bidirectional dc motor control

[MedievalMan]

I'm trying to do bidirectional dc motor current (torque) control using resistors as the current sensor.

I'm using a bidirectional pwm/bridge IC combo (see

http://www.solutionscubed.com),

and I've tried the sense resistor placed between the bottom of the bridge and ground. It works well for PWM based current control, using a simple PI controller, in either direction (by switching the sign of current based on which leg is being switched) This method doesn't work very well or at all when a switch between +'ve and -'ve current demand occurs, because of the inductive nature of the motor.

A single resistor placed in the middle of the bridge with the dc motor also does not work for the opposite direction, because one side of the current signal is references to ground; when the other leg is switched, the current signal has a floating reference.

One of the best documents I've found explaining this problem
"LOAD CURRENT SENSING IN SWITCH-MODE BRIDGE MOTOR DRIVING CIRCUITS" with possible solutions is:

http://www.st.com/stonline/books/pdf/docs/1668.pdf

I'm going to try one of the dual resistor configurations with opamps when I get the chance.

I'm wondering if anyone has any other suggestions to resolve this direction switching problem, perhaps even with a software solution (I'm using a Texas Instruments DSP for the controller).

Thanks in advance,

 

[Skogsgurra]

Those appnotes are about as complete as you can get.

For true motor current measurement, try LEM transducers. They are fast, insulated and not overly expensive.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[MedievalMan]

How would I deteremine what values of the capacitors in this configuration? The appnote says they are used to smooth the transition between one direction and the other.

 

[blacksea]

Yes, using motor armature current mearsurement directly is the better way. The resistor with special differential amp AD629 (Analog Devices) maybe suggested for low voltage application, or Hall based new sensor ACS750 (Allegro).
Both the above solutions are significantly cheaper vs LEM sensors.

 

[MedievalMan]

Thanks for the help guys. I'm getting a quote on the LEM sensors, although the ACS750 seems to provide the same electric isolation function as the LEM (to me, unless I'm missing something.)


That AD639 uses the same opamp high-common mode rejection difference amplifier configuration suggested in the AppNote fig 4. That would certainly cut down on discrete components... which I'm going to try to hook up tomorrow, anyway (gives me a good hint at what resistor values to use.)

 

[MedievalMan]

I tried hooking up the opamp/dual resistor configuration of Fig. 4, but realized I couldn't because the 2 lower transistors on the bridge are tied together in hardware.

So, I hooked up the circuit in Fig 5, using 50 kohm resistors and a 741 op amp. All I get at the output voltage (which is supposed to be proportional to current) is a constant voltage, regardless of the current flowing through the motor.
I'm wondering if this has to do with the fact that the sense resistors are in the middle of the bridge (in series with the motor).

I've verified all the connections... is there anything else I might be missing?

 

[MedievalMan]

I haven't had a chance to go back and double check that circuit configuration, but I think I may have found a better solution.

The LD 18200 is an Hbridge with built in current sensing: it works by composing the bridge transistors with many cells, each with equal share of the load current, and pulling out one of the cells: it's current is proportional to the total load current.