Ideas for entry level Ants

Please post all questions and answers in here. This way people can easily see if someone else has the same problem.

Moderators: BeligerAnt, petec, administrator

Post Reply
Ant Ipodean
Posts: 43
Joined: Mon Apr 24, 2017 3:31 pm
Location: Auckland, New Zealand

Ideas for entry level Ants

Post by Ant Ipodean »

Hi All

We are looking to get an Antweight scene going here in Auckland, New Zealand.

Whilst the N20 type gear motors with a dual ESC seem to be the optimal drive setup, I am also interested in the simplest, lowest threshold way to build a basic Antweight so that we may encourage as many people as possible to get started.
The most challenging aspect for many people seems to be sourcing and setting up (and paying for) ESCs, and mixing.

Following are what seem to be the options .. Please feel free to offer ideas, alternatives, etc?

1. Best setup: Gearmotors with “Rory” dual ESCs.
Problem: Rory advises that he will not post overseas.

2. Next best: Gearmotors with “DasMikro” dual ESCs.
Problems: Supply seems a bit erratic from Banggood.com, though the price is good at $US16.
No braking.
Also, they have a design flaw which will likely catch out many “home handyman solderers” in that the leads need to be soldered to small copper surface pads on the pcb, which easily break off. They really should have more robust through board holes for the leads.
A very small ancillary pcb could probably address the problem, but seems a rather clumsy solution.

3. Gear motors with 2 x RC car type ESCs, bi-directional. Cheap enough at under $US4 each.
[AliExpress - “RC Car 10A Brushed ESC Two Way Motor Speed Controller No/With Brake”]
Problem: These ESCs have a built in characteristic whereby you move the joystick up to go forward, centre for stationary, but the first time you move it down, nothing happens, then the second time the robot reverses. Apparently this is a safety feature for RC car drivers, but it is just plain confusing when trying to manoeuvre a robot.
Does anyone have any suggestions how one might eliminate this characteristic on these boards?

4. Modified servos. Some of you may have to think back a decade or so ;)
I have opened up a couple of small servos. An Emax ES08MAII and a TowerPro MG90S. ($US3-4 each)
The Emax seems more suitable as the pot can be removed without disturbing the drive train and speed hacking increases the gearing by 10 fold, whereas the TowerPro pot removal is not so simple and the speed hacking looks like it would increase the gearing 16 fold (As best I can count the teeth ;)
I have set a pair of the Emax on a test mule with 30mm wheels and the speed seems very good. Somewhere in between the fastest and the slowest gear motor driven ants we have built.
Problems: Instability to stay still when in neutral. i.e. the drives twitch or creep, despite best efforts to trim.
Initially I left the pots hanging out of the servos and adjusted them to stop the creep, then I replaced them with 2k2 resistors.
In both cases you can adjust the setup to stay still, but if you turn everything off then on again, the creeping starts all over.
So, here are my questions -
a) Is this instability likely a characteristic of cheaper servos, so more expensive ones might be better?
b) Are the 360 continuous servos offered on AliExpress likely to have this creep problem, or are they spec’d differently to somehow eliminate the problem? [https://www.aliexpress.com/item/DOMAN-R ... 0.0.9p6mwV]
c) Is there a way to broaden the dead zone between forward and reverse by just putting a capacitor and/or resistor across something?
d) After years of experience running servos for drive, can anyone confirm that there is a solution for this, or should we just accept it as a limitation?

5. Robbing ESCs from (cheap) servos to run gear motors.
I have not tried this yet. Will the “creeping” in neutral also then happen with the gear motors?

Mike.
User avatar
Shakey
Posts: 1119
Joined: Sat Mar 13, 2010 8:38 pm
Location: Reading

Re: Ideas for entry level Ants

Post by Shakey »

1. - Can't really comment here.

2. - The wires pulling pads off is not a huge issue to be honest, ants are small and wires are very rarely taking any significant force to pull them off a board. If they are taking that much force then you have other problems going on! Through hole mounting just isn't necessary. No braking is an issue but easily fixed by adding some resistors across the terminals: http://robotwars101.org/forum/viewtopic ... 504#p57290

3. - That characteristic is in the firmware of the boards, they'd need reprogramming. With that characteristic it is a no go.

4. - On servo drive the trim creep will never resolve itself well in all honesty. There's no way to increase the deadzone as that is software enforced not done by hardware, servos are normally single chips and reprogramming is not an option. In all honesty the servo drives are outdated in speed, torque, control and reliability. I'd go for the N20 gearmotor option, servo drive was left behind on the UK scene many years ago for a whole host of reasons and it's allowed a newer generation of far higher quality antweights through.

5. - It's an issue inherent in the electronics not the motor so will still occur.
Nuts And Bots - For all your components and ready built antweights!

Alex Shakespeare - Team Shakey / Nuts And Bots / Team Nuts:
AWS 44, 45, 49, 51 & 55 Winner - Far too many robots!
Ant Ipodean
Posts: 43
Joined: Mon Apr 24, 2017 3:31 pm
Location: Auckland, New Zealand

Re: Ideas for entry level Ants

Post by Ant Ipodean »

Hi Alex

Thanks for the quick reply.

On the DasMikro .. It seems to be the +5v pad that goes to the Rx that is really weak. I've lost 2 of them so far and whilst I may not be the best at soldering, I have been doing it for a few decades. I managed to trace it back to the next component leg and solder to that. A bit crude.

I've just had a little breakthrough with the cheap 10A ESCs.
The first ones we bought are "with braking" and have the nasty reversing characteristic.
The latest ones we bought are similar but different. They have a miniature slide switch on the PCB.
When I first tested it with the switch in each position, there seemed to be no difference and the same reversing problem.
But now I find that you need to switch the power off then on again after moving the slide switch.
In one position this eliminates the reversing problem.
I guess that the nasty reversing characteristic is what the vendors consider a braking feature :(

So it looks like we have our readily available "entry level" ESC + gear motor setup.
They will be a bit bulkier than an elegant little dual board, but when soldered together as a pair, several of the duplicated cables can be eliminated.

I had read your posting before about using resistors for braking, and have some 51ohm but yet to try it.

Re hacking servos, I guess I was being optimistic, but kind of knew the reality.
Thanks for confirming :)

Mike.
Post Reply