You need to apply Thevenin's Theorem.
So the equivalent drive to the bases of Q5 and Q6 is:
Voltage = 5(10/(10+3.3)) = 3.76V
Resistance = 10k//3.3k = 1(1/10k + 1/3.3k) = 2.48kThe Vbe drop is going to be about 0.7V. Therefore the drive current is:
(3.76 - 0.7)/2.48k = 1.23mA
As far as "calculate to get the resistor value", did you mean "how do I determine all the resistor values?"
I don't quite understand this part;
Voltage = 5(10/(10+3.3)) = 3.76V.
How do you get 10 and 3.3?
I cant find the current gain on the datasheet.
No nether can I.
Note this is not a very good circuit. Q1 will have a voltage drop of 2V and so will Q2 so therefore with 24 V on the input the motor will only ever see 20V.
Remember when calculating resistors like this you are calculating the maximum possible resistor value. It is normal practice to take this value and half or quarter it, or even more, to give you some margin.
Grumpy_Mike:
I cant find the current gain on the datasheet.
No nether can I.
Note this is not a very good circuit. Q1 will have a voltage drop of 2V and so will Q2 so therefore with 24 V on the input the motor will only ever see 20V.
Remember when calculating resistors like this you are calculating the maximum possible resistor value. It is normal practice to take this value and half or quarter it, or even more, to give you some margin.
So what should I do?
Btw how do you know the voltage drop is 2V? voltage drop is Vbe(sat) right?
jimmy12:
Btw how do you know the voltage drop is 2V? voltage drop is Vbe(sat) right?
No it is the Vce(sat) + 1.2V ( two transistor's Vbe)
So what should I do?
Follow the component values on the schematic.
No it is the Vce(sat) + 1.2V ( two transistor's Vbe)
I don't get it.
No. That 30mA base current is for the output transistors. For Q5 and Q6, their Base current should be 1/10th that. A 2N2222 isn't a particularly good switching transistor, so use 1/10th. I would -never- simply use the small signal current gain to calculate this for a saturated switch. Gain drops precipitously both with larger currents, and when you approach saturation.
But you are looking at the Thevenin equivalent of R3 and R4 (R5 and R6). Did you read the PDF I posted about it?
The gain of both transistors is on the datasheet, as is the rated voltage drop at the specified current.
polymorph:
No. That 30mA base current is for the output transistors. For Q5 and Q6, their Base current should be 1/10th that. A 2N2222 isn't a particularly good switching transistor, so use 1/10th. I would -never- simply use the small signal current gain to calculate this for a saturated switch. Gain drops precipitously both with larger currents, and when you approach saturation.
Meaning 30mA is to Q1 and Q2 which comes from the PN2222?
What type of transistor would you recommend?
Multiply 1 x 1000 (because it says K) and you get 1,000 ohms.
I followed the value shown on the given schematic. I replaced Q5 and Q6 with pn2222. But once I on the supply, one of my pn2222 exploded. May I know what is the reason?
I completely disagree the schematic values are optimum. With 3.3K on the base you will only supply roughly 1.5mA to the transistor if driven from a 5V pin. Even though that SHOULD turn on the 2N2222, there is no reason to not hit it harder. Operating the transistor in the linear range will cause it to go pop because it can't handle 24V x 24mA for long.
Take the 3.3K and replace it with 300 ohms, replace the 10K with 470. Really that 10K can be almost anything above about 200 ohms. The reasoning is it insures a saturated turn-on of the 2N2222 and by keeping the impedances low, improves switching times.
Personally, I don't care for this circuit, especially if it is to provide PWM control. Without separation between "high" and "low" side switches, you can't JUST PWM the low side, you are forced to PWM the high side as well. This is why most H-Bridge designs separate them. There's no reason to PWM both, it only increases your switching losses. On the other hand, using Darlingtons is prone to huge losses anyway, but as you said you are stuck with them.
What did you have in there before you substituted PN2222? Why did you substitute?
I agree with what rmetzner49 said.
The original schematic had a 2N3904 in that spot. The 2N2222 is actually a more robust device than the other both in terms of Hfe and IcMax. The 2222 gain is 35-100 so it should still have worked. Maybe connected incorrectly?
It works. It did not explode after I replaced the resistor value that was being told. For the diode, would it be OK if I use 1n4007 with a current of 1A? Or do I need a more current diode?
It's never a good idea to run any electronic component at its MAXIMUM. If you are doing a 3.5A bridge, I would use a 5A diode. Schottkys, when you are below about 60V are a good fit but since you are not going to PWM, speed is relatively unimportant.
Though for the price, why not use a Schottky? http://www.vishay.com/docs/88721/88721.pdf. These are $0.54 each at Digikey.
About HALF the voltage drop of regular diodes. Yeah, they don't do much most of the time, but when you change direction.
