PWM controlled DC motor issues

[Blumpkinson]

I've been designing a circuit to control a small DC motor which is attached to a butterfly valve. The motor assembly is completely enclosed and has some sort of logic governing the full open stop point and full close stop point. There are hard stops on the butterfly valve that once the motor/valve does it's learn function, it comes to a slow stop before reaching the stops to prevent damage of the motor. So I would assume there is some type of feedback to the motors control unit when the valve reaches the hard stop in both directions.

I've been going off the assumption that the signal wire to the motor/valve enclosure was a simple ground signal. It appears that the motor/valve is designed to operate off of a PWM signal , should have checked that right from the get go I know.

I can operate the valve with no problem sending the motor/valve a ground signal, the issue that I keep running into is the motor/valve does a "relearn function" of the stops every so often. I'm pretty sure the enclosure is potted, so gaining access to the control board is probably out of the question.

Is there any reason why a constant ground signal would work fine the majority of the time and only sometimes have the valve reset itself? I am unable to replicate the problem consistently which is the part that is bothering me. I have the valve and my control circuit sharing the same ground also.

 

[goldfist]

I would think that if its losing its memory on a learn function that it may be a internal cmos type battery that is weak, although without looking at the potted board its just a guess.

Have you replaced this with another to see if the problem goes away? To me it seems as though a fix to this issue may be bypassing the correct fix which is replacement. If the replacement acts the same way then its something wrong in its application such as electrical noise etc that is causing a reset condition.

There are many ways to check the valves position to tell the motor to stop vs bind at the limit of the valve. ( Limit Switches, Hall Effect Sensors, LED Break Beam, Pressure Switch downstream that shows that the valve has been opened enough and no longer needs to be opened further etc, encoder wheel on the motor to sense stop condition or stop after 50 turns etc, clutch that will slip when motor reaches its limit and is overtorqued for a fraction of a second and slip is detected through contacts between 2 plates ... and on and on )

 

[Fish4Fun]

I assume you mean a commercially produced "butterfly valve", NOT a valve controlled by a "butterfly" ( http://store.atmel.com/PartDetail.aspx?q=p:10500051#tc:description ). Sorry, only in this forum would you need to clarify that.

At the risk of being silly, have you contacted the valve MFG? Most reputable MFGs of automation controls have an engineering dept dedicated to resolving issues involving their products. That would be where I would start.

Fish

 

[Blumpkinson]

Goldfist- I've tried using another valve of the same type and the problem still occurs.
I'm probably going to need to look at using a different more simple motor and limiting the stop point myself.

Fish4Fun - I haven't tried contacting the manufacturer yet since they are a large OEM supplier to the automotive market and could probably care less about my issue. Probably worth a shot though.

I've looked into using limit switches on a simple DC motor, the only thing i'm not quite understanding is how to go from 0deg-90deg- back to 0deg again. I know the polarity needs to be reversed and was thinking about using a DPDT Relay.

Would there be an issue with reversing the polarity of both poles of the motor at the same time?

 

[Blumpkinson]

Thanks for the input guys.

 

[Blumpkinson]

Or possibly using a stepper motor that goes from 0-90-180-270-360. I have been unable to find one with that coarse of an adjustment. I am doing everything possible to keep this as mechanical as I can.

 

[(*steve*)]

You won't find a stepper with 4 steps per rotation. You would have to cause it to step several (or even many) steps to do 90 degrees.

Another option is a servo which can move from zero to 90 and back in one movement. It may be easier to drive one to do this too. The issue you'll have is the torque required (same incidentally for a stepper) but this is just a case of determining the requirements and getting a suitable device.

 

[Blumpkinson]

Would the only way to power the servo motor be with a PWM?

I'm currently looking at linear pull type actuators with a return spring to cycle the valve. Hopefully that'll work, if not, do you guys know of any cheaper rotary actuators that would do 0-90-0 deg?

 

[Fish4Fun]

blumpkinson,

At the risk of simplifying your task, have you considered a solenoid valve? http://stores.ebay.com/WIC-Valve?_trksid=p2047675.l2563

I mean if you are replacing versus repairing, why not make it simple? Look @ the gravity feed models for low pressure applications.

Fish

 

[Blumpkinson]

Fish,

That would definitely be a viable option, however i'm really trying to keep from using anything pneumatic.

Would using a DPDT relay/ 555 timer/ and a regular DC motor work? I'm thinking about having the DPDT relay flip the polarity to the motor when I power the relay on/off. I would use the two 555 timers to kill power to the valve at exactly the correct time so that it is positioned at 0deg and 90 deg.

Or possibly just have the motor jump in intervals of 90degrees each time I tell it to open/close. That way it would just continually rotate.

 

[(*steve*)]

It's time you told us what you're actually trying to achieve.

For example, "I want to electronically control the butterfly valve on my 1/4hp four-stroke petrol mower that I am converting into a combination skateboard and toenail trimmer."

 

[Blumpkinson]

Haha fair enough. It's for a butterfly valve which I want to put in my cars exhaust, which when open will bypass the muffler.


This is my first electronics project and admittedly I've bit off a bit more then I could chew.

I'm leaning toward the linear actuator now, bc I've invested way to much of my time in what was supposed to be a simple project.

 

[(*steve*)]

So this is some kind of DIY exhaust bypass valve?

Let's ignore the detail that it might make your car less than "street legal", and the fact that you can buy them off the shelf.

The three things you need to consider are the required torque, the operating temperature, and the speed of response you require.

If the bypass valve is close to the exhaust manifold (as it would be if you were bypassing a turbo for example) then the exhaust gas temperatures would be quite high and your actuator would need to be carefully chosen and/or designed to ensure it was not affected by heat.

If speed of operation is an issue then perhaps some type of solenoid could be used (this would require the valve to be open/closed, not held part way between). Be aware that a solenoid is not going to pull very far.

A stepper motor or a servo would be other options, however the servo may be a little less happy with high temperatures and a stepper motor may be slower in operation.

Do you have the valve already?

 

[Blumpkinson]

With the valve closed the car should still be theoretically street legal and I made the choice to not buy one off the shelf bc it would be "easier" to do myself. This has turned into an enormous pain in the ass but I'm determined to see it through now.

The valve is far from the exhaust manifold and my plan is to make an insulated chamber for the motor/actuator to live. The exhaust pressure isn't very high back there but I have yet to measure it.

Speed of operation isn't really an issue as long as it's not like 15+ seconds.

I initially got a valve from an Audi system that my friend replaced but ran into the pwm issue. I'm just going to remove the motor and make a new bracket to mount the servo, actuator or motor.

 

[Rleo6965]

How about use a stepper motor / dc motor with small gear 10 teeth and engage with a bigger gear with center connected to valve fulcrum around 60 teeth. Or a screw driven by dc motor to pull a lever to open and close valve. Then you can either place mechanical limit switch or optical wheel to count holes for feedback control. The dc motor turns ratio will determine control torque and valve closing and open speed movement.

 

[Blumpkinson]

I finally got the original valve to work. Apparently the switched power source I was using in the car wasn't a clean signal. I now have the valve hooked up directly to the battery and the PCB to a clean switched 12V signal. The valve only draws 29mA, so the battery shouldn't be killed to quickly by it considering I drive the car every day.

I appreciate all of the help and different suggestions given.