Thank you for your kind works of appreciation, Nhan! Unfortunately, I had to stop the development for a while, but I hope to be back on track in a couple of weeks, to finish the electronics. I'll be happy to return here to present the remainder of the project.

Good luck ING, I have a similar setup, and I will test it during this week - probably not the whole robot yet, but I will leave here some feedback..

The values from the encoder might change, but might be random, and not reflecting the real movement. I never tested these encoders with an undervoltage supply, but the recommended range is there for a reason :).

Just two questions: 1. Did you modify the switch limiters' voltage to 3.3V? 2. Did you install any level shifters on the CUI encoders outputs? Having 5V on the corresponding inputs of Teensy v4.1 is bringing the microcontroller on that board outside the safe limits, and may produce unexpected behaviours, or even permanent failure of the Teensy board.

I realize that this thread should rather be in the Robot Builds section, but I started it here, so probably I need to keep it here. My robotic arm build advanced nicely, even if slower than what I would had hoped for. But time is scarce, and nights are short. I finished the arm side, being ready to advance to the electronic box side. Just as background, I purchased from @Chris Annin the AR4 kit, and ordered, as everybody else, the AR4 stepper motors and driver kit from StepperOnline. Due to the delays still affecting the delivery of the stepper motors, I switched to the AR3 stepper motors and drivers kit, and that was delivered promptly by StepperOnline. I consider that the modifications required by the CUI encoders to be compatible with Teensy 4.1 to be quite simple, and nothing else is affected. There were some discussions regarding the limit switches, but I will explain later why this is not an issue at all. So, I don't know how to call this, an AR4 with AR3 electronics, or and AR3 with AR4 mechanics :). I also added a small camera on J6, hoping that in the near future I will be able to add a Jetson Nano with OpenCV for image recognition based activities. All the STL files are attached to this post. Here are some images of the implementation. - Breakout box - J5 enclosure side: - J2 enclosure side: - J2 back: - Camera front: - Camera top - View from the top: - Covers (these were re-designed to use Delrin as base, and the inscriptions were still done as 3D prints, insertedinto pockets of the same profile carved into the Delrin bases). All the plastics are CNC-ed from Delrin, apart from the inscriptions, J5 motor bracket, and a supplementary J4 Cable Bracket that I designed in order to keep the cable away from the J5 motor: Regarding the camera - I purchased the Arducam 8MP / 1080P autofocus USB camera, that is UVC compatible, and compatible with JetsonNano (https://www.amazon.co.uk/gp/product/B07YHJK4LN/ref=ppx_yo_dt_b_asin_title_o07_s00?ie=UTF8&psc=1, but this is available from a multitude of other sources). I removed the 38 mm pcb border frame, to reduce the size to 32 mm, and designed the adjustable angle enclosure that fits on top of the J6 limit switch, and also can support the J6 cable. The attached STLs are including the camera enclosure, and can be 3D printed as well, though the thicknesses might be to small for such process. Regarding the limit switches, I am still to test, but I believe that it is an overkill to supply the switches with both ground and a DC voltage. It should be enough to have a ground and the active switch connection (NC in the way Chris designed the Teensy sketch), at the other hand using the input pull-ups of the Teensy board. This way, the switches system would be compatible both with 5V and 3V3 systems, without further adaptation, and with a wire less for each limit sensor. More, this would also avoid the intermediary switching state, when the signal would be floating, As said, I will need to test to see if there are any unwanted effects, but any previsible effect can be avoided with a debouncing routine in the sketch. Finally, here is the archive with all the STLs I used for cutting the parts. I am happy to support any questions if somebody wishes to reproduce this work, I am sure that there are some mistakes, as this work is by far not as refined as Chris's work. I will continue posting the electrical implementation. All the best!

Hi All, The assembly and re-design for the parts that I decided to cut on CNC goes ahead, even if slower than I would like - yes, it is true, I have to burn the midnight oil for this project, and the weekends are invaluable for the progress. Nevertheless, there is some advance, which I wish to share here. Firstly, I just have to say how amazed I am to see the amount of work, the details and thoroughness that @Chris Annin put into this project – thank you so much, Chris! The slight modifications and additions that I made are not a critic, but rather adaptations to a different process, and some different perspectives approaches. By no means the big picture is not changed at all, all the functionalities will still be the same as designed. Form a different perspective, I know that Chris’ approach is cheaper and sometimes more efficient, but I don’t intend to show a cheaper design, but rather an adaptation to a somewhat different approach. Here we go… I made the “J1 base enclosure” cut from a single piece of Delrin. I added a bottom part / cover, with captive M4 nuts, so the J1 assemble can be fully closed. I also redesigned the break-out box section, to make it easier to be CNC processed. It is a tad smaller and narrower, but that means it will be more difficult to assemble the wiring – not a problem for myself, as I am dealing in my trade with microelectronics design and prototyping. I created a protector for J1 limit switch, to avoid damaging it in care the turret turns more than expected. I also modified the J2 protector, the “J2 Stop” detail was just a bit too complicated to be processed by CNC – of course, this is just an alternative way of achieving the same function. I will redesign the “J2 Stop” detail to make it decorative. I also advanced on the electrical assembly. I decided to install everything in a 6U wall mounted 19” rack, and I am just adapting the internals for now. I made DIN rail adapters for the drivers - ok, I know that it might be possible to find them ready made, somewhere, I just could not find something that was “off-the-shelf”. I also prepared the field for installing all the other components in the same rack, the image with the drivers is not showing that there are three DIN rails in the same shelf, plenty to support all that’s needed for AR4. I also adapted the connector’s panel for the rack. It might look odd that I have 8 drivers instead of the usual 6 or 7. In fact, I do have the configuration prepared to drive a 7th axis (the “rail” as described by Chris), for which I added a closed loop driver (CL57T) that will support a NEMA 23 closed loop stepper motor – this might look as an overkill, but I did have it at hand and decided to use it. The 8th driver’s story is a bit different: the DM542T drivers are able to support 50VDC supply, only that the DM320T, used to drive the NEMA 11 stepper motor supports 30VDC maximum. Using DM320T driver is the reason to use a 24V power supply only, when, in fact, the motors can benefit from a higher voltage – greater acceleration and higher maximum speed are achievable. One might ask the question on why not use a DM542T driver for the NEMA 11 motor – well, that’s impossible, as the minimum current that DM542T is configurable to is 1A, while the NEMA 11 motor accepts maximum 0.3A. My intention is to modify a DM542T to supply less current, but that’s not an action that Chris could recommend, nor the supplier of the drivers can support. I will try the modification, if that works I will then omit the DM320T driver, and let you know how that was done, but at a later stage. I’ll keep posting the progress here. Have great success with your builds! L.E. As promised, please see attached the STL files for the parts already made and tested.

Here is an image, and a 3D pdf of the breakout box and the base. I included the Aluminum J1 base plate, as delivered in the kit. There is a slight change to the structure: to make the box smaller I replaced the PG-21 cable glands supplied by @Chris Annin, with PG-19, as I will also use some thinner, flexible silicone isolated wires for the signals (AWG 30), and some thicker silicone wires for the motors (AWG 22). It looks like these will fit fine in the smaller gland nuts. Unfortunately, I will be not in the position to release directly the step files, I will let @Chris Annin to do this, if and when he will decide so. I might be able to release the cutting files for Mach3 based routers, as executable files, as Chris released the STL files, but again, with his permission, or let him sell these, as he will decide. The intended material to be used is Delrin / Acetal for the plastic parts, and double sided 1.6 mm FR4 bare PCB board for the thin sections. P.S. To see correctly the 3D pdf, you need to open it in Adobe Reader, not in the browser, and to enable the 3D content.

@gio s. I am in the same boat, I don't have yet the electrical package. To me, the re-routing of the limit switches to 3.3V was obvious, but thank you for the reference. The issue I am describing, with the encoders running on higher voltage than 3.3V is a bit subtle, and might not bite back immediately, but sending signals (A & B from each encoder) at 5V to the 3.3 inputs of Teensy v4.1 might damage the board in time (and, sometimes, immediately). That's why I will not even try to connect directly the CUI encoders' outputs to Teensy, but just through a level shifting circuit, like in below image (that has to be repeated for each of the encoders' outputs, 12 in total, two 74LS07 ICs will suffice): Please pay attention, the buffer used here as level shifter has to be open-collector, or open-drain. Other types of level shifters are available, but this would be one of the cheapest and most available. I will post the whole schematics and assembly when I will have it ready. Doing this, still needs adaptation for the limit switches, either supply them from 3.3V (the simplest and cheapest way), or add another level shifter for each limit switch. Regarding the steps to follow, you need to understand exactly what each steps means, Chris was very careful to describe everything in detail, but when combining the AR3 kit with the AR4 controller (aka Teensy), things might become a bit hairy if you don't know what you are doing. I might return the favor that Chris done to all of us, and create the step-by-step instructions on how to setup the combined setup (AR3 electrical with Teensy 4.1) - but this will take some time, unfortunately.

The CUI encoders have push-pull outputs, the high level voltage is actually higher than 3.3V - it could be very close to 5V. Teensy v4.1 is not 5V tolerant, so level shifters should be added to the phase signals (A and B, it looks like the index signal X is not used). I will post here the schematics after I will test the adapters, but this might take some time, as I don't even have the kit here. Mot likely a open drain buffer with pull up to 3.3V will work, but we'll see.

I just changed my order for AR4 steppers kit to AR3 steppers kit, I was lucky to be able to buy it from Germany, and together with the CUI encoders all was working actually cheaper for me. I am happy with the electrical changes necessary on supplying the CUI encoders, there's no rocket science there, and Chris did a terrific job explaining how to connect these. One question though, I might have to experiment with, the settings for the CUI encoders are for 512 steps for revolution, but might well be set for 1000 steps per revolution - that would keep them compatible with the AR4 encoders. The limitation of 512 might be due to the transmission wires, as I don't believe that the 7500 RPM acceptable for 1000 steps per revolution, compared to 15000 RPM for 512 steps per revolution would make any difference.

I just changed the order for AR4 stepper kit to AR3 stepper kit, bought from Mouser the encoders and connectors, and all should be right. I believe I can live with the bigger encoder right now, the issues with powering these encoders are simple to sort out, and Chris was good to detail them very well in the manual for AR3. And I trust a lot the CUI encoders, I used them a lot in other projects.