Lynxmotion SES V2 Hexapod Robot

Lynxmotion Legged Robots
, , , , , ,
255

Morning all,

I have not been making much progress on this, this last week. Been sort of distracted.

Been playing more with getting PS4 to be more reliable when it connects up to T3.6 or T4.x with Bluetooth. Still looking into some of this.

I have been wondering if maybe some of the time for others to join in and have things up and running, is due to me wanting to use a Teensy. Note: I personally game for most things, although personally I prefer to shy away from AVR 8 bit processors, like ATMega328 as the main processor. I personally prefer ARM either MicroControllers like Teensy, Some of the Adafruit, higher end Arduino Zero?, Portenta when released… Or Some full ARM boards, like: RPI, Odroid, Beagle… or when I was earlier playing with ROS, I liked using X86 based boards by UP, but there are others for example by Odroid, Intel NUC.

So also during this last week, I received an RPI4 8MB board, which I have now installed the latest Raspberry OS (32 bits), thought about trying the 64 bit beta. I have also updated the bios to allow it to boot from USB, and have it now booting off of an SSD.

So again not sure if others would feel more comfortable going down the RPI(3 or 4) route and use something to control the servos.

Could be the LSS adapter board, which all should already have.

Could be something like LSS210 board (I have one) Although…

Could be some other form of Arduino like board (Teensy,) Personally preferably something that has either two hardware UARTS and/or hardware Uart and built in USB. (i.e. you have connection to RPI or the like (by either USB or Hardware Serial port) and another UART for controlling the servos.

And again what to use as UI Controller. Again if something like RPI, that use Bluetooth, could use some form of controller. like PS3 or PS4… Could be something XBEE based, like Arbotix Controller, or?

Again the real question I am asking, is, is there another approach that would help many of you to get up and running?

Kurt

1 Like
256

Hi @kurte,

Wow, you’re going fast here :slight_smile:

I’m still waiting for my order to get in. And I’m aiming for the following setup:

  • LSS adapter board
  • Teensy 4.0
  • Xbee connected to my DIY remote (for now)

I think this setup will be the quickest way (for me) to get it up and running (or let’s start with walking). I’m planning to make the software modular so it will be easy to adapt to another controller later on. I do like to use my DIY xbee since it gives me more control (read precision) then a PSx controller. But a PSx controller will be easier to come by so easier for the public.

Same for the RPI, I would like to try that out, but first I’m going to get it to walk and explore the limits of the Teensy :slight_smile:

I got a message that the new battery will be delivered today. So I hope to do some calibrating this weekend!

I’ll keep you updated!

Xan

2 Likes
257

Hi @xan - Hope things are going well for you. sounds like you are making some progress. So far I have not worried about trying to mount a battery. When I get around to it, need to figure out probably where to mount a switch and the like.

I was using XBee before with the Trossen Arbotix Commander, which does all right. As for our old DIY remotes, yes the Joysticks are better, than many of the general controllers like the PS3 and PS4, but most of the time with my 10 thumbs, if I have more than maybe 4-5 sort of positions in each direction that is probably good enough for my usage.

I do like some of the other things we can do with things like the PS4, like the ability to use it’s built in Gyro and the like and you can play around with controlling things of robot by robot by tilting the controller and the like.

Another option that I have not tried in awhile is reasonably cheap RC receiver/transmitter. like one of the cheap a Flysky controllers where we can process the PPM or the like inputs from it.

But it will be great to get these up and running (or at least walking).

Kurt

2 Likes
258

May i suggest the RadioLink T8S with it’s PPM receiver ?

1 Like
259

That looks interesting. I don’t have that one. Earlier I picked up one of the others I mentioned as I think it was the user @jwatte of Trossen and PJRC mentioned them as an alternative to XBees. Especially for issues with XBees when you are trying to do bidirectional communications and at times ed up with a long time before a message is received… Also that it was better for things like Robogames as you don’t have as much issue with conflicts with the airwaves with other robots.

2 Likes
260

I hope everyone is doing well. We are still hunkered down here playing it safe.

Wondering if anyone has made any progress recently? And as mentioned wondering still what approach others are wanting to go.

I keep meaning to get back to playing. Been distracted with other things like figuring out issues with RA8876 display with images and SPI having problems. Solved this morning.

Trying to decide now if I should get back to trying to control using T4.1 and/or RPI.

Yesterday I also decided to pledge to an Kickstarter project, which campaign ended this morning.

https://www.kickstarter.com/projects/opencv/opencv-ai-kit/description

That looks like it can be interesting for this robot or my Turtlebot for doing stuff.

But I thought I would sort of ring out and see if anyone has made any progress.

Kurt

2 Likes
261

A quick follow up. If anyone else wishes that I had forwarded that kickstarter project earlier. I only noticed it yesterday as well.

And it looks like they may be allowing some pre-order stuff, not sure of price or the like, but the mention of it is up on their updates page:

1 Like
262

Interesting project - they are certainly doing well with funding.

1 Like
263

@cbenson - Yes it looked interesting. I did not notice that project until I received the kickstarter email that there was less than 24 hours left. Usually I toss most of these messages, but it looked interesting and I believe it will actually ship! Not sure about the 3d printer one I am still waiting on., which is only about 3 years late.

I forwarded the message to a couple of the people up on PJRC forum that I communicate a lot with, and one of them decided to go ahead and pledge… So I finally decided to do so also…

I find it is a lot a fun to work on projects like this, where ideas and code go back and forth. So hopefully will be a fun Winter project.

Right now back to enhancing some VL53L0x TOF library sensor code. I am trying to enhance the Adafruit library for this, to better support multiple sensors and doing non-blocking calls including starting up continuous ranging… Assuming we get them to work correctly may be fun to have one per leg… Note: the same person mentioned has these as well, so again fun to bounce it back and forth.

Now back to playing :smiley:

1 Like
264

Ah man. Missed that one but I did submit the form for the “I missed the Campaign” so perhaps I can still get in. I have an old Movidius stick and it was impressive…can’t wait to see what this can do with depth. We really need this kind of vision for our robots. :slight_smile: I would agree these guys will deliver and probably not too far past their date.

1 Like
265

I hope that they are able to add you on to the list. Sorry I did not post it earlier. At some point I will have to decide on which one… That is I signed up for the “D” one, but during the campaign they added some new variants of them. Including one with Wifi and BT. Also their FAQ talked about maybe some option, with UART, SPI, … Which will be interesting to find out more about.

But first hopefully get back to getting these guys to stand up and walk :wink:

1 Like
266

I’m only interested in the D version. I already have the equivalent of the entry level. My humanoid is missing a head, so the D one should look sweet with a little 3D printed T100 head…honestly the spacing of the D is a bit sloth-like, so maybe I should do Sid from Ice Age instead. Either way, this looks awesome. :slight_smile:

1 Like
267

For competitions, the 3D vision capability will be very important. LOL@ Sid from Ice Age…

1 Like
268

It indeed looks like Sid. haha

Very interesting camera setup! Thanks for sharing Kurt!

Here in the Netherlands the weather was extremely warm for a few weeks. Therefore I did not get much done. I do have all missing parts in so almost everything is connected and setup. I only need to replace the XBEE socket on the LSS adaptor board since one of the holes seems to be kluged up with solder or something. But for now I’ve connected the XBEE module using one of the sparkfun adapter boards.

I hope to have an update soon!

2 Likes
269

Just something small I wanted to share today…

WOW! What an enormous improvement going from the SSC-32 + Hitech to the LSS system.

Lynxmotion/Robotshop, you did an amazing job!

2 Likes
270

I’m glad you appreciate and thanks for the nice comment. We did put a lot of effort in this new series of products! :stuck_out_tongue:

1 Like
271

Glad you are starting to have some fun with it!

2 Likes
272

One question regarding the calibration of the servos.

I always like to calibrate the legs to be perfectly horizontal or vertical. which makes it easier to measure. I can move the move the leg horizontal en then write the current Origin offset to the servo. Perfect!

I see that it is also possible to configure the angular range of the servo. This is very useful since I use to have that as configuration in my code. The only thing I cannot figure out is if it is possible to configure the range, independent of the origin offset. I want to do this because the origin offset is not the middle of my range.

Thanks in advance!

1 Like
273

The only thing I cannot figure out is if it is possible to configure the range, independent of the origin offset.

@xan This exact point was actually discussed quite a bit during the design process (CLN and CLP commands to configure the negative and positive limits) were tested but not released with production firmware. This is the first time it has been raised, so we can see if it can be brought back in a future release. Although there are many questions which come up, the main two seemed to be:

Q1) If the servo receives a position command beyond the limit, would it ignore the command or move to the limit?

Q2) Why would the program send a position beyond the limit (limits to the position could be restricted software-side?)

2 Likes
274

I thought about it also as I have been playing around with the servos.

Yes you can (and maybe should) do software checks to make sure you are in range. But it is also nice to know that the servos are setup, such that if my software screws up, I won’t screw up the servos.

Again I am still more familiar with Dynamixel servos and also with these especially when debugging software another thing we might do is to set how much “Punch” the servos have. That is how hard will they try moving before they stop. I think that the LSS may have something similar with MMD command? I am not sure about some of the QMMD description?

Maximum Motor Duty ( MMD )

This command allows the user to limit the duty cycle value sent from the servo’s MCU to the DC Motor driver. The duty cycle limit value > can be within the range of 255 to 1023. The default value is 1023. A typical use-case for this command is active compliance.Ex: #5MMD512This will set the duty-cycle to 512 for servo with ID 5 for that session.

Query Maximum Motor Duty ( QMMD )Ex: #5QMMDD might return *5QMMD512This configuration command means the servo, when set to smart mode, will immediately move to 90.0 degrees upon power up. Sending a CFD command without a number (Ex. #5CFD) results in the servo remaining limp upon power up. In order to remove the first position, send no value, ex: #5CFD

I need to get back to experimenting!

1 Like