Posts made by sozkan

@dc42 @jay_s_uk

Dear All,

I have prepared a G-code reversing tool Console App just to use to reverse motion of the print to be able to test best position to rewind Using C#. Please feel free to use the source if you find it useful. for sometimes it will help us to work manually. But I still required the tool DWC plugin or SBC function.

https://github.com/tunerji/G-Rewind

While reversing code, the code blocks sharing Feed rates and Z height had to be critically analyzed. That's why it was a bit important to review the code a few times.

However, it is temporary solution to me and we still need to make interface buttons and functions to run drive back and forward. The method might need to different way of capturing feed-rate and Z values in real time, once reversed, it use that one.

I assume if we tag all line having G0 and G1 commands as F and Z height of each X and Y coordinates, it might be easy to produce reversing go code in real time at the client side.

However like my case, if paused, machine might lift head and we might need to disable macros like pause and resume too.

as you may notice, in the g-code, Feed rate and Z coordinate has to be shifted and placed on right place, by respecting their groups like G0 and G1 and F Values. My reversing app does that if you like to inspired from there.

Example G-Code Snippet

G1 F1500 X0.00 Y-550 Z1.25  ; Move to coordinates with feed rate of 1500
G0 F4000 X0.00 Y0.00   ; Rapid move to new coordinates
G1 F1200 X476.314 Y275      ; Controlled move to new coordinates
G1 X0.00 Y0.00              ; Move back to start point

Reversed G-Code Example
After processing with G-Rewind, the G-code might look like this:

G1 X0.00 Y0.00 Z1.25        ; Move back to start point
G0 F4000 X0.00 Y0.00   ; Rapid move to previous coordinates
G1 F1500 X476.314 Y275 ; Controlled move to previous position
G1 X0.00 Y-550         ; Reverse move to the starting coordinates

By the way, we have one more financial support candidate a friend of mine, who is wiling to donate 500 too. Therefore, we have total 1000 budget for now. We might rise more who is required to use it like my case. If you may accept, besides my method or code generation support, I can handover you my portion donation. Please feel free to reach me in private talk about it.

@jay_s_uk

I assume, it is not easy. You know better, hardware and software capabilities of the DUET, I am suggesting as user.

In my past I had a CNC machine 20 years old, it could find the position of the tool head and allow me to start from there, where I lost the tool and changed. I assume it is easier way. I know that I suggested complicated way that, but would maybe better and professional.

What if I offer preparing python or C# script could handle the tasks and you help to fix in the DWC interface? We could even start with script. I believe I will start making reversing post processor for now and revers the sequence geode. and as soon as satisfied, I will mark the position and second task the skip the initial code.

We really need to take it serious and rise a fund for that job. It could save tons of plastic globally.

@sozkan said in Reverse G-Code sequence to find last successful point.:

@jay_s_uk

So then,

May I request a new future update in the interface?

I can donate personally 500 Euros and invite people for donation and contribution of this future for the successful result.

Because large print outs such as my case and many people plastics are wasted to restart!

We can save our pockets and environment!

Feature Request: Enhanced Print Recovery via G-code Reversal

Overview

I'm proposing a new feature in the Duet3D firmware that would significantly enhance the ability to recover and continue a print after an interruption, such as an extruder clog. Currently, large prints can be difficult to restart precisely due to changes in temperature and positioning. To mitigate wasted material and time, I suggest the implementation of a sophisticated print recovery method that allows for the reversal of G-code to a previous successful layer.

Step-by-Step Implementation:

Capture the Last Successful Coordinate

• Automated Capture: Implement a feature that periodically logs the current XYZ position and the corresponding G-code line number.
  Manual Capture: Allow users to manually record the position and line number where the print was paused or failed.
  Parse the G-code to Identify the Last Successful Point

• Load G-code: Develop a tool or script to load the G-code file.
  Identify Closest Match: The tool should search the G-code for the coordinates closest to the last successful position recorded.

Reverse the Movement Sequence

• Reverse Movements: From the identified point, reverse the sequence of movement commands (G1) while excluding extrusion commands.

• Negate Coordinates: Reverse the direction of each movement command to create a G-code sequence that retraces the printhead's path.
  Prepare the Reversed G-code

• Coordinate Adjustment: Negate the movements in the G-code while keeping extrusion (E) commands commented out.

• Temperature Control: Ensure that the nozzle temperature is maintained or adjusted as necessary during this process.
  Control Movement Speed

• Adjust Feed Rate: Modify the feed rate (F) within the G-code to control the speed of the reverse movements.
  Stepwise Reversal and Verification

• Small Steps: Execute the reversed G-code in small, controlled steps, allowing for manual verification of the printhead’s position.

• Adjustment: Provide the option to adjust the G-code or manually fine-tune the printhead’s position as needed.. Adjustment like step or multi step coordinate jump back and forward. Manual position changes should be in the adjustment as well. for some case motor step or temperature change effect on sensors.

Resume Printing

Prepare G-code: Once alignment is confirmed, prepare the original G-code to resume printing.

• Skip to the Last Successful Point: Start the G-code from the identified successful point, ensuring all settings (temperature, extrusion mode, etc.) are correct.

• Continue Printing: Execute the G-code to continue the print seamlessly.
  Conclusion

This feature would greatly enhance print recovery options in the event of a failure, saving both time and material, especially for large or complex prints. I believe this enhancement aligns well with the innovative spirit of the Duet3D community and could be a valuable addition to the firmware. I'm willing to contribute to the development process and even offer a personal donation of 1000 Euros to support this feature's implementation.

@jay_s_uk

So then,

May I request a new future update in the interface?

I can donate personally 500 Euros and invite people for donation and contribution of this future for the successful result.

Because large print outs such as my case and many people plastics are wasted to restart!

We can save our pockets and environment!

Hello,

Is there any ways to return back without canceling present print, just return back where last successful print, It is kind of semi-automated process.

1. We will manually jog and find the desired or last successful point.
2. A Script finds the closes coordinates
3. Allowing a test to continue from the point.
4. Back and forward tests.
5. Reverse or forward the sequence by steps.
6. Continue running.

It could even be possible if the job seems finished but machine had issues during the print. We simply run G code in reverse to find last successful place and forward.

Best Regards
Serkan

@deckingman Not exactly all parameters provided by Dyze, because heaters on the same blocks influence each other, tuning was quite difficult and took many trial and time to find those settings. As far as I remember earlier time PID tuning was not possible together. I had to run one by one. if it is possible, I would definitely run PID tuning if it is possible to run together. However, bangbang mode especially on heated bed works problem free.

@T3P3Tony said in Cold extrusion even the one of the pair disabled:

@sozkan ok yes so those are defined as i would expect them.

It may be an issue with the temperature sensor, is it always sensor 1 that that has the issue? Maybe look at temporarily switching them?

I thought about, but Dyze made everything very tight with their pcb connection, It is impossible. However I have replaced the block completely with new ones, still the same.

@T3P3Tony said in Cold extrusion even the one of the pair disabled:

@sozkan are you saying both extruder heaters are in the same heating block? If so how do you have the tool defined? normally you would have that as a single tool, with multiple heaters.

Yes, two heaters on the same block with individual driver and sensors tool defined as single one as the dyze design`s recommended parameters :

; HotEnd 
M308 S1 P"spi.cs1" Y"rtd-max31865"   		 	; configure sensor 1 as PT100 on pin e0temp
M308 S2 P"spi.cs2" Y"rtd-max31865"             ; configure sensor 2 as PT100 on pin e1temp
M950 H1 C"e0heat" T1                               ; create nozzle heater output on e0heat and map it to sensor 1
M950 H2 C"e1heat" T2                               ; create nozzle heater output on e1heat and map it to sensor 2


M307 H1 B1 S0.36 ; 2.5 mm;
M307 H2 B1 S0.51 ; 2.5 mm nozzle 
 
M570 H0 P5 T5 S240 ; H heater, T tolerance degree, S second persist.
M570 H1 P30 T30 S10 ; H heater, T tolerance degree, S second persist.
M570 H2 P30 T30 S10 ; H heater, T tolerance degree, S second persist.

M143 H1 S480 ;Set temperature max temp:
M143 H2 S480 ;Set temperature max temp:

M106 P0 I-1   ; no fan so disable it.

; Tools
M563 P0 D0 H2:1 F0 S"Typhoon"                      ; define tool 0
G10 P0 X0 Y0 Z0                                    ; set tool 0 axis offsets
G10 P0 R0 S0 

@jay_s_uk said in Cold extrusion even the one of the pair disabled:

Thanks, I will definitely try this one. However, can I backup the existing firmware?

Dear Duet3D Community members and Admins,

Thankfully, I have successfully build and tuned nearly perfect situation a one of the largest Delta 3D printer with your contribution. We will build much more larger ones that present built was really good experience for us.

However, rarely having an issue in the middle of the printing one of the heater on the same heating core goes on emergency mode, eighter extreme or low temperature warning. But it is causing a big waste if happens.

The extreme temperature error as you may see in the photo that suddenly appearing and it disabling the one of the heater and continue printing with one! Since two heater set and specified in the config as single heating core it is cousing stress on the extruder motor and crushed the gears.

As you can see second heater shown 159.5 and printing stopped! Despite the one heater disabled, before stopping it, it continue printing until the second sensors shown low temperature. However during this time problems caused cold extrusion, drive gears and motors stress and and failed print output and few hundreds euros damage and several kilo waste of materials.

Things could be better if possible in firmware:

• I believe, printing should be paused if one of them disabled ( presently it is waiting for the low temperature warning triggering to pause it)
• Longer timeout before disabling the heater.

Could you please recommend what could be solutions to fix and avoid this problem?

• The heating Block DyzeDesign Typhoon
• Board: Duet 2 WiFi (2WiFi)
• Duet Web Control 3.4.0-b7
• Firmware: RepRapFirmware for Duet 2 WiFi/Ethernet 3.4.0beta7 (2021-12-16)
  Duet WiFi Server Version: 1.26

I still could not figure out why it is happening!
I have tried:

• Plug in and out connectors,
• Changed the cables,
• Replaced the hot ends
• Opened and tightened the sensors on the heating core.

@sozkan said in Disabling Heater Fault:

RepRapFirmware
Hi,
This is the latest firmware. I have just updated.

Heater 2 strangely spiked 480 celsius. have noticed this on long moves. Is this not sensing sensors during the long gCode interpretation?

Hello again,
During the 30 hours printing again fault again. it says Heater 0 fault, but a fault on Heater 1. Temperature drop to 203 at Heater 1 with no sense.
I don't understand, The fault detection can sense temperature from the sensor! The algorithm cannot activate power on time!

Isn't it possible to re-test and continue printing later?

Hello All,

I just wanted to share the successful result:

https://youtu.be/2QNMd8EgGmQ

The only target for me right now is filament detection. 5 kg filament consumed so fast with no sense.

@deckingman said in Disabling Heater Fault:

@sozkan What you describe is close to how bang-bang mode works. So you could try that instead of using PID (change the "B" parameter). It might be a better option for that particular hot end.

But IMO Dyze design are ducking their responsibilities by producing a device and not suggesting how to control it. The PID algorithm works fine for every other hot end - even those with multiple heaters ( as in my case using two heat zones, each with an 80 watt heater cartridge).

Hi, thank you for the all suggestions. I am very grateful. I simply change to B1 and a bit play with PWM. Now it works much more stable. Thankfully, the problem is solved now.

By the way, the 2.5 mm of dyze hot end seems cracked due to my last year's hard start trial. They say the resin-like leakage seems like occurred from the crack. Therefore, they suggest for return for repair.

Isn't that possible with an algorithm to play for real-time tuning?
like measured time 5 seconds at present or during the printing:

• Within the cut-in and cut-off, if more than 5 seconds simply duty cycle activates for longer duty until continues power.
• if heating is shorter than 5 seconds, the duty cycle lowers the power.
• If the temperature is out of the limit is due to the thermal expansions, lower the duty or cut-off thresholds.
  

It seems simple and more functioning to me. Measured time is even completely unnecessary. if the temperature goes out of the limit, simply auto-decrease the duty cycle.

@deckingman said in Disabling Heater Fault:

@sozkan said in Disabling Heater Fault:

These autotune things seem not really effective in my case.

Any suggestion?

Having seen those pictures, it is clear that the Dyze heater elements are unlike the usual cartridge heaters that the majority of people use. Perhaps there is some characteristic about their design which means the normal PID tuning algorithms will not work well. Have you contacted Dyze designs and asked them if they have any suggestions or recommendations?

Yes, clearly they say they don"t provide software support and suggested PID tunning and find a solution here!

However, I still believe there is another way to run. But, not a time-dependent algorithm. That's one of the problems! Since we have sensors and all advanced PWM things what's the point of PID tuning checking for death and uptime! Even the tuning was working on pause days when started printing failed. It is fact that plastics flow even act like cooling and ruin the PID tuning setup and safety margins start activates! It seems it is not seen on hobby-grade printers.

I start thinking to develop an external controller independent from the Duet hardware for this hotend. For that purpose, An extra Arduino controller, a new algorithm just for heater controlling purposes!

1- if the temperature of each individual heater is below the threshold, keep powering the heater until the cut-off threshold temp.
2- temperature start expands, it does not matter in 1 second or 100 seconds.
3- If the temperature rises more than expected, the decision for duty cycle or decrease the cut-off threshold temperature slightly. Which could be even real-time tuning!

Isn't that possible with the present controller?

I have installed another new hotend smaller diameter,
However this time H1 could not be calibrated at all. It took 3-4 hours to see the failed result. I think this time is not a technical issue. Somehow H2 is seemed okay. But, M303 H1 S220 result with failure:

Warning: heater behaviour was not consistent during tuning
Warning: Auto tune of heater 1 failed due to bad curve fit (R=5.433 K=4.650:0.000 D=14.85)

These autotune things seem not really effective in my case.

Any suggestion?

@deckingman said in Disabling Heater Fault:

@sozkan I'm wondering if there is/was a bad joint somewhere - either a bad crimp contact or a poorly soldered connector. It would explain why you were getting seemingly intermittent faults. Using a low PWM value would have effectively reduced the current and prevented the arcing which then occurred when you raised the PWM value and which caused the overheating at the connector.

Thank you for the follow-up. Sorry for the late reply, I was Covid-2 Positive.

here are some photos I have taken and a video: https://youtu.be/RfPiJUVuZdE.

It is well and durable built. But however, strangely there is some resin-like leakage. But I don't think that the fume source is the resin burn.

@phaedrux The damage is not really visible. But possibly weak connector clamping caused a weak spot on the wire to warm up. When I connect again and try again smoking! Therefore, the plastic insulation where below glass-fiber sleeves of resistant wiring seem burned away. But the glass-fiber sleeve is still in good condition. I have only seen ashes on connectors and heavy smoke dirtiness. It was flying while on disassembly, so I cleaned up a bit.

However, these hot-ends are designed above 400 celsius. Even the PWM 1.00, was not above 280 on the peak temperature for a few seconds. I don`t think something to todo with settings.

I regret now that I did not snap a photo before the ash cleanup but I will try to get better images to share tomorrow.

@phaedrux said in Disabling Heater Fault:

@sozkan said in Disabling Heater Fault:

I have tested all versions of the firmware now I am at the latest one.

Can you test this beta version as well?

https://www.dropbox.com/sh/i5vox3xmkd55gaz/AAC19mI0WEC5GmEjLOBRbKs-a?dl=0

Sure I have just updated. But now I gues the hothend H1 connector has burn! Because previously the smoke smell source has hotend PCB connector.