Prusaslicer: Temperature-commands while changing filament with MMU

[DanyboyInSwitzerland]

Version

1.41.0+win64

Operating system type + version

Windows 10 Pro, Version 1803, build 17134.345

Behavior

When using multiple materials with the MMU, at filament-change, the new temperature is instantly selected upon starting the filament-change.
Instead, the current temperature for the filament which is getting unloaded should be maintained until it has been ejected (detection by FINDA). Then the new filament should be loaded into extruder and only then, when the MMU has finished loading and the extruder takes over, then the new temperature should be selected and waited for - M104 followed by the appropiate M109 command.

_Is this a new feature request?_

[lukasmatena]

I don't have much experience with printing multiple materials with different temperatures. Could you please explain benefits of the proposed solution? As far as I know, the filament is out of the nozzle the whole time, so I don't see why not to use the time required for the filament change for changing the extruder temperature for the next one. But like I said, there may be something I don't see.

[DanyboyInSwitzerland]

Hello Lukas,
Well, during the filament-change, the filament may be out of the nozzle but is still in the hotend. It's getting inserted and retracted slowly two times before getting ejected. This is in order to get a consistent pointy shape that feeds best through the system.
The problem gets visible if you try to print with two filaments that require considerable different temperatures. Like PVA. If I can print PVA at 180°C, I will get a consistent sharp point. If I go with anything higher, it will get more and more stringy. If it gets retracted at the temperature of PLA, the result of retraction gets completely unreliable - sometimes a nice end, sometimes the end gets stretched and elongated. Once I had about 5cm of filament with only half the diameter. This inconsistency cannot be handled by the MMU2.
I was able to print PVA and PLA, PVA at 180°C and PLA at 215°C. Seemed fine. But if the nozzle starts heating from PVA to PLA while still forming the tip of PVA, the result is as unreliable as if I printed the PVA with the temperature of PLA to start out with.I managed to manually tune the temperature and thus keeping it low while forming the tip and the results were promising if not nice... Quite fiddly though to achieve it - I only made it through 5 changes... :D But without this tuning, filament-change usually failed on the third change, the second time from PVA to PLA.
Heating from 170°C to 215°C takes a mere seconds, definitely way less than it takes to retract the filament, move the selector and then insert the new filament. So if the temperatures are switched at the same time as the MMU2 starts retracting, the nozzle will definitely be up to temp once the new filament arrives at the extruder.By 'the MMU2 starts retracting' I do not mean when the retraction is initiated - that would also include the forming of the pointy tip - I mean the action where the filament is no more moved by the extruder but pulled by the MMU through the bowden-tube.
The other approach that would help printing PVA with the MMU2 is to have different procedures to form the pointy tip, depending on the type of material.I guess that would be a lot more complex than my proposal... ;)
Cheers,Daniel
-------- Ursprüngliche Nachricht --------Von: lukasmatena [e-mail] Datum: 18.10.18 23:17 (GMT+01:00) An: prusa3d/Slic3r [e-mail] Cc: DanyboyInSwitzerland [e-mail], Author [e-mail] Betreff: Re: [prusa3d/Slic3r] Temperature-commands while changing filament
  with MMU (#1339)
I don't have much experience with printing multiple materials with different temperatures. Could you please explain benefits of the proposed solution? As far as I know, the filament is out of the nozzle the whole time, so I don't see why not to use the time required for the filament change for changing the extruder temperature for the next one. But like I said, there may be something I don't see.

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub, or mute the thread.
{"api_version":"1.0","publisher":{"api_key":"05dde50f1d1a384dd78767c55493e4bb","name":"GitHub"},"entity":{"external_key":"github/prusa3d/Slic3r","title":"prusa3d/Slic3r","subtitle":"GitHub repository","main_image_url":"https://assets-cdn.github.com/images/email/message_cards/header.png","avatar_image_url":"https://assets-cdn.github.com/images/email/message_cards/avatar.png","action":{"name":"Open in GitHub","url":"https://github.com/prusa3d/Slic3r"}},"updates":{"snippets":[{"icon":"PERSON","message":"@lukasmatena in #1339: I don't have much experience with printing multiple materials with different temperatures. Could you please explain benefits of the proposed solution? As far as I know, the filament is out of the nozzle the whole time, so I don't see why not to use the time required for the filament change for changing the extruder temperature for the next one. But like I said, there may be something I don't see."}],"action":{"name":"View Issue","url":"https://github.com/prusa3d/Slic3r/issues/1339#issuecomment-431165356"}}}
[
{
"@context": "http://schema.org",
"@type": "EmailMessage",
"potentialAction": {
"@type": "ViewAction",
"target": "https://github.com/prusa3d/Slic3r/issues/1339#issuecomment-431165356",
"url": "https://github.com/prusa3d/Slic3r/issues/1339#issuecomment-431165356",
"name": "View Issue"
},
"description": "View this Issue on GitHub",
"publisher": {
"@type": "Organization",
"name": "GitHub",
"url": "https://github.com"
}
},
{
"@type": "MessageCard",
"@context": "http://schema.org/extensions",
"hideOriginalBody": "false",
"originator": "AF6C5A86-E920-430C-9C59-A73278B5EFEB",
"title": "Re: [prusa3d/Slic3r] Temperature-commands while changing filament with MMU (#1339)",
"sections": [
{
"text": "",
"activityTitle": "lukasmatena",
"activityImage": "https://assets-cdn.github.com/images/email/message_cards/avatar.png",
"activitySubtitle": "@lukasmatena",
"facts": [

]
}
],
"potentialAction": [
{
"name": "Add a comment",
"@type": "ActionCard",
"inputs": [
{
"isMultiLine": true,
"@type": "TextInput",
"id": "IssueComment",
"isRequired": false
}
],
"actions": [
{
"name": "Comment",
"@type": "HttpPOST",
"target": "https://api.github.com",
"body": "{\n\"commandName\": \"IssueComment\",\n\"repositoryFullName\": \"prusa3d/Slic3r\",\n\"issueId\": 1339,\n\"IssueComment\": \"{{IssueComment.value}}\"\n}"
}
]
},
{
"name": "Close issue",
"@type": "HttpPOST",
"target": "https://api.github.com",
"body": "{\n\"commandName\": \"IssueClose\",\n\"repositoryFullName\": \"prusa3d/Slic3r\",\n\"issueId\": 1339\n}"
},
{
"targets": [
{
"os": "default",
"uri": "https://github.com/prusa3d/Slic3r/issues/1339#issuecomment-431165356"
}
],
"@type": "OpenUri",
"name": "View on GitHub"
},
{
"name": "Unsubscribe",
"@type": "HttpPOST",
"target": "https://api.github.com",
"body": "{\n\"commandName\": \"MuteNotification\",\n\"threadId\": 397882037\n}"
}
],
"themeColor": "26292E"
}
]

[lukasmatena]

I understand the problem you are having, but I don't think that timing of the temperature change is the cause. This is what happens during toolchange, with some added comments:

.....
G1 X220.018 E0.3269
G1 F5526
G1 X197.010 E0.8147
G1 F5912
G1 X172.391 E0.8717 // this was the last ramming extrusion
G1 X170.750 E-15.0000 F5010
G1 X172.391 E-14.0000 F1208
G1 X170.750 E-4.0000 F649
G1 X172.391 E-2.0000 F466 // filament is now retracted 35mm from the nozzle. The tip should be formed by now.
M104 S205 // TEMPERATURE CHANGE
G1 Y129.030
G1 X229.250 E10.0000 F693 // 10 mm forward (to allow to cool a bit,) takes 5 s
G1 X172.391 E-10.0000 F346 // 10 mm backward, takes 10 s
G1 E-50.0000 F2000 // further 50mm of retraction
G1 Y128.860 F2400
G4 S0
M106 S255
T0 // MMU2 takes control
...

So - when the temperature change command is issued, the filament tip is already 35mm "uptube", and it won't get closer than 25mm during the cooling sequence. What is more important, the tip is no longer in contact with anything during the cooling sequence - the purpose of it is just to let the filament cool. So whatever temperature, the shape of the tip should not change. I cannot imagine that issuing the command later would change anything.

You mention that you improved the behaviour by tuning the filament temperature. It is quite possible, but it is something different than you request. Whatever temperature you tune is overridden at exactly the same spot as without tuning - at the M104 command in the gcode above. You would have to move the M104 just before the T command to prove it helps (the reason it is issued this early is simple: MMU1).

In my opinion, Slic3r behaves fine - you should just lower the temperature in Filament settings a bit. Different filaments from different manufacturers can behave a bit differently and the default values may not fit them all. This is more likely to be the cause of the problem, as far as I understand it.

The other approach that would help printing PVA with the MMU2 is to have different procedures to form the pointy tip, depending on the type of material.I guess that would be a lot more complex than my proposal... ;)

Actually, this is already possible in current Slic3rPE, at least to some extent. The details of the filament unload sequence are configurable in Filament settings->Advanced (from our experience, it is the ramming sequence that mostly influences the tip shape). The position of the cooling tubes etc. is configurable in Printer Settings->Single Extruder MM setup. The numbers in the gcode above are calculated from these config values, so you can experiment with them all.

[DanyboyInSwitzerland]

Thank you very much for taking the time and effort!
Most interesting, you are right, according to the code, the filament is indeed quite some distance from the nozzle....which leaves me wondering......obviously it was a coincidence that I was lucky to achieve my result...headscratch
...well...
Ok, I was not aware that the ramming was defined in the filament-section - I guess I will fiddle around with those figures... :)
I already did lower the temperature quite some - down to 180. Since the provided profile for the PVA sets 205°C, and I'm using precisely that filament, I guess I'm already maxed out on that end, which is why I'm looking in other directions...
Again, thank you for your help!
Cheers,Daniel
-------- Ursprüngliche Nachricht --------Von: lukasmatena [e-mail] Datum: 22.10.18 11:34 (GMT+01:00) An: prusa3d/Slic3r [e-mail] Cc: DanyboyInSwitzerland [e-mail], Author [e-mail] Betreff: Re: [prusa3d/Slic3r] Temperature-commands while changing filament
  with MMU (#1339)
I understand the problem you are having, but I don't think that timing of the temperature change is the cause. This is what happens during toolchange, with some added comments:
.....

G1 X220.018 E0.3269

G1 F5526

G1 X197.010 E0.8147

G1 F5912

G1 X172.391 E0.8717 // this was the last ramming extrusion

G1 X170.750 E-15.0000 F5010

G1 X172.391 E-14.0000 F1208

G1 X170.750 E-4.0000 F649

G1 X172.391 E-2.0000 F466 // filament is now retracted 35mm from the nozzle. The tip should be formed by now.

M104 S205 // TEMPERATURE CHANGE

G1 Y129.030

G1 X229.250 E10.0000 F693 // 10 mm forward (to allow to cool a bit,) takes 5 s

G1 X172.391 E-10.0000 F346 // 10 mm backward, takes 10 s

G1 E-50.0000 F2000 // further 50mm of retraction

G1 Y128.860 F2400

G4 S0

M106 S255

T0 // MMU2 takes control

...
So - when the temperature change command is issued, the filament tip is already 35mm "uptube", and it won't get closer than 25mm during the cooling sequence. What is more important, the tip is no longer in contact with anything during the cooling sequence - the purpose of it is just to let the filament cool. So whatever temperature, the shape of the tip should not change. I cannot imagine that issuing the command later would change anything.
You mention that you improved the behaviour by tuning the filament temperature. It is quite possible, but it is something different than you request. Whatever temperature you tune is overridden at exactly the same spot as without tuning - at the M104 command in the gcode above. You would have to move the M104 just before the T command to prove it helps (the reason it is issued this early is simple: MMU1).
In my opinion, Slic3r behaves fine - you should just lower the temperature in Filament settings a bit. Different filaments from different manufacturers can behave a bit differently and the default values may not fit them all. This is more likely to be the cause of the problem, as far as I understand it.

The other approach that would help printing PVA with the MMU2 is to have different procedures to form the pointy tip, depending on the type of material.I guess that would be a lot more complex than my proposal... ;)

Actually, this is already possible in current Slic3rPE, at least to some extent. The details of the filament unload sequence are configurable in Filament settings->Advanced (from our experience, it is the ramming sequence that mostly influences the tip shape). The position of the cooling tubes etc. is configurable in Printer Settings->Single Extruder MM setup. The numbers in the gcode above are calculated from these config values, so you can experiment with them all.

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub, or mute the thread.
{"api_version":"1.0","publisher":{"api_key":"05dde50f1d1a384dd78767c55493e4bb","name":"GitHub"},"entity":{"external_key":"github/prusa3d/Slic3r","title":"prusa3d/Slic3r","subtitle":"GitHub repository","main_image_url":"https://assets-cdn.github.com/images/email/message_cards/header.png","avatar_image_url":"https://assets-cdn.github.com/images/email/message_cards/avatar.png","action":{"name":"Open in GitHub","url":"https://github.com/prusa3d/Slic3r"}},"updates":{"snippets":[{"icon":"PERSON","message":"@lukasmatena in #1339: I understand the problem you are having, but I don't think that timing of the temperature change is the cause. This is what happens during toolchange, with some added comments:\r\n\r\n.....\r\nG1 X220.018 E0.3269\r\nG1 F5526\r\nG1 X197.010 E0.8147\r\nG1 F5912\r\nG1 X172.391 E0.8717 // this was the last ramming extrusion\r\nG1 X170.750 E-15.0000 F5010\r\nG1 X172.391 E-14.0000 F1208\r\nG1 X170.750 E-4.0000 F649\r\nG1 X172.391 E-2.0000 F466 // filament is now retracted 35mm from the nozzle. The tip should be formed by now.\r\nM104 S205 // TEMPERATURE CHANGE\r\nG1 Y129.030\r\nG1 X229.250 E10.0000 F693 // 10 mm forward (to allow to cool a bit,) takes 5 s\r\nG1 X172.391 E-10.0000 F346 // 10 mm backward, takes 10 s\r\nG1 E-50.0000 F2000 // further 50mm of retraction\r\nG1 Y128.860 F2400\r\nG4 S0\r\nM106 S255\r\nT0 // MMU2 takes control\r\n...\r\n\r\n\r\nSo - when the temperature change command is issued, the filament tip is already 35mm \"uptube\", and it won't get closer than 25mm during the cooling sequence. What is more important, the tip is no longer in contact with anything during the cooling sequence - the purpose of it is just to let the filament cool. So whatever temperature, the shape of the tip should not change. I cannot imagine that issuing the command later would change anything.\r\n\r\nYou mention that you improved the behaviour by tuning the filament temperature. It is quite possible, but it is something different than you request. Whatever temperature you tune is overridden at exactly the same spot as without tuning - at the M104 command in the gcode above. You would have to move the M104 just before the T command to prove it helps (the reason it is issued this early is simple: MMU1).\r\n\r\nIn my opinion, Slic3r behaves fine - you should just lower the temperature in Filament settings a bit. Different filaments from different manufacturers can behave a bit differently and the default values may not fit them all. This is more likely to be the cause of the problem, as far as I understand it.\r\n\r\n\r\n\u003e The other approach that would help printing PVA with the MMU2 is to have different procedures to form the pointy tip, depending on the type of material.I guess that would be a lot more complex than my proposal... ;)\r\n\r\nActually, this is already possible in current Slic3rPE, at least to some extent. The details of the filament unload sequence are configurable in Filament settings-\u003eAdvanced (from our experience, it is the ramming sequence that mostly influences the tip shape). The position of the cooling tubes etc. is configurable in Printer Settings-\u003eSingle Extruder MM setup. The numbers in the gcode above are calculated from these config values, so you can experiment with them all."}],"action":{"name":"View Issue","url":"https://github.com/prusa3d/Slic3r/issues/1339#issuecomment-431777896"}}}
[
{
"@context": "http://schema.org",
"@type": "EmailMessage",
"potentialAction": {
"@type": "ViewAction",
"target": "https://github.com/prusa3d/Slic3r/issues/1339#issuecomment-431777896",
"url": "https://github.com/prusa3d/Slic3r/issues/1339#issuecomment-431777896",
"name": "View Issue"
},
"description": "View this Issue on GitHub",
"publisher": {
"@type": "Organization",
"name": "GitHub",
"url": "https://github.com"
}
},
{
"@type": "MessageCard",
"@context": "http://schema.org/extensions",
"hideOriginalBody": "false",
"originator": "AF6C5A86-E920-430C-9C59-A73278B5EFEB",
"title": "Re: [prusa3d/Slic3r] Temperature-commands while changing filament with MMU (#1339)",
"sections": [
{
"text": "",
"activityTitle": "lukasmatena",
"activityImage": "https://assets-cdn.github.com/images/email/message_cards/avatar.png",
"activitySubtitle": "@lukasmatena",
"facts": [

]
}
],
"potentialAction": [
{
"name": "Add a comment",
"@type": "ActionCard",
"inputs": [
{
"isMultiLine": true,
"@type": "TextInput",
"id": "IssueComment",
"isRequired": false
}
],
"actions": [
{
"name": "Comment",
"@type": "HttpPOST",
"target": "https://api.github.com",
"body": "{\n\"commandName\": \"IssueComment\",\n\"repositoryFullName\": \"prusa3d/Slic3r\",\n\"issueId\": 1339,\n\"IssueComment\": \"{{IssueComment.value}}\"\n}"
}
]
},
{
"name": "Close issue",
"@type": "HttpPOST",
"target": "https://api.github.com",
"body": "{\n\"commandName\": \"IssueClose\",\n\"repositoryFullName\": \"prusa3d/Slic3r\",\n\"issueId\": 1339\n}"
},
{
"targets": [
{
"os": "default",
"uri": "https://github.com/prusa3d/Slic3r/issues/1339#issuecomment-431777896"
}
],
"@type": "OpenUri",
"name": "View on GitHub"
},
{
"name": "Unsubscribe",
"@type": "HttpPOST",
"target": "https://api.github.com",
"body": "{\n\"commandName\": \"MuteNotification\",\n\"threadId\": 397882037\n}"
}
],
"themeColor": "26292E"
}
]

[DanyboyInSwitzerland]

ok, so this is awesome...!

 

Thanks to you I am now already into my tenth layer (meaning twenty changes) and all works flawlessly...!

 

All I did was duplicating the settings from the stock Prusa PLA and only changed the number of cooling moves from 4 to 1. So far, always a nice consistently shaped filament-end, sometimes just a tiny short hair on it...

THANK YOU!!! :)

Cheers,
Daniel

 

Gesendet: Montag, 22. Oktober 2018 um 11:34 Uhr
Von: lukasmatena [e-mail]
An: prusa3d/Slic3r [e-mail]
Cc: DanyboyInSwitzerland [e-mail], Author [e-mail]
Betreff: Re: [prusa3d/Slic3r] Temperature-commands while changing filament with MMU (#1339)

I understand the problem you are having, but I don't think that timing of the temperature change is the cause. This is what happens during toolchange, with some added comments:

.....
G1 X220.018 E0.3269
G1 F5526
G1 X197.010 E0.8147
G1 F5912
G1 X172.391 E0.8717 // this was the last ramming extrusion
G1 X170.750 E-15.0000 F5010
G1 X172.391 E-14.0000 F1208
G1 X170.750 E-4.0000 F649
G1 X172.391 E-2.0000 F466 // filament is now retracted 35mm from the nozzle. The tip should be formed by now.
M104 S205 // TEMPERATURE CHANGE
G1 Y129.030
G1 X229.250 E10.0000 F693 // 10 mm forward (to allow to cool a bit,) takes 5 s
G1 X172.391 E-10.0000 F346 // 10 mm backward, takes 10 s
G1 E-50.0000 F2000 // further 50mm of retraction
G1 Y128.860 F2400
G4 S0
M106 S255
T0 // MMU2 takes control
...

So - when the temperature change command is issued, the filament tip is already 35mm "uptube", and it won't get closer than 25mm during the cooling sequence. What is more important, the tip is no longer in contact with anything during the cooling sequence - the purpose of it is just to let the filament cool. So whatever temperature, the shape of the tip should not change. I cannot imagine that issuing the command later would change anything.

You mention that you improved the behaviour by tuning the filament temperature. It is quite possible, but it is something different than you request. Whatever temperature you tune is overridden at exactly the same spot as without tuning - at the M104 command in the gcode above. You would have to move the M104 just before the T command to prove it helps (the reason it is issued this early is simple: MMU1).

In my opinion, Slic3r behaves fine - you should just lower the temperature in Filament settings a bit. Different filaments from different manufacturers can behave a bit differently and the default values may not fit them all. This is more likely to be the cause of the problem, as far as I understand it.

The other approach that would help printing PVA with the MMU2 is to have different procedures to form the pointy tip, depending on the type of material.I guess that would be a lot more complex than my proposal... ;)

Actually, this is already possible in current Slic3rPE, at least to some extent. The details of the filament unload sequence are configurable in Filament settings->Advanced (from our experience, it is the ramming sequence that mostly influences the tip shape). The position of the cooling tubes etc. is configurable in Printer Settings->Single Extruder MM setup. The numbers in the gcode above are calculated from these config values, so you can experiment with them all.

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub, or mute the thread.

[lukasmatena]

I'm glad to hear that. The whole unload sequence is unfortunately quite sensitive to material properties, which is why the stock profiles might not give the best results with your specific filament. That's actually the reason why we exposed all those parameters to the user.

I'm closing this issue, because Slic3r apparently does what it is supposed to. You can reopen it in future if you feel the need.

[lukasmatena]

I'll reopen the issue myself. The premature temperature change could be a result of gcode buffering and inadequate commands synchronization on the firmware side (Prusa i3 printers). A simple workaround is adding a G4 command before the M104, which forces synchronization. I have made an experimental Slic3rPE build that does just that. You can download the binaries here: https://drive.google.com/file/d/1Bz8ghiIMpzccv7ehyeetXzmfJI9jkCF4/view?usp=sharing

I'd like to ask everyone experiencing issues with premature temperature changes to test this build and provide feedback. Our testers can obviously test it, but an input from someone who actually suffers from the issue is more valuable. Thanks in advance.

[Builderdad]

lukasmatena The experimental Slic3rPE build you made corrected some of the problems with temperature changes during filament switching. The temperature change is occurring much slower. This is preserving the tips during retraction.
There are still problems with it in that the temperature cools down to the new set point before the new filament purge happens. This is causing purge block issues. The left-over hotter filament is not purging properly thru the colder nozzle. This is creating blobs,skips and adhesion issues at the purge block which is causing print failures.
The current test I have been running is with a 30c difference between materials. Have had 3 failures so far and am trying different setups after each one. Currently am slowly increasing the colder temperature filament towards the hotter to see if I can find a workable transition point.
Thank you very much for the troubleshooting and work you did.

[workinghard]

I was printing two different PLA brands 220C and 200C. With the release version I was able to see the temperature change during the ramming procedure. This was causing issues in both ways (cold to hot and hot to cold). So it must be buffering/synchro issue.

I would suggest to set the new temperature dependent on the transition. For example two filaments 250C and 200C

1. From hot to cold (250C -> 200C). This can be set after filament load. So during wipe movement it cools down anyway and old 250C filament is being purged hot.
2. From cold to hot (200C -> 250C). This can be set after filament unload. We could use the loading time to heat up the nozzle.

In both cases ramming and cooling would happen at the filament temp. Wiping would happen hot which should improve the change quality

[Mirarkitty]

I agree. There needs to be three temperature changes when switching filament to make this great:

1) temperature for retracting to form the tip. There should be an option to set this to have it at a lower temperature than printing.

2) temperature for purging the old filament. This should be set to the highest temperature of the two filaments. (Ie, ABS to PVA or PVA to ABS, it should be the ABS temperature.)

3) then temperature for printing and possibly a little more wipe to avoid the hot stringing.

[workinghard]

I've created a python script which has following logic and is currently under heavy testing:

Cold (200C) ==> Hot (255C)

Ram/cool
-> Stay cool, do nothing. If needed, set lower temp but don't wait
Unload filament
-> Set hot. We can heating up while loading. Save some time
Load filament
-> Nozzle might still warming up. Load to nozzle for smooth loading process
Purge
-> Before start purging, wait for destination temp
Print
-> We are printing with the stabilized temp. No further intervention required

Hot (255C) ==> Cold (200C)

Ram/cool
-> We need to stay hot because hot filament is still in the nozzle. If needed, set lower temp but don't wait
Unload
-> Stay hot, do nothing
Load filament
-> Stay hot, do nothing. Load to nozzle for smooth loading process
Purge
-> Before start purging, set cool temp. We can cool down during the purging process
Print
-> Before start to print, wait for destination temp. Most likely temp will bounce pretty hard

Here are the details in the forum

[workinghard]

Any chance this topic will get some priority? Let me know if anything is needed?

[VanessaE]

This affects more than just the MMU -- it also impacts simpler setups like two extruders feeding one hotend via a splitter. As a workaround in the meantime, put this at the top of your tool change g-code:

{if current_extruder == 0}
    M104 T0 S[temperature_0]
    M104 T1 S[temperature_1]
{elsif current_extruder == 1 }
    M104 T1 S[temperature_1]
    M104 T0 S[temperature_0]
{endif}

That's only enough for two extruders, but it'd be trivial to extend it to 5 or however many the MMU has. Of course, this won't address the issue of having the proper temperature during the purge, but at least it'll be correct for post-purge printing with the selected tool.

[mbordoni]

I understand the problem you are having, but I don't think that timing of the temperature change is the cause. This is what happens during toolchange, with some added comments:

.....
G1 X220.018 E0.3269
G1 F5526
G1 X197.010 E0.8147
G1 F5912
G1 X172.391 E0.8717 // this was the last ramming extrusion
G1 X170.750 E-15.0000 F5010
G1 X172.391 E-14.0000 F1208
G1 X170.750 E-4.0000 F649
G1 X172.391 E-2.0000 F466 // filament is now retracted 35mm from the nozzle. The tip should be formed by now.
M104 S205 // TEMPERATURE CHANGE
G1 Y129.030
G1 X229.250 E10.0000 F693 // 10 mm forward (to allow to cool a bit,) takes 5 s
G1 X172.391 E-10.0000 F346 // 10 mm backward, takes 10 s
G1 E-50.0000 F2000 // further 50mm of retraction
G1 Y128.860 F2400
G4 S0
M106 S255
T0 // MMU2 takes control
...

So - when the temperature change command is issued, the filament tip is already 35mm "uptube", and it won't get closer than 25mm during the cooling sequence. What is more important, the tip is no longer in contact with anything during the cooling sequence - the purpose of it is just to let the filament cool. So whatever temperature, the shape of the tip should not change. I cannot imagine that issuing the command later would change anything.

You mention that you improved the behaviour by tuning the filament temperature. It is quite possible, but it is something different than you request. Whatever temperature you tune is overridden at exactly the same spot as without tuning - at the M104 command in the gcode above. You would have to move the M104 just before the T command to prove it helps (the reason it is issued this early is simple: MMU1).

In my opinion, Slic3r behaves fine - you should just lower the temperature in Filament settings a bit. Different filaments from different manufacturers can behave a bit differently and the default values may not fit them all. This is more likely to be the cause of the problem, as far as I understand it.

The other approach that would help printing PVA with the MMU2 is to have different procedures to form the pointy tip, depending on the type of material.I guess that would be a lot more complex than my proposal... ;)

Actually, this is already possible in current Slic3rPE, at least to some extent. The details of the filament unload sequence are configurable in Filament settings->Advanced (from our experience, it is the ramming sequence that mostly influences the tip shape). The position of the cooling tubes etc. is configurable in Printer Settings->Single Extruder MM setup. The numbers in the gcode above are calculated from these config values, so you can experiment with them all.

@lukasmatena good point but temperature for having a good tip is rather different than the printing (specially for pva). It is a pity to have to set the tipping temp also for printing.
I would propose two solutions:
1- to define and set a tipping temp per filament so that the temperature could be changed (with m109) at the very beginning of filament unload (before ramming).
or
2- to add an event in print configuration to inject some gcode before ramming (in this case a temp tuning could be set in a parametric way).