Interfacing for dummies

[dreamproject]

Hi all,

I have gratefully accepted the idea of using a (not-so-hard-to-get)multi-turn pot , instead of a (hard-to-get) encoder for LM629 based DC motor servo control . But if you would be kind enough to look up its Data Sheets , you will find that it is designed for interface to an encoder and not for pot - output . What mods / circuits r necessary for this coversion ( pot --> encoder pulses) . What seems even more puzzling is that we need quadrature o/p's .How can a pot's o/p be made to simulate that ?. Can a stepper motor driven by the Dc motor produce these pulses ???

Thanks

 

[Nigel Goodwin]

How can a pot's o/p be made to simulate that ?.

Certainly not easily.

Can a stepper motor driven by the Dc motor produce these pulses ???

It wouldn't help if it could, because it wouldn't give absolute positioning.

The servo based on a wiper motor I mentioned earlier has the right kind of servo IC, one used in RC servos - as I see it that's all you need, a huge great RC type servo?.

 

[bmcculla]

I think your bets bet is to use a micorcontroller with built in A to D and PWM to run a PID controller. It's not that hard and will probably be much easier to get working than trying to get the LM629 to work with a pot.

I'm not sure how well the RC servo controller would work. They are designed to work with a big gear box so they don't have to worry about oscillation.

 

[panic mode]

dreamproject,

can you give us some details on this? what are you trying to do exactly?
if this is still "fly-by-wire" steering system for a car and if driver has
view of the road, he can adjust the small errors without need for encoder
simply by turning joystick or pot one way or another. It's another story
if you need to control the car positon blindly (using computerized system
and track position of the vechicle without GPS etc.).

If all you need is the first case (just manual control with very low effort
like for disabled person for example) then larger hobby servo could do
the job. Hobby servos use very simple control and thus have limited functionality which in this case could be suffient, at least as proof of concept. To generate pulses to control the servo, you can use simple 555 osscilator and potentiometer (open loop control).

Another common limitation is stroke which for hobby servos is reduced to less than one revolution. In my experience most cars have steering column designed for ca 3 revolutions end-to-end but you don't have to nececarily use all of it except maybe for parking.

Industrial servos usually use quadrature encoders. This allows to keep
track of position but has downside that each axis has to be homed upon
powerup to establish reference point. With absolute encoders this needs to be done only once but absolute encoders are more complicated and more expencive. One of the common workarounds is to use battery to
keep counter logic of the axis quadrature encoder powered. This acts
like absolute system which needs to be homed only when battery is changed and only if this is done while system is not powered.

Another thing to consider with hobby servos is speed (it's usually
very low - 0.25 sec to make 60deg turn for example) and lack of torque
control (can be very handy sometimes, like if you are driving up
against curb).

 

[dreamproject]

drive by wire-that's exactly what it is

Dear panic,

A drive-by-wire system is exactly what I'm trying to do. I want to converge all driving operations (steering , accn,gear-shifts,brakes,etc.) into a single joystick. The motivation----> as a part of my master's degree project.

I have already designed a system which recieves joystick input to control a stepper . But :

- this system is way too slow

- torque is a problem if this motor is to be geared for speed.

Mr . Nigel and others suggested a DC motor servo system . As you rightly pointed , an open loop is fine for this . But my primary need for an encoder is to make the LM629 work for DC motor PID servo control.

Added benefits include :

- closed loop control including holding torque.

- high rpm motor can be geared for torque.

- computer Interface (virtual steering display.)


My plan of action for other features like accn,brakes is to use steppers with a rack and pinion arrangement, to provide linear movement at the brake pedals.

Thanks to every one for their input

 

[tansis]

Why not use an ultrasonic or infra-red rangefinder
to ascertain the position / angle of the vehicles road wheels.