Here is my Fisher.

I can tell it looks modern and slick!







Console output of some Gcode commands on Duet Web interface

3:51:48 AM Disconnected.
3:45:20 AM G31
0
3:40:57 AM M119
Endstops - X: not stopped, Y: not stopped, Z: not stopped
3:40:48 AM M119
Endstops - X: not stopped, Y: not stopped, Z: at max stop
3:40:38 AM M119
Endstops - X: not stopped, Y: at max stop, Z: not stopped
3:40:19 AM M119
Endstops - X: at max stop, Y: not stopped, Z: not stopped
3:38:58 AM M119
Endstops - X: not stopped, Y: not stopped, Z: not stopped
3:31:01 AM Connection established!
3:30:54 AM Disconnected.
3:13:28 AM Connection established!
3:13:28 AM Page Load complete!

One modification I did already was I made a spool holder out of MDF and three bolts instead the spool holder.


Timelapse video of building Fisher Delta for three days.
[youtu.be]

Edited 1 time(s). Last edit at 08/04/2015 03:54PM by janpenguin.

Nice! Don't tell anyone, but I only finished mine last weekend... it prints very nicely, though!

Ian
RepRapPro tech support

Spool holder for 1 kg filament

Parts:
M6x100 bolt x1
M6 nut x2
M6 washer x2
67 mm length of Spacer
Two M8x100 bolts and six M8 bolts hold two triangle panel.





F623 bearing removes timing belt slippage on the belt tensioner.

I noticed the timing belt is always rubbing one side wall of the belt tensioner due to flat surface of 623 bearing.




Simple MDF base




I plan to replace all acrillic parts with laser cut steel and aluminium. I got the service quote from a metal fabrication company.
Total estimate cost is around 90,000.00 KRW (about 77.00 USD)

Steel parts - 39,300.00 KRW (material cost is included)
Base plate, 5T (5 mm)
Top plate, 5T
Front panel x2, 3T
Side panel, 3T

Aluminium parts - 36,400.00 KRW plus material cost

Print bed, 5T
Rod x12, 3T
Effector plate middle, 3T
Effector plate top, 3T











@droftarts

Thanks Ian. I'm anxious to see how the Fisher performs.

Yeh, I was debating whether to replace the idler bearings with a flanged type to stop the belt rubbing on the walls of the idler. I have little black rubbing marks on each of the idlers now.

Quote
janpenguin
I noticed the timing belt is always rubbing one side wall of the belt tensioner due to flat surface of 623 bearing.

Had the same problem with the belt, a layer of rubber dust on the carriage and bed. Redesigned the idler assembly with 4 screws so the idler could be angled to keep the belt in the middle.

Is that the typical way to mount the belt? Looks like upside down...
-Olaf

Quote
o_lampe
Is that the typical way to mount the belt? Looks like upside down...
-Olaf

Yes, that's how RepRapPro does it on the Ormerod too. It avoids the annoying clicking sound that I get on my Kossel as the toothed side of the belt runs over the idler pulley. There is a half twist in the return side of the belt so that it is the opposite way round on the stepper motor pulley.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

An alu fisher would be so cool looking.

---

----- Making the world smarter @ www.xetal.eu
----- Helping entrepreneurs @ www.fralke.com

I also used those bearings instead. I had used them on my full delta and kossel mini and they always did a perfect job of keeping the belt in line - I wondered why they had not supplied them for the fisher.

They are relatively cheap, I bought a pack of 50 for about $20 on ebay.

Generally, on our other machines, we only use one bearing for the idler. I don't think there is such a thing as a double flange 623 bearing (I can't find them after a brief look), they're all single flange. Also, I think a double flange wouldn't let the belt run between them. The Fisher is the first printer we've run with two 623 bearings, and we used what we had available. We'll look into using flanged bearings.

However, the belts should pull quite straight. Mine wander on the bearings a little, but generally don't rub on the edge all the time. Make sure the bearing is sitting in the printed part straight, and that it's straight into the belt clamp in the carriage.

Ian
RepRapPro tech support

Hello, I also faced some problem with excessive rubbing from the belt to the idler carriage and replaced the bearings by flanged bearings (they are the same as for Kossel, so easy to find on ebay or elsewhere).
However, I did not recommended this solution, because, if it is working now, I had some problems because I had initially not properly seated them in the carriage, and so, the belt climbed on the flanges and that was real bad. It was because as the space between the flanges was a bit short, I added a washer between the bearings and the enlarged ensemble was a too tight fit in the idler carriage.

Even if flanged bearind helped at the end, I think that the problem is not really flange or not flange.
With flanged bearing, you also have some rub on the flange, but for me the problem is that the design of the idler carriage make it flipping on the screw head as the belt load is above the screw head seats, then the belt go sideways, increasing the flipping. Depending the tolerances of you carriage within the support, the angle could be significant and so the side load of the belt on the carriage. I got flipping either inwards or outwards
So I think that the main point is to maintain the carriage well vertical. I got the idea to either enlarge the carriage to have a very tight fit or to slide on carriage sides (up to to top) slim sheets to well maintain the carriage.
The four screw idea allow also to maintain the ensemble well vertical, at the price of a slightly more delicate adjustment. This is a nice idea to maintain the tensioning in the same volume envelope and there is no need to resupply bearings.

About the question of idling the belt on its back, it shall be noted that the belts are not GT2-2 as on the Kossel for example, but they are T2.5 belts, which are slightly less costly. Due to the shape of the teeth, the T2.5 belts did not run smoothly on their teeth on idlers, hence the twisting of the belt to run on belt back, which works ok.

Edited 2 time(s). Last edit at 08/07/2015 07:49AM by PRZ.

Acrylic parts need careful attention.

I broke the middle effector plate twice where the hexa head bolts seat.


Alignment the base plate and side panel

Square wiring access port on the base plate must be right below the extruder motor's wiring harness terminal.
The instruction mentions it but gives confusion to a user because there is no clear picture or diagram.
I installed the base plate upside down and had to spend good six to seven hours to flip the plate.

Wiring diagram of Duet board is missing.

Duet connections only shows couple photos without wiring diagram.

Filament loading

The following instruction doesn't say about pushing down (or pulling up) the idler lever when loading the filament.

Quote

Push some filament through the slot in the side panel and into the extruder drive block. Using the thumbwheel, engage the filament and feed a few centimetres through. Hold the thumbwheel and try to pull the filament back out. If it slips easily, adjust the idler tension using the M4 button head screw.

Calibration and fine tuning information are important.

Hopefully it will come out soon.

Edited 1 time(s). Last edit at 08/07/2015 05:38PM by janpenguin.

Excessive friction or rubbing between the bowden tube and Y endstop cable popped out one terminal of Y endstop switch.



I printed the robot.

7:56:10 AM M32 Robot.g
File Robot.g selected for printing
7:54:03 AM G32
Calibrated 4 factors using 4 points, deviation before 0.640 after 0.000
7:50:46 AM Connection established!

Mine has outstanding retraction problem.

The idler bearing and hobbed gear shaft in extruder drive has to apply enough force that prevents slipping of filament during retraction and extrusion.
Once I adjusted the bolt tighter, around 18 mm long between the lever and bolt head, retraction problem gone.

layer height 0.2 mm, default Fisher config, 205C to 215C


It passed the retraction torture test I made.



layer height 0.2 mm, 205C


Best result from my Kossel Mini, layer height 0.1 mm


As soon as I finish dimensional calibration, I gonna use Fisher as a production machine.

Edited 1 time(s). Last edit at 08/09/2015 11:57PM by janpenguin.

Just completed my first rebuild of Fisher, replacing all the acrylic panels with polycarbonate. Looks more now like a crystal citadel rather than a printer.The panels are 3mm and the corner post 10mm. Much more solid now and no more bulging side panels.


Also did away with the microswitches and replaced them with Hall effect IC switches.
The XYZ offsets are now very small and the consistency in homing is better. Previous maximum variation about 0.08 is now 0.02



Calibration figures obtained

14:56:17
M666
Endstop adjustments X-0.39 Y0.44 Z-0.05
14:56:09
G32
Calibrated 4 factors using 4 points, deviation before 0.014 after 0.000
14:52:56
M666
Endstop adjustments X-0.41 Y0.44 Z-0.03
14:52:49
G32
Calibrated 4 factors using 4 points, deviation before 0.013 after 0.000
14:52:04
M666
Endstop adjustments X-0.40 Y0.45 Z-0.05
14:51:57
G32
Calibrated 4 factors using 4 points, deviation before 0.006 after 0.000
14:51:17
M666
Endstop adjustments X-0.39 Y0.45 Z-0.06
14:51:10
G32
Calibrated 4 factors using 4 points, deviation before 0.008 after 0.000
14:50:28
M666
Endstop adjustments X-0.40 Y0.45 Z-0.05
14:50:20
G32
Calibrated 4 factors using 4 points, deviation before 0.015 after 0.000
14:49:38
M666
Endstop adjustments X-0.40 Y0.46 Z-0.05
14:49:28
G32
Calibrated 4 factors using 4 points, deviation before 0.017 after 0.000
14:48:41
M666
Endstop adjustments X-0.41 Y0.44 Z-0.03
14:48:34
G32
Calibrated 4 factors using 4 points, deviation before 0.074 after 0.000

It looks fantastic with added features.

What did you coat or paint on all the ball joints?

Quote
janpenguin
What did you coat or paint on all the ball joints?

Just a bearing grease called Ambergrease EXL. The fact it is red makes it standout.

Nice! And bullet-proof, in polycarbonate!

Ian
RepRapPro tech support

johneato , did you think that polycarbonate is rigid enough to make a bed ?

Edited 1 time(s). Last edit at 08/15/2015 03:54AM by Greg_be.

Quote
Greg_be
johneato , did you think that polycarbonate is rigid enough to make a bed ?

I believe it wil be rigid enough, I have just ordered some 5mm poly and so shortly will be able to give you a definitive answer.
If not I will scale up the thickness and machine a 5mm rim around the bed to accomated the mounting and levelling

PS I have 10 acrylic connecting rods looking for a good home, anybody want them let me know.

Edited 1 time(s). Last edit at 08/15/2015 05:05AM by johneato.

Thanks !

I plan to buy also polycarb (in 5mm too) to replace my two broken beds but I'll cut it old fashion way, I don't have any "automated cutter" available.
Just to know, which machine did you use to cut yours ?

Quote
Greg_be I don't have any "automated cutter" available. Just to know, which machine did you use to cut yours ?

I use a CNC, nearly as much fun as 3D printing

Quote
johneato
I use a CNC, nearly as much fun as 3D printing

Thanks, it's my next plan, now that I have tasted 3D

Quote
Greg_be
johneato , did you think that polycarbonate is rigid enough to make a bed ?

Having now made the bed from 5mm Polycarbonate I can say its definately rigid enough. Certainly a match for the acylic.
I have also found it to be flatter than the acylic, first layer is much more consistent. Plus you will never break it.

Quote
johneato
Having now made the bed from 5mm Polycarbonate I can say its definately rigid enough. Certainly a match for the acylic.
I have also found it to be flatter than the acylic, first layer is much more consistent. Plus you will never break it.

Thanks for the feedback

OpenSCAD's DXF export removes basic geometry data that CAD program requires.

I had to run a CAD program to modify the original DXF files for aluminum and steel parts.

Base plate, Steel, 5 mm (thickness)
Top plate, Steel, 5 mm, it has big triangular access hole in the middle.
Front panels, Steel, 3 mm, one panel has ventilation and mount holes for 40x40x10 fan
Side panel, Steel, 3 mm
Rods x12, Aluminium, 4 mm
(The sheet metal company didn't have 3 mm aluminium plate. 2 mm plate felt too thin, so I chose 4 mm one.)
Effector plates, Aluminium, 4 mm
Print bed, Aluminium, 5 mm

I reduced all bolt spaces to 1.0 mm and submitted final DXF files to the company yesterday after hearing advice from a technician.
Below images are the areas I have concern about laser cut result.

Quote
janpenguin
OpenSCAD's DXF export removes basic geometry data that CAD program requires.

I had to run a CAD program to modify the original DXF files for aluminum and steel parts.

Base plate, Steel, 5 mm (thickness)
Top plate, Steel, 5 mm, it has big triangular access hole in the middle.
Front panels, Steel, 3 mm, one panel has ventilation and mount holes for 40x40x10 fan
Side panel, Steel, 3 mm
Rods x12, Aluminium, 4 mm
(The sheet metal company didn't have 3 mm aluminium plate. 2 mm plate felt too thin, so I chose 4 mm one.)
Effector plates, Aluminium, 4 mm
Print bed, Aluminium, 5 mm

I reduced all bolt spaces to 1.0 mm and submitted final DXF files to the company yesterday after hearing advice from a technician.
Below images are the areas I have concern about laser cut result.

laser_cut_draft_sc1.png

laser_cut_draft_sc2.png

laser_cut_draft_sc3.png

laser_cut_draft_sc4.png

That why I use Autocad or Solidworks. Special for dxf or submit for g code for cnc.

I disassembled Fisher and quick fitting test of frame parts: panels, base plate and print bed.

I coated aluminum and steel parts two weeks ago. Coating cost was $12 through business account.
I chose color yellow for aluminium parts but the company told me gold is available. So I chose gold. One day after when I got them back, as you see in the below aluminum parts were all orange. The company thinks gold as orange which is funny.

Laser cut quality of aluminium parts was below my expectation. 4 mm thickness was overkill to small parts such as effector plates. I gonna try another laser cut company for next time.



I modified the print bed to have fixed mount points.


Diagonal rods.


Top plate
Shallow coating was cheap and easy to be pill off by rubbing.


Base plate


One front panel
I replaced the RepRapPro icon with ventilation and mounting holes for 40x40x10 fan.


Another front panel


Side panel

Curious that you would have aluminium arms made, and may I suggest that when you make the effector again you remove the slots for the ball mounting screws and have that drilled and tapped for M3 studs.

[edit] sorry had you confused with PRZ who made printed arms, but still, not keen on the idea of alloy arms, would suggest you look at PRZs printed arms or my Carbon Arms instead

Edited 1 time(s). Last edit at 09/17/2015 02:50AM by bgkdavis.

---

RepRapPro Mendel 3 Tricolour
RepRapPro Fisher
-Carbon Arms
-Easy adjust Carriage+effector
-axis stiffness mods
HE3D -600 delta
-Duet 0.8.5
-PanelDue
-DC42 Height probe
-RobotDigg metal components
Simplyfy3D
RS Design Spark CAD