16F684 DC motor control help

[ciudatelul]

Hello

Hardware: 16F684

I have 2 buttons and I would like to give PWM, EN (Enable) and DIR(direction) to a motor driver to control the direction of my DC motor. What I want to do is from the buttons to give Start/Stop and change direction of the motor rotation.
What I want to is to capture also from a Hall Sensor the impulses given from the motor magnet (read the rotations period).
I have finished the schematic for all of this the only problem remains the programming of the PIC. I have looked thru the internet for examples and ways to do this. Found some and programing it thru CCSc in C made me thinks is the easiest way to do it.

This is what i wrote so far for my code. Still searching to add new things like debouncing or else. If someone has some ideeas the help would be welcomed.

The Start/Stop button is located on A0 pin
Direction button on A1
Enable output on C0
Direction output on C1
and PWM output on CCP1

[Mike Walne]

[ciudatelul]

Yes it's correct. A1 is for the button and C1 is for control of the direction.
The motor driver is an appropriate one (L9904). It has 1 input for PWM, 1 for Enable and 1 for direction.

The direction is changed by input high or input low on the driver pin.
The PWM freq I wanted to apply is of 20 kHz.
The motor driver has 1 pin of Enable and I wanted to control the on/off of the motor by controlling the input on that pin.

I have added for the use of internal OSC

[Ttelmah]

Comments:

1) Only select _one_ oscillator. Currently you have two - HS, and INTRC_IO. This is invalid, and will give unexpected results.
2) Do some reading on PWM. Higher frequency, allows smaller magnetics/capacitors, but increases switching losses, and requires higher current to drive the FET gates. 1KHz, is too low for a DC motor (noise), but 20KHz, is pushing the other end of the range, risking heating problems. Look at using the sort of frequencies other people use for the same application (most of them know what they are doing).
3) Remember that anything inductive, has significant energy when it is switched off. You have got trap diodes on your FETs?.
4) Worse though anything mechanical like a motor, also has significant mechanical energy present, which acts like a generator, when you switch off the FETs. Remember your trap circuitry needs to handle this, and you need to allow time, ramping the drive down, not just thumping it on/off, when you want to change direction.

Best Wishes

[ciudatelul]

Oh. Thank for the reply.
I modified and selected one oscilator.
The schematic i made is inspired from this kind of circuit . The motor driver(L9904) is a full bridged one. It suports PWM freq up to 30 KHz with FET`s that have trap diodes. Anyway i believe that for a small DC motor smaller freq would do to and i will change it. I still have to add some things in my code but still i get stuck . Anyway i apreciate any comments and help given it makes me understand better what i do and what i have to do .
Thanks

[Ttelmah]

I know the L9904, is rated to operate to 30KHz max, _but_ that is a chip limit, which in practice, is limited by the drive current into the FET's. Remember as FET's get bigger, their input capacitance rises. You need to sit down and look at your specific FET's, and what their input capacitances are, and what switching time this will give within the limits of the drive capability of the IC, and the total current this will require. Unless you are using very small FET's you may be surprised at how slow the switching becomes, and hence how much the power losses rise.
Don't rely on the internal diodes in the FET as 'trap diodes', _unless_ you specifically choose FET's that have proper diodes built in. _All_ FET's behave as diodes when reverse biased, _but_ the diodes this gives have much lower ratings than the FET itself, poor recovery times, and quite severe losses. Only a few very specifically designed FET's actually have separate diode structures deliberately built in (Hitachi for example do a range like this), and relying on the internal diodes, when they are not designed for the job, is a very quick way to kill your FET's....

Best Wishes

[ciudatelul]

The FET that I'm using is IRFR4104.

[Mike Walne]

Working above 20kHz is not, of itself, a problem. The great advantage is that it's ultrasonic. Hard switching power supplies at the kW level have routinely operated ultrasonically for over quarter of a century. The size reduction more than offsets the higher silicon losses. Getting ALL the right components in place is not trivial, but it can, and is done.

Mike

[Ttelmah]

Have a look at AN-944 from IR, about the drive requirements - depends on the actual voltage you are driving.
These look quite nice FET's, but the internal diodes have a fairly large forward voltage, so if significant generation takes place as you switch off the drive, there may be significant heating here. They are though properly rated for use as the inductive overshoot diodes.
Agree wholeheartedly about power drive being possible at high frequencies (last project involving this, I did a few months ago, was over 80KW, switching at 150KHz - with a PIC), but it is common to underestimate how complex good driving actually 'is', and just how important this is for efficiency/reliability....

Best Wishes

[ciudatelul]

Some updates on the code.
To release the energy stored on the motor. PWM and EN must be on 1. so i set up on intrerupt to set the PWM to full duty cycle and EN 1 for 0.5 second. After that to put EN on 0 and restore the PWM duty cycle. Direction don`t matter. The motor driver will see PWM 1 / EN 1 what will make GL1(gatelow1) and GL2 (gatelow2) enabled and release energy stored on GND.

[Ttelmah]

It sounds as if you may be talking about braking here.

Now, if this is what you want, then 'great'. However there are other alternatives. Basically your choices are:
1) Brake. Here you switch the drivers, so that either both the top drives, or both the bottom drives are 'on', which decelerates the motor quickly. This is what is needed if you are trying to control 'position' on the final system. However downside is the high loads on the motor/gearbox.
2) Freewheel. Here you cut the PWM drive, and allow some time for the motor to coast to a stop.
3) Ramped deceleration. Here you take time to slow the motor down at a rate that is smooth and comfortable. Advantage nice smooth operation.

Now both 2, & 3, are alternatives, according to what you want to do. In both cases some of the energy is dumped back into the supply rail, via the diodes in the FETs. However the amount involved is less, the more time you take to slow the motor. If you listen to things like modern washing machines, they do this from the spin, taking several seconds to let the drum slow down. Improves reliability, makes the kit sound nicer, etc. etc.. They also do the same when starting, slowly accelerating to full speed. Given the amount of energy actually 'in' the flywheel of the washing drum with water etc., in it, this is a much nicer approach.

If you want fast response, then braking is the way to go, but be aware of the electrical & mechanical shocks both fast acceleration, and deceleration can introduce.

Best Wishes

[ciudatelul]

Is a good idea to decelerate. The problem is i don't have the hardware built so i can`t figure out how it will work in real. Its a project for school and didn't had that much time to built it. Anyway I have added to change the pwm once 100 ms before breaking. I don't know if it is correct.

[Mike Walne]

Your original post

[ciudatelul]

I don't have time to finish it till i have to give the project, this is why i asked some help . Anyway i will continue to finish it afterwards. Thanks for the help.