Testing p.Brain and SMB with Lexapod

[Matt Denton]

Nice one Paul!

What kind of frame rate are you sending the PIP packets?

[paulp]

they're on a timer at present, 100ms so approx 10Hz

It will go quicker but I've got the baud rate at 9600 so might get overruns.

[paulp]

Testing the Auxilliary servos over PIP

To get this going I've linked my old OpenCV face tracking application to drive the p.Brain.

Its a bit slow on response because of the OpenCV code not the p.Brain. p.Brain responds nice and quickly

Again apologies for video quality and lack of action but it proves the functionality

[youtube=WkuI9fwob8o][/youtube]

[Matt Denton]

Nice one Paul!

I presume your sending a Body PIP followed immediately by a Servo PIP, or is there some delay between them?

I hope to have V0.4 ready for you to test by end of week, several new features along with a couple of new PIP commands.

[paulp]

Its servo then body....

I look forward to the update

[jaspinall]

Chomping at the bit here, waiting for the USPS to drop a lil box of servos through my letterbox - all these pesky update videos do is make me chomp some more! lol... Good work, Paul

[paulp]

Another test in place...

this time using the new 'V' command.

The video shows Lex and p.Brain interpreting the infamous roadrunner.tap file from Mach 3.

The NC interpretation is very crude as is the pen attachment but it proves the resolution of the translations.

[youtube=gpNkvu_4LZQ][/youtube]

[Matt Denton]

Great job Paul!

[florek]

I selected the leg, in this case leg 1 and start the refresh. This applied power to the leg. I then manually positioned the servo horns to get as close to the 90 degrees as possible. When I had screwed the horns firmly to the nearest position, I adjusted the 3 joints via the software to fine tune the position. The position starts at 0 and can be offset from -127 to +127.

So it's made on 8bit timer.. that gives 180 degrees /256 = 0.7 degrees resolution right? What's the minimum resolution a servo can get? (analogue, digital)? I am just curious.. I've never checked that..

[Matt Denton]

I selected the leg, in this case leg 1 and start the refresh. This applied power to the leg. I then manually positioned the servo horns to get as close to the 90 degrees as possible. When I had screwed the horns firmly to the nearest position, I adjusted the 3 joints via the software to fine tune the position. The position starts at 0 and can be offset from -127 to +127.

So it's made on 8bit timer.. that gives 180 degrees /256 = 0.7 degrees resolution right? What's the minimum resolution a servo can get? (analogue, digital)? I am just curious.. I've never checked that..

No the servo calibration trim is 8 bit, but the servo resolution is 1us, which equates to approximately 0.1 degrees per bit. However you will be hard pushed to get this resolution from most analogue or digital servos. You will probably see more like 2 to 3 uS steps, with digital servos performing slightly better.