555 PWM circuit

If you don't understand the function of a component in someone else's design it would be better to keep that component in. If you didn't understand the purpose of the 555 in that circuit, would you have left that out?

Hmmm, not entirely fine then.

See my note above about missing out components from a working design.

A 20Ah battery will probably peak at around 100A into a short circuit, so 30A is no problem. You will just get less service time from the smaller battery. This is probably caused by high frequency modulation. See my note above about missing out components from a wroking design.

That really depends on how far and fast you wish to travel.

--
My time machine is broken, but it was working tomorrow.

When I first tried experimenting with PWM circuits, my MOSFET (IRF530) overheated, and when it eventually failed, it acted as a short circuit: Motor Full Throttle On.

This would have been... inconvenient... if this had been used as the speed controller of an electric vehicle.

So put on some sort of emergency disconnect switch that's easily accessible.

Michael

Hi Michael,

I=B4ve also noticed that :) but thanks for your advice.

BTW Did you finished your project? Any more ideas?

Thanks (and escuse me for my english)

Thanks!

I will try with that capacitor, but, do you know what is for? and can I use a "non-polraity" capacitor?

The resistor between pin 3 and the gate is 10 Ohms, not 10K, right?

Yes, but the resistor which I had to remove was the one between the gate and ground (1K). If I use it, I never have current in the motor and I don't know if that is overheating the transistor.

Ouch! I had a mistake, in my circuit I am using a 10K resistor, and in the design it was10ohms. I will change it. THANKS!

It's for supply decoupling, i.e. preventing the switching current from causing noise on the supply lines.

It doesn't need to be polarised, but polarised electrolytics are normally cheaper than the alternatives at large sizes (i.e. >1uF).

You can connect FETs in parallel. Just connect all pins of the same type together (i.e. connect both gates to the 555, both drains to the motor, both sources to ground).

But you need to look into why you're burning out a FET rated at 33A while driving a motor which you say is only supposed to draw 10A.

If the motor is rated at 10A at no load, it could draw a lot more while spinning up. You should really think about adding current limiting.

More generally, don't expect to be able to take a circuit designed for a couple of amps and just scale the components up to handle ten times that. At lower currents, some things can just be ignored, while they may need to be dealt with at significantly higher currents.

At 10 ohms, you have a voltage divider of 1010 ohms (1K+10R) total, with

1000/1010 (i.e. almost all) of the voltage on the gate. At 10K, you have an 11K divider with 1/11 on the gate, so it's not turning on.

With the 1K removed, the 10K gate resistance will greatly increase the FET's turn-on time, as the 555 can't source/sink more than 12V/10K = 1.2mA to charge/discharge the FET's gate capacitance.

If you're using a relatively high switching frequency, the FET could be spending a significant amount of time in the linear region. It certainly isn't going to like passing 30A under those conditions.

If you have access to an oscilloscope, examine the drain voltage. It should look very much like a square wave. If the rise/fall times are significant (i.e. it looks trapezoidal or, worse, triangular or sinusoidal), the FET will be dissipating significant power.

This circuit is useless for your application. There are numerous circuits around based on motor control pwm chips.

You've noticed that too? I thought it was just my lack of skill. And I thought your MOSFET was a little tougher than my IRF530...

Put a heat sink on your MOSFET, regardless.

I didn't finish my project - partly because of difficulty involved with coupling my motor to the bicycle gears - I was thinking of using a system of belts, then I got sidetracked (this wasn't my primary project, after all).

Michael

Insufficient Gate drive voltage. that's my answer..

--
"I\'m never wrong, once i thought i was, but was mistaken"
Real Programmers Do things like this.
http://webpages.charter.net/jamie_5

"Vicente"

** Bad circuit for high power PWM use.

The frequency varies with the duty cycle.

This is a much better schem:

But change C1 to 10 nF to raise the frequency and use a 10 amp Schottky diode for D3.

Paralleled BUZ11s are fine with the old circuit drive component values.

...... Phil

The problem is that you have used a Transistor that is requiring more Vgs voltage to get it in the low ohm state.

This is the reasoning as to why it works when you remove that R. Also, more than likely this unit may have higher capacitance on the gate than the original and thus, is producing a longer propagation delay..

The 10 Uf you removed is most likely there to help stabilize the Vcc line during high current pulses when the 555 output pulses on which also can cause slower Slew rate for the gate bias, along with causing the 555 to not time properly.

On top of all that, the 555 output starts folding back around 100 ma if I remember correctly. If you were to view the gate positive transition via a scope, i'm sure you'll find the slow rate I am referring to. Also remember that a 555 sourcing on the output does not give you the Vcc rail voltage. I can't remember exactly what the drop is how ever, I can bet it's at least .6 or more volts below the Vcc. I guess if you were using a CMOS version you might not see this as much how ever, I think those versions also have lower output current handling. ( I could be wrong there, it's been a while since i've looked).

When selecting the CW frequency, one should consider the raise and fall time of the device that is producing the PWM on it's output. This all adds to the problem of driving a fet..

A nice high speed schmitt trigger driver would most likely help you out .

--
"I\'m never wrong, once i thought i was, but was mistaken"
Real Programmers Do things like this.
http://webpages.charter.net/jamie_5

Ok, I wil try this new design, but can you explain me a little more why this design is better?

Should I put a resistor between the gate and ground as I said in a previous post? (discharge the internal capacitor of the transistor) which value?

Thanks.

I am not sure about the motor ratings.

I tested it and I have this values:

12V No load: 1,5 A Normal use (conected to a gearbox and to a wheel and moving 80kg in a flat path): 10A

I saw the 30A peak while trying to climb a small step.

So what do you think is the best idea? Use a more powerfull transistor? Connect 2 BUZ11 in parallel? And what do you think about the new design that "Phil allison" posted some posts above:

Thanks.

"Vicente"

** The PWM frequency should be held constant ataround 1 to 5 kHz for best efficiency with small to medium power DC motors. ** LM555s can source and sink considerable current - so that resistor has little effect.

I designed and built several PWM drives for DC motors for RC model boats using them - used BUZ11s as power switches as well, just by coincidence.

Way back in the dim, dark, late 1980s.

They all worked just fine - long as you kept damn salt water out of them !!!

...... Phil

Add current limiting. If it pulls 30A while still moving, it could go even higher if the motor stalls. Even if you add enough FETs to handle the current, can the motor handle it?

AFAICT, that's meant to produce a variable duty cycle with a constant frequency. Probably not a bad idea, but it doesn't affect any issues on the output side.

That looks like a good choice, modified with BUZ11's in place of the TIP31.

Ed