Help with upgrades and Haydn arms

[zakfarias]

Okay so as I have said before I am new to the delta world, however, I have built plenty of CoreXY, Prusa style cartesian, and UM2 clone machines (have 6 running currently for production parts)
Delta though...they stop me like a brick wall when it comes to diagnosing and figuring out.
I have a Rostock MAX V3 that has had calibrations since day 1 when I built it a couple months ago. SeeMe should be sending some replacement part for the bed any day now, in the meantime though, I have had pretty naughty ideas in my boredom and ended up raiding my spare parts drawer to see what mods I could do...
I have enough spare parts to build: (note none of these have been installed yet (except the Haydn rods [failure?])
* either a 1.75mm E3D-V6 or 3.00mm E3D-V6
* Titan extruder leftover from when I put a Bondtech on the C-BOT
* plenty of spare NEMA 17 2.0A 76oz (don't think these will be needed really)
* I have a set of trick trucks on the way from trick laser
* Maker713 effector
* Berd air max kit (newest one)
* Spare accelerometerr board from SeeMe
* Haydn 360mm mag arms

What I would like is some opinions on the following
1.75 or 3.0mm?
EZR or Titan Bowden (or possibly direct???)?
Firmware changes for the E3D (all my other machines run DUET so I am not sure what is needed on the RAMBO (aka noob))
Upgrade routes to approach here, Since I have the parts is it worth it? Curently, until new parts show up the printer is pretty much none functional unless I go back to original arms and print with a very nice, thick, ugly raft

Issues I have found so far (this is where I say that Deltas confuse me)
I installed the Haydn arms while I was bored during hurricane IRMA over the weekend. I tapped and installed the magballs directly into the barbell screw holes for the stock carriage and cheapskates. I updated the EEPROM for 360mm rod length and set the radius to about 155 for testing. Went to calibrate and the effector goes way off of the bed and smacks into the X tower hard enough to knock a few rods loose/off when trying to probe. For the life of me I cannot figure out why? I would really like to use this machine as my personal use machine for fun since all my other machines are pretty much dedicated to specific tasks/projects 24/7


Sorry if everything is jumbled up and out of whack. I have a habit of typing my thoughts as they go, so pardon me if I had any kind of wild tangents

[Xenocrates]

If you don't already have a large volume of printers on 3.0, I would go for 1.75. The resolution of extrusion is better (Any given size of step results in less plastic, and thus more control at low flow rates), the heat transfer characteristics due to a higher surface area to volume ratio, and at this point more available and far better developed. But if you already have to buy 3.0 filament and stock it, but don't have to do so for 1.75, the ability to use any given spool on any of your printers is far more useful from my perspective (I believe it would be a major pain in the ass, considering that for my two printers, I have a total of 6 5-gallon buckets filled with different filament spools, to try to keep a stock of them, and keep them dry. For 6-7 printers, it's probably even worse. )

The general consensus is usually that a shorter (Or completely eliminated) bowden tube is better), so if you have a pancake stepper in that pile, you might want to use it with the titan for either a flying (Suspended from the trucks above the hotend), or direct extruder.

Firmware changes for the E3D are easy, it's pretty much just retuning the PID loop and changing the sensor type to 8.

Since you have a 713 Maker effector, I would definetely use it, and with the Berd air Max, you might find my design here useful, since it allows you to use a pair of M3 grubscrews to hold the berd air loop in place to the 713 effector, and looks not entirely like garbage.

[zakfarias]

Darn it! I was hoping to have a good excuse to use my 3mm setup. haha. You're right though, stocking filament would become a much bigger hassle... having to have 2 sizes of ABS, PETG, and the like around. (all of my other machines have been converted to run 1.75mm)

Is anyone perhaps running a direct drive titan on the 713 maker effector? any links to necessary files perhaps? I have a couple 1.8 degree pancakes and a spare E3d 0.9degree pancake with two boogered-up mounting holes (probably just re-tap to a 4mm) If not, links for a flying setup that would work in this scenario?

And yup definitely collecting the file you linked for the Berd.

[dc42]

If you are going for the Haydn arms and Duet electronics, you may also wish to consider the Smart Effector and carriage adapters, see https://www.duet3d.com/DuetAddons/DeltaSmartEffector.

We don't yet have a solution for mounting a direct drive extruder on the effector, but some users have mounted the Nimble remote direct drive extruder on it.

[zakfarias]

@dc42 - I have heavily considered that route. I thought I would wait a bit to see the crazy people do their crazy things with the smart effector setup(and see what mods are born), that and I just installed my last spare duet WiFi a couple weeks ago. Production machines will have to recuperate costs some before I can get more boards. And then the nightmare of rewiring for 24v and 0.9 degree stepped would begin again. (Considering I just recently put the Rostock together) But it is the only machine i have in the stable that doesnt run duet

[dc42]

@dc42 - I have heavily considered that route. I thought I would wait a bit to see the crazy people do their crazy things with the smart effector setup(and see what mods are born), that and I just installed my last spare duet WiFi a couple weeks ago. Production machines will have to recuperate costs some before I can get more boards. And then the nightmare of rewiring for 24v and 0.9 degree stepped would begin again. (Considering I just recently put the Rostock together) But it is the only machine i have in the stable that doesnt run duet

You don't have to use 0.9deg motors or 24V with the smart effector, it's compatible with 12V systems too.

[zakfarias]

Okay so an update:

I installed the Maker 713 effector with an E3D-V6 hotend (can't seem to fit the hot end cooling fan on it though)
Also newly installed
360mm Haydn magball arms
trick trucks from trick laser
New Rev 10 heated bed

The machine now heats up past 60 and gets 21- 100 in about 6-8 minutes which is a vast improvement over the 45 minutes to reach 70 with my original heat bed.

On the haydn arms I had to use 10mm magballs for the rods to attach to the trick trucks since they were tapped for M4 (haydn magballs are 9.57mm and M3) I used Haydns magballs to attach to effector since it was pre-tapped for M3.

Haydn rods were marked as being 360.21mm So I put that into the firmware +0.22mm to account for magball radius difference at the top end. I got a vague horizontal measurement of 134.8mm using good ole triangle maths using the diagonal rod length + distance of slider magball center to glass bed with nozzle at z zero. I input this into the eeprom, ran g28 and g29 and HOORAY it ran calibration without crashing. Horizontal rod radius ended up being re-calculated to 140.8. Then I ran both the octoprint delta calibration plugin as well as SeeMeCNC's online delta calibration wizard. The best corrected error I could get was 0.41

What I found was that at the interior probing points of the bed as well as the Z tower I would get values ranging from -0.04mm to 0.05mm (pretty good) However, starting with the first point probed after the Z tower I would get 0.45mm followed by -0.49mm. It follows this alternating up and down pattern around the entire circumference; the positive values range from 0.44-0.46mm and the negatives range from -0.48 to -0.51mm

I changed the diagonal rod length to 1% longer and the values changed such that the center area was 0.21mm and the circumference went 0.00mm followed by 0.88 mm repeating much the same as before.
I ran through the calibration wizard somewhere in the ballpark of 20 times last night. Best final error deviation was 0.41 with the typical being 0.44