Hello,
For the L298 motor driver, I see that there is a tab used to attached a heat sink to.
The L298N (vertical standing) and L298HN (horizontal laying down) have a tab that is connected to GND (Pin 8) of the chip. When it comes to the SMD package (L298P) it doesn't state if the thermal pad is GND. I see pictures that the L298P has a thermal pad but the datasheet doesn't a) show it and b) indicate its connection.
Does anyone have experience with this specific chip that can help me out?
Daniel
I don't know the answer to your question, but if you are building your own board using SMD components, why on earth are you using an L298? There are lots of mosfet-based SMD motor driver chips available that are far superior to the L298. In particular, they have much lower voltage drop (and therefore they generate much less heat), and most of them have built-in current limiting.
dc42:
I don't know the answer to your question, but if you are building your own board using SMD components, why on earth are you using an L298? There are lots of mosfet-based SMD motor driver chips available that are far superior to the L298. In particular, they have much lower voltage drop (and therefore they generate much less heat), and most of them have built-in current limiting.
The first revision of my board utilized the VNH5019, I found those to be extremely efficient but also pretty tedious in soldering them. I hear a lot of horror stories with the L298 being a fat power dissipating radiator of heat. Fortunately for my application, I hooked up the L298's H-Bridges in parallel boosting up my current capacity, hoping this will keep the driver and a decent level of heat. Essentially each motors is being driven by one L298.
For my application, I do not expect these motors to pull anywhere past 600 mA - 800 mA of current. With the L298 being use in parallel, it can output 4A, this is giving me the OKAY to use these chips. Unless you have another reason in which I haven't looked into, I want to use the L298 for it's simplicity.
Any suggestions?
Daniel
Don't forget the voltage drop though: I've never read the datasheet of the sm version, but assuming it's the same as the others, the drop is 1.8 at least, even under no load. So just remember to throw that many extra volts at it......
The L298 has a voltage drop of 2.55V typical, 3.2V max @ 1A. So at 800mA load, each L298P will have to dissipate around 2W, perhaps even 2.5W. You will need a lot of PCB area fanning out from the thermal slug to get rid of that heat. Even if you solve that problem, you will need to provide a higher battery voltage to allow for the voltage drop.
If the VHN5019 has too many legs for you to solder easily, how about using the TB6612FNG dual motor driver, if your supply voltage is 13.5V or less? It can drive 2 motors at up to 1.2A continuous, with a typical voltage drop of 0.4V @ 800mA. So you will only have to get rid of 1/6 of as much heat compared to the L298. You could connect both channels in parallel for even better performance.
btw if you have been hand-soldering the VHN5019, then I strongly recommend that you look into other SMD soldering techniques. The hotplate method is a good place to start.
PS - if you want to experiment with the TB6612FNG you could get this http://www.amazon.co.uk/Motor-Driver-1A-Dual-TB6612FNG/dp/B007XEVHLC.
dc42:
The L298 has a voltage drop of 2.55V typical, 3.2V max @ 1A. So at 800mA load, each L298P will have to dissipate around 2W, perhaps even 2.5W. You will need a lot of PCB area fanning out from the thermal slug to get rid of that heat. Even if you solve that problem, you will need to provide a higher battery voltage to allow for the voltage drop.
http://www.amazon.co.uk/Motor-Driver-1A-Dual-TB6612FNG/dp/B007XEVHLC.
dc42, I appreciate your suggestions and I believe I will be switching my driver based on what you have provided me. I did not know that there such a significant voltage drop on the L298 that would contribute to heat production. I am assuming you calculated that 2W from 2.55V (voltage drop) * 800 mA current draw = 2.04W dissipation.
dc42:
If the VHN5019 has too many legs for you to solder easily, how about using the TB6612FNG dual motor driver, if your supply voltage is 13.5V or less? It can drive 2 motors at up to 1.2A continuous, with a typical voltage drop of 0.4V @ 800mA. So you will only have to get rid of 1/6 of as much heat compared to the L298. You could connect both channels in parallel for even better performance.
http://www.amazon.co.uk/Motor-Driver-1A-Dual-TB6612FNG/dp/B007XEVHLC.
I am powering my system through a 9V regulated from 12V battery supply, so the TB66 seems efficient enough to drive the motors. The motors in particular is linked below:
As you see the current draw is low and with the torque provided, my application will not load much further than 2-3 times the free-running current.
dc42:
btw if you have been hand-soldering the VHN5019, then I strongly recommend that you look into other SMD soldering techniques. The hotplate method is a good place to start.
http://www.amazon.co.uk/Motor-Driver-1A-Dual-TB6612FNG/dp/B007XEVHLC.
The pin count isnt the biggest trouble for me. My first revision of the board, I really didn't think about how I will solder the thermal slugs located on the bottom of the chip. I did not know about using vias to feed in solder from the bottom. I had to use a hot air gun with solder paste to attempt to get this to work. It was a large mess and it kind of forced me away from the chip. Our lab has no other tools, like a reflow oven to appropriately do this. I like this TB66 because it seems to be efficient without a thermal slug.
Additionally, I am using the Arduino Due, and the TB66 seems to allow CMOS level operation voltage (>2.7V).
Something is concerning me, the breakout board doesn't have fly-back protection diodes, and the datasheet doesn't specify that it is included. I am sure they are necessary.
Thank you for your suggestions, I appreciate them.
Daniel
dgelman:
I am assuming you calculated that 2W from 2.55V (voltage drop) * 800 mA current draw = 2.04W dissipation.
Correct.
dgelman:
I am powering my system through a 9V regulated from 12V battery supply, so the TB66 seems efficient enough to drive the motors.
It's unusual to power motors from a regulated power supply.
dgelman:
As you see the current draw is low and with the torque provided, my application will not load much further than 2-3 times the free-running current.
However, the stall current is quite high, looks like it will be about 3.75A @ 9V. So I suggest you either use both channels of one chip in parallel per motor, or (if you do not require the full torque), connect a low-value resistor in series with each motor to limit the current. It's probably also a good idea to use soft-start.
dgelman:
Something is concerning me, the breakout board doesn't have fly-back protection diodes, and the datasheet doesn't specify that it is included.
All mosfet-based motor driver chips have the flyback diodes built-in. They are shown on the TB6612 datasheet.
dc42:
It's unusual to power motors from a regulated power supply.
Yes I agree with you. For my application, I am using a custom battery and it is important that I maintain a constant DC external motor supply. Regulating the 12V to 9V with a low-dropout high-current voltage regulator should do the job.
dc42:
However, the stall current is quite high, looks like it will be about 3.75A @ 9V. So I suggest you either use both channels of one chip in parallel per motor, or (if you do not require the full torque), connect a low-value resistor in series with each motor to limit the current. It's probably also a good idea to use soft-start.
I doubt I will ever get close to the stall toque, but just to be same I will nonetheless use one driver for one motor. More current capacity cant hurt anything.
Again dc42, thank you for your suggestions.
Daniel
When DC motors start from a standstill, they draw the stall current for a short period of time. If you suddenly reverse the direction of rotation, the short-term current draw is even higher than the stall current.
jremington:
When DC motors start from a standstill, they draw the stall current for a short period of time. If you suddenly reverse the direction of rotation, the short-term current draw is even higher than the stall current.
Hello jremington,
Yes the inrush current will a concern since the driver will not be able to provide that current, ~3.75A at 9V. Additionally, my power supply is not rated for the inrush current. I will need to develop a current limiter or just add a resistor to enable a soft-start.
Daniel
So, to move past all the "oh, don't use that part" stuff, could anyone answer the OP's question:
What connection for the thermal pad?