Mark's Antweight build

All things antweight

Moderators: BeligerAnt, petec, administrator

Post Reply
User avatar
MarkR
Posts: 375
Joined: Mon Dec 18, 2017 12:46 pm
Location: Reading Hackspace
Contact:

Re: Mark's Antweight build

Post by MarkR »

That's quite a nice H-bridge board, it doesn't require a second 5v supply, could run directly from the Lipo voltage, which is nice. I'm not sure about the current capability though.

The h-bridge chip SN754410 is apparently quite good (and supports the 3.3v logic level, which is important) - but needs a 5v supply too (=more wires)

I spent some time last night shortening and heatshrinking wires to try to fit everything in. It almost fits (also I made the robot a bit bigger.. still ok for the 4 inch cube)
Robots: Betsie - RaspberryPi controlled flipper bot with gyro stablisation - too clever for her own good?
Stacie - tidy flipper; 4wd driven by hair bands
User avatar
MarkR
Posts: 375
Joined: Mon Dec 18, 2017 12:46 pm
Location: Reading Hackspace
Contact:

Re: Mark's Antweight build

Post by MarkR »

Ok,

So the good news - the DasMikro did not melt, and runs the Raspberry Pi, and the Pi can drive the motors.

The Pi now sort-of fits in the robot (which fits in the cube, I think). The wheels stay on, the battery will fit.

I've got some 300mAh lipos which are clearly massively overspec, I would definitely try to get smaller packs in future.

Surprisingly, my remaining parts from China arrived about 2 weeks early, so I now have all of the 6x motors required. There is still a lot of fiddling to do with motor-mounts and wires.

I am honestly surprised at the amount of faffing around with wires involved. it seems mostly soldering heatshrinking and other kinds of fannying around to get small parts with tight tolerances to fit.

Still, keeps me off the street.

Software is another issue. I got the cheapo ripoff PS3 controller, and I can't make it speak Bluetooth to the Rapsberry Pi. This is not, however, a disaster, as I can employ a 2nd Raspberry Pi as a radio relay and use my own software on both sides.
Robots: Betsie - RaspberryPi controlled flipper bot with gyro stablisation - too clever for her own good?
Stacie - tidy flipper; 4wd driven by hair bands
Paulmchurd
Posts: 227
Joined: Sun Dec 10, 2017 4:39 pm

Re: Mark's Antweight build

Post by Paulmchurd »

Nice, maybe you’ve received my motors that I order 1 month ago now 😫...

Did the dasmikro work just like attaching a h-bridge. You just send it PWM and it worked?

Pictures?

Sounds a fun project. Feels more rewarding then attaching an esc and reciever together.

Paul
User avatar
MarkR
Posts: 375
Joined: Mon Dec 18, 2017 12:46 pm
Location: Reading Hackspace
Contact:

Re: Mark's Antweight build

Post by MarkR »

A really quite bad picture.

The white heatshrink is convering a MPU6050 IMU board. The Dasmikro is in black heatshrink stuffed underneath that. The white wires were the ones which came with Dasmikro ESC, but shortened a great deal.

The other wiring is basically invisible in this picture.

It will be 4wd, but only the rear wheels are fitted at the moment. The plate is to mount another pair of motors for the flipper. I've little idea how powerful it will be, but I'm hoping- very :)
Attachments
robot20180316.jpg
robot20180316.jpg (101.69 KiB) Viewed 11196 times
Robots: Betsie - RaspberryPi controlled flipper bot with gyro stablisation - too clever for her own good?
Stacie - tidy flipper; 4wd driven by hair bands
User avatar
MarkR
Posts: 375
Joined: Mon Dec 18, 2017 12:46 pm
Location: Reading Hackspace
Contact:

Re: Mark's Antweight build

Post by MarkR »

Not all good news, had a frustrating morning trying to diagnose a short circuit.

Discovered that one of the Chinese N20 gear motors had a dead-short circuit internally despite appearing normal otherwise.

My advice for anyone wiring up a robot with several motors in parallel on a ESC is:

* Check the ESC isn't a dead short before you start
* Check each motor isn't a short circuit (mine were 6.6 Ohm normally, < 0.7 short)
* Check your wiring loom for shorts after soldering but before power-on.

It saves a lot of time and possibly some destroyed electronics.

I didn't manage to destroy anything though. The microcontroller spontaneously rebooted (presumably due to overcurrent condition dropping the power out), which immediately turned off the power to the failed circuit.

Also, conveniently, I was testing it with a limited-current bench supply instead of a LiPo battery, that was good luck.
Robots: Betsie - RaspberryPi controlled flipper bot with gyro stablisation - too clever for her own good?
Stacie - tidy flipper; 4wd driven by hair bands
Paulmchurd
Posts: 227
Joined: Sun Dec 10, 2017 4:39 pm

Re: Mark's Antweight build

Post by Paulmchurd »

Atleast nothing blew up!

I found another nice looking h-bridge. The dasmikro seems to have been taken off banggood.

https://m.ebay.co.uk/itm/Stepper-Dual-M ... =547517332
User avatar
MarkR
Posts: 375
Joined: Mon Dec 18, 2017 12:46 pm
Location: Reading Hackspace
Contact:

Re: Mark's Antweight build

Post by MarkR »

That H-bridge looks very tidy.I like the extreme low profile and that it's official rated to work with 3v3 logic levels. also benefits from actual holes in the PCB, unlike the Dasmikro which is compact but fiddly.

I could use one of those for weapon control.
Robots: Betsie - RaspberryPi controlled flipper bot with gyro stablisation - too clever for her own good?
Stacie - tidy flipper; 4wd driven by hair bands
User avatar
peterwaller
Posts: 3213
Joined: Fri Feb 15, 2002 12:00 am
Location: Aylesbury Bucks
Contact:

Re: Mark's Antweight build

Post by peterwaller »

I used those chips in my previous speed controller board I used one chip for each motor if you parallel the two drivers in each chip you get 2.4 Amp drive. The main problem using the chip direct was the are fairly large with 24 pins on a very narrow spacing so difficult to solder. The MX612 are much smaller with only 8 pins although they are limited to 9.6 v.
Paulmchurd
Posts: 227
Joined: Sun Dec 10, 2017 4:39 pm

Re: Mark's Antweight build

Post by Paulmchurd »

Wouldnt using two of the chips be far to much for an antweight? When only 1 is needed.

I've ordered one off banggood. Cheaper then the one I posted. I'll give it a test and report back ... in a few weeks
User avatar
peterwaller
Posts: 3213
Joined: Fri Feb 15, 2002 12:00 am
Location: Aylesbury Bucks
Contact:

Re: Mark's Antweight build

Post by peterwaller »

Back when that controller was designed some of us were using the pololu high power motors which had a stall current of 1.6 amps at 6 v.
So if you were running them on 8.4 v that's 2.2 amps and even higher on 3 cells hence the doubling up.
Post Reply