I'm looking for advice in regards to using either an analog or digital (micro processor) circuit for this project. The goal is first durability, second ease/speed of assembly, and third low cost to make in low volumes (100 at a time).

The goal is a servo emulator for the automotive market. The servo to emulate is very basic, just a 12v DC motor and a 5v pot for position feedback. Inputs are +12v to either side of the motor from an external H-bridge, and +5v to the pot. The only output is 0-5v from the pot to indicate position. Slew rate needs to be about 4v/second and it needs to hold the same output value when no voltage is applied to the motor.

This is super easy to make using a micro processor, but then someone has to program it during assembly. With an analog circuit I'm thinking it would be some sort of integrator circuit using opamps, but I'm not sure how well it will hold value between inputs due to capacitor leakage? Please share your suggestions and advice!

 

I would vote for the processor version. After assembly, it will need to be tested anyway, and the programming only takes a few seconds. Then, when you find a mod is wanted, as you will, a program change is a lot easier then a hardware change needed for an analog solution.

 

I am missing something here. I don't understand why an H Bridge is in the picture? A servo motor generally has V+ and Ground or Common. There is also a signal line making for a 3 wire motor. The signal line delivers a pulse and the width of the pulse determines the servo position. Normally the pot used for position sensing is internal to the servo. When power is applied less a position signal the servo will either full CW or CCW or just sit and hum. When I think servo motors I see pretty much what is covered in this article with much larger or smaller motors.

Now if this is going to be a basic DC motor driving for example a gear box with a position sensing pot out there it becomes a different sort of animal. A servo motor I would opt for a little 8 pin micro-controller but the latter gets a little more complex. If I were doing the latter and driving a DC motor for position I would run with a circuit like this. Using drivers like the L149 gets you to about 3 amps and chips like this are made for what you seem to be looking to do. I wouldn't really look to use an H Bridge unless you are really set on it. The H Bridge will be jerky as it is FWD or REV and either On or Off where a beefy servo driver will likely give you a smoother transition. Depending on max motor current neither should be very expensive to produce. This is also just my thinking on this.

Ron

 

You might investigate a digital servo for easy control with uP.

 

I think Reloadron, that all servos have an H bridge in them.
Those "RC" type digital servos do, but include a digital to analog converter so the digital pulse train is converted to an analog voltage that is compared to the position feedback pot. They are just a sub set of the term Servo, but have become what a lot of people think of when servos are mentioned.
This is more what I think of.. See the motor pictures.
https://www.embedded.com/print/4014222
The H bridges are usually driven by a direction and PWM signal so the speed it well controllable, and the servo output signal to the controlling processor allows ramp up, run and ramp down speed be programmed in for best performance.
A lot depends on the size of the motor as to what H bridge is used. In one of my products I have a SMT 8 pin chip that is good for 40V and 3.5Amps. Pretty amazing I think, compared to what I used in the past. And no heatsink, other than the board!

 

To me there is a confusion between RC servo's and a commercial servo.
RC servo's do not need a H bridge they are just a 50hz PWM with just two pulses for either direction and are 0-180 or 0-360 with mod. You can make one with a simple 555.
It helps immensely when someone posts to mention if it is a simple RC servo they have in mind as the difference is quite large.
Also if it is a simple 0-360 deg positioning or motion control, position and velocity.
There is no mention in the OP what kind of torque is required?
Max.

 

I think Reloadron, that all servos have an H bridge in them. Those "RC" type digital servos do, but include a digital to analog converter so the digital pulse train is converted to an analog voltage that is compared to the position feedback pot.
The H bridges are usually driven by a direction and PWM signal so the speed it well controllable, and the servo output signal to the controlling processor allows ramp up, run and ramp down speed be programmed in for best performance.
A lot depends on the size of the motor as to what H bridge is used. In one of my products I have a SMT 8 pin chip that is good for 40V and 3.5Amps. Pretty amazing I think, compared to what I used in the past. And no heatsink, other than the board!

I like that. I was thinking back to circuits driving a servo on older welding machines where a motor was driven in a servo fashion to maintain an arc gap or what they called AVC (Arc Voltage Control). I guess for the most part it is how much current the motor will draw. I like the Allegro chip you linked to.

Ron

 

RC servo pulse width for position.
Max.

 

Thanks guys for the replies. It is a very basic servo as I mentioned, all of the smarts and control is external. This is just a DC motor and a pot. The servo is commercially manufactured, I need to emulate it so that the control computer does not throw an error when the servo is physically removed from the system. I'll get more details and post back.

 

Here is a snip from the schematic. I'm sorry for the poor quality, it was a poor quality schematic to begin with. As you can see, it's just a DC motor and a pot in a box. The DC motor operates on 12v, and the pot operates on 5v.

 

A possible outline for a servo control. A quad op-amp. IC is suggested, A, inverting summing amp., B invertor,
C, as comparator, D, comparator as invertor. R1 is position select pot, R2 is motor position pot. If R2 wiper is more - than virtual ground , then output A is + & = to absolute difference. The inverted signal insures that the pot differences are always + to control the PWM generator. C & D are open loop, acting as comparators to give direction control to the H bridge.

 

Eleven posts in and I'm still lost. You say you need an emulator. Does it need to act like the actual servo, physically moving something to a desired position? Does it just need to behave electrically like the servo, sort of "tricking" the control system into thinking there's a servo there?

If it's the latter, you basically just need an analog 0-5V output that can be incrementally increased or decreased based on the application of +/-12V "motor" signals?

I feel like most of the responses so far aren't addressing what you really want. Can you please clarify what you actually want?

 

If it is anything like this, doesn't look much to it, they want $50.00 on Amazon.
http://www.skutr.net/exhaust-valves-and-servo-motors-explained/
https://www.woodcraft-cfm.com/cgi-bin/commerce.cgi?search=action&category=SKSB
Max.

 

@MaxHeadRoom has it, that's pretty much exactly what I'm making. The system as-is has a servo, but for various reasons the servo will be physically removed from the system. The computer will throw an error unless the servo is replaced with an emulator so the computer thinks the servo is still there. It's easy to make a digital solution, I'm just hoping for some feedback in regards to whether there is any advantage in trying for an analog solution as opposed to digital.

 

Am I to assume you do not want the motor function, IOW use the valve wide open or removed and simulate that it has moved by way of the electronics?
Close?
Max.

 

Yes you are correct. Mechanical parts on the system will be replaced and the servo will no longer have a physical place to reside, mechanically it is out of the system entirely. Electronically the computer must think that it's still there and operating normally.

 

Yes you are correct. Mechanical parts on the system will be replaced and the servo will no longer have a physical place to reside, mechanically it is out of the system entirely. Electronically the computer must think that it's still there and operating normally.

Can't you just uncouple the servo to whatever it is actuating? For instance, just remove a sprocket attached to its shaft, or a belt? That way the servo can stay in place and fully functional, but performing no physical task.