The post is still under construcion
look also 3 post below for the rest of the description !!!

This will take some time to explain with photos (i must still produce some of them) and descriptions... So for now I open the 3d.
I'll use this post as reminder for me, to do not forget all the steps I made and , maybe , it will be useful for other adventurer decide to puts ends on a such big delta printer.

1st and important... The Tevo company is terrible. I mean the back office for sure. In particular Kelly Zheng took 3 month for NOT send what she promised (and I paid). Paypal cover the funny joke.

2nd... the printer for became something close an accurate tool need lots of work and money to put in.

Inspirational :
https://www.akeric.com/blog/?p=4325#Upgrades

List of upgrade and modifications....

Hardware:

1. DuetWiFi + Duet LCD 7" (last firmware stable) + heat sink (with thermal pad T412) on parts could became hot.
2. Nema motors 0.9 (for 3 motors + extruder) + dumpers + 4 12v noctua 6cm for cooling with lamptron fan hub (with resistance for lower the fan speed)
3. Extruder (Bondtech)
4. Original D3D Hotend + Volcano extension + copper nozzle (Nickel coated) + silicon cover.
5. Original BLTouch v2
6. Belts with iron core
7. Full enclosure
8. End filament guide/sensor
9. Custom Borosilicate glass (41,5cm X 5mm) + adapters block (i'll provide the STL)
10. Cartridge modified for better precision and less torque.
11. Starting for the hot chamber solution... Double plexiglass (aluminum + double plexiglass + pvc + .... ) .

Software:
Slicer used ... Firmware Settings and so on...
For now I'm using Ultimacker Cura and Matterhacker doing tons of test for the better result. The Firmware I'm using the last stable 2.02. And I'll try to explain why I used those settings.

Filament used :
At today I have used 3 filament.
1)Formfutura Easyfill ABS
2)Patona ABS
3)Formfutura ApolloX (ASA)

I can say the Easyfill ABS was pretty good but suffer as all the ABS of retraction and therefore warping.
The Patona was pretty bad. Pop any time at any temp. Warping like hell. Cost half the Formfutura. But a shiny finish.
The ApolloX is a ASA material that have less problem of warping of ABS at the cost of a little less beautiful finish (small percentage against the Easyfill).
Less warping not mean none. So for object that have a big amount of layers ,placed as a corner , and a little surface of contact with the bed , you can experience some warping at the extreme corner. Any way very less than the other two ABS.

1. The pouring of all the cable was not so fast as I imagined. I little scared of cook everything at the end i succeeded.
   I putted the two mobo close each other and with a laptop on the side started slowly to move all the cables using white plastic scotch to sign what i was moving.

The two schema:

With lots of patience I toke a very small screwdriver and moved the cables from the old connectors to the new duet ones. Some one have catted come plastic on the old ones and plugged on the new board but instead of this I preferred to use the connectors that can stay blocked when inserted.
Other tricky stuff...the optical sensor have strange color wire. All different. So I looked at the schema and the polarity of the endstop (as in picture) and considered one endstop at ones as per se cabled.

2)The nema motors 0.9 . I soldered all the cables and covered it with thermo sheath . Some hot glue for blcok the cables that enter in the motor.
Applied a heat sink on each one with under a 3M thermal pad.
Recovered from old back covers of some computer case , 3 metal plate. Of course drilled.
Then I linked all the fans (also the mobo one , changed with a 12v one) to a 12v fan hub. And between of them I putted a potentiomenter for regulate the fan speed as i like. (the noctua at half speed are really silent).
Between the motor and the support I applied the classical nema dumpers.
I had to smooth the metal pivot (was full rounded) for a better block of the gears.

1. The Bondtech extruer. I have to say the bowden connection is a little lose, so make sure you insert the teflon hose well to the end.
   I just added on the top a little piece of hose for a better guide and for a convenient calibration. (using the digital caliper im far enough to do not touch anywhere during the operation)

1. Also if the chinese copy of the volcano D3D hotend was not that bad , I decided of put the original one. The TLM already have a volcano copy inside but with a small metal core and an old thermal sensor. All tensions are 12v.
   
   
   

2. The original BLTouch v2 is slightly different. It have a sturdy shell despite the copy is a little more thin. The original trigger is in plastic and not in metal. The copy writing are a little more faded. The original one should be more accurate.

3. Because of this article:
   https://www.reddit.com/r/3Dprinting/comments/61j9et/i_tested_some_gt2_belts_to_failure_on_a_tensile/
   
   I choose to use the iron cored belts. They look more rigid but this absolutely dont impact in any bad mode to the movement. For Delta printer is suggested use those ones despite the standard ones just because the head is always in tension on the belts.
   

7)As I started with the intention of build ABS parts i read about the necessity of build an enclosure for keep the worm.
I used some aluminum bar . For the big windows I used 3mm plexiglass and for the smaller window I used the PVC material. All those parts , with the relative screw and rubbers are available in common DIY stores.
For the small windows I tried to use the plexy but it crack for vibration and torsion.
now with the black panels is more cool ^_^.
For the closing mechanism I used dome adhesive magnet.
I have done some holes on the top and bottom frame for block the corners that sustain the enclosure.

8)End filament guide/sensor , it is well explained here:
https://www.thingiverse.com/thing:2629228
I just used the remix version for add the Teflon tube for a better guide and less friction.
Just keep attention at wall close the mechanical sensor. The must be super flat and i added also some grease for a perfect slide.

As told I added the top insulation.
And also added some clip because now the sponge create too much pressure on the magnetic closure.

The next step is build the air mechanism for the hot chamber

Dear absolute !!!! Maybe you nailed it !

Look at this , the six rods:
1> -0.25
2> -0.25
3> -0.15
4> longhest
5> -0.1
6> -0.6 !!!!!!!!!!!!!!!!!!!!
Poor auto calibration , It cant reach those parameters.

From the Duet guide (and your linked pdf too):
https://duet3d.dozuki.com/Wiki/Calibrating_a_delta_printer
All 6 rods must be the same length, measured between bearing centers

If you are building the rods yourself, make them one by one in the same jig. If you are buying them, check that the supplier guarantees a maximum difference in rod length, which should be less than 0.1mm. It is particularly important that the rods in a pair are of matched lengths, because any difference will cause the effector to tilt.

I lost damn ages because I trusted the manufacturer REALLY My BAD ... I had should check this hundred time before all the attempt I made for calibrate it.
My mistake was : the product is already made , it will be good. Such a sheep.

Now what? Buy hundred of them to make sure found SIX equal ?!?!?!?!? Damn Tevo.

I feel so bad for wasted your time... All the people had helped here , damn ... sorry.
I should had to think as the printer was made by me and check ALL the steps as reported in the guide. Inexperience here docet.
Now I'll go check all the rest of the hardware hoping that the frame didnt have absurd variations

I'll keep you informed about the evolution of the story...

Here the adapters :
https://www.thingiverse.com/thing:2498606
I make all the 3 pieces identical in dimensions. D9.58mm , H30mm , L48mm.

Trying to upload images (jpg)

Error
Input file contains unsupported image format

Finally got some material... (in Italy almost twice the price DAMN+shipment+taxes+taxes+taxes+taxes ecc ecc )
400mm bars and effector , lets see if this trap can work properly...


I'm printing with an other printer the supports for the 3 plates...

Thanks anyway at all that had participate and helped...

Dear absolute !!!! Maybe you nailed it !

Look at this , the six rods:
1> -0.25
2> -0.25
3> -0.15
4> longhest
5> -0.1
6> -0.6 !!!!!!!!!!!!!!!!!!!!
Poor auto calibration , It cant reach those parameters.

From the Duet guide (and your linked pdf too):
https://duet3d.dozuki.com/Wiki/Calibrating_a_delta_printer
All 6 rods must be the same length, measured between bearing centers

If you are building the rods yourself, make them one by one in the same jig. If you are buying them, check that the supplier guarantees a maximum difference in rod length, which should be less than 0.1mm. It is particularly important that the rods in a pair are of matched lengths, because any difference will cause the effector to tilt.

I lost damn ages because I trusted the manufacturer REALLY My BAD ... I had should check this hundred time before all the attempt I made for calibrate it.
My mistake was : the product is already made , it will be good. Such a sheep.

Now what? Buy hundred of them to make sure found SIX equal ?!?!?!?!? Damn Tevo.

I feel so bad for wasted your time... All the people had helped here , damn ... sorry.
I should had to think as the printer was made by me and check ALL the steps as reported in the guide. Inexperience here docet.
Now I'll go check all the rest of the hardware hoping that the frame didnt have absurd variations

I'll keep you informed about the evolution of the story...

Fine , It will be the first check. TNX again.

bouldnuts thanks for the support. I'll read that document deeply and for several times. As you linked it several post above. I have to dismantle the printer and check as stated in the document. I have all the tool for do that.

At the end did you changes for a smart effector with ball bars? Or you finally found a good calibration?
Did your frame or bars had uneven issue?

Probably the linear rail with smart effector and ball bars could finally stop this damnation.

Never mind... I reopen the 3d. Unlucky. After age of attempts I'm still here
As I have changed so many parts for reach more accuracy I still can't find a barely good calibration.

If I try to manually to calibrate the printer following this video :
https://www.youtube.com/watch?v=Xk7remEhUx8
I reach some good stuffs , but absolutely not perfect and still full of mistakes.

I noticed: i could accomplished to make a perfect equilibrium of calibration in front of the 3 tower using the screws on the cartridge.
So now if I run this 3 commands.
G28
G1 X-121.24 Y-70.00 Z0 F5000 - tower X
G28
G1 X121.24 Y-70.00 Z0 F5000 - tower Y
G28
G1 X0 Y140 Z0 F5000 - tower Z

I can reach all the points at the same distance from the bed.
Hurra!!! I though... WTF , absolutely not.
If from the center G1 X0Y0Z0 , I go to X-100 the nozzle press toooo much against the bed. Fine , maybe the radius culprit ? Is settled as the original parameter as the rods as all the rest.

After i found the "right" offset for the 3 towers I left to zero the M666.

Now the drama, if I move the nozzle in the points in front of the tower all is good , but 5cm on the left or right all is wrong. No matter I try to change slightly in radius or rods DAMN !!!!!

So the last resort:
The last resort was the AUTO calibration. BUT HELL if that was able to adjust a thing ! All the time produce uneven results.
I tried all the permutation on the Esher bed generator. Tried all the factor from 3 to 9. NOTHING make that damn bed flat and perpendicular to the towers.
Few point or all the possible to use.
The auto cal change the end-stop offset despite is the only thing that seems to manually work. Dramatically declinate the bed. (this was the FIRST problem I face of when I start to calibrate this devil machine)
The H parameters are not useful . They change the height JUST and exactly in the points mentioned and dont declinate all the bed.
I hoped that setting up the offset manually I could help the auto calibration to make a better result for the bed inclination . BUT NO. NOTHING.

So, I'm lost (not in space but in my house looking this metal trap) .

I mounted the nema 17 , 09 degree and properly settled in :
; Motors 0.9 degree
M569 P0 S1 ; Drive 0 goes forwards
M569 P1 S1 ; Drive 1 goes forwards
M569 P2 S1 ; Drive 2 goes forwards
M569 P3 S0 ; Drive 3 goes forwards
M350 X16 Y16 Z16 E16 I1 ; Configure microstepping with interpolation
M92 X160 Y160 Z160 E920 ; Set steps per mm
M566 X1200 Y1200 Z1200 E1200 ; Set maximum instantaneous speed changes (mm/min)
M203 X18000 Y18000 Z18000 E1200 ; Set maximum speeds (mm/min)
M201 X1000 Y1000 Z1000 E1000 ; Set accelerations (mm/s^2)
M906 X1800 Y1800 Z1800 E1800 I60 ;

I changed the belt with the ones with iron cores for less stretch and more accuracy.
They are well tied . All sound pretty the same (they sound more or less "like" a"Mi" of a guitar)
The carriages are shimmed with Teflon for avoid torsion ad it work damn fine.
All the structure is pretty sturdy.
The glassed custom build made by a professional industry have tolerance in order of cents.

I have even tried to rise the bed using springs. I thought : if the automatic calibration see my bed declinate I'll correct manually the inclination of the bed. SOOOOOOOOOO ?!?! The auto calibration no matter what end declinate the bed !!!!!

NOW: I noticed something that some good souls could interpret for give me a light spot.

1)If I go to G1 Z0.2 (just for dont touch the bed) and then I move the head to X-100 , when turn to X0 the nozzle dig on the glass. As it was unable to turn in the same position. (belt tied much more or much less dont make difference - motor are at full power and very well cooled)
2)Without the automatic calibration the movements stay consistent inside the 100 radius (more or less). But between 100 and 140 the nozzle rise also of 0.2 (near). I suppose this is the effector shift . but it is really tiny till 100 and rise much more from 100 to 140. I'm avoiding 150/155 because the rising became much more evident.
3)When manually i Try to rise 0,05 sometime those commands are not received. I have to push twice the command for make one result.
When I press the button I can ear different sound form the motor. As first a whistle and then the movement. (the belt are tied , but I suppose that for a so tiny movement maybe tolerances of reverting the turn of the belts are the culprit? ) The motor are set to 1800mA and 160 micro steps.

Any suggestion apart throw everything from the windows?

ps: the Akismet spam filter is REALLY annoying (even if I just want correct a typo), no other alternatives ?

@dc42 said in [SOLVED] Setup the 3rd (virtual) thermistor:

Thanks for the further tip !

Ok, I dare try ...
It was as stated by David.

Usefull links:

https://forum.duet3d.com/topic/11679/thermistor-read-temp-only/4

https://forum.duet3d.com/topic/2098/multiple-combined-thermistors-for-a-single-tool-heatbed/19

; HOTEND 01- Nozzle Heater H/P 1
M305 P1 R4700 T100000 B4725 C7.060000e-8   ; Set thermistor + ADC parameters - E3d Uses - Semitec 104-GT2
M143 H1 S280 ; Set temperature limit for heater 1 to 280C


; HOTEND 02 chamber- Heater H/P 1
M141 H2 ; heater 2 is the chamber heater
M305 P2 R4700 T100000 B4725 C7.060000e-8   ; Set thermistor + ADC parameters - E3d Uses - Semitec 104-GT2
M143 H2 S280 ; Set temperature limit for heater 1 to 280C


; HOTEND 03 enclosure- Heater H/P 1
M305 P103 X3 S"Enclosure" T10000 ; 

So the virtual thermistor is named P103. P100 ,P101, P102 are reserved so for additional heaters must used those ones.
X3 is needed to map the heater on the channel 103.
With only this setup the thermistor in the "EXTRA LINK" , near "tools" report 2000 Celsius.
So I added the 4.7K resistance between the ADVREF to the positive pin35. The negative is VSSA pin 40. The temp dropped down to 23.2 as all the others thermistors.

Please don't shot me if is not related but missing here your bed.g can you try to give a look here:
https://forum.duet3d.com/topic/10455/solved-bl-touch-deploy-only-one-time-with-2-03rc1

If you have this in your bed.g:

M98 Pdeployprobe.g ; 

Try removing this as stated in that link.

Your welcome buddy
When I start my adventure with 3D printing here I found lots of help so I try to give it back as I can...