Need Help!- Totem pole output to drive a motor

[yervatn]

Hi,

I'm trying to use the following circuit to driver my motor (currently LEDs), but it seems the transistors just can't work in saturation mode.

Inputs (IN1 and IN2) are from MCU (maximum 8mA each port), 3.3V. Transistors are MMBT3904.

Would somebody please tell me what's wrong with this circuit? Or suggest an alternative?

Thanks and best regards,
yervant

[DaveSNH]

Can you send or post the attachment in a format that can be printed? Then I'll do the bias calculations for you and see if the transistors are in saturation.

What is the current and voltage of the motor you are trying to drive? 2N3904s are pretty light duty for most motor applications.

[bharbour]

The top two transistors won't saturate, they are running in emitter follower mode. The highest that the emitter of the top two transistors could be in an unloaded condition is 3.3-0.6V or 2.7 Volts. If you were trying to pull 100mA out of them (not much for a motor), it would drop down to around 1.2 Volts from the base current dragging down the voltage through the 1.5K resistor (assuming a current gain of 100 which is worst case for those transistors as I remember).

The lower transistors will saturate at low current. If TK1 turns on, the voltage at the base of Tk3 will be about 0.6V. You are dropping some voltage across the CE junction of TK1 and the 1K resistor to ground gets some of the current, so use Ib3max = ((3.3 - (0.2 + 0.6)) / 1500) - (0.6/1000) = 1.7mA - 0.6mA = 1.1mA.
Using 100 as the current gain again, TK3 will pull out of saturation around 1.1mA * 100 or about 110 mA.

At 100mA for a motor load again, the high side would be about 1.2 volts and the low side would probably rise to 0.6 volts, giving you about 0.6 volts across the motor. Not much.

Also, you need to add current limiting resistors between your driver circuit and the bases of TK1 and TK6. Those two points will not rise much above 1.2 volts (Vbe1 + Vbe3). If your driver can output 3.3V at 8mA, RB1 = (3.3 - 1.2) / 0.008 = 263 ohms or 270 for the nearest 5% value.

You also need to add protection diodes to protect the transistors from the inductive spikes form the motor. Connect one diode with the anode on the emitter of TK3 and the cathode on the collector of TK3. Connect another diode with the anode on the emitter of TK2 and the cathode on the collector of TK2. Repeat with two more diode for TK4 and TK5. These diodes will prevent the motor voltage from rising above or below ground when the transistors switch.

Connecting the top of the 1.5K resistor to +5V instead of +3.3V will help, it should buy you more output swing. If you re-designed it to use a PNP and an NPN as output transistors, and changed it to run completely off 5V, you might get 4.6 volts across the motor at low current.

Good Luck,
BobH

[yervatn]

@DaveSNH,
Yes, 3904 is a little light duty for 'most' motor applications (maximum Ic is about 200mA), but it just suits mine. Thanks anyway.

@BobH,
Your calculations helped me a lot in finding problems in this circuit. Thanks! Your suggestion on protection diodes is appreciated, too.


I did try some other parameters to this circuit, and try to add more transistors to make it work. Finally I gave up, for now. I'm just new to electronics and it's way too complicated for me to implement an H bridge driver. I think I'll do it later.

Anyway, I choose to use a mosfet driver IC, TC4426 from Microchip, to do the job. Since my little gadget works at 5V and draws about 50mA, I think I don't need any more MOSFET to drive it. The spec says power dissipation is 470mW for SOIC package, and Ro is 10 ohm maximum, so the output current is limited to about 216mA, which suffices my application. However, there's one concern. Can I use this IC to continuously drive my gadget?

Another question about driving a MOSFET from MCU I/O:
I know it doesn't make sense by connecting the I/O directly to the gate, but why? What's the point to add a series resistor? Doesn't the MCU output have its limit?

Cheers,
yervant

[bharbour]

Hi Yervant,
It is hard to answer whether that chip can drive your motor long term without more information. If your steady state current is 50 mA, your power dissipation in an H bridge with 10 ohm RDSon will be 2 x (0.050 ^2) x 10 = 0.050W.

If something stalls the motor, your current will likely be much higher. This is typically where motor drivers get into trouble.

Another considerations is the current required to start a motor can be around 5X the run current. Your driver needs to be able to supply that current.

A resistor is used between the gate driver and the gate on a FET for two reasons. It slows down the switching of the FET, which reduces radiated interference and large inductive spikes. Also, it helps prevent the FET from oscillating during and after the switching.

If you are driving n-channel FETs from a logic output, make sure that the source terminal on the FET is grounded and that your FET is a "sensitive gate" device. Normal n-channel FETs need to have the gate brought about 10V above the source terminal to switch on fully. A "sensitive gate" device will work with 5V or so. If your FET's source terminal is not grounded, there are parts called high side gate drivers that will translate logic levels to higher voltages. If a FET does not turn on fully, it may have much higher resistance and dissipate a lot more power, leading to loss of smoke.

BobH

[yervatn]

bobH, thank you for the explanations.

The spec of TC4426 says its peak output current is 1.5A, so I think it's OK. Anyway, I'll find out as long as I get the chip.

Best regards,
yervant