Removing Voltage Spikes from an Inductive load

Thread Starter

navidelec

Joined Jul 12, 2014
29
I want to drive a servovalve, the servovalve is a dc type with linear movement which it's position controlled due to a feedback, as it knows from dc valves the current is fixed and I reverse voltage across the valve to change the way of movement, I mean going from one end to the other end as a linear movement, so current is dc and voltage is being reversed to control the linear movement.
Would you mind tell me how can I remove SPIKES?

11-1-2014 8-56-55 PM.jpg 11-1-2014 8-58-20 PM.jpg
 
Last edited:

Papabravo

Joined Feb 24, 2006
15,731
Show us the schematic. Would it be too much trouble to label the voltage and time axes and provide us with an expanded view of either edge where you see the spikes. Can you give us the impedance of the actuator and the mechanical properties of the load?
 

Thread Starter

navidelec

Joined Jul 12, 2014
29
Show us the schematic. Would it be too much trouble to label the voltage and time axes and provide us with an expanded view of either edge where you see the spikes. Can you give us the impedance of the actuator and the mechanical properties of the load?
The resistance is about 100ohm and I'm not sure about the inductance maybe 100mH.
Just the way of removing the spikes is the case for me, Thanks.
 

Papabravo

Joined Feb 24, 2006
15,731
It doesn't work that way. You give us the information we request so we can give an authoritative answer. Anything we do without that is just guesswork. I guess we'll see if anybody else steps up. I'm goin' QRT.
 

MaxHeadRoom

Joined Jul 18, 2013
22,831
Are you controlling it with a PWM signal? many are controlled in a DC linear fashion.
Some of the Bosch hydraulic proportional valves use a 16804 IC, the app note for it may help.
Max.
 

Dr.killjoy

Joined Apr 28, 2013
1,196
I could be wrong but shouldn't there be a diode on the inductance load to help prevent voltage spiking since the inductor can't change its voltaged instantly when turn on and off ???
 

Papabravo

Joined Feb 24, 2006
15,731
I could be wrong but shouldn't there be a diode on the inductance load to help prevent voltage spiking since the inductor can't change its voltaged instantly when turn on and off ???
Not in all cases, that's why we need the schematic.
BTW it is current that cannot change instantaneously in an inductor, not voltage.
 

Thread Starter

navidelec

Joined Jul 12, 2014
29
and dear friends
here is the Multisim 11.0 simulation file of the schematic
just change the .pdf extension to .ms11
Thanks
 

Attachments

Thread Starter

navidelec

Joined Jul 12, 2014
29
Is there any different type of Power amplifier, I mean opa544 which I used that could help to drive inductive loads and removes spikes, I chose the opa544 for it's high voltage and high current output. Is there any better alternative?
 

Thread Starter

navidelec

Joined Jul 12, 2014
29
At which point in your circuit are you seeing the spikes?
Can you post a link to the valve spec/datasheet?
I just test it with an impedance load, 100ohm and 100 mH.
and saw the spikes which are above.
Up to now I didn't connect the practical circuit to the valve, because I afraid about damaging the valve with spikes.
I saw the spikes at the output to the valve.
 

Papabravo

Joined Feb 24, 2006
15,731
Looking at the output circuit, I'm wondering if you are seeing phantom spikes as a result of uncompensated probes. Most oscilloscopes have a calibration signal on the scope for the purposes of compensating the probes. There should be a small setscrew adjustment on the probe. You adjust the probe to get a nice looking square wave. What you are adjusting is a small trimmer capacitor across the 1 Mohm impedance of the probe. Too high a value and you see phantom spikes and too low you see rounded edges on the waveform. If you could expand the time scale it would help.

1st Edit:
Why did you not follow the recommended power supply bypassing arangement?
2nd Edit
Why did you not employ the output stage compensation mentioned in Figure 4. 1Ohm in series with 0.01 uf?
3rd Edit
Why do the LM324 opamps have undefined power supplies?
 
Last edited:

Alec_t

Joined Sep 17, 2013
11,999
Simulation shows that the spikes:
1) are due entirely to the dither voltage and valve inductance, and
2) do not exceed either supply rail voltage of the driver amp.

Why do you have the dither voltage?
Provided the valve voltage rating is less than the amp supply voltage there should be no harm to the valve IMO.
The spikes can be suppressed by use of a snubber across the valve. 100 Ohms in series with 10uF(non-polarised) more or less eliminates the spikes. A lower cap value is somewhat less effective.
 

MaxHeadRoom

Joined Jul 18, 2013
22,831
Is there any different type of Power amplifier, I mean opa544 which I used that could help to drive inductive loads and removes spikes, I chose the opa544 for it's high voltage and high current output. Is there any better alternative?
I implemented a proportional valve a while ago that had ±10vdc analogue control and simply used a Nat. LM759 to control it, it has worked trouble free for a decade now.
Max.
 

Thread Starter

navidelec

Joined Jul 12, 2014
29
Simulation shows that the spikes:
1) are due entirely to the dither voltage and valve inductance, and
2) do not exceed either supply rail voltage of the driver amp.

Why do you have the dither voltage?
Provided the valve voltage rating is less than the amp supply voltage there should be no harm to the valve IMO.
The spikes can be suppressed by use of a snubber across the valve. 100 Ohms in series with 10uF(non-polarised) more or less eliminates the spikes. A lower cap value is somewhat less effective.
So helpful thanks
 

Thread Starter

navidelec

Joined Jul 12, 2014
29
I implemented a proportional valve a while ago that had ±10vdc analogue control and simply used a Nat. LM759 to control it, it has worked trouble free for a decade now.
Max.
I want to have 1 ampere at valve terminal, but LM759's output current is 325mA, Do you have any better choice than opa544 that I chosen.
 
Top