L298N low torque

@jonnyr

Perhaps start off with a high PWM to get it moving and then back to the speed you want which if the slowest possible is something you can find out with trial and error.  One way to increase torque at slow speeds is with a variable gear box.

Thanks @robotbuilder. I'm just using one of those cheapo plastic motors that come with the robot chassis kit. When I test the motor with a power supply it is very difficult to stop with my fingers. But when I test with the motor controller and just a battery supply, it is quite easy to stop with my fingers. I have tested with a jumper on ENA and ENB.

@jonnyr 

By "power supply" I assume plug into the wall? Voltage and maximum current rating?
What type of battery, voltage and maximum current rating?
Voltage required to give the motors maximum grunt?
The battery probably cannot deliver the same amount of power to the motors as the other power supply.

Given the numbers you can calculate the requirements.

@robotbuilder 

Sorry, that wasn't very clear. By power supply I meant battery without the motor controller. I think I should probably try a higher pwm. I was using an esp32, which uses 3.3v logic for pwm. Do you think I should use a shifter?

Battery is 2x18650 Li-ion. So, voltage around 8v.

@robotbuilder 

Also, do you think I might get a better result with a MOSFET type motor controller?

@jonnyr 

What is your source for the project?

I see Bill uses a mosfet h-bridge.

@robotbuilder I just used an L298N because I had one lying around. I think it's probably a bit faulty. I will see if I can test with another one. I think the BJT has a larger voltage drop then the mosfet.

Are you using an L298N directly? or via a controller board module? It can be trickier to use correctly than is commonly appreciated if you're using it directly.

I would recommend poking around with a multimeter to see what the L298N is actually doing, regardless of the form.

Thanks @jker. I tested the voltage across the output terminals and it was ok. I got a new L298N and it seemed to work much better. Perhaps it just wasn't able to draw enough current.