I made a little tool to help you plan the arm lengths:
https://github.com/bondus/5barscara/tree/master/JS
Fetch the git and open the html in a browser. It's a very crude little tool where you can experiment with arm lengths and work modes, it renders the possible work area and you can mouse in the work area to see the arms.
I have build a five bar, it actually works very well. The double arms makes it very strong in the XY-plane, exactly what we need when FDM printing.
There are a few tricky parts building a five bar.
You need to gear down your motors a lot to get enough steps/mm. And you need a gearbox with almost no backlash at all. The normal gearboxes you often find with steppers have way too much backlash. Ideally you would want a harmonic drive or possibly a cycloidal gearbox. I used gigantic pulleys (260t) and belts on my machine.
You have to be very precise on the length of the arms, and the angle of your homing sensors. A small error will result in warped parts. This is the similar to a delta printer, except that the firmware have amazing calibration algorithms to compensate for errors. The Five Bar firmware has no such features.
It can also be tricky to make the arm move in a perfectly flat plane. It's important to make sure all bearing rotate in the same plane. It is actually an over constrained system and if not built perfect there will be tensions in the arms.
A bed probe can compensate for small errors ofc.
I'll bring my 5 bar scara that was supposed to be shown two years ago.
Just have to get the thing moving again..
Sorry for barging in. But...
If you were to describe the robot as a number of transformation matrices (https://en.wikipedia.org/wiki/Denavit–Hartenberg_parameters) you could describe any serial robot with any number of axis. Even a serial scara could fit in this (except the bed).
Matrices are very nice to work with.
Combine this with the matrix defined cartesian and you can describe any combination of linear cartesian and serial rotary kinematics. What a dream 
@iamthebest22 No updates I'm afraid. It works perfectly fine, I use it to print various small parts for other projects. The accuracy is not good enough for big mechanical parts, we need a calibration method for that.
It's actually a very simple mechanical construction. Some parts from my CAD files:
https://a360.co/3dNKGjZ
You have to make your own Z axis
@arhi If you do not return true from IsContinuousRotationAxis() that is exactly what you get. You can still rotate multiple rotations, it only affects the automatic wrapping.
@arhi It's supposed to be used with polar/arm printers, if you move 360 degrees you end up at the same place.
M92 sets steps/deg in this case and not steps/mm. For example a 0.9deg stepper with 16 microstepping and a 20:1 gearbox will be 16*20/0.9 = 355.556 steps/deg
@pgoellner It worked when I played around with it.
When IsContinuousRotationAxis() returns true the firmware should always find the closest direction to rotate to reach the goal.
It has to be a G0 move for it to do this.
I'll probably sort it out. Just venting a bit
A "build clean" and another attempt to build gave another error:
23:29:55 **** Incremental Build of configuration DuetM_RTOS for project RepRapFirmware ****
make -j16 all
make: *** No rule to make target 'src/GPIO/GpioPorts.o', needed by 'DuetMaestroFirmware.elf'. Stop.
23:29:56 Build Failed. 1 errors, 0 warnings. (took 1s.23ms)
Eclipse and the RRF build system is very fragile and frustrating.
I avoid to update my local repo, each time I get these seemingly random errors.
18:15:43 **** Incremental Build of configuration DuetM_RTOS for project RepRapFirmware ****
make -j16 all
make: *** No rule to make target '../src/libc/strptime.c', needed by 'src/libc/strptime.o'. Stop.
18:15:44 Build Failed. 1 errors, 0 warnings. (took 1s.66ms)
It's quite some work keeping the repos up to date and in the right branch..
A hint for everyone building a parallel scara:
Before mounting the head move the arms around in all kind of strange positions and make sure that the distal arms are at the same height (or at the offset you designed them to have). Especially try to reach the same spot with different arm-angles. If they do not line up everywhere it will be impossible to get flat XY movement in the z-plane.
It can be quite a lot of work to get it all right but it is at the same time a kind of self-verification of the machine that all joints rotate perpendicular to the same plane. A very nifty feature.
I had to use a mixture of gentle filing on some plastic pieces and brute force bending and twisting the aluminium tubes to get this machine straight. Alu-extrusions from the local hardware store are not made for precision machinery.
1mm! But a machine does not have to be more precise that what you need. And as long as you keep the corners of the printed parts round the ghosting will not affect the prints.
It's easy to become a machine snob, chasing those last 100ths at any cost
@oliverracing, nice woven carbon tube
Have you measured the arm drop with the hotend in the inner vs outer position?
I see you have piezos under the bed, it should show up in the bed leveling mesh.
It's quite brave to do direct drive with such a long arm, you must be getting some ghosting in the outer position?
I love your build even if I sound critical.
Fans and heaters are switched on the ground side. But you can have one common supply cable, ie, 24V or whatever you got. If you use a special supply for the fans they would need their own cable.
I have tried to tune this beast for a while now to be good at general printing. But that long bowden tube is hard to tame.
Others seems to be ok with this, I can't do it. Always plastic where you don't want it or nozzle clogs.
I made my delta flying extruder and my CoreXYs direct or nimble.
A direct drive is minimum 300g, which is a bit heavy for this mechanics to carry in the z axis.
And what is the consensus for a good part cooler? My homemade ring cooler does not work very well.
There should be a line in your config.g looking something like:
M950 F1 C"fan1" Q500 ; create fan 1 on pin fan1 and set its frequency
you can change "fan1" to "fan2" and it will use that connector instead.
It can be repaired, you can swap out the broken mosfet component. Some SMD soldering skills required.
But as you say there are two more to burn use. M106 configures the fans in RRF2 (and some M950 in RRF3)
Those IGUS bushings have pretty tight tolerances on how they are supposed to be mounted. There is a lot to read on their website.
I have tried them a few times, but as many of you say they either bind or have play.
They seem to be popular in the industry, "lubricated by dirt".
Awesome with another odd printer.
How is the rotational stiffness of your "tower"? It looks like it's one single 2040 which seems a bit skimpy. Those alu extrusion have very poor torsional stiffness, the rectangular ones like the 2040 you have are better. A round or square tube is a lot stiffer. It looks like you plan to add another rail on the other tower, probably a good idea.
How are you gearing down 1:30? Two step belt gearbox?