I ended up installing CuteCom on my linux Desktop and followed the instruction. Configured the Duet IP to DHCP, which didn't worked from windows(I don't know why).
With linux, it worked like a charm.
Now I can connect to the DWC from the linux Desktop, windows laptop or via IPad Wifi.
Ok, I did some G32 (4 time), here are the consol result
I homed, then send a G32 from DWC.
(mean, deviation)
before (-0.737 , 0.437) after (-0.000 , 0.014)
(-0.001 , 0.019) after (0.000 , 0.015)
(-0.009 , 0.012) after ( -0.000 , 0.012)
I homed again, then send an other G32.
(-0.018 , 0.011) after (-0.000 , 0.009)
Then I send a G29 (61 points mesh)
min error -0.052
max error 0.043
mean -0.010
deviation 0.022
Also, in config-override:
M665 L(I set manually) R... H... B... X-0.077 Y0.451 Z0.000
M666 X0.355 Y0.115 Z-0.470 A0.00 B0.00
Ok, sorry for the delay. I had a lot of stuff going on, but I came back to say that it work like a charm.
I ordered the thermocoupler you mentioned and instead of passing it's wire in the looming with all other wire, I decided to install it in parallel.(didn't take a chance)
It work amazingly well to be honest, the reading is spot on and precise.
Thanks.
By the way, I don't know how to tag this post as "solved". Can you explain it to me so I can do it. ?
Big thanks dc42, your help is always efficient!! 
I was having spike all around, so I changed the entire wire for an insulated wire (strainless braided). Which mean, from the thermocoupler to the effector, it's a k-type, stainless braided wire and from the effector to the Duet2, same thing. I still have spike all around.
I've checked the isolation of the thermocouple from it's casing and there is continuity, which mean the probe inside the casing is touching the case... isn't supposed to be like that to allow thermal transfer(since air is an isolation by itself).
I've tested an other model of K-type thermocoupler and it also has the same conductivity between the thermocouple and it's casing. Isn't normal?
What should I expect it to be ?
Is there a model you can recommend that is electricaly isolated ?
Note that my printer is a Delta, with a semi-floading stepper for the extruder. And the signal that "pollute" my thermocoupler is when there is a retraction/pression of that stepper, so I pick-up the magnetic flux from it's wiring, despite the fact that all the wiring of the thermocouple is stainless braided.
Also, starting the heatsink fan or the cooling fan doesn't create spike, neither the heater (I've tested them all), which lead me to ask: If all these doesn't interfere with the thermocoupler (specially the heater), how can the thermocoupler(non-isolated) be affected by the extruder stepper and not by the heater nor the fans ?
I'll try to remove the thermocouple wire from the sleeve spiral wrap and see if it pick up the stepper signal again, if so, maybe it pick up electromagnetic flux from the motherboard ? <-- is that even possible ?
Sorry for the long text, I shoot in every direction, trying to figure out the cause...
Any input/advice are welcomed!
Side note: I finished a print of 3h30 and have no problem. There was spike(+/- 45 to 60sec interval), more or less, but still I was able to print and the temp was always spot on (+/- 0,3 °C), which is "normal" for me condition here.
@ReVeRb Hi dc42, glad to hear from you! 
I finally get the board. it's tiny, but I can read on the back:
MAXTempTC v0.2
K Thermocouple Board
Design by: ...
...
...
www.Duet3d.com
So I assume it's a OEM part, is that correct ?
If so, since it's the v0.2 I can assume the capacitors are already on it, is that correct ?
I joined a photo of it.
Should I consider adding a shield too or is it enough to deal with the problem ? The thermocoupler will have approximately 2 inch(or more) of distance mostly everywhere. Except from the thermocouple to the connexion(under the effector) but this part is shilded.
Hi everyone, hope your doing great!!
I've been given a thermocoupler K1000, with what appear to be a third-party board or maybe the board is simply old? I dunno, but it is supposed to work, which is the important part!! I should receive it in few days so I'll be able to confirm the board model then. Anyway, I proactively did some research and found all the information I needed, except one thing about the capacitor...
Here, dc42(administrator) said in 2017: " The mod consists of adding a 0.01uF ceramic capacitor between each thermocouple terminal and ground."
https://forum.duet3d.com/topic/1270/thermocouple-issues/13
And by searching more, I found that one: "If you purchased a third-party thermocouple interface board, check that there is a 10nF or greater capacitor in parallel with the thermocouple terminals."
https://docs.duet3d.com/User_manual/Connecting_hardware/Temperature_connecting_thermocouples
If the board is a Duet, I assume it will have the capacitor. Or maybe not... In any case, I'll look for them on the board and if they aren't, here come my question:
Does a 0,1uf would do the trick ? dc42 suggest a 0,01uf(10nf) and in the manual it is recommended to install a capacitor with "at least" 0,01uf, which open the door to a bigger one, but how much bigger is my main concern.
I already have plenty of common 104 ceramic capacitor on hand(0,1uf) and few 103Z(0,01uf), but they are +80%/-20% tolerance and I don't know if they would be good enough for such application.
Any info or input would be greatly appreciated
thanks
@dc42 Ok, it's been a while, but I finally had the time to build the PCB capacitor bank and install it. It work good and I have approx. 1.5 - 2 sec. before the board shut down.
I'm not sure I understand how to implement it in the firmware, neither how to program it for my needs and limitations...
This is the code provided on the documentation:
M911 S21.0 R23.0 P"M913 X0 Y0 G91 M83 G1 Z3 E-5 F1000"
As far as I understand, I simply copy that line of code info config.g and configure it for my needs?
I have a Delta, with a 12V setup pushing around 13.5V... so it would be:
M911 S12 R13 P"M913 X100 Y100 Z100 M83 G1 Z3 E-5 F1000
Which would raise the nozzle 3mm from it's actual position and retract 5mm all motors(X,Y,Z,E0) at F1000 speed right after it detect a power outage ?
Then it will save the position ?
Does it have to home before it save ? If not, can it just raise the nozzle and retract a bit, then save the position... And when the power is back, I simply home the printer and re-launch the print and it will move where is stopped ?
I don't understand very well how the whole process work all together to be honest and some explanation is welcome!!
Yes, my PSU have a potentiometer to adjust the voltage. As for the maximum rating when everything is turned on, I have no certainty, just a vague idea...
Duet2 board: 12V, 0,5W
Panel Due: ??
Heat bed: 12V, 120W
Heating cartridge for the nozzle: 12V, 70W
stepper(extruder): 600mA max
stepper(x,y,z): 800mA max each.
1 fan: 5V 0.15A
I doubt I'm over 20Amp max... I have plenty of room.
So I could just connect my capacitor in parallel to reach 20 000uF, and install the capacitor apparatus in parallel with the PSU and I should be fine ? I just want to be 1000% sure that I won't create a ShortCircuit to the PSU because I have no idea if the PSU has a SC protection and I'm too low on a budget to buy a replacement.
About the ESR. On the other post, it is mentioned that low ESR cap draw a lot of current from the PSU and the more cap. are connected in parallel, the higher the current draw is... I though the Rippler Current spec was the max current draw of the cap, but apparently no. How to I determine that ? (I just like to validate as much as I can with some formula before I solder anything in place).
How does turning the PSU voltage up will affect the setup? The potential difference will increase, passing from 4V =(10.5V to 6.5v) to 6V = (12.5V to 6.5V), so the delay before everything shut down will increase a bit. On the other hand, that's the reason I bought 2x 10 000uF. I though that it would be well sufficient to add some delay while staying at 12V+/-5% on my PSU...
What is the main advantage of raising the PSU voltage slightly ? How will it affect the 5v supply from the Duet board ? Does anyone tried that 14V supply instead of 12V and if so, how does that affected your bead heater, your heating cartridge, fans, etc.
Thank you guys
Re: [Resume from power outage](really only with 24V?)
I recently faced an unpleasant situation like many of you... A 3 day print that ended after 2 and a half days due to a loss of electricity...
So I decided to consult the information on the forum to add capacitors on my 12v setup.
Being on low budget, I acquired 2x 10,000uF 25V, low ESR, Ripple current:2.9A@120Hz-3.335A@10kHz. and will salvage other parts from different electronic part I already recycled/ kept in a box (like the resistor if needed).
I saw this post and inspired myself from it:
But, I don't know how to properly install everything(which configuration to be optimal for my setup) to avoid burning my PSU (Chinese 12v 360W, 30A).
Also, I'm in no way an engineer and only have basic technical knowledge of electricity, but capacitors is out of my knowledge so I double checked and the first formula seems to be wrong. Isn't C=Q/V instead of C=Q*V ?
In the post, they talk about the ESR, I have no idea what it represent nor how it work in my case. As I understand it from the reading, if I connect my capacitors in parallel, the low ESR of my capacitors will play a significant role, but I don't understand how to determine if I'm "ok" or not, if I absolutely need a resistor or not and if I do, which kind...
How to I determine the proper resistance value I need to add in series with the capacitor. Anyone have a formula, maybe?
As far as I understand it:
With 2x 10 000uF capacitor. I could store plenty of energy, enough to save on the sdcard, that's for sure, but could I also be able to move the nozzle away (assuming that cooling fan, the bead and nozzle heater are shut down, and maybe even the nozzle fan).
Thank you guys
I tried 800, 700, 600, they all are worst... 550 just get slightly better with a peak at 65%, but the fan stop passed 90% and the second serie I tested is at 450, 430, 480, 470 the fan try to kick-in if I jump from 0% to 100% but the fan just "wiggle" a bit and stop after 2 sec. Also, the fan power is way weaker at any level... but it will kick-in if I jump from standstill to 99%... not at 100%...
What are the frequency range I can work with without damaging the fan or the motherboard ?
-EDIT- I also tested 525 and get the highest peak at 60%, with the same symptoms... from 0 to 99%, the fan kick-in but is weak... from 0-100% it just "wiggle".
-EDIT 2- I also tested 1000 and beside having a fan blowing even harder on the whole spectrum until 60-65%, I have the same behavior overall. At 1% we clearly hear the high pitch frequency noise from the fan. The noise is lesser the closer to 60-65% I get... passed 70% we clearly hear the bearing rattling... passed 80% the performance drop significantly... at 0-100% the fan wiggle again, but wont kick-in and at 0-99% is will start but the air flow is weaker as usual...
Did the same test at 2000Hz... same result but a slightly higher flow between 0 to 60-65%, weaker flow passed 70-75%, the bearing rattle is stronger and the initial high pitch sound is slightly stronger...
I guess that if I keep raising the Hz, I'll just damage/destroy the fans...
So what is my solution?
Is there a way to set the maximum fan speed to 60% for the PID calibration?
@phaedrux Sorry for the delay...
I did all the verification/calibration except one and everything is mostly the same. The only problem remaining (I had it before) is that when I calibrate the nozzle heater as a tool, the cooling fan never turn on. The walk around I used last year was to manually start the cooling fan during the PID calibration.
Known issue with cooling fan: 60% is the real maximum speed. From standstill jump to 100%, the fan just don't kick-in... From standstill jump to 60%, the fan will kick-in... If the fan is running already, it will increase in speed until around 60%(like it should do), then the higher I go(passed the 60% limit), the slower it get... At 100%, it's like being at around 20%-25%? (hard to tell, but it's definitely slower). If I jump from anything lower than 80% to 100%, the fan will stop spinning(I didn't test higher than 85%)... The only way to reach 100%, is to slowly increase, wait a bit, and repeat... Maybe that's the reason why the fan never spin during the PID sequence ?
For info: I connected 2x 3031 5v fan in parallel.
Also, my bowden is around 210mm long, straight... and the pressure advance need to be set way out of the parameters suggested in the instruction provided on duet3d.
The Bowden setup is made from a PTFE tube and a BMG clone, suspended by a gt2 strap. There is no bent or curvature on the PTFE...
Before it start acting crazy, I normally used (as you can see in the config file) a way lower step/mm for the extruder...
But I now have to use a way higher e-step/mm, otherwise, I'm under-extruding.
I have absolutely no clue what is suddenly happening to my printer. I worked on machine for more than 10 years of my life as an electro-technician of automated system and built myself 2 delta printer and never had this kind of randomized/weird behavior from a machine. I know it's not the case, but I felt like I was working with the machine and someone was reprogramming it every night... 
I'm clueless and I ran out of expertise.