Current limit circuit help

[fahque99]

I have been working on a bipolar stepper driver and I have the mosfet / driver part done, but now I am working on the current limit portion.

I have built this circuit on a bread board with the PWM comming from a pic 18f2550 with a freq around 20khz and a 3uS on time. I think at this freq it shouldnt make any noise but I can still hear it in the motor. Not sure if its cause of the breadboard or from lack of filters...

So I am looking for any advice for adding filter circuits to my schematic.

I am using the 3uS to enable a NOR flip flop and also add dead time with the same 3uS.

Here is the schematic.

[vger]

Hi,

Looking at your circuit design.
1. Any time your 3us pulse is high your enable will be high and your motor current will not control the enable line.
2. Any time your motor current passes the threshold set by the pot, your enable goes low, if your 3us pulse clock line is also low.

Is this what you intended?

Steve

[fahque99]

The LM393 goes low when the current it higher then the set point.
The 3uS pulse sets the flip flop (enable on), and this stays set untill the reset sees a high signal (enable off).
This high signal comes when both lines on U1C are low (the clock is low after the pulse so its just waiting for the comparitor).
The 3uS is also used to keep the enable on for the 3uS period as a blanking period by giving U1C a high.

I think I got that right...

[neilw20]

The lower frequency you hear will be due to 'cycle skipping'
An inductor, by nature resists a change in current.
When you turn off the drive the inductor current falls, but may last much longer than you think.
So the next cycle, there is still current, and no current is commanded.
This is why we us high voltage to drive steppers so we can force the current to change in a shorter period of time, and hence run at higher frequencies.

[irving2008]

You also need to look at the decay paths which release the stored energy in the motor coils. You dont say which driver chips you are using but different products handle different decay paths, 'fast decay' offloads the energy back into the power rail, sometimes by part reverse driving the coil earlier. Slow decay just leaves it to decay through the L/R constant and the reverse diodes in the MOSFETS. These modes and the switching between them can greatly change the characteristics of the driver as will the L/R constant of the motor.

[fahque99]

Well right now I am just testing the motor with 5 volts and a constant coil on (no motion on the motor) so the low voltage may cause the slow decay. I will be running the driver with my 48 volt supply when I know it works.
I am using the IR2102 driver.

[vger]

Attached is a scope trace of the voltage across an inductor with a 14ms 10V pulse applied. There is also a resistor in series with the inductor that is about 1 1/2 times the DC resistance of the coil. The 0 V is at 2 divisions below the top of the screen. You can see the voltage jumps to the applied voltage and then decays over time to about 4 volts at which point the inductor is acting as a pure resistance. When the pulse ends, you can see the negative pulse (caused by the rapid collapse of the field in the coil) that goes to about -25V. The current at that point trough the coil is opposite of the direction of the original applied current. The larger the inductance, the larger this negative spike will be. This is the basis for creating high voltage pulses in ignition coils in cars and in stun guns.

Sound.... is created when something is moving. A speaker cone is one example. The magnetic field in a motor can actually cause the frame (and core) of the motor to deform, or bend, very slightly. This minute movement will generate sound. There is also the rotor of the motor that is free to turn and has inertia. When forced into a specific radial positon by a magnetic field (as in a stepper) there is also and opposite force placed on the frame of the motor. That is, if the rotor is accelerating clockwise, the frame will be accelerated counter clockwise. Even an AC motor operating on very clean sine wave current will generate noise (a little hum) and if you run that motor on a VFD (which uses a higher frequency PWM) you will most likely hear the "whine" generated.

Hope I didn't get too far off subject...

Steve

[irving2008]

Another example of a trace...

[Mariss Freimanis]

My comments: Ground referenced short-circuit detection is simple but it doesn't address the most common kind of drive fault; a short from a motor output to ground. That requires high-side (sensing +Vsupply current) short-circuit detection.

Mariss

[fahque99]

High side would be nice, but I really don't know how to go about conditioning the high voltage signal to something that can be read reliably without getting some expensive purpose built chips...

[fahque99]

I think I am going to try to use fast decay mode too.

[irving2008]

High side would be nice, but I really don't know how to go about conditioning the high voltage signal to something that can be read reliably without getting some expensive purpose built chips...

Same way as the low side... series resistor, a low pass filter and a comparator... but vref for the comparator relative to the high side rail and well decoupled... need to use an opamp/comparator with a high common-mode voltage rating, or run the opamp off the high side rail and a ground offset. Something like the diagram...