I am wondering if I am overthinking this one. I would like to protect the logic device from excess voltage. Would the Zener diode I have shown suffice?

Thanks
(this site and forum are just awesome)

-Jim

 

The zener needs to be in parallel (not series) with the IC, so that it allows an alternate path to ground for the excess voltage and current. And it needs to be placed in the circuit "backwards" so that it normally conducts nothing, until it breaks down under high voltage. You need a zener at the right reference voltage, and one that can handle the anticipated current from the motor. On that latter point, I'm not sure you'll easily find what you want. It may be a lot of current for a zener.

 

I think I follow what you are saying. Perhaps I am incorrectly assuming the Zener will "breakdown pass" ONLY 24V , when it will actually pass anything in excess of 24V. . . .

Maybe what I need is a DC-DC regulator rated for 1-2A with a wide range input and 24Vdc output.

 

It would help a great deal if you would describe this "logic circuit" a bit better.

If this circuit is in a vehicle, you might experience transient spikes far exceeding 100v. Even in 12v automotive systems, transients up to 60v occur during load dumps, like turning off headlamps/brake lamps.

The "24v battery" will likely measure around 25.6v when fully charged, and a typical charge voltage would be around 28v, so with that alone you are at the bleeding edge of your "logic circuit" limits.

There are devices called TVS diodes (TVS = Transient Voltage Suppression), some of these can withstand surge currents (spikes) in the hundreds of Ampere range, far more capable than your typical Zener diodes.

Zeners need to have their maximum current limited, or they will pop like popcorn.

 

You are not over-thinking; this is a common problem. Back EMF as well as motor noise can wreak havoc with a digital circuit, especially if there is a microcontroller involved.

You want to get a Schottky diode (much faster than a standard diode) and place it directly across the terminals of the motor, not on the logic circuit. The cathode goes to the (+) side of the motor, the anode to the (-) side (this "steers" the back EMF into the positive voltage leg). You want the leads on the diode to be as short as possible (soldered directly across the terminals if possible). This keeps the inductance down in the leads (a good thing). A capacitor in parallel with the diode cannot hurt either. Make sure you rate the capacitor voltage higher than the motor running voltage and the current about 2x higher. For instance a SBR12A45SD1-T (available from Mouser for $1.22) is a 45V, 12A (200A peak) chioce.

EDIT:
I just looked at your circuit diagram. Something like the MBR1100G might be more appropriate; 100V, 1A (50A peak). $0.39 from Mouser.


Mike T.

 

It would help a great deal if you would describe this "logic circuit" a bit better.

I guess I should first say that the "logic circuit" is actually "IO Power". Let me explain further:

The motor is actually a self-contained 24V industrial servo (PM BLDC) motor with integrated drive. The power to the unit is fused at 15A for the main power, 2A for logic power, and a separate 1A for IO Power.

Here are the specs on the main and logic power:
"24-48 Volts DC nominal, 20 -60 Volts continuous operating range.
Under voltage trip 19V, Over voltage trip 85V"

Here are the specs for the IO Power:
"24V nominal, 30V max, 12V min"

The unit will be running in "torque mode" so the drive will be essentially acting as a current controller. The motor will normally be acting against a load that is very close the torque it is putting out. In other words, the motor will be sitting at 0-speed (which is OK for this device). If the force it is acting against increases, the motor will be "overdriven"; if the force it is acting against decreases, the motor will overcome the force.

I guess you could picture an elevator with one person in it in a shaft and the motor at the other end of the elevator is in torque mode. If the person gets out, the elevator begins to rise; if another person enters, the elevator begins to descend.

The motor Ke works out to 24Vdc/850rpm. When the motor is overcomming the force, the drive will automatically clamp the output speed at 850rpm. However, if the load overcomes the motor (over-running), the drive won't trip on over-voltage until 85Vdc.

There are cases when the motor may be overdriven to 1000-1500 rpm. I am trying to protect the IO Power from seeing that higher voltage.

You are not over-thinking; this is a common problem. Back EMF as well as motor noise can wreak havoc with a digital circuit, especially if there is a microcontroller involved.

You want to get a Schottky diode (much faster than a standard diode) and place it directly across the terminals of the motor, not on the logic circuit. The cathode goes to the (+) side of the motor, the anode to the (-) side (this "steers" the back EMF into the positive voltage leg). You want the leads on the diode to be as short as possible (soldered directly across the terminals if possible). This keeps the inductance down in the leads (a good thing). A capacitor in parallel with the diode cannot hurt either. Make sure you rate the capacitor voltage higher than the motor running voltage and the current about 2x higher. For instance a SBR12A45SD1-T (available from Mouser for $1.22) is a 45V, 12A (200A peak) chioce.

Mike T.

Hmmm . . . based on my description above, would the Schottky diode still be an option (albeit a slightly large one)?

 

I guess I'm a bit thrown now... if the motor has an integrated drive you never actually "see" the motor; it's drive electronics do. You are only dealing with a control signal which "shouldn't" have any of these problems. Secondly, from my experience the I/O cards (I use these often from Galil and Emerson) are usually opto-isolated. In this case the control side is protected. Even if it's not protected I don't think you need to worry about this.

You might have been right to start with; you are over-thinking this...

Mike T.

 

I guess I'm a bit thrown now... if the motor has an integrated drive you never actually "see" the motor; it's drive electronics do. You are only dealing with a control signal which "shouldn't" have any of these problems. Secondly, from my experience the I/O cards (I use these often from Galil and Emerson) are usually opto-isolated. In this case the control side is protected. Even if it's not protected I don't think you need to worry about this.

You might have been right to start with; you are over-thinking this...

Mike T.

The "bus" will rise when the motor is overdriven. Based on the specs (requoted below), the IO Power needs to be kept within a lower range than the motor can tolerate.

Here are the specs on the main and logic power:
"24-48 Volts DC nominal, 20 -60 Volts continuous operating range.
Under voltage trip 19V, Over voltage trip 85V"

Here are the specs for the IO Power:
"24V nominal, 30V max, 12V min"

At this point, I am guessing a DC/DC converter/regulator might be the best solution. I was trying to get "crafty" and come up with simpler solution than something like this:

(10-30Vdc in and 24vdc out - 1A rating)
http://rd.phoenixcon.com/products/interface/powersupplies/files/2866284.pdf

 

Hmmm - this might work, too:

I could put a big diode on the main motor power leg. I found this one:
DPG30C300HB
Peak Reverse Voltage: 300V
Forward Cont Current: 15A
Max Surge Current: 150A
http://www.mouser.com/ProductDetail...=sGAEpiMZZMtldOIVIVqa3R9CyC09MSR8Sl787zu/MG4=

 

All that diode is doing (the way drawn) is dropping the voltage by one diode drop (about 1V).

What are the pieces of equipment. Can you post links for them?

 

But if the bus rises above the battery voltage than the diode will prevent the battery from seeing that higher voltage, right?

 

Yes. I'd still use a Schottky; they are much faster than a standard diode and have a lower forward voltage.

Mike T.

 

Taking this one step further - I wonder if I could put a 30V Zener and some dynamic breaking resistors in series between bus power and power common. The Zener would pass voltage over 30V and the DBR would absorb the power?

I know this would probably be a pretty beefy Zener (or set of Zener's in parallel).

Unlike normal DBR being used to control the load, I want to absorb the extra voltage, but let the load generate all the voltage it wants to (remember that the drive is acting in torque mode (as a current controller)

Hmmm. . . .

 

Why not clamp the motor leads to the power supply? That should keep it in range.

 

Why not clamp the motor leads to the power supply? That should keep it in range.

I'm sorry, I don't follow you.

The motor is integrated with the drive. If the motor is over-running above 850 rpm while the drive is trying to maintain constant current/torque, the bus will rise above 24V.

The diode already shown (and it will be a schottky thanks to other advice here) prevents "greater than battery voltage" being seen by the battery IO Power (which can't take more than 30Vdc).

The latest idea is to deal with that excess voltage if it gets too high for the drive/motor. I thought a large 30V Zener in series with a large resistor bank - positioned between motor power and ground, would give that voltage somewhere to go.

Help me understand what you are saying. I do appreciate the help.

-Jim

 

OK, then the clamping is covered.

Look at TVS diodes. They are like turbocharged Zener diodes.

 

I'm searching out the TVS devices and seem to be having trouble finding one with decent continuous current ratings.

Let's assume my motor/drive is in torque mode at 10A. If the bus voltage goes over 30V and I have a TVS with a break-down around 30V, I am assuming that I'll be shunting 300W or more to the Dynamic Braking Resistors.

Should I plan on a a bank of TVS in parallel, or is there a better way to do this? In my automation world we'd use a shunt regulator or drive with built in DB Transistor. I just don't have a shunt regulator for this "low" voltage (~30Vdc).

 

I also think that for the schottky blocking diode - for mounting, heatsink, and current rating purposes - I might be better off using a bridge rectifier design that I can screw down to a backpanel and use push-on tabs to connect. I'd connect V+ to one of the AC connection points and then feed V+ on the BR to the motor.

Still wondering if I can find a TVS or Zener to use as a shunt regulator. . . .

 

You can get Schottky's in a TO220 package. The numbers I gave were for devices to connect directly across the motor (when we were talking about that). Have a look at Mouser, they have a bunch (not pushing Mouser, Digikey, Allied, Newark are all good)...

Mike T.

 

I found some nice "stud mount" styles of diodes with very high continuous current ratings. I'll probably use one of those.

Thanks again for the help.