Not necessarily at 500C but I like to have a bit of tuning margin. The nature of my use is highly experimental. The custom 3d printer build in itself is part of the experience. For the high temperature I want to print Peek and the Hotend is in preparation for that.
Interesting. Iâd love to see pictures of your printer and also reports of how well it went. Iâm working on HT printing myself but TBH with very mixed results.
BTW, it sounds kind of strange to go for a filament where one kg costs 600⏠onwards but complain about 20⏠for a sensor 
Only the grease you will need for linear rails in the temp range of 100°C chamber temperature is about 90⏠for 60g (taking aside the fact that you need HT specified rails)
1 Like
Not that I need to justify myself, but as I mentioned, this is preparation for that. In the meantime I print with ABS, Nylon and whatnot. Beyond the issues I already mentioned with SPI, I am slowly building up my capacity to have the flexibility to offer a service that can benefit from this, specially on industrial prototyping. I am on a budget, and this is not a source of income yet, but I want it to be. This is also a learning experience for me and if I can keep costs low for where I am heading, I will. Grease that can take 100C is not expensive, at leat not here and I already have it. The heated chamber is still on my to-do and I plan to DIY it. For now, I want to get this working and learn while I am at it.
I will post some pictures later, but it is mostly looking like the Frankensteinâs monster at the moment.
So, gonna share some update and would like to thank you again for the helpful attitude.
Been silent because I got sucked into a rabbit hole with all free time I had. A few conclusions:
1 - The K thermocouple is super sensitive to any voltage noise, like @Sineos mentioned. With the osciloscope I was able to determine that the grounded frame had a noise delta of around 100mV. That was enough to throw it off. So when I wired the extruder to the frame, the temperature just went bananas.
2 - I then removed the wiring from the extruder and wired the gantry to the frame and that completely stabilized the sensor and the K thermocouple would have almost zero fluctuation. All went beautiful (or so I thought), Until the extruder went into action. When that happened. puff. Print stops immediatelly with huge temperature drop.
3 - So, if I wire the extruder to the frame I am screwed from the start, if I only wire the gantry to the frame, I get screwed on nozzle priming (extruder starts operating). I figure due to an all metal extruder, there must be some runaway current. Maybe even an inductive current from the stepper motor. So thatâs a no-go. The extruder needs to be grounded.
4 - When I removed the grounding from the frame, I got rid of the noise and that would be a possible way to go âifâ I didnât have a mains heated bad. It is a safety hazard to no the grounding, I even got stung by the mains during this playthrough.
A solution to this would have a capacitive filter. Had a nice read here: Negative Effects of Grounding (Earthing) a DC Power Supply - Technical Articles (control.com)
5 - Got annoyed pulled a PT100 I had laying around that was way too long. Ans since I had nothing to lose I thought, what the heck. I will just try to cut the darn thing. So I got my dremel, cut the head to size:
I had no idea this would work, but to my good surprise there was just more wire and in the end I didnât ruin the sensor.
6 - Got to war with Manta M8P SPI again. And turns out nowhere in the entirety of the internet, there are clear instructions on how to use a Manta M8P a BTT max31865 v2.0 and a three wire PT100. So was punched to the cornered, but eventually triumphed. For those people victim of horrible documentation from BTT/biqu (making biqu biqu setups is more challenging than anything else), here is how you do both the wiring to the stepper slot and the soldering on the board (which I had to magnify, so ridiculously small it is):
So, in short, remove the default bridge between 24 pad and the middle pad. then bridge the middle pad to the pad marked with number 3 on the right. Then above is the wiring on the manta board. And here is the kilpper config for manta m8p v1.1:
[extruder]
...
# The SPI params below are specific to the manta m8p v1.1, at least these values are well documented for eitther v1.0, v1.1 or v2.0. You can check those on the manual.
spi_software_sclk_pin: PA5
spi_software_mosi_pin: PA7
spi_software_miso_pin: PA6
sensor_type: MAX31865
sensor_pin: PD10 # For v1.1, I used Motor 8, which on V1.0 is PC4. For other slots, check the manual or the example on klipper repo.
spi_speed: 5000000
rtd_nominal_r: 100
rtd_reference_r: 430
rtd_use_50Hz_filter: True # If your AC power frequency is 50Hz (like EU countries), otherwise false.
I accept 3D printing supply donations from anyone that will get to keep their hairs after I pulled out all of my head.
PS: I had to disable the ADXL, because of the SPI BS this board gives me and I was still unable to get around.
8 - After all of this, guess what happened? Extruder starts operating and the same behavior happens. I figure that after messing so much with so many times with small experiments I have something wrong in my wiring. So what I will do next is rewire everything, make all neat again as right now it is looking like spaghetti. I probably have something touching what shouldnât. I will also take the opportunity to add a third Z axis as my print bed is as flat as the earth is for flat-earthers. The third axis will come in handy with the Z-Tilt, which is already very helpful with only 2 Zeds.
Feel free to make suggestions, ask questions, add comments, make fun of my shenanigans as long as they are respectful or not charged with condescending vibes. I quite enjoy discussing this stuff.
1 Like
Sorry youâve gone through all this.
The big suggestion that I would make is that before you do any wiring changes draw it out (pencil on paper is probably fine) so you know exactly what you are doing. I also make up a net list detailing which pins each wire goes to. It seems like a lot of work but it can save you time with fewer problems later.
Iâd like to understand exactly how you are grounding things. You must start at the power supply and use the âV-â or âGNDâ output as your system ground. This should not be connected to the AC âEarthâ. This will give you a ground connection to your main controller board and I run my external device grounds from there - there should be a power line as well, so I use both there.
Now, your metal frame should be connected to the AC âEarthâ and there should be no contact between Earth and GND.
Good luck!
1 Like
As always good to have a second pair of eyes to cover some blind spots. Thank you.
Yes, thatâs exactly how itâs done. Please check if this drawing makes sense to you:
I have printed with K sensor before and it didnât give me problems albeit, not as sharp as PT100, the fluctuations were not significant. Plus, the driver is only connected to the board ground (although I did test pulling a parallel to the frame, for some testing)
I canât remember exactly when I started having issues with it and I think was when I got the silocone AC bed heater.
From the drawing do you see any issues? What I did notice that caused problems is the braided wire getting in touch with the frame (and thatâs what I mentioned about losing the ground reference). So I tried isolating everythin (kapton on the bed, for example). The problem is that the sensor head touches the heatblock, that touches the nozzle and heat break, that touches the extruder, that touches the stepper motor and all of these are conductors. And it makes sense that made the sensor go bananas when I grounded the extruder (or better earthed it). Oh wait⌠was this my mistake? I should have connected the extruder to the boardâs GND. 
Thanx for the drawing - I donât see any issues either but I donât see how youâve wired the toolhead.
Could you provide that wiring?
The toolhead wiring is no big deal. I made all wires terminated in 3 sections for ease of maintenance (already fried the PT100, and replaced it in less than 10 min):
- Toolhead to harness
- Harness
- Harness to control board.
This is how it looks like (a bit messy, but I plan to make tight grouping once I stabilize the setup):
The only wire the doesnât go directly to the board is the new addition of ground wire attached to the toolhead via the extruder fan screw. Well, in correction, I have now changed it to go to the CB1âs board GND pin:
The dashed line is how I had it. And I am pleased to say that, so far, three prints down after this change, I had no issues. Iâd probably have succeeded with the K, if I had realized my mistake before, and you were critical to that realization.
You were also very right about drawing it out. It took way less time than I was procrastinating about and would have saved me a lot more time than my usual ad-hoc way to do things, if I had done it before. Which is one of my biggest known shortcomings. Lesson learned (again).
Thanks for sharing your beards of experience. Really learned a significant amount of lessons on this thread alone. I hope I can buy you a cup of coffee someday.
PS: Now I can go to the next thing. Which is having the ADXL for input shaping working, with this boardâs SPI hell (you can understand what I am talking about, if you look at the issues in the github repo for this board).
1 Like
You are blocking the coldend fan.
Not a good option!
2 Likes
The picture at this angle may make it look like that, but there is plenty of room for airflow. This is the spot I found the most stable for the probe. Other places I have to frequently recalibrate Z offset and so far it has remained cool enough to touch it.
Even if the angle might change the picture this thick inductive probe is in the way and reduces the effective diameter/cross section of the fan inlet that creates resistance to the air flow!
The coldend fan is crucial for good printing at least when printing with PLA and stuff like that. Just a well-intentioned advice.
And now back to topic. 
1 Like
Not complaining, thanks for the advice.
I did check how it would affect the temperature of the extruder after you mentioned it, and after a 5 hour PLA print with a flow rate averaging 10 mm3 the extruder body was still cold and no signs of heat creep.
Ps: the reason I think the effect is very limited is due to the construction of the H2 extruder. Itâs an all metal extruder. There is a tiny heat sink inside the body where the heat break sits (I also have thermal paste applied to it) and I think the heat quickly dissipates into the body. The fan that came with it is not very powerful because I assume it doesnât require too much cooling.
Will have to reevaluate when I put an enclosure in and possibly have an external duct pulling cold air from the outside and another to blow the exhaust out (if it comes to that for higher temperature printing). But thatâs a project of its own.
I havenât done a lot of research into heat sink cooling but, based on what I have seen, surprisingly modest fans and cooling fins are required to avoid heat creep. Donât forget that the filament itself absorbs heat that is in the heat break (and heat sink) which is then pushed out into the model with the filament.
This doesnât mean that you should deliberately design your hot ends with minimal fans and heatsink fin area but it does mean that if you make sure there is as much as possible heat sink cooling youâll very probably be okay.
I trust your judgement. I am using stock fan that came with the extruder. The heat sink this extruder has inside is the size of my pinkie fingernail.
If you look at the design of this extruder you will see that the air doesnât even flow into the body it cools am external hemera like heatsink. So it does seem to rely mostly on heat dissipation from the body. When I touch the extruder whilst itâs printing it feels cold, so I assume the cooling requirement is low. If cooling wasnât being sufficient the body would heat up or there is a fundamental design flaw I canât do much about.
Please donât call my previous comments as a âjudgementâ. At best theyâre an âobservationâ.
I would always recommend that people design their hot ends to provide as much cooling air to the heatsink as possible; heat creep can turn your hot end into scrap if there isnât enough cooling air.
Looking at your image above, it looks like you should have good convection cooling as well as the forced air from the fan.
I can confirm this. On one of my printers with a dragon hotend, heat creep really is an issue and I had to modify my hotend geometry and use a stronger fan to get rid of it.
It would otherwise 100% block at fast printing PLA with 240°C, latest after 3h
2 Likes
Guys, I am not questioning the need to prevent hear creep. But have you seen how the construction of biqu H2 is? There is not much to do here, or design or modify.
The thing looks like a safe. I could probably hammer it and it would still remain with the same shape. The only thing is mine there is the support for the probe.
@mykepredko by judgement,.I mean your wits. Donât worry, I am not taking it as a judgement of my setup.
I really donât know what I could do here in respect of Hotend design. I donât have machining equipment, which would be required if I wanted to do any modification to this extruder. I am printing PLA at this moment at 230C. Itâs been hours, the thing still cold to the touch. If internally it is getting hot, then I have no idea what to do if not get a different extruder. But at the moment I see no signs of heat creep. When I retract the filament itâs mostly intact, except the very tip that is still soft from the nozzle:
Edit: This extruder is am all in one. The only parts that stick out of the package is the neck of the heat break and the heat block with the nozzle.
Just to be sure I am not shooting myself in the foot I bumped the temperature mid print to 240C. Then, tried to measure the temperature at the heat sink of the heat break. I had to stick the temperature probe at that tiny little hole that gives access to the heat break screws, itâs otherwise sealed from the outside world.
I donât really know what would be an acceptable temperature for PLA there, but it if the heat break is close to that temperature, Iâd say PLA can take it for a few seconds. Thoughts?
If youâre using a unmodified commercial hot end then youâre probably fine.
Only modification is the heatblock (same material) from V6 to Volcano