An alternative to the L293D dual H Bridge IC

westaust55

Moderator
A little late for my recent project work involving some 50 off L293D dual H-bridge IC's (some folks may also know of the approx. equivalent the SN754410) however I have just come across an alternative dual H-bridge IC being the LB1909MC.
http://www.onsemi.com/pub_link/Collateral/ENA2037-D.PDF


This is only available in a 10 pin SMD SOIC package format which is still solderable using "normal" soldering irons with care. I have soldered SOIC chips from 8 pin to 32 pin successfully in the past.

There are seemingly a few advantages including:
1. No need for a separate 5Vdc logic supply
2. One input controls both sides of a H-bridge - so no need for an inverter to the other side
3. Far lower internal current draw than the L293D (ie 25 mA when En is high compared to 44 mA)
4. Output saturation voltage is lower (ie 0.5 V @ 400 mA compared to 1.2/1.4V @ 600 mA for L293D)

There is only 1 Output Enable pin to control both H-bridges.

So certainly an alternative IC that I will be considering and maybe experimenting with in the near future for bi-polar stepper motors and small low current DC motors requiring bi-directional control.
The only source I found at this time was Mouser.
 

Technoman

Senior Member
Thanks for the info.

It is an interesting product, with some improvements and cheaper than L293 (Farnell).
A point to check is the need or not for an heatsink.
I know, one day, I'll have to switch to SMD devices...
 

PieM

Senior Member
Thanks,

Certainly an alternative IC for small bi-polar stepper motors, but for small DC motors, no possibility to stop only one motor, or to sent pwm1 for motor 1 and pwm2 for motor 2.
 

westaust55

Moderator
Thanks,

Certainly an alternative IC for small bi-polar stepper motors, but for small DC motors, no possibility to stop only one motor, or to sent pwm1 for motor 1 and pwm2 for motor 2.
True PieM.
The LB1909MC is not going to replace the L293D for every application.

The application I had recently was for ~100 Cobalt "stall" motors ("Tortoise" is another brand) each with a small DC motor and gearbox for slow motion output and a typical stalled motor current around 20 mA.
No PWM and as the motors are of the "Stall" type they are continuously energised - just needing to change direction when required.
 

inglewoodpete

Senior Member
Certainly an alternative IC for small bi-polar stepper motors, but for small DC motors, no possibility to stop only one motor, or to sent pwm1 for motor 1 and pwm2 for motor 2.
If the price was right, I'd consider using one for each motor (They are only 5mm long and 6mm wide!)

Edit: Just checked the price - typically under 1GBP/US$2

@WA55: Are you exhibiting a layout at next weekend's Model Train Exhibition?
 

westaust55

Moderator
@WA55: Are you exhibiting a layout at next weekend's Model Train Exhibition?
Hi IWP,
The answer is no. My layout is still under construction though might look close at a cursory glance - maybe another year to go. Still no people on the streets, platforms or in vehicles. Adding controlled lighting and (static) people and lighting to carriages slowly.
Station Lighting1s.jpgStation Lighting1s.jpg
PICAXES used for accessory decoders and controllers for building, street and platform lighting and yard swiveling crane plus semaphore signal control (SMD 08M2's plus other SMD chips in the base of the semaphore signals though still to install 90% of the signals).

Albeit off topic, the Perth (Australia) Model Railway show run by AMRA apparently made a loss last year. Down to 1 pavilion for 2016. If it makes a loss in 2016 it may be the last AMRA model railway shown in Perth.:eek:
 
Last edited:

westaust55

Moderator
I have now identified a further alternative to the L293D which is also manufactured by OnSemi.

This is the LV8549MC
http://www.onsemi.com/pub_link/Collateral/ENA2039-D.PDF

The pinout and control is the same as the LB1909MC mentioned in Post 1.
The differences are:
1. a newer device (2014 vs 2013)
2. Using MOS output devices instead of BiPolar transistors
3. Higher efficiency (standby mode and internal operating currents about 10% of that for LB1909MC)
4. 1 Amp per channel as opposed to 0.8 Amps
5. Ron (high + low side) total is typically 1 Ohm at 1 Amp, so with 200mA the voltage drop is 0.2 V or better and thus equvalent or better than the LB1909MC and far better than the ubiquitous L293D.

For those who do need to control 2 DC motors and need to stop each individually there is the LV8548MC.
Still in an SOIC10 package but with 4 separate control inputs.

I have placed an order for a quantity of the LV8549MC for some experimenting/trial purposes. Also some SOIC10 to DIP10 adapter boards for these initial testing.
Lower cost and readily available from my local (RS Components/Online) suppliers
 

Technical

Technical Support
Staff member
How do you control the 2 DC motors individually?.... you need 4 control pins to give full forward/reverse/stop control of 2 motors.
 

westaust55

Moderator
How do you control the 2 DC motors individually?.... you need 4 control pins to give full forward/reverse/stop control of 2 motors.
As discussed earlier, YES, the LB1909MC cannot stop one of the two motors alone. It is a case of both stopped or both running with independently selectable directions.
For some applications that may be adequate. Certainly in my personal case with many "stall" motors being controlled I do not turn off the motors and the L293D enable pins are all permanently high.

As mentioned in post 8 above, "For those who do need to control 2 DC motors and need to stop each individually there is the LV8548MC."
http://www.onsemi.com/pub_link/Collateral/LV8548MC-D.PDF

The LV8548MC is still an SOIC10 device but has 4 inputs and 4 outputs. The two extra inputs use:
  • pin 4 which is not used for the LB1909MC and LV8549MC
  • pin 2 which is the ENAble pin on the other 2 devices.

Not having an Enable pin is not seen (by me at least) as a significant problem as the LV8548MC can stop the two motors using the 4 input control pins and the overall internal power consumption is far less than the L293D.

In terms of internal power consumption, by way of information from the datasheets:
For the L293D:
  • the Total Quiescent Logic Supply Current (Iss) at 5 Vdc with the Enable pin high is typically 44 mA to a max of 60 mA per chip. For the many L293D chips I am using it is proving to be closer to the 60 mA per chip.
  • the Total Quiescent Supply Current (Is) = at the motor supply voltage is typically 16 mA.

For the LV8548MC and LV8549MC:
  • there is no separate control logic supply required
  • the Total Quiescent Logic Supply Current (Icc) at with the Enable pin high is typically 1.7 mA to a max of 2.3 mA.

That 5Vdc power reduction alone equates to ~2.5 Amps (or more) reduction in total 5Vdc power for 50 dual H-bridge controller/driver chips if I change all from L293D to the chips I have mentioned above.

The three dual H-bridge chips I have mentioned in this thread may still not cover everyone's requirements as an L293D alternative but certainly can go a long way for many folks.
 
Last edited:

westaust55

Moderator
To conclude my investigations into the use of LV8549MC (and LV8548MC)dual H bridge drivers for small stepper motors and DC motors, below are some data I measured after constructing and installing a couple of LV8549MC based driver boards in lieu of L293D based boards for control of small DC "Stall" motors. These small DC "Stall" motors are nominally rated 8 to 12 Vdc and have a nominal current draw of approx. 20 mA.
Each board comprises two H-bridge driver chips allowing control of 4 DC motors (or 2 stepper motors).

For an earlier produced L293D motor driver boards:
- Logic supply voltage at board terminals: 4.98 Vdc (nominal 5 Vdc supply)
- Logic supply current: 61 mA (this is lower than some of my past measurements on other boards)
- Motor supply voltage at board terminals: 8.98 Vdc (nominal 9 Vdc supply)
- Motor supply current with 4 motors connected: 90 mA
- Voltage at motor terminals: 7.55 V for all 4 motors
- Voltage drop within the L293D = 8.98 - 7.55 = 1.43 V (with nominal 20 mA load)
- Typical motor currents: 16mA (M1,M2,M4) to 25 mA (M3)
- Estimated Motor supply quiescent current: 90 - (3*16 +25) = 17 mA
These boards also used a 74HC14 hex inverter chips (4 of 6 inverters used) to invert the input signal to one side of the H-bridge so that only one control signal was required per motor driver.

For a LV8549MC based motor driver board:
- Logic supply voltage at board terminals No separate supply required
- Logic supply current: No separate supply required
- Motor supply voltage at board terminals: 8.98 Vdc (nominal 9 Vdc supply)
- Motor/logic supply current with 4 motors connected: 93 mA
- Voltage at motor terminals: 8.93 to 8.94 V for all motors M1, M2, M4 and 8.88V for M3
- Voltage drop within the LV8549mc = 8.98 - 8.93 = 0.05 V (with nominal 20 mA load)
- Typical motor currents: 19mA (M1,M2,M4 )and 30 mA (M3)
- Estimated Motor & Control supply quiescent current: 93 - (3*19 +30) = 6 mA
No additional chips required to control each motor with a single control signal per motor.

My proto-board based layouts for the LV8549MC also required only around one third of the area that my earlier L293D based boards needed and there was far less wiring required. Wiring is significantly reduced as the inputs are all on one side and the outputs are all on the other side. This permits the connections from the driver output pins to terminal strips on a simple one to one connection and for many proto-boards no actual wires/jumpers are needed to achieve the connections.
For my model railway application PICAXE and PIC microcontroller chips are driving 16-bit shift registers (2 x 74HC595s) but whether directly driving from a PICAXE chips or via a shift register, needing only 1 pin to control the motor direction is an advantage with stall motors.

As mentioned earlier, for full forward/reverse/stop applications the LV8548MC driver chips will need to be used with 4 control inputs to achieve the stop state.


A further note about the outputs.
For the L293D, if we consider that channels 1 and 2 are used for a H-Bridge and we consider input 1 as the control signal with input 2 being an inverted level of input 1 then the state of Output 1 will follow the input 1 and output 2 is the inverted/opposite voltage level).

For the LV8548MC, the situation is more like the L293D except there is no enable pins. If both inputs are low the two outputs are off (high impedance) and if the two inputs are both high both outputs are driven low to achieve a motor braking effect for faster deceleration. One should still read the specific datasheet.

However for the LV87549MC part, this situation is different. The LV8549MC with a single control input per H-Bridge has Output 1 at an inverted level compared to the input signal.
Thus when the control input 1 is taken high, output 1 is low and output 2 is high.
Where a drop in replacement board is used for existing L293D boards, it may therefore be necessary to swap the wires to the motor. Again please read the specific datasheet.
 

Attachments

Last edited:
Top