Not getting what the OScope is showing me???

[photomankc]

Hello,

I've been working on building a drive based around the SLA7062M. I've now got a prototype board that is working. The 270oz/in motor is turning my jerryrigged test axis screw and the pointer moves to the correct positions on the scale. I'm running it with a 19V supply and around 2.0A to keep things safe.

For grins I hooked up the scope probe to the MOSFET pins on the IC and I was a little shocked at what I saw. There was a squarewave of a few microseconds, but it was the voltage the scope showed that puzzled me. So the MOSFET switches on and the voltage drops to near zero while it conducts, then it switches off and the voltage leaps to over 112V for several nano-seconds. I take it that is back EMF but I was not expecting to see that number from a 19V input supply. But anyway that's in the datasheet as expected. Now after this it settles down but not to 19V... instead to almost 40V? What??? That's where it loses me.

The scope trace insists it's a steady 40V after the back emf event is over. Then the MOSFET switches on, the signal grounds, and the cycle repeats. When the motor is disabled and chopping stops then you see just the expected 19V. A multimeter on the wire at the same time shows 19V. I'm not sure I understand what I am seeing there at that point?

Anyone know what I'm missing?

[H500]

Keep in mind that a stepper act more like a transformer than an inductor. When the fet switches off, the current flow transfers to the other phase. This induces a voltage in the coil you're driving. This voltage sums with the power supply. Hence your fet sees a voltage twice as high. If you take a look at the resistor current for the phase, you will notice that it reverses direction when the fet is off.

[wildwestpat]

First point to note is that the SLA7062M is a unipolar driver chip. Each winding has its mid point tied to the motor positive power rail. This chip uses output devices that can tolerate modest amounts of overswing voltage as the coil is turned off. As one half of the coils is being disconected the other is being connected to the supply negative and this provides magnetic flux which helps dissipate the back emf. You can think of this as an auto transformer constraining the voltage to twice the rail voltage but if your oscilloscope is fast enough you will see fine spikes on the switching edges. (This point about the auto transformer action has also been made by H500.) With the chip you are using there is a limit to the power that the output stage will take before the chip is broken. The amount of power in these spikes that is ok is stated in the data sheet but unlike some other stepper driver chips over voltage can be tolerated provided the spike power is not exceeded. For this reason the motors should be low inductance and used of course in the unipolar configuration.

If you are concerned external zener diodes could be used to clamp the chip terminas at arround 2.5 times the motor rail voltage. Bar end of the diode to the chip output and the other end to the zero volt earth point of the chip as it is the chip you are protecting wire the dodes direct to the chip pins with as short a lead as practical.

Hope this helps

[photomankc]

Ok, that makes sense. I wasn't considering any kind of transformer action raising the voltage in the other coil.

I knew there would be some large spikes because, as you say that is mentioned in the datasheet. If I calculate it correctly (114*.5*2.2*0.0000220) it's currently around 3mJ so it's well inside the specs listed. I'll work my way up towards 40V and see what I get from that. It will be interesting to see if that spike voltage is linear or not.

I've been staying away from any motor with more than 3mH anyway just to keep it somewhere close to the optimum voltage that I could run these at. I may look at the zeners though, if it would make the board more robust they would be cheap insurance.

[wildwestpat]

Hi

The scope you are using and the measuring probe will have an effect on the voltage spike you are viewing. The actual spile is probably a lot bigger unless you have access to a very fast storage scope. I don't think the supply voltage will have that much effect on the voltage spikes as these are the coils reaction to the current being switched off. The chip is a pwm current limited device so the current will be similar. The increased supply voltage will cause the current to rise faster and hence the speed of acceleration should improve with the higher psu voltage.

Hope this helps

Reagrds Pat