Fade DC Motor Without PWM

I’m wondering if it’s possible to fade a 12v 1.5 amp motor in and out when you connect it to electricity.

I know how to do this with PWM, but I want to use a LM317 to drop the voltage/rpm of the motor, and a mosfet to turn it on and off. I’d like to have it ramp up the speed when it’s on, and slow down when you turn it off, instead of instant on/off.

The reason I want to use a 317 is because I don’t want to hear any PWM noise from the motor, and when you go to a lower duty cycle, there is a greater loss of torque with PWM opposed to analog voltage.

I’m looking for a simple, non code driven way to ramp up and down the motor.

Worst case scenario I’ll PWM up and down from the LM317 analog voltage, but if there’s a simple way to do it with electronics, then I can also use digital pins to control the motors. I’m connecting about 20 of these big motors.

Dropping the voltage not only reduces the speed but also the torque. With a PWM the torque is not reduced nearly as much. The motor needs to be under some load in order for PWM speed control to be effective.

xxmamakinxx:
The reason I want to use a 317 is because I don't want to hear any PWM noise from the motor

Then use a PWM frequency that's outside our hearing range.

That's why computer CPU fans use 25kHz PWM for speed control.

Leo..

Grumpy_Mike:
Dropping the voltage not only reduces the speed but also the torque. With a PWM the torque is not reduced nearly as much. The motor needs to be under some load in order for PWM speed control to be effective.

Yeah, there will be some load on the motor, and I've been in situations where I've tested a lower voltage to slow the motor on a bench top power supply and successfully carried the load, then moved to PWM with the same voltage reading and failed to carry it.

xxmamakinxx:
then moved to PWM with the same voltage reading and failed to carry it.

That does not make sense. PWM is always full voltage (or no voltage).
Can't measure PWM with e.g. a DMM.
Leo..

Wawa:
That does not make sense. PWM is always full voltage (or no voltage).
Can't measure PWM with e.g. a DMM.
Leo..

Yeah, don't know what to say. I'm not measuring PWM with the meter, I'm measuring volts, and when it's reading the same voltage as would be on a bench top, the PWM does not carry the same amount of torque. I've just chocked it up to loosing torque on the 0v from the duty cycle.

You cannot use a DMM measuring volts while using PWM. PWM is either 5V or 0V, nothing in between. As for torque, you need to ensure the power supply is capable of supplying the current, and all the leads are sized correctly.

Your missing the point of my question, but you can measure the overall voltage from PWM with a multimeter.

All of the wiring is fine, and I’m on a 15 amp power supply.

I have a question (actually a couple)

  1. how quick is you ramp from stopped to your running speed.
  2. same question on slowing down.

Are they 1/2 second 5 seconds, a minute......

What are your load dynamics?

i.e. 1) A fan motor will decrease torque requirements with the drop in speed.
2) A pump feeding a regulator will require near constant torque even as speed drops.

John

xxmamakinxx:
...but you can measure the overall voltage from PWM with a multimeter.

Digital to Analog Converter using PWM - YouTube

An external RC circuit without load will average perfectly.
But do you know what's inside the DMM you're using?

An LED used without CL resistor puts the guy instantly in the "do not trust him" corner.

PWM motor control is the best thing since sliced bread.
Yes, you can use a linear regulator to control a motor.
Not sure if a single LM317 is up for that task (1.5Amp).
It could run stinking hot, depending on output voltage and motor load.
And will need a heatsink the size of a fist.
Leo..

JohnRob:
I have a question (actually a couple)

  1. how quick is you ramp from stopped to your running speed.
  2. same question on slowing down.

Are they 1/2 second 5 seconds, a minute......

What are your load dynamics?

i.e. 1) A fan motor will decrease torque requirements with the drop in speed.
2) A pump feeding a regulator will require near constant torque even as speed drops.

John

For ramp up/down, I'm shooting for something between .5 - 1.5 seconds.

While it's spinning up and down, I'll avoid putting a load on it, however it is a large screw drive motor, so any momentum it could pick up in operation is stopped by the gears when I cut power. Also, I'm driving each motor between 60 and 190 rpm, and there isn't enough forward energy to keep the worm gear spinning, so it stops almost immediately.

Wawa:
An external RC circuit without load will average perfectly.
But do you know what's inside the DMM you're using?

An LED used without CL resistor puts the guy instantly in the "do not trust him" corner.

PWM motor control is the best thing since sliced bread.
Yes, you can use a linear regulator to control a motor.
Not sure if a single LM317 is up for that task (1.5Amp).
It could run stinking hot, depending on output voltage and motor load.
And will need a heatsink the size of a fist.
Leo..

I'm using a pretty standard multi meter, nothing special. That's just been my experience in the past, pwm has given me less torque than a regulated power supply. I hear you on the heat issues, the 317 was the first thing that came to mind.

xxmamakinxx:
Yeah, there will be some load on the motor, and I've been in situations where I've tested a lower voltage to slow the motor on a bench top power supply and successfully carried the load, then moved to PWM with the same voltage reading and failed to carry it.

What decay mode though? The voltage measured by a multimeter is only meaningful in synchronous
rectification mode.

Use PWM, use 16 or 20kHz.

If you need fine control at low speeds, you need an encoder and control feedback loop.

That’s just been my experience in the past, pwm has given me less torque than a regulated power supply.

Well either they have changed the laws of Physics without telling me or you are misinterpreting what you see.

As others have said, the LM317 is not likely a good choice. Assuming your power supply is 12V the LM317 will max out at 9V and only drop to ~1.3 Volts minimum.

You could make what you are looking for with a couple of op-amps and a P channel FET or PNP transistor.

Regarding the PWM. In a normal DC motor operation a PWM drive will work. However you mentioned you are driving these motors at 60 to 190 RPM. Normal brush type DC motors do not work that well at those RPM’s. Is there something special about your motors?

John

JohnRob:
As others have said, the LM317 is not likely a good choice. Assuming your power supply is 12V the LM317 will max out at 9V and only drop to ~1.3 Volts minimum.

You could make what you are looking for with a couple of op-amps and a P channel FET or PNP transistor.

Regarding the PWM. In a normal DC motor operation a PWM drive will work. However you mentioned you are driving these motors at 60 to 190 RPM. Normal brush type DC motors do not work that well at those RPM's. Is there something special about your motors?

John

Thanks for the suggestion of op-amps and FETs/transistors.

I've worked a lot of PWM with motors, and am competent of driving motors with PWM. Even if I end up using it for this project, I'm still very interested in a hardware solution that could fade in and out. Being able to do this without an Arduino is still useful.

Do you have any specific references for setting this up with op-amp/FETs/transistors?

(for anyone following this thread, can we stop the PWM talk, that's not what I'm asking for help with)

BTW, this is the motor. It's a simple worm gear drive.

Hi,

The first stage could be an integrator. Pretty simple with an opamp, resistor(s) and capacitor. This circuit will generate a linear ramp each time the input is switched. Google can show you many examples.

The 2nd stage could be like an opamp driven emitter follower (or source follower for a MosFet). But it will need an extra stage so the you can saturate the output device for low "on" voltage across the power switching device (else it will dissipate excessive heat). I don't have a reference right now.

Do you have a preference whether the high side of the motor is switched or the low side?

I think the motor you have is likely spinning at 3000 to 5000 RPM, with an output gear drive lowering it to your stated 60 - 190 RPM.

Often it is important to know the right words to search on. In this case what you want to put into googel is
soft start circuit
then look at the images tab.
This is one example

OMG. Soft start circuit. Thank you. That's exactly what I was looking for.

If you want to control the speed of a dc motor accurately you need to build a 'velodyne' ( as it was called in my days of learning control theory).

You need a means of measuring the motor speed instantly ( eg a tachogenerator) .

Then compare that with the wanted speed and apply a correction to the motor drive voltage to get it to the wanted speed. A PID algorithm - whether in software or hardware - will achieve this. The response time is limited by the physical constraints of your motor's torque , the moment of intertia of the system, and the applied load.

Once that is working, you just adjust the wanted speed according to any pattern you want.

Smoothed PWM may be useful here.

Allan

I've tried the PID thing and with PWM, and I just haven't gotten the algorithm correct. The last time I tried it was with an encoder at the end of my motor shaft. Not sure how to setup the relationship between encoder readings, velocity, and PWM output using code. I know there's an elegant solution there, but I'm not quite sure how to code for it.