Driving a proportional solenoid valve with no flyback diode

Thread Starter

rrpilot

Joined Jul 16, 2014
9
As the topic suggests, I would like to drive a proportional solenoid valve without using a flyback diode and looking for some industry experience/feedback.

First off, I'm using an off the shelf "smart" low side driver (ST Micro VNL5030J-E) which has the necessary avalanche rating for this to be possible without overheating and damaging the the device. For example, one of the valves that I might be driving would have specifications similar to:

Max current: 1.6A
Coil Resistance: 7.3 Ohms
Coil Inductance: 20 mH

If you look at Figure 15. Maximum demagnetization energy for the VNL5030J-E, these solenoid specifications fall below the maximum rated for the device.

I'd like to mention that I want to avoid the flyback diode for some very specific reasons and it makes life a lot easier. Either way, I'd like to investigate this option.

With that aside, I'm concerned about 2 things that I've identified during Spice simulation.

1) Following the MOSFET avalanche, there is some high frequency ringing as the MOSFET is turning off and the inductive energy has been exhausted. Is this representative in real life situation? Is there some sort of snubber circuit that I can use to minimize EMI?

2) When compared to using a flyback diode, the current profile during the off phase is significantly different. For instance, the flyback diode seems to follow a very similar discharge profile compared to the charge profile (both look like identical RC time constant type waveforms). With no diode, the discharge time is reduced significantly. I'm worried this will affect actual positioning of the valve. Has anyone successfully driven proportional valves like this, can you provide any experience?

Thanks for any insight.
 

alfacliff

Joined Dec 13, 2013
2,458
a proprotioning valve is made to open and close incrementaly, not on- off like a regular valve, you cant use a diode across it, and it should not be needed if the valve is used properly. spice dosnt know what the use is, only modeled a coil. does an output transformer use a flyback diode?
 

Thread Starter

rrpilot

Joined Jul 16, 2014
9
Maybe I'm getting my terminology mixed up. We control the position of the solenoid with the average current through it, so we are using a PWM signal with variable duty cycle to control that. The diode would traditionally be used for the back-EMF when turning off the switch.
 

MaxHeadRoom

Joined Jul 18, 2013
28,617
Also Proportional valve can also be bi-polar -ve one dir, +ve the other.
Even if it is uni-directional, the open and closure current is generally of a gradual nature not on/off as in a direct acting type.
I have never come across a BEMF diode used in a proportional valves.
If no feedback, how do you know the position point has been reached?
Max.
 

crutschow

Joined Mar 14, 2008
34,281
What you want to do makes no sense. PWM is used to give varying DC or AC currents (which requires a flyback diode with an inductor). If you don't want that, then what are you attempting with the valve?
 

Thread Starter

rrpilot

Joined Jul 16, 2014
9
What you want to do makes no sense. PWM is used to give varying DC or AC currents (which requires a flyback diode with an inductor). If you don't want that, then what are you attempting with the valve?
We are controlling the position of a solenoid valve using the duty cycle of a PWM signal which allows us to vary the average current through the solenoid. This is a very common thing to do.

My question is around removing the flyback diode and allowing the MOSFET to avalanche, giving a path for the inductor current instead of through the flyback diode. If you take a look at higher current MOSFETs you'll notice they start to specify what their avalanche capabilities are. There have been papers written around doing this and how it affects the MOSFET. My question isn't whether or not this is possible, my question is what are some of the practicalities around EMI and positioning of the valve when driving it in this fashion.
 

Thread Starter

rrpilot

Joined Jul 16, 2014
9
Here are some spice waveforms to clear things up.

One image shows the MOSFET driving an inductor load with a flyback diode and the other without. AM1 is the current through the inductor, DRAIN is the MOSFET drain terminal and VDRIVE is the MOSFET gate terminal.

On the image without the flyback diode you can see the drain terminal rise in voltage until it avalanches at around 40V and dumps all of the inductor energy to ground. Power MOSFETs have specific ratings for this.
 

Attachments

Thread Starter

rrpilot

Joined Jul 16, 2014
9
Every 3-terminal MOSFET has an internal diode.

If you've never seen a solenoid that's driven with a PWM signal that's fine, I'm trying to find someone that has. Thanks for your time.
 

crutschow

Joined Mar 14, 2008
34,281
I still see no reason for trying to do PWM without a flyback diode. :confused: It just wastes energy, creates EMI, and stresses the drive transistor. Just saying you want to do it is not sufficient reason.
 

Thread Starter

rrpilot

Joined Jul 16, 2014
9
I still see no reason for trying to do PWM without a flyback diode. :confused: It just wastes energy, creates EMI, and stresses the drive transistor. Just saying you want to do it is not sufficient reason.
I appreciate the feedback. I am trying to create a discussion around some industry experience. I have a very specific reason for wanting to do this which is irrelevant to the discussion and why I didn't try to explain it.

I've done my research on this topic and there are a number of really good papers written on it. I can't remember off the top of my head which manufacture (I think it might have been Infineon) wrote this one paper where they actually tried to test whether or not repeated avalanche events actually degraded the MOSFET and they found that it didn't. As long as the device was operated within the safe operating region it lasted just as long as if it wasn't avalanched.

If we never think outside the box and try to go against the norm, how will we innovate? A lot of research is done without understanding all of the practical applications it might provide.
 

crutschow

Joined Mar 14, 2008
34,281
So this is some proprietary device you are working on and you can't discuss what is does?

If you are concerned about spikes and EMI then you can add a series RC snubber across the driver. The optimum values are determined by the inductance of the load. Typically you design it to be a critical damped LRC tank.
 

Papabravo

Joined Feb 24, 2006
21,159
A couple of "back of the envelope" calculations.

L/R for the coil:

20e-3 / 7.3 = 2.74 ms.

This is the time it takes for the voltage across the inductor to reach 63% of the applied voltage. You don't want your PWM frequency to be near this value of 365 Hz.

The power dissipated in the coil will be the square of the current times the resistance, so for your 7.3 Ohms

(1.6A)^2 * 7.3 ≈ 18.7 Watts

Since there will be no discontinuities in the waveform you should not need the diode. By increasing the series R you can reduce the L/R time constant and operate your PWM in the range of 100-700 Hz. The drawback of this approach is that you will dissipate large amounts of power in that series resistor. The other approach is to use a high voltage supply with current sensing to regulate the PWM.

Is that what you were looking for?
 

Thread Starter

rrpilot

Joined Jul 16, 2014
9
crutschow, the details aren't necessarily proprietary but its just a small component of a larger system. It has to do with reverse polarity protection. I don't want to sway from the discussion but I'm doing reverse polarity protection for a maximum of 50 A at a system level. These flyback diodes on low-side drivers cause a lot of grief when it comes to this protection under some fault conditions.

Thanks for the note about how to size the snubber.

Papabravo, we do use current feedback for closed loop control and maybe that answers my question regarding how this will affect the positional control.

I was hoping someone might have tried this in one of their projects and had some "gotcha" they could share. Maybe an experience where they wished they hadn't done it. I see lots of research on the topic but I'm having a harder time finding projects/products that do it.
 

Thread Starter

rrpilot

Joined Jul 16, 2014
9
rrpilot, how about the 2n7000? the 3n128? mosfets without diodes.
Excerpt from Wikipedia:
It can be seen in figure 1 that the source metallization connects both the N+ and P implantations, although the operating principle of the MOSFET only requires the source to be connected to the N+ zone. However, if it were, this would result in a floating P zone between the N-doped source and drain, which is equivalent to a NPN transistor with a non-connected base. Under certain conditions (under high drain current, when the on-state drain to source voltage is in the order of some volts), this parasitic NPN transistor would be triggered, making the MOSFET uncontrollable. The connection of the P implantation to the source metallization shorts the base of the parasitic transistor to its emitter (the source of the MOSFET) and thus prevents spurious latching.

This solution, however, creates a diode between the drain (cathode) and the source (anode) of the MOSFET, making it able to block current in only one direction.
Any 3-terminal MOSFET you buy (including the 2n7000, the 3n128 is a 4-terminal device) will have an internal connection between the substrate and source terminals which automatically means there is a diode. It's unavoidable.
 
Last edited:

ronv

Joined Nov 12, 2008
3,770
If you look at Figure 15. Maximum demagnetization energy for the VNL5030J-E, these solenoid specifications fall below the maximum rated for the device.
This is true, but make sure you have a good heat sink as the clamp current is now in the FET.

1) Following the MOSFET avalanche, there is some high frequency ringing as the MOSFET is turning off and the inductive energy has been exhausted. Is this representative in real life situation? Is there some sort of snubber circuit that I can use to minimize EMI?
I suspect it is real. A little RC snubber may fix it up. I'm surprised spice modeled the avalanche is it an ST model?

2) When compared to using a flyback diode, the current profile during the off phase is significantly different. For instance, the flyback diode seems to follow a very similar discharge profile compared to the charge profile (both look like identical RC time constant type waveforms). With no diode, the discharge time is reduced significantly. I'm worried this will affect actual positioning of the valve. Has anyone successfully driven proportional valves like this, can you provide any experience?

Thanks for any insight.
That's one of the side benefits of clamping in in this fashion. The stored energy is dissipated much faster because the voltage is higher.
It's usually used where quick release of a valve is required like fuel injection.
I'm not sure how this will work for you as if you keep the frequency the same the current will fall to zero so it may creep back on you or buzz.
 
Top