Circuit, to move the actuator of an HDD in a variable ANALOG! fashion.

Thread Starter

Trackmann

Joined Sep 7, 2013
10
I am experimenting with an HDD actuator arm to control its movement with current supplied by a transistor capable of sustaining a constant collector current of at least 300mA. I want to be able to vary the base current with an analog signal from a potentiometer or other analog source. The aim is to use the actuator as a variable force lever. The experiment I have conducted so far consisted of applying 5 VDC to the coil at 300mA and keeping it that way for 24hours with the coil getting barely warm. I would appreciate help with a circuit that will do that.

Sincerely,

Ralf
 

Sensacell

Joined Jun 19, 2012
3,432
You will need to add a light spring to counter the coil's force.
HDD electronics uses position feedback to control the movement, without this complex system, the arm will just whack from end to end in an uncontrollable manner.

If you add a spring, it becomes like a meter movement, the force of the spring and coil will balance.
I suspect it will jiggle around quite a bit unless you apply some damping to it.

Harder than is seems...
 

Bernard

Joined Aug 7, 2008
5,784
A simple way would to use an emitter follower to drive coil. Something like a darlington D1326 ,8A, 60 V, B 2000, which might be a little over kill. Control signal needs to about 1.5 V higher than coil supply V.
Other choices: TIP121, 122, 130.
 
Last edited:

Thread Starter

Trackmann

Joined Sep 7, 2013
10
A simple way would to use an emitter follower to drive coil. Something like a darlington D1326 ,8A, 60 V, B 2000, which might be a little over kill. Control signal needs to about 1.5 V higher than coil supply V.
Other choices: TIP121, 122, 130.
Hi Bernard,

Sorry for the delayed reply, I had to find a schematic for an emitter follower first.
Could you please look at the attached schematic and tell me if that is what you suggested?
I drew a motor because I don't know what the symbol for an HDD actuator looks like.

Sincerely,

Ralf
 

Attachments

atferrari

Joined Jan 6, 2004
4,764
Hi Bernard,

Sorry for the delayed reply, I had to find a schematic for an emitter follower first.
Could you please look at the attached schematic and tell me if that is what you suggested?
I drew a motor because I don't know what the symbol for an HDD actuator looks like.

Sincerely,

Ralf
Too big in my tablet I cannot get it of a smaller size.
 
Last edited:

Sensacell

Joined Jun 19, 2012
3,432
Here is what I would try.

This circuit should give you controllable current that varies between 0 and ~350 ma.
If you say the coil draws 300 ma with 5V on applied, the coil resistance must be around 16 ohms.
The diode should be installed across the coil, not the transistor.
You need some voltage "headroom" for the circuit to work, that's why it should run from 12 Volts.

The 10 Ohm resistor should be rated for 2 watts, and the transistor needs a heat sink.

Current sink.JPG
 

Tonyr1084

Joined Sep 24, 2015
7,852
Peanut gallery here - I don't "KNOW" the solution to this problem, but I was wondering if PWM might be something to consider. Using a high frequency and varying the modulation somewhere around the center might result in moving the meter in two directions at once. Yeah, yeah, I know - that's not possible, but my thinking is to have the BEMF counteract the EMF. At a given PWM the meter you're creating might slowly swing from one end to the other and at a slightly lower PWM the meter might swing back again.

Maybe not. Like I said, I don't know the solution to this problem. Just poking my nose in where it probably doesn't belong.
 

Thread Starter

Trackmann

Joined Sep 7, 2013
10
Peanut gallery here - I don't "KNOW" the solution to this problem, but I was wondering if PWM might be something to consider. Using a high frequency and varying the modulation somewhere around the center might result in moving the meter in two directions at once. Yeah, yeah, I know - that's not possible, but my thinking is to have the BEMF counteract the EMF. At a given PWM the meter you're creating might slowly swing from one end to the other and at a slightly lower PWM the meter might swing back again.

Maybe not. Like I said, I don't know the solution to this problem. Just poking my nose in where it probably doesn't belong.
Hi Sensacell,

I will build your circuit and report back.

Hi Tonyr1084,

PWM is actually the way an HDD actuator works in a factory HDD. However, the actuator always operates at full torque. The reason I want to operate it as stated in my original post is that, I want to control the torque of the actuator. In my application the actuator just applies a variable force within its rotational range, which is approximately 30 degrees.

Sincerely,

Ralf
 

Papabravo

Joined Feb 24, 2006
21,157
It is unlikely you can accomplish what you want. Most HDD spindle motors are BLDC motors. You have to provide the commutation signals in the correct sequence to produce rotational torque. The feedback will normally come from three or more hall effect sensors that tell you which pole to energize next. I'm afraid that PWM won't do much for you if that is the type of motor you have.

Sorry: You did say the actuator.

For the actuator you cannot run it open loop. It depends on feedback from the disk itself to hold it's position over a track or to seek another track that is some distance away. As the head moves it picks up track addresses on the fly so that the control algorithm can slow the arm down to prevent overshoot. Again PWM is hardly the method used for head positioning.
 
Last edited:

Tonyr1084

Joined Sep 24, 2015
7,852
Papa: The OP wants to swing it slowly from one position to the next. He's not looking to move it to a specific position. He's taking something engineered for a specific purpose and applying a whole new application. I'm not sure of the effect he's after, but from his description he wants it to swing (from left to right for instance) slowly, like a second hand of a clock, maybe a little faster, maybe not. I haven't read any specifics on that which I can recall.

I agree, I think it will peg one way or the other with any kind of current. A spring mechanism might be required to resist full deflection and pull it back to its original position. Depending on how much power is applied, the spring will allow deflection of a specific distance. Greater spring resistance would take greater electrical energy. I just assumed (with no prior knowledge) that PWM might be sufficient to vary that level of energy to suit the desires of the OP.
 

Papabravo

Joined Feb 24, 2006
21,157
PWM is going to make it move back and forth. The higher the frequency for a fixed duty cycle, the more it will try to move in one direction. Three steps forward, one step back for example. If that is what you want then go for it. At very high or low duty cycles you will peg it to the stops. You might even get it to chatter against the stop.
 

Thread Starter

Trackmann

Joined Sep 7, 2013
10
PWM is going to make it move back and forth. The higher the frequency for a fixed duty cycle, the more it will try to move in one direction. Three steps forward, one step back for example. If that is what you want then go for it. At very high or low duty cycles you will peg it to the stops. You might even get it to chatter against the stop.
Hi Bernard and Sensacell,

I have built the circuit suggested by Sensacell. It worked just fine.
I then experimented and I found that the force I needed from the actuator arm is much less than I anticipated, in the range of one to three grams. I then changed the supply voltage to five volts, experimented with the resistor values and changed the transistor to a 2N3903 and to my surprise it still worked. Using the potentiometer, I can now vary the coil current from 20mA to 80mA which produces the actuator forces that I need. Neither the transistor nor the coil get warm. Sensacell, could you please look at the attached circuit and see if it needs fine tuning of the resistance values?

Actually, there is more to the final circuit that I need. I did not mention that in the beginning, for fear that no one would respond if my request for help was too complicated. No dishonesty was intended.
I would like to amplify the output of an Allegro Micro linear Hall effect device, using an inverting single supply differential op amp and apply that to the base of the above transistor circuit.
The end result would be that, a magnet sliding over the hall effect device would control the current through the HDD actuator coil.
Any help toward that end would be very much appreciated.

Sincerely,

Ralf
 

Attachments

Top