Can anyone recommend a speed controller to vary output on a starter motor, or an easy way to do it? Only need it to turn one way, no regen or anything posh like that, only need to be able to vary the power output without anything burning out. Anyone who clocked my last post will know where I'm going with this...

rs sell a few diffrent controllers or strip one out of a cordless drill and mod it.
adam

For a decent speed controller for a DC motor I would try and find a fork truck engineer. Older electric fork trucks were all DC although few were 12v. Smaller independant outfit probably best bet.

Davie

You need a Pulse Width modulation control circuit PWM....

Basically applys full power in varying pulses so you can get full torque at low RPM....

They are not the worlds most complicated electronic circuits...

You would need some fairly large drivers of some descriptio to be able to power the 60A? motor....

Deffinately doable for less than £30 if you know how to solder

You'll need some nice big chunky MOSFETs and the controller will need quite good cooling considering the amps you're going to be driving. Have a look on robotics websites. Quite a lot of the (not very good) Robot Wars robots used starter motors because they're cheap and available.

Disadvantage of them for robots is that they use up quite a lot of current to get the magnetic flow going so they're not very efficient. Obviously not a problem for your situation.

Your only going to be running the controler under load for short periods of time, so cooling wont be such a huge issue, but its still going to be a problem...

Starting current just to turn the motor alone is IMMENSE. Only real DIY way to do it with affordable components is using ressitors as on the old Trams.

My old man used to work in a forklift plant, might be a good plan. Not sure he'd be that much help though as he did the 40 ton jobs, all 12lt diesels and hydraulics.

So - resistive control. Is that easy to do? Is it as simple as a whacking great variable resistor to vary the voltage to the motor? Would the fact that it's less efficient mean a lot more load on the bike alternator?

Believe me, resistive control is definitely NOT the way to go.

A simple pwm controller is easy to build, but not as cheap as some may want you to believe.

You will want some big mosfets(300 amp, read expensive), and the appropriate mosfet drivers to suit.

I've got the same circuit driving my mig welder feed. The only difference for you is the mosfet drivers and the mosfets themselves.

And it WILL need cooling. A computer fan will do the job, rigged to a thrermostat type switch.

Cheers,
Syd.

quote:

---

Originally posted by tegwin
You would need some fairly large drivers of some descriptio to be able to power the 60A? motor....

Deffinately doable for less than £30 if you know how to solder

---

Ask tks on here - I know he had a go at one.
series wound starter motor will behave different to PM type.
Control - voltage control - easy & cheap but might be a bit twitchy
current control - ideal behavior & bomb proof but complex
I've been meaning to do one for a couple of years but never got round to it....

Seems that your bec is getting more heavy and complicated than necessary

quote:

---

Originally posted by nitram38
Seems that your bec is getting more heavy and complicated than necessary

---

You could just gear it for 2~3mph , and get a boost in torque at the wheels, and forget about any regulation at all.

Cheers,
Syd.

As an aside, the speed regulators for the electric car I was working on, are pwm based, and very simple. They are used on the electricarts in europe. The trick is to get the frequency not too high or low, and then use the biggest mosfets or IGBT's you can get. The mosfets appear to handle the higher frequencies better, and the igbt's appear better at high currents.

Welll thats lucky!! That i found this...

i was searching for something else.....

Anyway i used 6x 100amp (200amp pulsed) in parallel. I used NPN type in a boottrap situ.

i have used severall drivers but none of them could do what i wanted drive 6x a fet with huge power, so i ended up building my own driver. I used a 10watt resistor and a pnp mosfet and some transistors to build it.
it was by far stronger then all the dip ones...
(and cheaper) (it didn't blow in the end)

Also i used 2 shotky diodes of 60Amps..
(one to ground and one to battery voltage)

I mounted all the fets on a CPU fan (just thread them up with M3 holes and screw the fets to it. then make a hole with M6 and screw the battery terminal to it.

Then i used a LM opamp to measure the voltaje drop over the fets...because this is the only simple thing you can measure to know a bit how much amps there are bleeding....if the

Mine worked okey, the sound was a tipical whinening PWM sound like a golf car...

You say you aint gonna use regenbraking, but you don't have much option. If you use a permanent magnet type motor once you accelerated it and you backoff, the weight of the car puts some serious torque on the engine and it will generate power.

If you don't cope with that fact the voltaje across the fets will be to high and you will blow some of them and the driver....

If i would do it again:
I think i don't ever do it again. Because i also had some serious problems with the bendix releasing itself.

Think about it, the reason the bendix shoots out is clear, the reason it turns back in is because the engine turnsover faster then the starter does (when its off) and then auto you turn it back in (on the splined/turned shaft).

But in reallife if you go backwards and you stop. Then the power isn't applied to the bendix and it only will come loose if you push the car a bit backwards... (to turn it up the shaft) in reality its complicated......

the reason the normal starter circuits work is because they have speed when they release the button, and then it works okay (since the car is still moveing)

I used a 6mm lasered gear of normal iron maybe it was to weak, but its surely warped
(think 8mm will do)

saying that my reverse worked, altough with allot of noise and when i backed off it blew up a diode.. it did 10meters and the cpu fan alloy block was already warm (not hot).

The pwm signal is best bet to start with a low frequenty until you have resolved the rest and its working i won't up the switching frequency... the faster the switching the more heat in your fets....i think i had 3000hz

- I used a 10.000uF cap for the logical supply to the board.
- my starter motor was rated 1200Watts (it was printed on it)
- my current limiter was set to 600amps and i deffo needed it/them...


rememer that the starter motors aren't build for this purpose so i think that the speed of the car was manageed between 80% and 100% duty cycle.. (if you apply 80% it will start moveing the car a bit)

the reducion i used was huge!! my ringgear had 92 teeth
the motor itself 11
the diff 3.92.

the max speed dry was 6km/h.
the max speed wet when tested was walking pace (i think slower) but hey it can work. But the bendix thing is a problem!!

My best bet is to use a different type of motor and a different type of attachment to the car. and because thats difficult to archieve i would just now (with my cnc mill)
build a reverse box (a proper unit)

Tks