555 PWM circuit

Hi,

I've build a PWM control following this design:

http://www.pablin.com.ar/electron/circuito/varios/pwm-cc/index.htm

But I made some changes:

1. I don't know what the 10uF capacitor is used for, so I am not using
it.
2. As I want to use a high current motor (10A), I've changed:
2.1: IN5408 > BY229-600
2.2: IRF830 > BUZ11 with HEATSINK
3. I guess that the 1K resistor in the BUZ11 gate is used to discharge
the internal capacitor of the transistor (avoid over-heating), but if
I use it there is no current in the MOTOR
4. Capacitors on pin 6 and 5 are 10nF instead of 100 nF

I will use a 12v 10A motor, and a 12v 20Ah battery, but for testing
purposes I am using an standard battery car (12v 50Ah).

The circuit is working fine but I am experiencing some issues:

a) When the motor is not working at maximum I hear a high frequency
noise.
b) With the 50Ah battery I saw a 30A peak current, and the transistor
burned out. I am thinking about using a more powerfull transistor. Do
I need to change it or not? Cause I will use a 20Ah battery in the
final design.
c) I will use this PWM in a three-wheeled electric vehicle. What do
you think about the general project itself? It is enough with that
battery and motor? Any other advice?

Thanks in advance for your help.
Vicente.

 

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.

 

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´ve 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?

On the other hand, what do you think about using 2 parallel BUZ11 (30
A peak burnt out one)? How do I have to connect them?

 

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..

 

"Vicente"

** Bad circuit for high power PWM use.

The frequency varies with the duty cycle.

This is a much better schem:

http://www.cpemma.co.uk/555pwm.html

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 .

 

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:

http://www.cpemma.co.uk/555pwm.html

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