Electric rideable kid's car soft start

[CatalinaWOW]

Just get a heavily over-rated ESC.  I see one rated for 320A for under $20.  You will still be dealing with the fact that the foot control is a switch.  The commercial ESC's expect a standard servo control PWM input.  An Arduino or other microcontroller can provide this with whatever ramp up you find appropriate, including non-linear things like square law.  Setting up a proportional throttle can be a mechanical challenge, but look into sewing machine foot pedals.  Some are electronic and based on a pedal operated potentiometer.  Under $20 and an easy hack to microcontroller input.

[Gyro]

Is that under load? With a kid in it going up a steep hill the current is likely going to be substantially higher.

My friend's daughter has a similar kiddie car that he retrofitted with a 4 cell LiPo, it will do a burnout now   Ok so the solid plastic tires provide very little traction so that helps too.

Is anyone else slightly uncomfortable with the idea of retrofitting LiPos into sit-in kids plastic toys? Granted, SLAs can be a pain with Sulphation, weight, etc. and they can source a lot of current, but properly fused, they're not known for their spontaneous flammability. 

[ebastler]

My friend's daughter has a similar kiddie car that he retrofitted with a 4 cell LiPo, it will do a burnout now  

Is anyone else slightly uncomfortable with the idea of retrofitting LiPos into sit-in kids plastic toys? Granted, SLAs can be a pain with Sulphation, weight, etc. and they can source a lot of current, but properly fused, they're not known for their spontaneous flammability. 

Well, as james_s stated, the "burnout" is actually the point of this conversion... 

[Gyro]

Hmmm,  If they're lower impedance and sourcing more current than the original SLA then it sounds as if they could well be bare cells (unprotected).

Edit: Sorry OP, I'm drifting off topic.

[james_s]

I'm not advocating the use of LiPos but those of us who have been using them for years generally have a pretty good sense of how to use them safely so it's not really an issue. I've seen fires caused by lead acid batteries too, not the batteries themselves but they can deliver enough current easily to set wiring on fire. LiPo fires are almost always due to improper charging where one or more cells gets overcharged. Mechanical damage can also do it, we had a guy crash a large electric RC heli at the field I fly at, the thing smashed into the ground at high speed and the battery erupted into a fireball, it was pretty impressive. Fuse the output, charge them properly and don't smash them and they're fairly safe.

Heck when I was little the lucky kids had gasoline powered go-karts and scooters, fires due to fuel leaks or spills have got to be as common as LiPo fires. Years ago one of my neighbors set his lawnmower on fire trying to pour fuel into it while it was hot.

[VelOBor]

Sorry for not replying earlier. Busy time here!
Did not measure current through the motors directly, just the voltage across motor terminals and plugged it into Ohm's law. I am now looking for ESCs on Aliexpress, they have acceptable pricing. Looked around my local electronics shops and anything they have over 30A is "commercial grade" stuff - for industrial motor controls and such. And priced accordingly. Will keep looking locally, as I am not a fan of waiting for a month for my goodies to arrive... However the price wins here. So far the plan is - hack a cheap-o car remote and retrofit with arduino and NRF24 (with pair in the car) for RC, that will give PWM to the ESCs for proper speed control. Retrofit the stock steering with a proper servo (I'm thinking around 20kg/cm torque should be more than enough), but if I do something like that - I need to consider some sort of power steering (I doubt that turning this servo by hand will be possible for a small child) for when the RC mode is off, in manual mode so to speak. This is easy enough (a pot mechanically linked to the steering wheel and controlling the arduino's output to the servo), of course there is risk of loss of steering with controller failure. But then again - if the arduino kicks the bucket, the whole thing will just stop. As for the gas pedal - I'm thinking a slide pot mechanically linked to the pedal. I'll have to stick a spring in there to have the thing return to "idle", but that's easy enough. I'm okay with the mechanical skills. As far as the battery goes - I'm on the fence. If I go with LiPo - I will have to make sure to design proper protection, not too keen on the idea of a kid riding on top of a firecracker. If I stick with the stock lead/acid pack - I'd definitely want to boost capacity by adding a couple more (stock version has two 6V batteries in series and with 15A fuse, I'd just get two more sixers and parallel them up to the stock ones, but deal with the doubled re-charge time).
So far that's it for the mechanical side. The rest is pure aesthetics. Headlights, perhaps MP3 player, as it already has some audio thing on-board, but it's garbage - has 4 tracks recorded in 10-15 second loops that play for a minute and a half, and cycle through the tracks with a button... The aesthetics are probably going to be left for the winter-time tuning though.
Sorry about the long-winded post, but the second white russian in me is taking its toll.
So far so good. I'm starting to get a solid understanding that this is going to be a fun project, and I just need to get all the parts together in one place (preferably my workshop) and get started.

[james_s]

Look at the RC hobby suppliers like HobbyKing and such, they should have everything you need, the speed controllers will be a "brushed ESC". If you really want radio control, I highly recommend a radio system made for RC models rather than trying to hack together Arduino type stuff. A friend of mine is into making remote controlled robotics like R2D2 replicas and such which he takes to hacker/maker/robotics club events. He's another RC modeler so that was his natural choice for the radio link and he said it's always the guys trying to use various radio modules with Arduinos and such that have problems. Meanwhile a $25 radio set for an RC car or boat can be remarkably dependable.

Robotics suppliers are another good place to look for motor drives that are cheaper than industrial stuff. I'm not in that scene but I'm pretty sure I've seen ~100A dual motor drives for big robots. Pololu is a place that comes to mind, not the cheapest but they have a good selection of stuff.

[CatalinaWOW]

I will repeat my suggestion of a sewing machine controller for the foot pedal.  They contain a pot and all of the springs and mechanics to make it move and to make the pedal follow your foot up and down.  You throw away or disconnect the light dimmer style electronics, bolt or glue the controller to the floor of your car and wire the pot into whatever your control mechanism will be.  These controllers are widely available on EB and thrift stores for very low prices, often because they are not working.  The pot is sometimes the failure mechanism, but more often it is not.  Even new they aren't all that expensive.

[james_s]

Just make sure it's a modern one. Older sewing machine pedals were typically a wirewound rheostat that directly controlled current to the motor.

[CatalinaWOW]

Just make sure it's a modern one. Older sewing machine pedals were typically a wirewound rheostat that directly controlled current to the motor.

Even the older pedals will work.  Add a series resistor on each end and you have the voltage divider you need.    The ground end series resistor can be several times the rheostat total resistance of the rheostat and after the A-D you can scale as required.  The other resistor is chosen to make the top of the rheostat match the high end of the A-D range (usually 5 or 2.5 volts).  The two added resistances reduce the current in this leg a few percent of the motor current, thus not high power resistors and not a big drain on the battery.  Not ideal, but when you are going for cheap you look where you can.

Making a foot pedal operated variable resistor that is reliable and resistant to users is a bigger project than the whole rest of the build put together, which is why I suggest avoiding all of that mechanical work.

[VelOBor]

Yeah, I think I'm convinced of using "tried-and-tested" gear such as ready-made RC controllers and sewing machine foot pedals. Makes sense not having to reinvent the wheel. I haven't done much RC modeling since high school (and that was nigh 20 years ago), but if I remember correctly the controller basically has PWM outputs that you connect straight to ESC or servo. That makes things easy. I just like my arduinos and want to find use for them
The main problem is - the cheap-o remote that I found comes without the receiver module, and I'm too cheap to pay more than I found initially. Although by the time I hack the 'duino into the remote, the whole kit will run me at least as much as a decent 2-channel ready-made RC set...
Now I just have to sell this idea to the wife, and I'm set... She's not too keen on me spending on my "toys". And let's face it - this project is as much of a toy for me as it is for my son
Pretty much found all the bits and pieces from local suppliers (as far as ESCs go and all that stuff). The "gas" pedal will be the only thing left.

[CatalinaWOW]

You are right, that is how most of the RC servos work, PWM just like the Arduino.  The receivers put out this PWM. 

You will have to decide whether you want to incorporate the remote control idea.  Steering will require a high power servo ($$$) and some pretty extensive hardware mods.  You may also want to check whether you are breaking any local regulations putting a human payload in a remotely controlled vehicle.  Any laws that exist won't have been aimed at an application like yours, but that won't help when some busybody gets involved.

Check out the used market for RC gear.  The technology has gone a long way over the last two decades and older gear is often given or thrown away.  Just have to ask the right folks.  I have some like that, but I suspect shipping costs to you would be more than buying new locally.  PM me if you are interested.

[VelOBor]

The car already has RC function, but the controls are "digital". In that they are either ON or OFF, or ALL THE WAY RIGHT or ALL THE WAY LEFT. No analog steering or acceleration. The whole idea started with the car being too jerky when it starts moving, so the idea of ramping up the motors evolved into full-on RC upgrade.
And yeah, shipping ANYTHING half-way across the globe will cost more than getting some cheap-o or used gear locally, but thanks for the offer! Really appreciate it.
Found this servo from a local electronics shop: KJH KS-3620 for about 20 bucks US (which fits easily within budget for this part of the project), that claims about 20 kg/cm of torque @6V. I'm thinking it's probably closer to 15 kilos, but I think that should still be enough. I will have to power it separately from the control board I'm guessing. Also one of the issues with using ready-made RC module is "local" steering control. When I put the car into "manual mode" (where the child driving it has control of accelerator pedal and steering) - he will have to overcome the servo in the steering. And turning a deeply geared motor can be tough, and also not too healthy on the reduction gears. I was thinking either a servo tester just linked mechanically to the steering wheel and feed the signal into the micro, or use a pot and again feed the signal into the micro.

[CatalinaWOW]

Or put an electric clutch between the servo and the mechanical system.  Deactivate when in local control.  Really cheap ones can be salvaged from automotive systems like the air-conditioner, but are usually large and current hungry.  I don't have any suggestions off hand for smaller, more appropriate sizes, but there might be something out there for the nitro/gas powered RC car market.

[VelOBor]

First of all thanks all for helping me out so much! Y'all are AWESOME!!!
And as for the clutch - I'm thinking of staying away from clutches, I'd rather have a servo-driven power-steering thing I think. Seems like it would be easier to implement a pot linked to the wheel, rather than trying to jam a clutch in there somewhere... I dunno, I'll look into it. I don't have a machining shop (just some woodworking tools and such), so hacking that into the steering system may prove more difficult than it seems.

Okay, so the project calls for a standalone atmega328 (for some lighting and such, maybe an actual MP3 player...). Now I'm experimenting with powering this thing (and because I don't want to blow it up, I'm using a motor as load instead of the chip) with an LM7805 and a few caps (electrolytics) and a lab PSU.

Input: 12V and something like 0.08A with the motor running
Output: 5V

So far I have everything on a breadboard, and I have an issue. Perhaps several, not the least of which is me being a total NEWBIE!...
Let's start with the fact that all of these caps are salvaged from an old PC PSU, which seemed okay, except one transistor having had a very bad day. So bad in fact, that it left a brown skidmark on the PCB and left it smelling toasty.

The input capacitor is 16V 1000uF, across the LM7805 (GND and OUT pins) I have a 50V 4.7uF, and the output has a 50V 10uF. All the polarities seem proper. The motor is a chinesium-grade hobby motor out of some dead toy, with resistance measured at around 8 Ohm.

I hooked up my scope (don't ask why I have one... had the opportunity to get it basically for free, so I did) to the input (channel 1, yellow) and output (channel 2, blue). And here's what bothers me: the output voltage has a spike at around 10V (as seen in the photo attached, note the timescale). Am I reading something wrong? Is this normal for an inductive load? Is it an issue with improperly selected capacitors voltage-wise? Capacity-wise? Electrolytics, where I should've used ceramics? Dead capacitor somewhere? Or am I just an idiot and should give up this whole electronics hobby and take up knitting and gardening (or perhaps something safer where I can't hurt myself or others)

What I'm getting at is I don't really understand capacitors. Not like how or why they work, but rather how to use them properly. Where resistance and current and voltage make sense to me, capacitors not so much. I mean I understand that they are in a way like a battery and a temporary reservoir of short-term power and all that, but the whole thing of which caps (voltage and capacitance, electrolytics VS ceramic VS film) to use in what cases - that's a mystery to me. Photos attached.

[james_s]

For the servo, look at what guys are doing for robotics, if you want I can ask my friend that builds the droids and stuff what he uses for big servos.

Inductive loads are notorious for creating spikes, why this happens is beyond the scope of this topic but I encourage you to read a book like The Art of Electronics, that will help you understand what's going on. Looking at your breadboard you have the load way down at one end and the source at the other, this means you have a good bit of resistance as well as parasitic inductance and capacitance. A small ceramic capacitor across the motor as close to the motor as you can get it may help. Generally you also want to keep connections as short as you can, especially ground, ideally all of the grounds should converge at a single point. This is going to be more important the larger the load you start dealing with.

[VelOBor]

Awesome! Thanks!
I'll get that book, been looking into it for a while.
Also will experiment later today with some different loads. Perhaps a couple LEDs or something of the sort, to avoid inductance. See what that gives me. Also will try a smaller breadboard and/or shorten my connections and grounds.

[Circlotron]

How about adding a resistor (a car headlight bulb may do but these also have a considerable inrush current) in series with the motor and bypassing it with a timer relay?

Even better, use an NTC thermistor with timed bypass relay. The kind of thermistor used in SMPSs for inrush current limiting. The bypass relay will enable the thermistor to cool down once the motor is running otherwise the next start will be a hard start because the thermistor would still be hot and so low resistance.

Edit -> changed PTC to NTC.

[james_s]

There's no need for anything like that, if you have PWM control of the motors then soft start can be done in the control. It's already built into most ESCs so you don't strip the gears in geared drives.

[Circlotron]

^^ Of course. I’m thinking though if the OP doesn’t have pwm. Trying for the least worst solution.

[WB8PFZ]

I need to take apart an old variable speed drill and see how they did  it.

[Nicton01]

I also have the same problem with my granddaughter's ride-on electric car. It is far too jerky both when starting and stopping. I have been considering a number of potential solutions and the one I have decided on is as follows:-

 I am going to leave all the controls as is apart from the motor drive. I shall disconnect the current relay motor drives from the motors and replace them with a 12V PWM CW CCW Reversible Switch Motor Speed Controller, from ebay. All inputs to the current board will be left intact.

 I will put together an Op Amp RC circuit between the pedal and and the Potentiometer input to the PWM board. The circuit will have a time constant of 2-3 seconds and ramp down time constant of about a second. I shall connect the ON/Off switch and direction switch to
the new Op Amp  board + PWM board.

I shall be starting this as soon as all the parts arrive. Has anybody else had similar throughts on this problem?

[james_s]

[Haenk]

"= fun" ?

Just a sidenote, some of these can go very fast, even unmodified. And most likely, the kids do not know how to properly handle the toy at higher speed, even less, when panicking. So *please* let them wear a helmet.

[james_s]

Doesn't seem like a big deal to me. When I was a kid I knew people that had little 50cc gas powered trail bikes and four wheelers made for children, hardly anyone even wore helmets back then although I do think those are a good idea. We would ride them around the field behind my friend's house unsupervised and go home when it was time for dinner. Kids are not as fragile as people seem to think, I feel bad for those growing up these days that are forced to live in a bubble and not engage in any potentially risky activities.