z_thermal_adjust

[therm_virtual0y]

I’m curious if anyone has Z_Thermal_Adjust in use?

More specifically, I’m wondering where you take your adjustment temp measurement from.  
also, there is an article that talks about moving your bed mesh zero point reference to a more stable corner too.  Just seeing if anyone has any experience implementing it or is it worth the hassle. 

[Dirk]

The only z thermal adjust I know of, is the one mentioned in Ellis' tuning guide in the thermal drift article.

It references a GitHub resource

- https://github.com/alchemyEngine/measure_thermal_behavior

And a few klipper documentation references about thermal measurement and adjustment.

 

Funny thing is that it looks like nobody really trusts this method. Not even the author of the github page.

I think mainly because there are easier and more reliable and less complex solutions( like buying a set of 80€ titanium backers) 

But if you genuinely want more information, I suggest the github reference and it's author alch3my 

Edited February 15, 2024 by Dirk

[Dirk]

I realise I did not answer your questions.

I do not use it. I have not seen it mentioned previously.

A more stable /reliable reference point would be close to the corner of two rails. So positions (0,0) or (350,350)(and the other two corners). The bed movements during heating is most extreme at the center of the bed. The corners move least. Take a look at the thermal graphs from the initial research 

Something different where thermal adjustment scripts were also mentioned, is the cartographer3d probe. On their discord I saw the author mention developing scripts for thermal adjustment, a-work-in-progress. 

[therm_virtual0y]

This is what I have seen also.  Thanks for the confirmation that thermal adjustments based on frame temperature is really not mainstream enough for me to mess with at this point.  
as far the zero point reference, are you using a corner or the center?  If I do 0,0 then all calculations for z end stop, Probe z offset and auto Z will be in the corner.  This seems like something I might try this weekend.  

[Dirk]

I don't use those scripts

I bought two linear 400mm rails (for both sides) and a titanium backer for the center. (Total cost +-40€)

But it you look at those pictures I linked and the story from whoppingpochard, then your bed becomes a 'taco' when heated compared to an unheated bed.

When heated the side rails move becoming a C. With the highest point in the center.

https://github.com/VoronDesign/VoronUsers/blob/master/printer_mods/whoppingpochard/extrusion_backers/images/no_backer.png

This means your bed becomes like a turned down taco. The corner points of your bed are relatively fixated.

And the center moves up... On a 300mm bed, 0.4mm in 1-1.5 hours. The more you go to the edges, the less the moving up is.

So if you are going to try it out, the corners would be the best place. I think you need to adjust 0,0 to a number that can be measured by your probe. 

I'll be interested in your tests I think many others as well

[Dirk]

oh... and just saw this post: 

 

pointing to another effort in thermal correction: https://github.com/JiMcNulty/FDC

 

[therm_virtual0y]

I finally read through this white-page ( https://github.com/VoronDesign/VoronUsers/blob/master/printer_mods/whoppingpochard/extrusion_backers/images/no_backer.png), and decided to buy new Y rails for my Voron 300.  I spent a little extra on the HoneBadger Mgn9 and used the old ones as backers.  I had to cut the z belt holders on top so the old rails would fit but that wasn't too bad.  I'll just reprint later.  Anyway, the results actually surprised me.  Attached is a before and after of my bed mesh and the results speak for themselves.  Let me first say that I had a regular bed taco that I thought was the problem.  I still had pretty good prints despite the 0.1 to .15mm height difference from front to middle of the bed.  Doing this was because I have my Voron totally finished, and this was the last mod (for now) LOL.  

[mvdveer]

Well done - just fitted titanium backers to all my "big printers".  Similar results.

[therm_virtual0y]

I thought about the titanium backers too, but also wanted better linear rails. Figured I would try and see if it would work.  It seems to have done well, so I’m happy with it.  I also am using Cartographer probe with my BTT SB2209/2240.  Still trying to dial in the z-offset though.  It seems too squished on the first layer but when I put z to .1mm and use my .1 feeler gauge, I barely feel the nozzle.  It’s new so I will figure it out.  

[wowpalompa]

On 2/14/2024 at 11:37 PM, therm_virtual0y said:

I’m curious if anyone has Z_Thermal_Adjust in use?

More specifically, I’m wondering where you take your adjustment temp measurement from.  
also, there is an article that talks about moving your bed mesh zero point reference to a more stable corner too.  Just seeing if anyone has any experience implementing it or is it worth the hassle. 

Hi,

 

Your post was from a while ago, so I hope the moderators don’t mind me reviving it and read the reasoning for allowing this further ahead.

But first, a little background: I was skimming through a video that’s over two years old when I noticed a strange .cfg file. A quick Google search—bada bing, bada boom—and here I am.

 

I have no idea what was available a year ago, but it looks like the conversation back then was centered on the XY-axis. The consensus was (and still is) to just get the backers and forget about it, which honestly sounds fair and true even today.

For 2025, as of my reply, the discussion seems to be focused on the Z-axis specifically. With a little digging, I found posts, commits, and development around this exact topic as recently as last month. This means the topic is fresh and ongoing.

The question is about the Z-axis, specifically where to place the thermistor. I have exactly the same question. Based on the goal (measure vertical extrusion temperatures) and checking the [Z_Thermal_Adjust], it looks like just one (1) NTK 50K thermistor is used. I'm guessing one that's anywhere between 19 - 39 " so you can comfortably route it as needed. As for the placement...

Now, In terms of redundancy: "Two is one, and one is none". Therefore, I'm incline to place 8 thermistors. Two on each vertical extrusion, one inside the vertical extrusion channel to directly measure the extrusion temperature and the other facing inwards like an antenna. At a glance, this looks like a great place for such a configuration:

"Hot air expands, and rises; cooled air contracts – gets denser – and sinks".

I have not done any testing / math and I do not own any app that will allow me to do a simulation, but based on this basic principle I imagine there are some challenges to manage when placing the thermistors up here. This is specially so if you have 5015 fans under and/or you have dedicated chamber heaters.

Btw, all of this is SUPER early information. It's FOSS/RepRap - take everything with a grain of salt. I've done no testing, no math, no simulations. Just trying to make an educated guess, if we can call it that.

The second part of your post is an entirely different topic, so forgive me if I focus on just the one.

Cheers! 

---

EDIT: Wishful thinking aside, my setup has 3/5 thermistor ports in use. This leaves my setup with just two probes for the [Z_Thermal_Adjust]. This probably means I will have to average the Nitehawk chamber thermistor (GPIO28) with the thermistor that will be in the chamber like an antennae. That leaves me with just 1 thermistor to be placed directly on a vertical extrusion. Far from my speculated ideal scenario.

 

 

I've reconsidered the placement, because I'm a sucker for symmetry. I'll try to design something like this ASAP and order the thermistors.

Edited 3 hours ago by wowpalompa