My take on the @home shows. Plus TMI.

Finally getting back to it. Obvious first step is an ominous and overly dramatic video.

The lack of progress on this project is due to the passing of my precious friend Shadow. I was lucky(?) enough to be with him at the end. It was very peaceful and Shadow mercifully spared us from an agonising decision. Shadow was adopted from the RSPCA and enjoyed a very long life with a loving family. These photos show just a few of so many wonderful moments with the eternally patient Shadow as he shared growth into adulthood with our boys (now in their 20’s), and us.

Experimental camera gimbal for SpotCorgi’s head. It works OK but is still very rough. And, yes, I know you saw the Band-Aid. Guess who thought he’d get away without his cut resistant glove for this one little job. As my motorcycling colleagues say, ATGATT.

Every Corgi Needs A Body This is the first pass at completing SpotCorgi’s body. I hear you ask, why suddenly switch to finishing the body when SpotCorgi was so close to walking? Well it happened like this, as testing progressed various gremlins appeared. First, structural weaknesses lead to SC looking like it had a crippled leg, so I had to do some repairs. Then, having the Mega hanging from the body was messing things up with wires shaking loose, etc. I decided it was time to complete the elec

Quick teaser. I’ve got SpotCorgi walking again but still need to tweak some variables to smoothen things out. Thankfully, the long time I’ve spent on the code and remote means these can be done quickly on-the-fly. As I said previously, with a least six inter-dependant variables to adjust, avoiding the edit, compile, test routine meant it was time well invested. Next task is to code for turning and then the code for the running chassis will be almost complete but I’m sure I’ll get diverted and

SpotCorgi finally performs! The Chassis including parts list is on Thingiverse at
: https://www.thingiverse.com/thing:3900070 Since this design has gone through so many iterations there may be some gremlins in the design. Please let me know if you find any so I can correct them. The Phone Tripod Mount is also up. It fits a Galaxy S8 but could be easily modified to fit other phones:
https://www.thingiverse.com/thing:3899863 And sorry/not sorry for the intro. I was so embarassed about the power

Overview of my redesigned Spot, the Spot-Corgi. All new parts except for the reduced-length legs from the original KDY model. I’ve created a new base plate with 18650 battery holder on the top and circuit board holders on the bottom. It’s designed to mount PCA9685 servo controller, voltage regulator and Arduino Nano. Currently it’s running from an off-board Mega due to my frying my Nano. Also new bearing supported hips. Servos :
https://www.aliexpress.com/item/32820041789.html?spm=a2g0s.9042311

I missed my prediction of a new running chassis by about 24hrs but am still making progress on the redesigned Spot. Changes include: – shorter legs to reduce servo load – removal of hip servos to simplify and lighten – electronics mounted to the underside of the plate – changed Mega + Sensor Shield to Nano + PCA9685 + power regulator (IR and/or Bluetooth, ultrasonics, camera and accelerometers to come, probably mounted in the snout) – shorter and narrower body – inclusion of 18650 battery holder.

I’m continuing work on an updated design to simplify design and reduce weight. Includes using a Nano & PCA 9685 , adding on-board battery holder and power regulator but mainly shorter legs and removing hip servos (in the video I mention removing two servos but it’s two per side).

During testing you will sometimes (often?) drive a servo past its movement limits and it will stop responding. There are two main reasons for this: the servo is just stuck and can’t move back, or the motor has been damaged by pushing too hard against the stop for too long. This video shows how to correct the first situation and determine if the damage is more terminal.

Or maybe he can’t. But he does struggle bravely to slip the surly bonds of earth.

The walking routine is sorted with leg angles and sequence pretty much what I was aiming for. But… way back when, I decided without too much thought to use hi-speed servos over hi-torque which is why the motion is quite jerky. I did say at the time that I may have to revisit that choice and, well, that time has come. But I’ll get to that in the next video. Overall despite, or maybe because of, all my mis-steps I feel the build has turned out OK. Alright, at this time it’s just a static model

The missing vision from my previous video. I know, to be truly scientific I should do a demonstration without the glove to provide a comparison but I suspect I would run hard into Youtube’s blood and gore ban. Do I need to say DON’T try this yourself without the glove? Anyone who knows how sharp X-acto knives can be will easily see that if I did this demo without the glove I would have cut enough veins and tendons to permanently cripple my hand. However, while you can see this glove provides

Poorly framed (again sorry) video about setting servo park angles before installation and, trust me, cut resistant gloves plus closing in on the correct leg angles. The camera-man has been sacked for gross incompetence. https://www.aliexpress.com/item/32921813316.html?spm=a2g0s.9042311.0.0.27424c4djgYsZP

Sorry about the terrible framing, I was eager to get back to work! Although I did come up with a mathematical formula using sines and arcsines to calculate leg angles it is too dependent on the individual servo angles. The monkey who assembled my Spot didn’t get everything lined up perfectly and I guessed that tweeking the formula would take longer than setting the angles by sight. So that’s what I’m doing now after more sidetracks with faulty servos, etc. With 12 servos you can count on at leas

In my last video I made two predictions: failure and walking. Turns out more of the former than the latter. So, more failures to report. As a prelude to Spot confidently striding around, I decided to do a simple sit and stand demo. The sit was more a collapse and unlike me initially, eagle eyed viewers will see why Spot wouldn’t stand. Still, progress is being made and I’m learning more. One thing I did learn as took a break after this failure was that my walking model was wrong. I’d programm

Almost time to cut the strings and go k = walkies(steps); The code to do the actions demonstrated runs to just over 300 lines though a lot of it is logging/monitoring dross that won’t be needed in the final version. Happy to answer any questions or share parts of the code. These are the parts I’m using : Servos Mg996r Mg996 Servo Metal Gear (I think I’ve ended up with a mix of these types). https://www.aliexpress.com/item/32636102294.html?spm=a2g0s.9042311.0.0.27424c4dyVbnTj Servo cable extend

Quick teaser. After many delays, distractions and side-tracks it’s back. The code for the walking is 75% done, just need to add in the paced servo moves from my manipulator experiment. Hilarious when testing, I got the angles wrong and it tried to kick itself in the nuts but only succeed in breaking it’s rear left hip. Again, weakness due to printing with 10% fill to get a quick build. The PCA9685 is great, can control all 12 servos with 2 wires (plus power). Plus, you can daisy-chain other devi

So many failures! But finally, a decent working prototype. As I said previously, while I was waiting for parts for the Spot Micro I went back to working on the arm. Wow, so many iterations of hardware, electronics and software. Firstly, hardware. I’m zeroing in on a reasonable design for the arm itself. I’m happy with the base rotation mechanism and the mounts for the gripper servos in the upper arm but that’s it. The upper arm elevation works OK but is still very messy around the servo attach

A lot has happened but unfortunately none of it what I’d call progress. The very first thing I ordered at the start of this project was a PCA9685 16CH PWM Servo Driver. I’m still waiting for it to arrive, grrr. While I was waiting I tried to press ahead with work-arounds but the results were not very satisfying. https://www.aliexpress.com/wholesale?catId=0&initiative_id=SB_20190531060945&SearchText=PCA9685+ I tried setting up a breadboard to run the servos off the Arduino power but after adding

As I said previously, my first build was definitely a prototype. Hence, I opted for speed over solidarity. This shows why my legs fell off; can’t remember which but obviously only 10-20% infill. I also created but have yet to print and test a side cover incorporating an opening for a 12v power supply to use during testing: https://www.thingiverse.com/thing:3644573

These are the servos I’m using:
https://www.aliexpress.com/item/4PCS-MG996-MG-996R-MG996R-Metal-Gears-Digital-RC-Servo-Motor-High-Torque-Helicopter-Car-Boat/32954064687.html?spm=a2g0s.9042311.0.0.27424c4dGyutWk. I got the Hi-Speed for the legs and the Hi-Torque for a planned manipulator arm. Time will tell if I need to rethink this choice. I found that some of the servo nuts vibrated loose so used a drop of blue Loctite on them. I also attached an external 12v power supply for use during testi

Making some progress. New servos came so I completed partial chassis assembly. Mega arrived too so I built a quick control test bed with one pot each for hip, thigh and knee joints (on both sides). Results were somewhat hilarious. Full disclosure, despite many years experience in IT and basics of several programming languages, I’m a complete newbie to Arduinos and robotics. I purchased the Arduino Beginners kit but only got as far as Project 5 before deciding to build Spot. I think I’m doing OK

Very rough first draft of a SpotMicro manipulator arm using very old, very weak servos (hence the upside-down gravity assist). Built while waiting for the proper servos for the SpotMicro to arrive.