Anyone Used The L9110S Motor Driver Chip?

[JonSea]

The L9110S motor driver IC (click for data sheet) looks interesting for an application I have. I have 2 groups of 4 LEDs to control. Currently, a pair of micro pins control a high side switch for each group of LEDs. All of the anodes are connected in common to an n-channel mosfet controlled by a pwm pin. This allows me to switch on either or both groups and control the brightness using pwm. Another set of LEDs is controlled the same way using the other pwm of a PIC18F25K22.

It seems to me I could use one of the L9110S H bridges to control each group of LEDs. The output of L9110S can control >800mA, which is 10× or more current than I need.

This has a couple advantages for me. First, it makes each group of LEDs more independent for easier circuit board routing and saves some space on the board. It also cuts costs by about 75%.

Potential problems - the high side switch looks like it drops a couple volts, which I should be able to compensate for with proper LED series resistor collection. The other problem - when using pwm, the LEDs will be reverse-biased when off. I believe the LEDs will be ok being reverse-biased on a 5 volt system.

My schematic is shown below. The system operates from 5 volts and the LEDs draw about 10mA each (maybe as much as 20mA).

Does this look workable?

 

[JonSea]

Just a note - the truth table for the L9110S is a bit different than most. If both control inputs are high, the two outputs are low.

This scheme maintains the same logic as my original design, meaning no code changes are needed, which makes maintenance a lot simpler.

 

[rjenkinsgb]

As you don't need reverse polarity, you could just run the LEDs between one output and ground.
That should halve the voltage drop in the chip and reduce power dissipation within it.

The drive logic remains the same, just without the possibility of reverse bias.

 

[JonSea]

Sorry, that misses two of my points:

¤ ] need 4 LEDs drawing between 40mA - 80mA total. A PIC18F series can supply a maximum of 25mA

¤I need to pwm either or both groups of 4 LEDs using a single pwm on the PIC, which again can source or sink only 25mA (max load of 160mA).

I am not looking to reinvent the wheel. My proposed solution replaces a working solution with a implied layout and cheaper BOM.

 

[Beau Schwabe]

Why not just use a transistor and an I/O from the micro that way? or ChrliePlex the LED's?

 

[JonSea]

Charlieplexing has its place, but this isn't it. What I'm doing isn't complex - I want and need two groups of 4 LEDs to be on or off, with variable brightness controlled by pwm. Each group will draw between 40 and 80mA so the pwm needs to handle as much as 160mA. I don't want the complexity of multiplexing as my micro is plenty busy already.

I already have a working solution using a dual high side switch and an N channel mosfet for pwm modulation control. I am specifically considering the L9110S to reduce costs and simplify circuit board layout with at most only minor changes to my firmware.

So my question remains: as I've shown in this schematic, will the L9110S work well for my application?

Thanks,

Jon

 

[rjenkinsgb]

Sorry, that misses two of my points:

No.. cross purposes!

I mean use the L9110s as power drivers as per your diagram, but only one output of each & ground.
(eg. LED cathodes to ground instead of OA).

You reduce the voltage drop and cannot apply reverse polarity.

Your L9110 circuit is otherwise unchanged.

 

[Beau Schwabe]

"I already have a working solution using a dual high side switch and an N channel mosfet for pwm modulation control " - You could probably cut half of your costs by eliminating the high side driver section and high side mosfets all together. Just tie all of the LED anodes to +5V and modulate the LED cathodes using a LOW side mosfet or simply a BJT. Post your "working solution schematic" so we can see what works rather than try to guess at an alternative solution. What are you paying for the L9110 ? You can get a couple of BJT or single mosfets for $0.06 a piece.

 

[JonSea]

Sorry for the confusion. I was asking for comments on a specific idea, not for a discussion about alternative suggestions. As one of my coworkers used to say, there are many ways to skin the electronic cat and I'm sure other solutions would work as well or even better than my idea but that's not what I am asking for.

The cathodes of each group of LEDs must be switched to control each group of 4 LEDs independently. The anode common to ground is controlled by a single pwm output to control the brightness of both groups of 4 LEDs. Both groups will be at the same brightness if both are active.

There are actually two groups as described above. Each of these two groups must have independent pwm modulation control. For my PIC18F25K22, that uses the available pwm channels. IF I had 4 channels of pwm, I would not need to switch the cathodes of each group of 4 LEDs, but in the range of hardware I wish to use, this isn't an option.

Again, sorry for any hard feelings caused, but I want to evaluate the proposed solution rather than come up with something different.

Thanks,

Jon

 

[rjenkinsgb]

I was only suggesting you change one connection in your diagram - the devices are perfect for the job, it optimises their function.

The LEDs will be on under exactly the same input combination etc., just with less voltage drop in the driver IC and without the ability to reverse voltage them, which was the thing you were querying.

eg.

 

[JonSea]

Ahhh... Now I see what you meant. That may do the trick.

If B is high, and A is low, Output B is high and the LEDs are on. If A is high during pwm and B is high (to activate the group of LEDs), Output B is low and the LEDs extinguished. If B is low to turn off the group of LEDs, A is don't care. Very nice.

I think this is going to invert my pwm signal. Presently, I'm using an N-channel mosfet to switch to ground, so a high signal turns the LEDs on. That's a minor firmware change to make. I could add a bit in the PIC's eeprom to control the sense of the pwm.

Thank you!

 

[alec_t]

Judging by the rated tolerances of the IC, in the worst case you could be dropping about 2V in each bridge. Depending on your LED specs, that could leave very little headroom for equalising LED currents with the resistors if you are using a 5V supply.

 

[JonSea]

Yeah, I wish the datasheet provided some greater detail about that! If I make the change that rjenkinsgb suggested, I'll only have the drop of the high side bridge (~1.5v) so I think I'll be ok in that regard. I am using red and yellow LEDs - I believe Vf of the yellow is about the same as the red, around 2.2v.

 

[alec_t]

I think I'll be ok in that regard. I am using red and yellow LEDs

That should be fine then.