Hello, I've made this scheme to control a DC motor:
(Not allowed to post images and urls 
until posing one "normal message")
P.S. This was a "normal message"
And there's a little problem I can't understand the cause.
When I put 4V on the output pin(PWM), the transistor is almost full-open and the motor (no-load) consumes about 500mA. And everything seems perfect. But when I connect some load to the motor (e.g. propeller, wheel) and turn the pin for about 1.6V, the motor consumes same 500mA, but this time transistor goes hot very quickly and starts fuming =.
The transistor is TIP41C (6A maximum current)
Can anyone tell me what's the reason for this effect. Is there anything more important than current and voltage? Thanks 
I think you'll need to turn that PWM into an actual DC voltage before feeding it into the transistor. You'll just need a simple RC low pass filter for this
The transistor needs to be fully on or fully off, not somewhere in between, that is one reason why pwm is used to drive it.
The base resistor value seems too big for the gain of the transistor to me, try a lower value resistor bearing in mind the maximum output pin current of the Arduino.
When I put 4V on the output pin(PWM), the transistor is almost full-open and the motor (no-load) consumes about 500mA
That is a puzzling statement.
When you say the transistor is fully open do you actually mean fully closed or fully on?
Anyway that transistor has a very low gain, according to the data sheet it is less than 30 at half an amp. So you need more base current or more gain. It is best to have more gain so:-
You need another transistor, small signal high gain to make this transistor into a darlington. Put the second transistor to the firsts collector. Disconnect the original transistor base and put it to the new transistor emitter. Finally treat the new transistor's base as if it were the original transistor.
I think you'll need to turn that PWM into an actual DC voltage before feeding it into the transistor.
Couldn't you explain then why PWM is not sutable here and what will change after converting the signal?
The transistor needs to be fully on or fully off, not somewhere in between
Wait, what do you mean..? I've seen several project where people use transistors to smoothly control the speed of motor/light of a LED.
The base resistor value seems too big for the gain of the transistor to me
Actually, I first connected an 1K1 resistor, but could hardly turn the transistor on. As this is the smallest resistor I have, I connected the transistor directly to pin (the maximum base-emitter then current makes up 60mA)
When you say the transistor is fully open do you actually mean fully closed or fully on?
I mean fully on, sorry
Thanks a lot for your replies, but, however, I still have no idea why a load on a motor makes the transistor fuming (with only 500mA collector-emitter current). While without any load and the same current it doesn't even become warm.
The problem is that everything works just as expected. Right until I put a wheel on a motor =.
Sorry I misunderstood what you were trying to do. Ignore my comments above. Just out of interest - have you measured the current consumption of the motor under your load condition in a simple circuit?
I connected the transistor directly to pin (the maximum base-emitter then current makes up 60mA)
Bad because the arduino pin should not be driven that hard 30mA at most.
with only 500mA collector-emitter current
How are you measuring this?
The only way for the transistor to get hot is if the current increases or the collector / emitter voltage increases. Try measuring the collector / emitter voltage under load / no load conditions then measure the current again under load / no load.
It could be that you power supply is flagging and causing an overall voltage drop thus reducing the base current.
have you measured the current consumption of the motor under your load condition in a simple circuit?
Well, actually, no :). Wait an hour, I have to wait for the glue to dry. It's not easy to connect a load here
Bad because the arduino pin should not be driven that hard 30mA at most.
Hm, I've read about 40:
DC Current per I/O Pin 40 mA
I understand that it is bad, but I do not have an appropriate resistor yet, and I never let to run those 60mA for more than 2-3 seconds.
How are you measuring this?
Like this.. mm. How can I measure this other way?
Try measuring the collector / emitter voltage under load / no load conditions then measure the current again under load / no load.
Yes, I'll do this as soon as possible, thanks.
have you measured the current consumption of the motor under your load condition in a simple circuit?
Well, actually, no . Wait an hour, I have to wait for the glue to dry. It's not easy to connect a load here
I've measured this. Powered from 4.5V power supply, it consumes 200-220mA without load. With load.. 1700mA..
Powered from 9.5V - 470-520mA without load and with load I'm afraid to test because my power supply is only for 2000mA
Is it normal?
and I never let to run those 60mA for more than 2-3 seconds.
Doesn't matter about how long you are damaging your arduino pin.
m, I've read about 40
40mA is a stress rating not an operational rating, do you understand the difference? If not read the processor's data sheet, the section on electrical characteristic. You will damage your arduino if you run at 40mA or above.
200-220mA without load. With load.. 1700mA.
So that is why your transistor is getting hot. Yes it is normal for a motor to draw more current when it is under load. The ultimate load is a load that is enough to stop the motor, then it draws what is known as the stall current.
40mA is a stress rating not an operational rating, do you understand the difference
Yes, I do. Didn't though it's so important. Thanks, won't do like this any more.
200-220mA without load. With load.. 1700mA.
So that is why your transistor is getting hot. Yes it is normal for a motor to draw more current when it is under load. The ultimate load is a load that is enough to stop the motor, then it draws what is known as the stall current.
Yes, I've already learnt about this stall current. But I still cannot understand something.. Firstly, the transistor is estimated on 6A. Lets assume even on 3A. Secondly, I've put 1.8V on PWM specially to reduce the base-emitter current, so that the collector-emitter current is also low. with 1.8V on pin, I have 20mA base-emitter current and 500mA (as my multimetr shows..) collector-emitter current. So with the motor loaded I only let 500mA run through the transistor.. Not 6A or even 3A. Where am I wrong?
P.S.
1700mA is the current in a simple curcuit, not limited by transistors or other stuff.
Whenever I've tried to read PWM signals with my DMM - I may as well have guessed the figures. It generally ends up being some weird average, typically it goes up kind of gradually (but jumping around) before suddenly maxing out.
I've put 1.8V on PWM specially to reduce the base-emitter current,
No you haven't, you still have the full 5V unless you smooth the PWM signal which you don't want to. If you put less base current the transistor will not turn on fully and so will drop power in the transistor.
So with the motor loaded I only let 500mA run through the transistor.. Not 6A or even 3A. Where am I wrong
It gets hot because of the power not the current. The power is voltage times current. So even at less than the maximum current the voltage may be such that the power rating of the transistor is exceeded. Even at the minimum voltage (Collector emitter saturation voltage) you can't run the maximum current because the power would be too high. In a continuous mode of operation it is the power that limits the maximum usable current on this device.
See these sites:-
http://www.thebox.myzen.co.uk/Tutorial/Power.html
Whenever I've tried to read PWM signals with my DMM - I may as well have guessed the figures.
I actually thought it's not a good way to measure PWM values with DMM, but while experimenting with a LED, it showed quite likely results, so I started believing it.
which you don't want to.
I didn't say I don't want to, I'm just trying to make out all this electronic things. And I won't be very glad just filtering the PWM to make it as simple DC voltage without understanding the reason for doing it.
It gets hot because of the power not the current.
Oh, this might be the thing I wanted to hear from the very beginning.
As I understand now, during those periods of time, when PWM signal is 5V, the transistor is fully on and lets the huge collector current to run through it. Multiplied on 9.5V (power supply of a motor), it hits the maximum power dissaption (65W) and makes the transistor boil. If it is so, everything falls into place :).
Thanks for links also, will study them..
If it is so, everything falls into place
Sorry no. when there is lots of current there is only a small voltage across the transistor. Look at the data sheet for the exact value it is known as Vsat of Vce. However it is this small value that is giving you the few watts that istoo much for this transistor without a heat sink.
The point is that when the transistor is off there is no current flowing so zero times nine volts is zero. That point allows the transistor to cool a bit.
If you use a FET there is much less voltage dropped across that when it is on and it runs cooler for a given current.
The DVM (a cheap one) will only show normally an average voltage and not very well at that.
Sorry, was cut off from the internet.
If you remember what we were talking about..
Sorry no.
:-/ Oh, how complicated the world is.
when there is lots of current there is only a small voltage across the transistor
Finally, I've measured this, and it appeares to be true.
If you use a FET
I've read about them before starting building schemes, and it seemed quite remarkable thing to me, but when I saw other people's things made with transistors, they all were using usual bipolar transistor, I don't know why.
By the way, I bought a new transistor, a more powerful one (120 W), but it becomes hot all the same.
What would you do in my shoes? I've heard about some heat sinks, may be this is the way to go..?
Thanks a lot.
a more powerful one (120 W), but it becomes hot all the same.
The rating of a transistor is how much it will take before it is damaged. This is not related to how hot it will get when you pass a certain current through it. That is fixed by the circuit you are using it in. If you have 1700mA through a transistor with a saturated on voltage of 1V then you are going to have to burn 1.7W of power in the transistor. This is the same amount of heat irrespective of what transistor you have.
Yes a heat sink is the way to go if you are within the current rating of the transistor.
they all were using usual bipolar transistor, I don't know why.
Transistors were around before FETs and so were cheaper and more widely used. They can also be switched with smaller voltages / currents. Typically a FET will need 10V to turn it on but now there are logic FETs that turn on at 4.5V but logic FETs that handle higher current are not so common.
... but logic FETs that handle higher current are not so common.
Maybe so, but there are some potent devices out there.
Take a look at this one:
Rated at 195A continues drain current, 2 milli Ohm on resistance and max Vgs threshold as low as 2.5V.