty good print quality. I was surprised to learn that some people to push their Vorons much harder than that.

Edit: Octopus will act as can interface. Research shall continue... This seems like an up to date resource:
https://canbus.esoterical.online

Edited March 19 by sebster

[sebster]

I have learned a lot about canbus and Katapult. I thought I wasn't making any progress and suddenly most of the essentials were together. It almost looks like a printer:

Unfortunately it seems something if off with my BTT EBB SB2209. I get an error in Klipper:

MCU 'EBBCan' shutdown: ADC out of range
This generally occurs when a heater temperature exceeds
its configured min_temp or max_temp.

As per Klippy:  target=0 temp=1119904.1

This does not change when I disconnect the thermistor or connect a 1k resistor to the pins directly. The PT1000 seems to work. I has the expected resistance and the crimps are conducting. Unless there is an issue with my config, which is copied from the official repo, I don't see any explanation other than a faulty device.

Any suggestions from people with experience?

[extruder]
step_pin: EBBCan: PD0
dir_pin: !EBBCan: PD1
enable_pin: !EBBCan: PD2
microsteps: 16
rotation_distance: 33.500
nozzle_diameter: 0.400
filament_diameter: 1.750
heater_pin: EBBCan: PB13
sensor_type: PT1000
sensor_pin: EBBCan: PA3
control: pid
pid_Kp: 21.527
pid_Ki: 1.063
pid_Kd: 108.982
min_temp: 0
max_temp: 260

I'm out of ideas...

[Penatr8tor]

On 3/18/2024 at 12:01 PM, sebster said:

Stealthburner is done, I hope. Those 1.25 crimps are comical but they seem to be connected.

Is this the correct wiring for the extruder motor? I'm not sure how to read the spec sheet and this is my best guess in consultation with GPT4

 

Check out this thread... This is the correct wiring order for the extruder stepper.

 

[mvdveer]

2 hours ago, sebster said:

MCU 'EBBCan' shutdown: ADC out of range

That generally means there is a loose connection between the heater and the board - bad crimp or broken wire? Often only occurs after printer movement like homing or QGL (force on the cable) Worth while checking before replacing the board. But then @ChicagoKeri did a very good thread on static electricity damaging the thermistors on these boards. (SB2209)

[sebster]

20 minutes ago, mvdveer said:

That generally means there is a loose connection between the heater and the board - bad crimp or broken wire? Often only occurs after printer movement like homing or QGL (force on the cable) Worth while checking before replacing the board. But then @ChicagoKeri did a very good thread on static electricity damaging the thermistors on these boards. (SB2209)

I triple checked the crimps and in the end connected a resistor directly to the two pins (1k ohm) on the board. The reading in klipper was the same regardless of what and if anything was connected (Temp in the klippy log was around 12,000,000 which makes sense to me if resistance is infinitely high). This leads me to believe it is the board or I'm reading the wrong pin. I assume if the sample config was wrong about the heater pin it would have come up somewhere.

I've searched far and wide, I can't see anything I could try. If crimping a resistor to the pins doesn't cause a reading I don't see what might. Is that not a conclusive test? The only options I see are "wrong pin" and "broken board".

I read @ChicagoKeri thread. I don't see how it applies to this scenario. Am I missing something?

It stills seems more likely that it's a user error but I don't see it...

Edited March 21 by sebster

[Dirk]

which CAN toolhead board is included with your formbot kit?

[mvdveer]

16 minutes ago, sebster said:

I've searched far and wide, I can't see anything I could try

You obviously had a look at this

Two wire Pin connection:

 

 

[sebster]

@mvdveer I actually looked at their github and the pdf manual which provides different information regarding the switches. It's also really not clear to me what the MAX31865 is. I know it's an amplifier but not how it influences the board's config.

Does anybody have and example config for BTT SB2209 with a PT1000?

I tried changing the switches as suggested by the wiki. The temps in the log are unchanged. Does anybody know if this is correct?

sensor_pin: EBBCan: PA3

 

[Dirk]

1 hour ago, Dirk said:

which CAN toolhead board is included with your formbot kit?

so I assume the SB2209 (RP2040), which is a completely different PCB from the SB2240-2209... 

I would suggest taking a look here, to get you more confused... or maybe telling us which PCB it is so we can point you in the right direction...

 

[mvdveer]

30 minutes ago, sebster said:

Does anybody know if this is correct?

That is the sensor pin for the non PT1000 thermistor.

From what I can see the PT1000 use a SPI connection:

I found this configuration that seems to work on reddit:

 

[mvdveer]

From a comparison with your config it seems you have a different pin - this will depend on your wiring - where the two wires are connected. 

The sensor type here is set to Max31865

You are also missing the SPI section (In the red rectangle.

 

Hope this helps.

[sebster]

11 hours ago, Dirk said:

which CAN toolhead board is included with your formbot kit?

I have the EB2209 (non RP2040).

From what @mvdveer posted I was indeed reading the wrong pin for my PT1000. It would have been a great solution but unfortunately reading PA4 using max and SPI gives me a different error and a temp of 0 instead of millions:

UI: MCU 'EBBCan' shutdown: Thermocouple reader fault.

klippy.log:
Receive: 96 40.176043 39.985331 14: seq: 1d, thermocouple_result oid=7 next_clock=2730634858 value=1 fault=65
Receive: 97 40.255110 40.254514 11: seq: 1f, clock clock=2716537012
Receive: 98 40.476072 40.254514 14: seq: 1f, thermocouple_result oid=7 next_clock=2749834858 value=1 fault=65
Receive: 99 40.476149 40.254514 12: seq: 1f, shutdown clock=2730637526 static_string_id=Thermocouple reader fault
Transition to shutdown state: MCU 'EBBCan' shutdown: Thermocouple reader fault

I now understand the sample config. It has the part for max31865 commented out so that was my initial mistake. It seems like I am one step closer but still off.

This is my config now:

sensor_type:MAX31865
sensor_pin: EBBCan: PA4
spi_bus: spi1
rtd_nominal_r: 1000 # tried 100 and 430, makes no difference
rtd_reference_r: 4300
rtd_num_of_wires: 2

Connecting a resistor to the pins didn't change the error or reading btw which I hope rules out the PT1000 cables. I have checked the dip switches again since that was what another user reported as the issue.

Thanks for your help. I'd be completely lost without it.

Edited March 22 by sebster

[mvdveer]

45 minutes ago, sebster said:

I have checked the dip switches again since that was what another user reported as the issue.

Depends on where the wires are connected. If it is on the middle two pins, then the sensor_pin would be PA5 or PA7. Worth trying these and see what happens.

   ---------->.       

 

 

[sebster]

I wasn’t sure how to read the diagram. PA5 and 7 gave me an error. Something along the lines of reserved for spi. I’m not home so I’ll give that another try. 
 

https://github.com/bigtreetech/EBB/issues/58

That looks like PA4 should be correct. That thread is about 2240 so maybe not. 
 

this guy uses gpio which is an approach I have not tried:

https://forum.vorondesign.com/threads/bbt-sb2209-pt1000-problem.1184/
 I’ll report back 

[Dirk]

2 hours ago, sebster said:

I have the EB2209 (non RP2040).

I would do it this way (and have done so)...

Use the BTT EBB github to get the relevant files.

There they have

- a sample config. (https://github.com/bigtreetech/EBB/blob/master/EBB SB2240_2209 CAN/sample-bigtreetech-ebb-sb-canbus-v1.0.cfg)

- The pinout: https://github.com/bigtreetech/EBB/blob/master/EBB SB2240_2209 CAN/SB2209/Hardware/BIGTREETECHEBB SB2209 CAN V1.0_IO.pdf

- The build manual, which also contains the pinout and the wiring: https://github.com/bigtreetech/EBB/blob/master/EBB SB2240_2209 CAN/Build Guide/EBB SB2240 2209 CAN V1.0 build guide_20240219.pdf

 

The error you had at the beginning, is indeed because there is a thermistor signal out of bounds. If you already have been printing a while, this might be because of a broken / shorted / earthed thermistor. But with a new build, it is usually a wiring error.

 

You need to decide if you really want to use the MAX chip (better accuracy and speed with PT100 and PT1000 sensors), or the regular  thermistor port (non PT sensors). You can in fact use both. If you are having troubles with the MAX port, why not use the simple thermistor?

I see you have been busy setting up the MAX now for two days, so lets do that. It is all in line with what @mvdveer has suggested...

1- use the settings in the sample config. Not ones from forums. Comment out the normal thermistor

 

## sensor_type: EPCOS 100K B57560G104F
## sensor_pin: EBBCan: PA3
 

and uncomment the max.

sensor_type:MAX31865
sensor_pin: EBBCan: PA4
spi_bus: spi1
rtd_nominal_r: 100
rtd_reference_r: 430

rtd_num_of_wires: 2
 

 

 

2-  Use the pinout schematics AND / OR Page 13 of the Build manual, to determine which pins to set for your thermistor. Apparently you have a 2 wire PT thermistor, so make sure to set only pins 3 and or 4.

VERY nice drawings on page 13 of the manual / installation guide.

This same drawings & text shows exactly how you have to connect the two wires from your thermistor to the GH 1.25

 

And if you have done everything correctly, it will work.

If it does not work, contact BTT or the Kit maker, if they can assist you.

 

 

Good luck.

[Buurman]

On 3/15/2024 at 4:25 PM, sebster said:

Purple Posting. I like

oooh, I am too late to this party.... 

[sebster]

Thanks for the replies everybody! I really appreciate it.

@Dirk Those were the documents I used initially. I did not know how to read the diagrams well enough which is how I ended up on PA3. Now:

I still get the same error despite the following:

1. I checked continuity between PT1000 wires and the pins at the back of the board. That rules out the crimping.

2. My config is for the MAX31865 on PA4 as per the sample config with the PA3 sensor commented out. That fixes my initial mistake and I get a different error so that did change something as expected. That rules out a config mistake, I hope.

3. The dip switches are in the correct position for a PT1000

4. The PT1000 reads 1089ohms according to my multimeter which seems about right. AFAIK that rules out the PT1000 as source of the issue.

(@mvdveer reading PA5 and PA7 gives me SPI errors).

I will contact Formbot & BTT and see what they say.

If anybody has other suggestions I'm all ears.

 

EDIT: I just powered the printer back on and now it's reading the temps as expected.

Eric Zimmerman spotted this in the logs:

 

Max31865 RTD input is shorted
Max31865 RTD input is shorted
Max31865 RTD input is shorted

Multimeter confirms continuity between Pins 1-2 and 3-4 on the PT1000 ports. I'm just confused and will see if this survives another restart

 

Edited March 22 by sebster