This is a temporary lash-up to check the mechanics - the electronic parts are not fitted yet. For the first time ever I have no noticeable effector tilt on my delta. This is after upgrading to Haydn Huntley's magnetic arms, PCB effector and carriage adapters, and linear rails instead of wheeled carriages.
I hope to get the electronics assembled and tested before the end of this week.
Duet WiFi hardware designer and firmware engineerhttp://www.escher3d.comhttps://miscsolutions.wordpress.com
It looks great David. I have been meaning to ask, but haven't yet as it might be a silly question - is a PCB (and they don't look that thick) stiffer than metal? What do you see as the main advantage to using these?
Last edited by DjDemonD (14 March 2017 15:52)
Yes the PCB is very stiff. I see the following advantages of using PCBs for this purpose:
1. As with CNC'd metal parts, they are fabricated to precise dimensions, so that you can get the rod spacing the same at top and bottom by design. But they probably cost less to mass-produce.
2. They are lighter than a metal effector for the same stiffness.
3. We can mount connectors and electronic parts on them, including the electronics for piezo or other nozzle-contact sensors.
Looks lightweight, compact, shouldn't flex (due to how compact it is), and precise. Nicely done.
Thanks, that makes sense. I'll definitely try some as/when you have them for testing.
Is that a custom E3D Heatsink David?
Yes, it's a special made for us by E3D, hopefully to become a standard option when both companies are happy with the design.
i like the spirit level, I guess you'll replace it with a BMI160 or something !bmi160s on the carriage adapters would be ace as well 😁4 IMUs to control delta calibration, vibration issues and provide z probe facility, whooohooo that'd be nice ! ;-)
Last edited by lolorc (14 March 2017 20:18)
We considered using an accelerometer for both bed probing and effector tilt measurement. But the word I have is that accelerometer based probing doesn't work well on softer surfaces such as PEI.
You can count on me to order a set, that looks fantastic! (I have nearly the same hardware down to the 360mm arms)
Being able to mount things to it with standard holes is a huge plus. Like fan brackets and the differential IR sensor
Last edited by firstname.lastname@example.org (15 March 2017 12:36)
I'am interested too!
Any time frame when these come available? Also, can you tell something about how you envisioned to attach fans?
Definitely interested in a set as well.
That looks promising!
Ditto. Looks clean and light!
On the effector pictured, I envisage a printed part to support fans being attached below the effector using the tails of the ball studs. Or you could clamp a fan assembly to the hot end heatsink. We also have another design prototype with additional mounting holes in it.
Is there an eta on availability/approximate cost.
Guess I know what I am buying in the near future then ;D
I got the built-in nozzle contact sensor working today. If I move the nozzle to 0.1mm below Z=0 then the probe sometimes triggers, and if I move it to 0.125mm below Z-0 then it always triggers. I think this is probably a reasonable sensitivity, because I am using a PEI bed. PEI is known for being soft and causing difficulties with accelerometer-based sensors because of that. I've already identified some design improvements to make the next prototype more sensitive.
We don't yet know what the cost will be. If we can avoid the expense and risk of drilling piezo disks - as we hope to - then it will help keep the manufacturing cost down. At present the component cost is reasonable, but the manufacturing cost may be a little high because there are SMD parts on one side and through-hole connectors on both sides. So at least 2 of the connectors will have to be hand-soldered.
I've also been talking to E3D about a slightly modified hot end mounting method that puts less of the hot end below the effector than the current setup.
As for ETA, we need at least one more round of prototyping and some beta testing before we go to manufacture. Given the turnaround time on PCBs, that makes it at least 8 weeks away from general availability.
I think it's reasonable to sell something like that with the hand soldered pieces separate, and charge a small fee to solder them. That's not totally uncommon (Headers on many breakouts, for example) and people at a place in this hobby to want to buy your effectors should be able to handle basic through-hole soldering.
I'm fine with no piezos. And let me know if you need any prototypes on this side of the pond- there's a few PCB houses in california that have done PCB prototypes for me for very cheap.
Any chance you or T3P3 will make custom length Haydn Huntley magnetic arms and sell it all as a kit with the PCB?
I used the onboard nozzle contact sensor to do auto calibration and mesh bed levelling on my 300mm diameter bed delta today. After I remembered to remove the H corrections from bed.g and put in the new diagonal rod length, the deviation for both was around 20um. I'll do more tests tomorrow to check how reproducible the sensor is.
We will certainly explore collaborating with Haydn on purchasing his mag arms in larger volumes to bundle with the effectors, once we get closer to a production version. That makes sense for people in the UK/Europe at least (it makes less sense to import them from the US just to ship them back).
Duet Wifi Hardware Designerwww.duet3d.comwww.think3dprint3d.com
I'll keep my eye one this!
I'm curious about that custom e3d heatsink. Does it have the piezo disk integrated? I'd love an e3d hotend with simply three screws to mount it rather than the groove mount, but I'm not certain that I'd like to experiment with piezo disks. Can you clarify what that is in the photos?
I don't like the custom heatsink, not easy to leave a standard.