Tool change issues after firmware update 1.19.2

jacksat

Best Regards
MT

Ps if you intrested something I have tested with great success ,the use of compressed air for cooling the part during printing , for small part worked miracles only problem was the amount of air used and the noise, my neighbor became crazy , the compressor was working to much and for to long :-). so reverted to the traditional fan,

the advantaged was less weight on the head and vibration was reduced ,not that there is a lot of vibration , but any improvement is an improvement how ever small

Hi,

I used a aircompressor to ,but now I use a airpump from my pond for part cooling.
All you hear now is the sound of air at the nozzle the pump is super silent.
The one I use is even too much airflow, you probably can use a smaller Hiblow 40hp or similar but this is what I had and they are not cheap.
The air regulator I use too dry the air with silica (cat litter),I replaced the oil in the oiler and and the watercatcher with silica, so no condensation can drip
out at the cooling end and ruining your prints.
You can go with cheaper brands but I know from experience they are also noisier.

And I use watercooling to cool the extruders ,so my setup is really quiet.

dc42

Thanks for the tip! A few of us have been looking for quieter air pumps.

Scachi

I have updated to 1.19.2 , the issue still exists.
I have the checkboxes active for tool change, I can see it start to move but it isn't working like it should.
To me it looks like the commands in this file aren't executed one by one, after the previous line has finished.
It seems they get executed without waiting for the previous one to complete.

I have emptied my free[0,1].g files and tried to only use the pre[0,1].g files.

[[language]]
; Absolute positioning
G90
G1 X244.0 ;F3000 ; move to right
M42 P6 S0 ; right magnet off

G1 X-34.0 ;F3000 ; move to left
M42 P5 S1 ; left magnet on

With this file, I can hear the relay controling the magnet click right after the first move has started, not after it has finished !

Maybe the free and pre files are getting executed without waiting for all the commands to complete, one after the other. This would explain why some of my moves where stopping in wrong places when using both files (free & pre).

dc42

I don't know why that is happening, because M42 commands should be synchronised so that they occur at the right time. I'll take a look. Meanwhile, try inserting a M400 command between the G1 command and the M42 command.

Edit: I just checked the code. Looks like code synchronisation is disabled for commands coming from the web interface, USB and PanelDue inputs. So that explains why it is happening. I think it would make more sense to always synchronise commands that were read from a macro file, so I'll change this. Meanwhile, M400 should work.

Scachi

M400 fixes the relay timing. I'll test if it is working now correctly with the merged tfree and tpre files.

There is still something strange happening to the moves when using tfree and tpre files.

In tfree0.g I am using a negative x position -34.0 to drop of my T0 tool, I have set the minima limit of x to -34.0 correctly for that.
In tpre1.g I am using x position 244.0 to pick up tool T1.

When I put everything into tpre1.g (drop tool T0 and pickup tool T1), the moves are correctly reaching the positions.

When I put the drop of T0 into tfree0.g and the pickup of T1 into tpre1.g it will only reach X 0.0 on the free move of tool T0.
For the T1 tool drop of it only reaches X 210.0 … it looks like the -34.0 gets calculated in there somehow or there is somewhere a limit set for the free moves to limit the x movement to min 0, max 210 .. but I am not able to find this limit in any file.
And I have put M400 in there, too. This isn't fixing the wrong drop off positions.

dc42

The tpre file is executed before the new tool is selected, therefore any tool offsets you define using G10 commands will not have any effect. It's usually best to put commands in tfree and tpost, and leave tpre empty.

Scachi

Oh, that explains my issue, as my tool offset is -34mm / +34mm.
Thank you very much !

MrT

Dear Jacksat

Very nice build !!!!! I also thought of using a much bigger airpump , but the cost is fairly steep , after all the testing I did , using compressed air is best for small parts as for biger parts the fans work beter of course it depends how much air you have available , during my testing i managed to use both and that gives really nice result , quick cooling near the nozzle and even cooling on the rest of the part .

I really like you water cooling , you have a much heavier build then me , I used the same material as the ultimaker 2 to build my printer , it s a much lighter build ,

Do do have an idea how much extra weight it adds on the head ? And do you get resonance to the frame from those big fans from the cooling bloque ?

I am asking cause for my build i opted for to have as light as possible heads since I use belt drives that is the reason I added the magnetic tool change in the 1st place , after tying many printer from my friend and so I figured out that ultimaker got very nice and repeatable prints , I figured that one of the reasons is the lightnest of the axes and head

I am not sure but from ur pictures looks like I using like a buildtalk on you table , I am wondering how good is the result with the dc42 ir prob ? I am using the prob and got a buildtalk plate but have not tried it yet
Best regards
MT

MrT

Dear dc42
merci ,thank you, dank u ,xie xie
I am back printing !!!
Best regards
MT

jacksat

=topiolm@hotmail.com]Dear Jacksat

Very nice build !!!!! I also thought of using a much bigger airpump , but the cost is fairly steep , after all the testing I did , using compressed air is best for small parts as for biger parts the fans work beter of course it depends how much air you have available , during my testing i managed to use both and that gives really nice result , quick cooling near the nozzle and even cooling on the rest of the part .

I really like you water cooling , you have a much heavier build then me , I used the same material as the ultimaker 2 to build my printer , it s a much lighter build ,

Do do have an idea how much extra weight it adds on the head ?

Hi,

At the moment i am rebuilding my printer from a dubbel direct driven printhead to 4 direct driven.
On 2 heads i had no troble to go up to 70 a 75mm/s printspeed befor the printer start to walk arround on the bench ,but i print mostly on 40 a 50 mm/s.
The printer is so sturdy(stupidly overbuild) that when you go at these high speeds you don't see anything in the prints ,so the wheight is not the limiting factor.
The dubbel head weight is arround :280g x2 for the pancakes ,80g x2 for the Titans and 200g -250g for the hotends, heatblocks and carriage.
So that's very, very heavy , and you can easely dubbel that for the 4 head version.All the above + 2 servo's.
If that becomes to heavy i will go to a 2 direct and 2 bowden setup.

I use syncroflex T2.5 belts instead of the cheap black T2 "rubberbands", that made a huge differce ,all steppers are monted on vibration dampers ,that makes the
printer a bit quieter ,but in my mind i think also less accurate on X and Y ,not a problem on the Z axes.

The watercooling does not add any extra weight i think comparred with a stock schroud and fan.(maybe the silicon hoses in the cable chain)
It is just 2 rubber O rings and a piece of polycabonte 30 mm/out 24 mm/inner dia tube and silicone hose.
Just it is alot quieter than fans.

And do you get resonance to the frame from those big fans from the cooling bloque ?

Don't really understand this one ,if you mean the fan on the radiator on the back off the printer ,then no.
This fan does not even need run on 2 heaterblocks ,i can keep the temp solid up to 235C on a 6 hour ABS print.
The coolant temp went up only by 12°C from ambiant temp ,but i will need the fan when i will run 4 heatblocks (i think).
But for the moment i get away with only 1 fan running and that is to cool the drivers on the Duet
It is a very silent Noctua NF-F12 industrialPPC-24V-3000 Q100 IP67 PWM

I am not sure but from ur pictures looks like I using like a buildtalk on you table , I am wondering how good is the result with the dc42 ir prob ? I am using the prob and got a buildtalk plate but have not tried it yet

Good eye ,no problems with the Buildtak and DC's probe.
Best regards
JS

MrT

Dear jacksat

Thanks for your nice explaination. very nice build, if I may ask one more question what slicer are you using ?

when checking your pictures , I say you have Belgian plate number what area you from ? If i may ask don't want to intrude I am from Antwerp

Best regards
MT

jacksat

@topiolm@hotmail.com:

Dear jacksat

Thanks for your nice explaination. very nice build, if I may ask one more question what slicer are you using ?

when checking your pictures , I say you have Belgian plate number what area you from ? If i may ask don't want to intrude I am from Antwerp

Best regards
MT

Hi,

I use mostly Slic3r prusa version these days ,before that i used S3D.
But i prefer Slic3r, don't know why,i think the GUI S3d looks old.

And yes i'm Belgian ,i live in Oostende.

Kind regards.
JS

dc42

I had the problems described in this thread on my list to review, and I've just spotted a firmware issue that could explain them. The fix will be in firmware 1.20beta9. Anyone still having problems with commands in tool change files not being executed properly, please try that version when I release it.

EDIT: the workaround for now is: if any of your tpre, tpost or tfree files contains a movement command, add command M400 at the end of that file.