Linear Actuator

[electroRF]

Hi,
I'd like to order this leaner actuator - it is meant for pressing a machine's Power Button - i.e. Printer, PC, Etc.

**broken link removed**

I'd like to use the BBB (BeagleBone Black) Digital I/O pins to control the Actuator's position, by a PWM signal.

I believe that 12 mm stroke length will be enough for pressing the button.

1. There're several options - which option is required for the application of pressing a button usng a PWM signal - I or R?
(I'm not sure what is the difference between R and I options.

B - Basic 2-wire open-loop interface, no position feedback, control, or limit switching. Positive voltage extends, negative retracts.

S - 2-wire open-loop interface (like B option) with limit switching at stroke endpoints.

P - Simple analog position feedback signal, no on-board controller.

I - Integrated controller with Industrial and RC servo interfaces (see L12 Controller Options section). Not available with 10mm stroke length configurations.

R - RC Linear Servo. Not available with 10mm stroke or 12 volts.

2. It is said in the datasheet: "Stalling the actuator under power for short periods of time (several seconds) will not damage the actuator.".
What does it mean please?
I'd wanna have the Actuator pressing the Button for more 5 seconds - i.e. keep it in the same position for more than 5 seconds.
it is OK?

 

[ozgur84]

I believe, you can crash and break any button which you want to press, if you use this one. h have you seen how powerful it is?

 

[jpanhalt]

As I read the specification, you will have at least 40 N (about equal to the pressure of 4Kg). Do you really need that much? With a standard servo arm, you can use a "servo saver" to spring load its action. It is a torsion spring, and you can set that to exert whatever maximum pressure you want. If you are committed to using these linear actuators, I suggest that you put a small compression spring on the end. That will allow the actuator to over-ride the end position and limit the force you are putting on the key.

As for your other questions, the servo or linear actuator you show is one wire for signal. There is no feedback. You set the throw by the pulse width. As for stalling, the spring I recommend will prevent that.

John

 

[electroRF]

Hi guys,
Thank you very much!

John,
You're absolutely correct.
I tired looking less powerful linear actuators but did not find.
Could you please direct me to one that would suite this application of pressing a PC's button?

Thanks a lot.

 

[ozgur84]

2. It is said in the datasheet: "Stalling the actuator under power for short periods of time (several seconds) will not damage the actuator.".
What does it mean please?
I'd wanna have the Actuator pressing the Button for more 5 seconds - i.e. keep it in the same position for more than 5 seconds.
it is OK?

if you stall the actuator (if you want it to keep the same position for a while), it will not damage the actuator. The reason of a possible damage is the high switching magnetic fields in the coils of the actuator. Something like in the stepper motor coils. However, it can somehow create a kick-back current ripple to your I/O pin if you connect directly. That's why, you had better isolate the actuator from the I/O pin of your microcontroller via a mosfet or a 4 diode kick-back circuit for isolation. I would go for the mosfet solution.

 

[electroRF]

Oh thank you very much.
They do not mention it in the Datasheet.

It is said in the datasheet that:
"When the actuator is powered up, it will repeatedly scan leads 1, 2, 4 for an input signal that is valid under any of the four supported interface modes. When a valid signal is detected, the actuator will self-configure to the corresponding interface mode, and all other interface modes and input leads are disabled until the actuator is next powered on."

However, what happens if I power up the Actuator before I set a PWM signal in the MicroController? (i.e. the BBB)

I won't be able to use it in PWM mode if I power up the Actuator before the BBB?

 

[ozgur84]

Oh thank you very much.
They do not mention it in the Datasheet.

It is said in the datasheet that:
"When the actuator is powered up, it will repeatedly scan leads 1, 2, 4 for an input signal that is valid under any of the four supported interface modes. When a valid signal is detected, the actuator will self-configure to the corresponding interface mode, and all other interface modes and input leads are disabled until the actuator is next powered on."

However, what happens if I power up the Actuator before I set a PWM signal in the MicroController? (i.e. the BBB)

I won't be able to use it in PWM mode?

I also checked the datasheet but it does not tell much about the internal circuit. That's why, I could not understand that whether the coil ends are directly connected to the connection leads (to which you supply pwm) of the actuator or there is an internal circuitry which does the job. They mentioned that there is an option like buying the actuator with an internal controller, but if you pick an actuator without the internal controller option, the issue which I mentioned before, can be a problem.

If you connect the actuator via a mosfet array, you can still use the pwm mode but the power to the pins (let's say current for a better understanding) will be supplied by the external power supply instead of the microcontroller. Mosfet will follow the pwm signal and switch on an off the channel. Of course if you choose the mosfet as a fast switching logic level one.

I also checked the current-force plot in the datasheet and the current values stay under the limits of I/O pins, if just pressing a button is concerned. However, imagine the button is stuck or something went wrong and the actuator tried drawing higher currents from the microcontroller. Because of all these reasons, I would use the mosfets for the channels to be on the safe side and keep everything safe...

 

[KeepItSimpleStupid]

OK, my turn:

I and R

I gives your 3 interfaces: 4-20, 0-5 and RC, R only gives you RC.

RC may be cheaper and if you used a 8 channel RC interface like I showed earlier you would not need D/A converters. 4-20 is very common in industrial environments. the reason it's used is there is no length issues or ground loop issues and it's easy to provide isolation.

===
Stalling the actuator is like being between a rock and a hard place. A carefully selected spring on the end of the actuator with current protection (possibly fuse( would work just fine, but it does mean the spring has to be selected.

Remember the force on a spring is kx where x is the extension or compression of the spring. So, take a spring and your finger and determine the deflection. So, let' say that you could move the button when the spring deflects 1/4 of it's length. Now let's suppose that full travel of the button is reached at 3/4 of the spring length and now suppose that the force of the actuator is not exceeded at 100% of the spring length.

Your finger acts as a feedback device. You know when to back off the pressure, when the button doesn't move. Making the actuator have a spring-tipped would basically accomplish that. The actuator movement would have to be longer than the button movement.

So, as the actuator moved, the end spring would start to compress and eventually the button would move. Now, you can move the push button distance. Then there should be some spring left over, so the actuator won't stall. e.g. It might take 125% of the push button distance to move the button 100%. Stall might be reached at 150% (no more spring to compress).

With the tip of the actuator being a spring of the right length and k, you would be able to push the actuator without stalling it. A stall should activate some sort of protection and it might be a fuse. The reason fuses are bad is because you can replace them with a bigger one.

 

[jpanhalt]

Th datasheet describes one mode of operation to be just like a model servo. An RC servo can hold a position forever, so long as voltage and signal are applied. It will simply idle at the position determined by the pulse width. The current drawn will be the idle current, which is low. Now, if you apply pressure to change that position, the current increases. At some point, if you apply more pressure than the servo can resist, the position will change, current will go very high, and the servo will be said to be "stalled." That is the condition you want to avoid for long periods.

If you work in the RC mode and the actuator presses a key until there is a physical stop, then the servo/actuator is likely to be stalled. You can adjust the end points quite closely using digital control, but there will still be a little variation. That is why I suggested a spring that would allow the actuator to apply pressure without depending on a physical stop. A medium to light weight foam rubber would also accomplish the same function.

I did not read the datasheet that closely, but I suspect the other modes have some mechanism for maintaining a position without stalling the actuator.

John

 

[KeepItSimpleStupid]

I didn't look at the datasheet closely enough, but the center COULD be the rest position. So, check the specs.

 

[KeepItSimpleStupid]

Power up;

probably the first to send a non-zero signal.

Key point: How is rest defined? 0 to 5 = 0 to 100%?

RC mode is it +- relative to center position?

 

[MrAl]

Hi,

Wayyyy too complicated to merely press a PC power button. Buy yourself a 12v electric door lock actuator and power it with 12v to push the button. Cost about 5 dollars or so.
You dont need feedback either, just make sure the mounting is flexible so it gives a little or else use a heavy compression spring tip, or simply limit the current to the motor (with a resistor) which controls the force.

 

[electroRF]

Hi guys,
I thank you very much!

You helped me out very much!

Regarding the "Stalling"

My intention was to set the 12mm-stroke actuator close to the Power Button, with a soft tip, so only 6mm would do to press the button.
i.e., I'll set 50% duty-cycle PWM wave, in PWM Mode.
Is it consider stalling?

There's one thing I'm not sure about.
What happen if I connect the 5 (red) Motor V+ (+6 Vdc for 6 V models, +12 Vdc for 12 V models) and 6 (black) G Ground wires to constant Power Supply, and I'll apply the PWM wave only after the Actuator is powered-up - would the actuator be in PWM mode?

As it says:
"When the actuator is powered up, it will repeatedly scan leads 1, 2, 4 for an input signal that is valid under any of the four supported interface modes. When a valid signal is detected, the actuator will self-configure to the corresponding interface mode, and all other interface modes and input leads are disabled until the actuator is next powered on."

i.e. - Must I need to connect the 12V Power Supply to the Actuator via a Relay?

 

[ozgur84]

Hi,

Wayyyy too complicated to merely press a PC power button. Buy yourself a 12v electric door lock actuator and power it with 12v to push the button. Cost about 5 dollars or so.
You dont need feedback either, just make sure the mounting is flexible so it gives a little or else use a heavy compression spring tip, or simply limit the current to the motor (with a resistor) which controls the force.

I agree with that if you can find such a door lock actuator. then I would use one 555 timer with two buttons, first to do the 5sec push and another button for the short push by changing the resistor with a button, bob's your uncle! no microcontroller no worries about burning the I/O pins of microcontroller and so on...

555 timers can handle relatively high voltages like 5V to until 18V(I am not sure the max value but I think it is around that value) instead of 5V and microcontroller stuff, so you have more flexibility...

 

[KeepItSimpleStupid]

read the specs. I finally did. The PWM mode is different than the RC mode.
PWM 0-100% controls 0-100% extension. It's a control signal.

RC mode uses the RC servo pulse widths with the range corresponding to 0 to 100% extension. If I remember, the standard RC servo uses a +- rotation scheme.

50% PWM is a control signal for position. If it hit an immovable object, yes, it would be considered a stall.

Lots of times the power switch on a PC isn't ON/OFF. OFF requires you to push for like 5 seconds or so. That's where the spring comes into play.

The datasheet suggests that it would hold it's position if the actuator is powered off with the control signal present.

 

[KeepItSimpleStupid]

Their actuator datasheet doens't quite match up with the controller datasheet: **broken link removed**

Two things I immediately saw is 3.3 V and 5 V tolerant. USB is an option.
Limit adjusts are possible and it includes stall protection.

The company has a Pet Peeve of mine: Any interface to the outside world on any product that one interfaces too should have an equivalent circuit in the instructions.
In this case, what does the PWM input look like electrically?

 

[electroRF]

Hi Guys,
Thanks a lot again!!

KISS
I did read the **broken link removed**of the linear actuator- several times

As you suggested, I should connect a spring to the tip of the Actuator, right?
Could you recommend on such a spring that is used for such application?

I'd like to make the Order - I summarized your suggestions in the following scheme

Would it work?

I did not connect the 12V Power Supply through a relay - Does it need to in case the BBB will create a PWM signal AFTER the Actuator was powered up?

 

[jpanhalt]

Ordering springs can be very difficult, as there are so many types, and the minimum quantity may be large. Most important, we don't know the dimensions of your device and keyboard setup. Here is an example of a small selection of springs: https://www.mcmaster.com/#compression-springs/=pyj8pk

For your prototype, I suggest a short piece of small rubber tubing mounted so the axis of the tubing is perpendicular to the long axis of the actuator ram. A piece of medium density foam rubber would also work. Something like the stuff used to insulate pipes or "backing" rod for caulking should work.

John

 

[MrAl]

I agree with that if you can find such a door lock actuator. then I would use one 555 timer with two buttons, first to do the 5sec push and another button for the short push by changing the resistor with a button, bob's your uncle! no microcontroller no worries about burning the I/O pins of microcontroller and so on...

555 timers can handle relatively high voltages like 5V to until 18V(I am not sure the max value but I think it is around that value) instead of 5V and microcontroller stuff, so you have more flexibility...

Hi,

Yeah that sounds pretty good. And these actuators dont need to turn off when they hit something, just stall then turn off, so yes a timed 'on' period would probably work just fine. And current limiting is easy.

These devices can be found on Amazon for example. Probably a lot of other places too that sells automotive parts, but dont pay too much, around 5 dollars or maybe a little more should do it. They come with some mounting parts too.

 

[electroRF]

Thank you folks.

jphan

Is it OK to connect the 12V Supply directly to the Actuator, and not via a relay?