Gruber
HobbyMasters
M.L. Toys
KidsWheels
Gruber
HobbyMasters
M.L. Toys
KidsWheels
HobbyMasters M.L. Toys
Gruber
KidsWheels
Use this area as your personal project pages or blog! Each thread belongs to one user as a project page!
By TinCanTomCat
#130952
I bought this for my daughter, choosing one with R/C as she's still too young to grasp the concept of a steering wheel :lol:
I also chose to have 2wd drive, so I have a base for a bit more power later.

Stock pic:
shopping.jpg
shopping.jpg (10.7 KiB) Viewed 1145 times


Sadly, the R/C in these are a flat on/off - making steering near impossible (slack in the column means you just oscillate the poor kid as you attempt to keep the car somewhat on the path).

The accelerator is no better, with full power or a dead stop.

Having a spare R/C model car, I set about scaling the electronics to control the Merc.


This week, my family was away visiting parents and I was too ill to work on the real car. So cracked on with this.

I had a 2ch transmitter and receiver already.
I bought 2x Cheap "320A ESC" modules:
UT8fgOxXcVXXXagOFbXb.jpg
UT8fgOxXcVXXXagOFbXb.jpg (40.11 KiB) Viewed 1145 times


First step was to test the driving motors on the ESC's. They're cheap and get a fairly mixed review, but at worst case, they are a base driver for a custom circuit.
Ititially, I fitted an ESC per motor with the signal wired sharing a channel - worked fine as expected, though there was an offset on an ESC at slow ramp up (i.e. a random ESC at power on would be fractionally offset to signal input - so one ESC would begin to turn a wheel 3-5 degrees before the other!).

This probably would make no difference under load, but as the ESCs were stone cold after 2-3 minutes at various speeds, I paired the motors to a single ESC. Again, driving it for 5 minutes - ESC was still cold.
The benefit is that I can keep the drive wiring stock for the time (so the fast/slow switch still works, pedal, etc).

The steering would be another problem, as I wanted to convert the existing motor to a servo.
So I marked the lock-to-lock points of the motor (do this under power, as their is movement in the mount!), then removed it.
I stripped down a model servo and reversed the pinouts of the H-bridge, then attempted to pair that to the ESC.
I removed the control board from the ESC, leaving a single board with FETs and other unused parts.

That seemed straight forward, so after wiring and testing, connecting the servo FET control to the ESC FET controls, some magic smoke came out of the servo FETs. Oops.
A double check had shown that I took the wrong pin from one of the tiny FETs and shorted them!

A look online showed examples of using the motor feeds onto a secondary H-Bridge - but that's too late for this...

The ESC has an LM339 quad comparator, and the 4 input pins are nicely tucked at the bottom of the board.
I soldered Pin 3 of the LM339 to 12v.

Wiring the control pins to the LM339 inputs made the motor jitter, but everything was inverse - the motor stopped when a signal applied, but jittered when the POT was centered.
The ESC also become VERY hot!

With a fresh brain the next day, I looked over the 2 boards.
Digging some more into the ESC's I had also showed that all the ESC's FETs were of the same type (not NPN and PNP style like most H-bridge designs),
so this left me to reverse engineer the ESC and Servo a bit more.

The servo IC had no useful markings, but using a multimeter, I found only 2 pins of the 4 that drove the FETs doing anything at all.
(Note: the 3rd and 4th pins 'may' actually be useful for anti-crowbar, so will be revisited at some point)

Taking one of the pins and one of the LM339 inputs actually gave me a sweeping output when a signal was applied!
A bit of tweaking later, I found it was simple enough to bridge the inputs of the LM339 1&3 and 2&4.
Joinup.jpg


Connecting the motor showed this worked great!
The ESC gets warm, but not unable to touch.
Centering the POT gets a little jitter, so I may change the POT to give the dead-zone a fraction more margin for error..

The motor, as I said was marked for end-to-end, so centering this I fitted a tall self-tapping machine screw for moving the POT arm.
The POT is hot glued to a spare mount on the motor casing and an arm is made from a coat hanger wire.
20150331_204221.jpg


Once I'm happy, i'll either replace the POT for a larger type, or drill the existing one for a screw to hold the arm in place.

Testing worked perfectly.

Only issue (if at all) is that the servo whines on center - I believe this was always true on the original model motor, but now it is amplified through a larger motor and chassis.


The 320A ESC's are rated for around 7.6v, but the somewhat neat power circuit on board should cope with higher - I'm running them at 12v fine.
The do also backfeed the receiver with 7.6v even though the power feed is higher.
But I blew something when I crow-barred the Servo, thus now the receiver is powered via a 7806 vreg (a 7805 is probably too low, but this works fine).


Next:

Non-R/C mode: upgrade the pedal to use a servo test tool
Lights!
Modify the R/C transmitter to accept a Wii nunchuck, so I can keep my hand in my pocket for winter months!
Last edited by TinCanTomCat on Wed Apr 01, 2015 7:44 am, edited 1 time in total.
User avatar
By toycrusher
#131442
:? :? :? :?: :? I'm obviously quite uneducated in RC circuitry :roll: :lol:

I like what you've done. Nothing like getting your hands dirty and creating something great from what was once mediocre! :D
#140188
Forgot about this!

Project is still ticking along, changes so far:

New battery.
New motors.
New wiring.
New MOSFETS.

Yes, in that order - the motors draw significantly more current, which the wiring couldn't cope with.

It was very fast prior to failure! The motors are cheapo 23 turn types, which swapped over perfectly into the original gearbox.

The new wiring exposed a flaw in the ESCs causing them to release their magic smoke.
(Basically, the ESCs were in a floating state until the receiver livened up. This caused a short. The wiring acted as a shunt for those few milliseconds.)

After rewiring to handle the new motors, the 'shunt' effect disappeared - poof goes the ESCs.

This lead me down the rabbit hole of MOSFETs and such.

The cheap Chinese ESC actually have a decent set of MOSFETs (though still not 320A, peak was 200+ and continuous was around 120ish)

I'm still not sure how a 220 package can handle such currents via their tiny legs, despite any data sheet saying so.

I've upgraded to IRLB3034 - just because they're cheap and common (apparently very popular with the 'vape' and 'airsoft' crowds)
These also live on individual heat sinks (glued together) with pull down resistors to prevent floating issues.


As a backup, I have a nice mobility scooter axel and ESC, but I've not studied it enough to add R/C control yet. (And I really want it for build #2)
#142551
Another hiatus (life, decorating, etc.)

This weekend, the wife was very adamant about 'filling up the conservatory with all my stuff' and '<daughter> will not stop asking about her car'.

So I decided to drop in the 24v setup instead.
Originally, I held off using the mobility scooter parts due to not having time to find points to allow RC override.

But for the weekend, I just used a long rope and let her do the driving.


Having recently acquired a 'multi cutter' tool, I set to trimming some plastic to fit the mobility axel.
For the time, I've also used wood (for studding) bolted in for the axel to fit to.
This left just enough room to squeeze two 12v 22ah batteries in.

The mobility ESC is fitted under the faux bonnet (hood) with all wiring moved up there too.
Anything non-required was cut off. The mobility scooter also boasted a 'built in' charger - which was a regular mobility
charger strapped in. So that's separated and a lead will be made up.

The 'wig-wag' was cut in half and a self-tapper was fitted the cut side to limit the use to press-down only.
This now lives where the foot pedal was - just needs a pedal fitted.
Under the front wing is a second pot to limit max speed.

After a whole day with my daughter driving it around a local nature park (up hills, across fields, etc.), only when returning home did it even hint at low battery!

Us adults also had a sneaky spin, but I kinda killed the toy front wheels (oops). Another side project!


Next stop:
- Find a suitable point to override the ESC input to the MOSFETs.
- Fit bushings/bearings to steering rack/front wheels.
- Rebuild the steering servo electronics for the 24v electrics.
- Update the dash to drop the 24v to 5v for lights, etc.

Pics later.
#142787
I was going to post something detailed about the progress in driving the speed controller, but that'll come later - the state with that is mixed (but not over).
As I want the RC stuff to override where needed, I've settled on using a uC to do some heavy lifting.
When doing some tests, I dropped a lead which fell on a 24v pin, killing my usb-rs232 adapter and weakening the 5v feed I was using for the uC (the speed controller still works, but can't handle the extra current draw on it's 5v line).

The simple 'todo' here is to use a separate 5v reg.

I also found the pot input is 1) biased, 2) the +/- are offset. (i.e GND is not direct GND, + is not exact 5v) and 3) the +/- is monitored for the 5k pot (so a break would disable the SC).
Before killing my tools, I had limited success getting some control using a low-pass filter, but due to the weirdness, i'll likely need to use an op-amp.


On a physical stance, I "upgraded" the steering rack:
The rack consists of a horizontal box-section with a tube welded vertically on each end.
An 'L' bar each end fits through the tubes and wheels, with a little thread each end to hold in place and hold the wheel.
Each 'L' bar has a tab welded on with a connecting bar across the two, which a large 'J' bar joins the centre to the steering wheel.

(If you follow, that's pretty much the rack).

However, the two tubes are oversized by 2-3 mm, so the steering is very slack (and noisy!), plus the bar connecting the tabs are joined with thin rivets.

The rivets are replaced simply with bolts, washers and lock nuts.
The tubes were measured and suitable bushes fitted.
One tube was re-welded (we broke it last week playing on it!)

The result is a very sturdy, but smooth steering mech.


Secondly, I've started to re-build the steering servo.
Prior, I fitted the servo-pot the the motor housing, but due to slack, the steering never quite returned exact, plus the pot-arm was too delicate.
Now, I've fitted a bigger, metal pot to the chassis and is controlled directly via the steering rack. (This may open up oscillation issues, but this should be easily addressed).

Thirdly, when spelunking on eBay for a wheel set, I happened across another almost-identical mobility scooter for £10! (2-3 years older, so some parts were revised). Sadly the wheels are not the same, but the SC is the same (ok, a few revisions older, but I can re-produce for... two #2!)


Stay tuned...
#142923
Small update:

Springs!
As a simple self-centring mechanism, I fitted some springs I had spare from a set of drum brakes.
Easily over-powerable by the motor or child, but sturdy enough to pull back straight.

The 'J' bar has a small spring to tension the rack.

The steering motor and the position pot/bar has one each, to limit oscillation in servo-mode.

merc_steering.jpg


It's a little 'sproingy', which hopefully will either bed in, or I'll just sleeve them.

The 'J' bar-to-motor cog is now fixed to the gearing, as each section had a few degrees slack, which accumulated to nearly 25-30 degrees!


The servo circuit is being re-done. I'm using the PCB as my prior post, with some nice new MOSFETs (as I stupidly popped the other new ones :roll:).
I may drop the servo PCB and drive the MOSFETs direct from the uC, though that becomes 5 pins instead of 1, though I 'might' be able to squeeze that down to 3.
#147640
Tiny update:

- Removed the steering motor now - my Daughter can control the car, and the motor just added drag.

- Steering rack fortified - the factory welds were useless and finally gave way. The rivet were shonky and the control bar was easily bent. So I welded up the tubes, replaced the rivets with bolts and welded, and replaced the 2mm bar with 5mm bar.

- Wheels replaced with black and pneumatic all-round (it had plastic on front and grey/beige on back).

- The controller is a two parts - the ESC from a mobility scooter, and a custom wifi uC.
This uC offers a simple control 'web page' - with configuration (speed limit and manual/remote control), and a joystick.
It reads the pedal input and scales it to the speed limiter - useful if we are in unknown areas.
Initially I used a 'wireless Wii nunchuck', but the range was 10 meters and would hold last position for 5 seconds - not good!

So far - totally silent and cut through steep hills like a knife through hot butter :lol:

Next stage: tidy up ancillary circuits & work on project #2.
User avatar
By toycrusher
#147643
Thanks for the update!
Newbie in the UK

Interesting. That looks like an older (red painted[…]

Converted from 6v to 12v. Added ESC with twist th[…]

Search around the website. there is all kinds of p[…]

Your gonna need flotation tires in soft sand. The […]

HobbyMasters Udemy Course