Help With TL494 Motor Speed Controller?

Thread Starter

kh602

Joined Jun 28, 2011
25
Hi there,

I am hoping someone could shed some light on my project i have going on for a motor controller.

So far i have tested a few example circuits using the tl494 pwm chip and tweaked the circuit slightly.

The problem i am having is motor stability in terms of constant speed. Its not as stable as i would like and keeps dipping or rev'ing a little when i set the speed using a pot on the dead time control.

I have tested with a different frequencies ranging from 200Hz-20Khz to see if this would help the stuttering of the motor but still no joy.

I would imagine i need some form of compensation or something but i am not sure how to apply this.

See schematic for what i have so far.

I do also have UC3637 & UC3678 motor controllers from TI but these not a great deal of info on how to use them either.
 
Last edited:

shortbus

Joined Sep 30, 2009
7,678
Is your circuit on a PCB or solderless breadboard? Some breadboards will get an intermittent open when things are moved. If your pot is connected with wires and you pick it up to adjust it, maybe it is causing an intermittent open?
 

Thread Starter

kh602

Joined Jun 28, 2011
25
It use to be on breadboard but now i moved it over on to veroboard and its the same. works but not as accurate as i would like.

When using an adjustable regulator is dead smooth with loads of losses, using this circuit sounds like the motors ticking over maybe missing a few poles from the pwm signal controlling the motor or something making it not as stable.

Using a bourns trim pot for adjusting the pwm. I am not entirely sure if this is the correct way for adjusting the width/speed of the motor using the deadtime control or using a pot to the comparator input
 

Thread Starter

kh602

Joined Jun 28, 2011
25
i tried that circuit without current limit/transistor part and its still ramping the motor up and down like its drifting from the setpoint and back again.

How does that current limit part for the circuit work and is it needed for a stable output as i am confused as to why it goes to pin 13?
 
Last edited:

Alec_t

Joined Sep 17, 2013
11,277
How stable is the power supply for your circuit? Heavy motor current could be pulling the voltage down and it might take time to ramp back up. Or the IC could be shutting down periodically if it's overheating.
The current limit works by developing a voltage across a current-sense resistor and applying the voltage to one of the comparators in the IC.
 

Thread Starter

kh602

Joined Jun 28, 2011
25
I have been using a 12V 9aH battery fully charged 13-14v. I do want to use some sort of smps for the battery or something to keep it at a fixed voltage.

current draw is 1.8amp with load on the motor using an LDO regulator
and i have it down to 1.6-1.7A with the TL494 its just not stable/smooth enough.

I don't have any heat issues with this circuit. Mosfet is barely warm. room temperature i would say. mosfet driver and everything else is cool and the other thing thats a little warm is the 470R emitter resistor sinking the current. Thats about it
 

Thread Starter

kh602

Joined Jun 28, 2011
25
Yeah I have a scope. I have been checking the signals
All the time. The pwm signal is pretty clean and square for sharp turn on and off
To the gate. However I do seem to have a large spike on the rising edge (corner) to the MOSFET gate.
 

BC107C

Joined Apr 3, 2011
14
The circuit doesn't have any kind of feedback.Powering the motor with constant voltage( or duty cycle) is not going to guarantee a constant speed.Can you get a tacho output from the shaft ? If not, with good results, you can try stabilizing the back emf.
 

RichardO

Joined May 4, 2013
2,271
I don't think it is related to your problem but you need to connect the input of the unused driver in the TC427 to either power or ground. Leaving a MOS input floating can damage the device.
 

Thread Starter

kh602

Joined Jun 28, 2011
25
Not to worry the driver input is grounded. I made sure of it even though
I haven't put it up.

I don't have a tacho on my brushed DC motor however I do on my brushless motor, how would I hook that up?

How would I stabilise back emf other than using fast switching shottky diodes?
 

Thread Starter

kh602

Joined Jun 28, 2011
25
I have just attached some scope shots of whats happening. Green/CH1 is the gate of the mosfet and yellow/CH2 is the drain of the mosfet.

You can hear the motor hunting up and down slightly yet its variable revving sound and not something thats a repetitive sequence.

I have set the speed i want using the pot tide to the feedback pin rather than using the deadtime control.
Freq at 20Khz Approx

Supply = 12.01V dropped to 11.95V/11.94V fluctuating between the two.
Current draw from supply = 1.71A/1.74A Fluctuating between the two.

IRLI3705N Gate Voltage = 8.96V/8.97V Fluctuating between the two.
Drain Voltage = 2.93V Stable

Motor supply voltage = 11.88V Stable
Motor- to Drain current draw = 2.18A/2.22A Fluctuating between the two.
 

Attachments

Thread Starter

kh602

Joined Jun 28, 2011
25
The battery supplies the power to the motor and tl494 chip. I have tried all sorts such as bench top power supply, 12v 8A mains smps and the battery. All pretty much doing the same thing and stuttering the motor slightly.

Does that mean there's interference and have to isolate both parts for the circuit and motor?

I changed my circuit so i use the feedback control to vary the speed rather than using the deadtime control. Both still do the same thing pretty much.
 
Last edited:

Alec_t

Joined Sep 17, 2013
11,277
Since you have no feedback from the motor to stabilise the speed, my guess is that motor friction variations (e.g. as the bearings warm up) result in the speed and current changes.
 

Thread Starter

kh602

Joined Jun 28, 2011
25
Since you have no feedback from the motor to stabilise the speed, my guess is that motor friction variations (e.g. as the bearings warm up) result in the speed and current changes.
How do I go about putting feedback into this system? If I current sense the motor and feed
It back to one of the comparators, what signal do you compare it with? I kind of know what I need to do but
Not sure how to go about it.

Cheers for the reply
 

Alec_t

Joined Sep 17, 2013
11,277
I don't see how controlling the current will stabilise the speed, any more than providing a given PWM duty cycle does at present. Won't you need a shaft angle sensor or back-emf monitor?
 

Thread Starter

kh602

Joined Jun 28, 2011
25
I don't see how controlling the current will stabilise the speed, any more than providing a given PWM duty cycle does at present. Won't you need a shaft angle sensor or back-emf monitor?
I need something to at least feed back to the motor to compare against to stabilise the speed which is what i thought current or voltage sense did to close the loop and stabilise the output that bit better.

I dont have any room to attach anything on to the shaft as there are con-rods attached as the motor is a motor pump assembly. I am after a stable air flow which at the moment its not stable enough and is hunting up and down slightly
 

BC107C

Joined Apr 3, 2011
14
There are few methods that do not imply tacho signals known to work great with brushed DC motors.
They have been implemented in few ICs, and the controll is linear, ie some power is inherently lost in the process.All of them are maintaining the back emf as stable as they can.
The chips to look for are TDA 1059B,TDA7274 and for sure other manufacturers have released their own flavor.

Attaching a PWM into it ( current mirrors) should not be a huge exercise, but a pleasant one.

Another variant was released by BB, the application sheet is attached.The reason the RPM is stable derives from the same ecuations as described in 1059 for example, just that the BB schematic can be much much easier implemented using a PWM driver and power element.
 

Attachments

Top