Hi to all,
I'm designing a motor driver circuit. The H-bridge is already ok. I want now to stop the motor if it’s current reach 100mA. So, I’m thinking about a circuit that sense the 100mA and send a signal to the 5V microcontroller. I’m searching for a simple circuit, with 1 transistor that amplifies the shunt voltage of a resistor. I don’t need a lot of precision. I don’t want to use opamp.

Any idea ?

 

Because you are using an H bridge, the motor current can go both ways. The usual solution is to connect the sources of the bottom two transistors together and connect these to ground through a low-value current measuring resistor. No matter wich way the motor is going, the current through this resistor is always same direction.

For example you could make this resistor 2 ohm. At 100mA there will be 0.2V drop. Feed this into a comparator that switches at this value. The output of the comparator goes to your micro-controller.

 

Yes I have already think to make a single resistor.
The issue is about the voltage amplification of R shunt. I don't have the AD converter, I need a 5V signal or a open collector, when the current go over 100mA.

Because you are using an H bridge, the motor current can go both ways. The usual solution is to connect the sources of the bottom two transistors together and connect these to ground through a low-value current measuring resistor. No matter wich way the motor is going, the current through this resistor is always same direction.

For example you could make this resistor 2 ohm. At 100mA there will be 0.2V drop. Feed this into a comparator that switches at this value. The output of the comparator goes to your micro-controller.

 

Yes I have already think to make a single resistor.
The issue is about the voltage amplification of R shunt. I don't have the AD converter, I need a 5V signal or a open collector, when the current go over 100mA.

That is what a comparator will do for you.
Take a look at an LM339 quad comparator or LM2903 dual comparator.
Datasheets are available at http://www.national.com

 

I know opamp, and is easy to do this with them. My need is to build the circuit with a transistor.

 

This is starting to sound like a homework assignment.

Is it?

 

This is the project need that I have in my circuit.

 

This is the project need that I have in my circuit.

That doesn't quite answer my question, so I'll re-state it.

Was your project assigned to you by a teacher, instructor or professor as part of a course/class requirement?

 

No, this project is about my work.

 

OK.
Well, you could use just a single transistor, but it will be very inefficient.

Use a 6.2 Ohm resistor from the bottom of the H-bridge to ground.
Connect the high side of the 6.2 Ohm resistor using a 100 Ohm resistor to the base of an NPN small signal transistor, like a 2N3904.
Connect the transistor's emitter to ground.
Connect the collector to Vcc using a 4.7k transistor.
When current through the 6.2 Ohm resistor exceeds about 100mA, the transistor will turn on, causing Vce to decrease.

 

Op amps are so cheap and simple that avoiding them for a one off circuit is not worth the bother.

 

I'm agree with you, russ. But I don't have space on the board and the application is very low power. Every uA is important. If I use an opamp I have to power off it. And the precision isn't important.

 

Well, the single transistor suggestion I gave you will waste 62mW via the 6.2 Ohm resistor when the motor is running at 100%. Keep that in mind.

 

Right, but the motor run only few second a day. An opamp waste power 24h/day. I already do some calculation, in my application the problem is to reduce the continuous (24H/day) current. The momentary current consumption isn’t a problem.
The issues of the circuit that you suggest, is the value of the Rshut that cause a 0,62V loss in motor voltage. I have to verify this.
And I have to test also the precision and repetitiveness.

 

They make pretty tiny OpAmps these days.

 

You could use something like a LMC6762 dual micropower comparator. It has push-pull outputs, which wouldn't require a pullup resistor on the output. 10uA current draw max per channel. You could power it using a uC pin set high. Set the pin low before putting the uC to sleep.

You could then use a 0.5 Ohm or less resistor to sense motor current. You'd still need a couple of resistors to use as a voltage divider. That could also be powered by the uC output pin.

 

Here's a very small, micropower single-channel push-pull comparator by Micrel:
http://www.micrel.com/_PDF/mic7211.pdf

 

What neat little comparators, Sgt.I've got to have some.

 

It's already done with 1 transistor;
http://www.romanblack.com/current.htm

 

Roman,
Ran your circuit in simulation; it didn't perform very well. Using a 2N5088 helped, but I was not able to get a reasonably accurate window.

Adding the load of the LED and current limiting resistor really throws things out of whack.

I've attached a simulation that shows your basic circuit, plus another using a channel from an LM339 comparator.
The b and a 'scope markers correspond to 95mA and 105mA through R1, respectively.
Much of the time, the output from your circuit is at an indeterminate voltage level for TTL or CMOS.