Malum IK

**jwatte** · 05-19-2019

I clicked the "version 1" option when downloading R+, so I'm hoping that's good?
The "scan" button first tries a version 2 scan, which doesn't find these servos.
And the "installing firmware" log ends up saying " ... Failed" at the end. (It also has an intermediate step talking about "making sure dynamixels are active" (or "awake?")
And the firmware recovery process doesn't work -- they're not seen by the recovery mechanism.

I'll try the version 2, too, to perhaps avoid having to send them back, but the failure messages are not encouraging.

What's the right Robotis forums these days, btw? It used to be that weird Korean software in English-language mode...

**jwatte** · 05-19-2019

Actually, R+ Manager 2.0 doesn't list MX-64 as a supported product.

**KurtEck** · 05-19-2019

Originally Posted by jwatte

Actually, R+ Manager 2.0 doesn't list MX-64 as a supported product.

I don't have any MX servos (unless you count the MX12w I have around here somewhere) so can not say very much... But, If I run R+ Manager 2...

And choose Dynamixel as product it shows support for MX-28, 64, 106 but only protocol 2...

**jwatte** · 05-20-2019

When I click "Dynamixel 2.0" it only lists:

XL-320
XM430-W210
XM430-W350

Maybe it can actually support others, too?

Anyway, I have an email in to customer service, we'll see what they say.

**KevinO** · 05-20-2019

Weird. I have MX64s and I flashed them to protocol 2.0. They have been working great.

**jwatte** · 05-22-2019

Glory shot:

**jwatte** · 05-24-2019

I'm in discussion with Robotis tech support about the servos that died when trying to update them.

A few things I really need to do, based on learnings:

1) A gun feeder that doesn't jam. I have a few ideas about how to go about it. We'll see.

2) A battery system for the controller that is slightly less janky. Right now, I'm using a 10 Ah "power bank" USB output taped to the back of the controller. (I could also put 4 AA batteries in, but those don't last that long ...)

3) A better/more portable video system. The 12" diversity receiver/panel I was using just ... failed. It showed data in a weird mosaic pattern. I had a 7" display with a separate receiver as back-up. However, something about this set-up (probably the ancient Boscam transmitter) still was janky and flickery. I need a larger screen, easier to set up for the control station. And I'm going to try with one of the "SmartAudio" (really, 4800 bps UART) transmitters that let me set the transmission frequency remotely. This might also be useful if we get "color based" channel allocations and need a simple way to change.

4) Various tightening up of not-quite-finished things. My gun barrel is still "floating" (the bucking + hop-up is all that holds it in) so the front clamp would be nice. There are a couple of add-in boards (driver for agitator, regulator for video transmitter) that are just taped for isolation and hanging free. The lens for the aiming camera is pretty heavy, and just suspended by the PCB of the camera sensor chip.

5) I'd like some control over zoom-in versus overview video mode. My current camera gives me very little peripheral vision, and indications are we might be limited to one camera per 'mech in upcoming events (channel crowding is a thing.) A way to switch zoom level, or switch between zoomed/wide cameras, would be nice.

6) Better PWM for the agitator. Currently, I'm using the Teensy SoftPWM library, which uses a PWM frequency of 30 Hz, because I didn't route enough PWM pins out to where I could get at them. Re-doing the board might be the best way. If I do, I could also integrate the agitator motor driver and the video switcher and the video transmitter buck converter on the same board.

7) Bumper wires to protect my side panels against killing myself by bumping into things. (This happened during fighting last week-end.)

8) Longer wires for the scoring panel harness. The wires were short just a little bit of slack, making installation a lot more frustrating than necessary.

More cray-cray things to consider:

- anodizing the aluminum with dye would look good
- an RPLidar would give me a good idea of where the sides are
- video overlay, get a hackable FPV OSD, tell it what to display (voltages, servo errors, distance measured to sides, detected hits, ...)
- Xbee remote command line interface (not that hard to make, but I didn't really need it much, so it seems superfluous)
- printed shell of some sort to look cool
- move gun, and perhaps BB system, onto a stabilized gimbal, for move-and-shoot capability

The good news is that the next event will probably be at maker faire New York in September, so I have lots of time to vascillate and postpone, until the next two-week deadline panic appears!