Boost solenoid driver circuit

[luey02]

Hi,

Does anyone know how to built a circuit to drive boost control solenoids?

http://www.amazon.com/gp/product/B005Y294QK/ref=oh_details_o02_s00_i00

I believe it's a pin valve that opens/closes at the specified frequency of 30 Hz (unlike proportional solenoids where the pin is held open at different heights).

However, I tried to drive the solenoid with constant current pwm using a ssr but the valve will not open and close at any frequency between 10-500 Hz that I tried. It just stays open and I cannot modulate the flow. The circuit I used is attached.

Anyone has experience working with these valves? Any help is appreciated!

-Kevin

 

[CDRIVE]

No I haven't. That said, what I do know about the nature of solenoids, I don't think you'll be able to do what you're attempting.

 

[KrisBlueNZ]

Does anyone know how to built a circuit to drive boost control solenoids?

No, but I may be able to help. Until now, I didn't even know what a boost control solenoid was, and when I read that "The kit includes all the necessary fittings to plumb your wastegate", I started to wonder whether it was something I'd be best to avoid - maybe a euphemism for colostomy surgery!

I believe it's a pin valve that opens/closes at the specified frequency of 30 Hz (unlike proportional solenoids where the pin is held open at different heights).

Right, as I understand it the solenoid opens and closes at a rate of 31 Hz with a duty cycle from 10% to 90% to allow the desired amount of airflow through it. The drive circuit needs to provide 12V at several amps, with a rectangular wave at a frequency of 31 Hz with a variable duty cycle. Is that right?

However, I tried to drive the solenoid with constant current pwm using a ssr but the valve will not open and close at any frequency between 10-500 Hz that I tried. It just stays open and I cannot modulate the flow. The circuit I used is attached.

You say you used "constant current" using an "ssr" (solid state relay), but the circuit you attached is a constant voltage driver using a transistor...?

So does the solenoid change state at all? From what I've read so far, in these boost controllers, the solenoid's default (un-energised) state is open, so you seem to be saying that it never closes at all. Is that right? Have you tried connecting 12V directly across the solenoid terminals to check that it closes? It is designed for pulsed operation but this shouldn't damage it. If in doubt, connect it just very briefly. It should go clunk and switch the airflow to the other port, and when you disconnect the 12V supply it should go clunk and switch the airflow back.

Once the solenoid is responding properly to 12V DC you can look into the driver circuit. Here are some thoughts I had on the driver for you to think about later.

I don't know how these solenoids are typically driven. It's normally all internal to the ECU. I know they want 12V DC and they aren't polarised. Based on my limited experience with solenoids, I think the schematic you posted should be suitable except for the suppression diode. You see, when the transistor turns off, and the solenoid is supposed to "drop out" (return to its default, de-energised state), the inductance in the solenoid "tries" to keep the current flowing. This causes a "back-EMF" or "inductive kickback" that tries to pull the transistor collector above the positive supply rail. If the collector voltage goes too high, the transistor could be damaged, so it's normal to connect a diode as shown in your schematic, to absorb this kickback pulse and prevent damage to the transistor, but the presence of this diode significantly increases the amount of time taken for the magnetic field to collapse in the solenoid, and therefore the length of time before the solenoid drops out. This will probably be significant if the solenoid is being activated and deactivated 31 times per second. At 90% duty cycle the OFF-time is only 3.2 milliseconds. The effect will be that the solenoid will stay activated, especially at high duty cycles. The answer is to use a different type of suppression that doesn't allow the current to flow so easily, such as various combinations of resistors, capacitors, diodes and zener diodes. The suppression circuit allows the transistor's collector voltage to exceed the positive supply rail, but not so much that the transistor is damaged. This reduces the current flow and allows the magnetic field in the solenoid to collapse quickly. Also, you make sure the transistor has a suitably high voltage rating. This is nicely explained in http://relays.te.com/appnotes/app_pdfs/13c3311.pdf in relation to relays, which are a type of solenoid. In this case I suggest a diode and resistor in series, connected across the solenoid:

transistor collector ------------------- solenoid coil --------------------- +12V
. . . . . . . . . . . . . . . . .| . . . . . . . . . . . . . . . . . . . . . . . |
. . . . . . . . . . . . . . . . .---------|>|-------------------\/\/\/\/\/------
. . . . . . . . . . . . . . . . . . . . diode . . . . . . . resistor
. . . . . . . . . . . . . . . . . . MBR340 . . . . . . 22R/5W

It would be better to have a zener diode in there as well, but then you need more information about the solenoid so you can calculate power dissipation. If you have an oscilloscope that would be very useful for optimising the suppression circuit.

 

[luey02]

Thank you, KrisBlue! I think adding a resistor make a lot of sense.

I did check my pwm signal with a oscilloscope before and after connecting the solenoid valve. What I found was that the pwm become a CV signal once the solenoid is connected. I think you have explained it nicely!

To add, a CC power supply replaced the CV on the diagram...and I switched out the MOSFET with a SSR.

 

[KrisBlueNZ]

Thank you, KrisBlue! I think adding a resistor make a lot of sense.

It's a start, at least. If you read that article I linked, you'll see that the best solution is a zener in series with a diode, to clamp the spike to a specific voltage above the supply rail. In the example they clamp a 12V relay coil to 24V above VCC. But you get significant power dissipation in the zener diode. At the start of the spike, the current in the suppression circuit is equal to the current drawn by the coil while energised, which is a few amps. You will be able to see how quickly the current falls away using your scope. That, in conjunction with the repetition rate, tells you the average zener current and therefore the power dissipation.

Have you measured the solenoid current? I'm interested to know what it is.

I did check my pwm signal with a oscilloscope before and after connecting the solenoid valve. What I found was that the pwm become a CV signal once the solenoid is connected. I think you have explained it nicely!

Are you talking about the collector voltage waveform? I would still expect to see a squarewave there; it's the CURRENT in the solenoid that should turn into a triangular-looking waveform. Have you double-checked that your frequency is correct?

To add, a CC power supply replaced the CV on the diagram...and I switched out the MOSFET with a SSR.

Were you trying to get better behaviour by using a constant current to drive the solenoid? Is that when you saw the solenoid voltage being "smoothed out"? What advantages did you think you would get from using constant current drive to the solenoid?

 

[CDRIVE]

So Kris, when all is said and done do you think that you can make this solenoid operate at 500Hz? I think the electromechanics of a solenoid make it highly unlikely.

Chris

 

[KrisBlueNZ]

It operates at 31 Hz. It's designed to operate at 31 Hz. It just needs a suitable drive signal.

 

[luey02]

So Kris, when all is said and done do you think that you can make this solenoid operate at 500Hz? I think the electromechanics of a solenoid make it highly unlikely.

Chris

CDRIVE, there are proportional valves that operate > kHz. Check clippard.com for those valve

 

[luey02]

It's a start, at least. If you read that article I linked, you'll see that the best solution is a zener in series with a diode, to clamp the spike to a specific voltage above the supply rail. In the example they clamp a 12V relay coil to 24V above VCC. But you get significant power dissipation in the zener diode. At the start of the spike, the current in the suppression circuit is equal to the current drawn by the coil while energised, which is a few amps. You will be able to see how quickly the current falls away using your scope. That, in conjunction with the repetition rate, tells you the average zener current and therefore the power dissipation.

Have you measured the solenoid current? I'm interested to know what it is.

The solenoid impedance is 27 ohm, so about 450 mA at 12 VDC. I will try what you've suggested on the oscilloscope.

Are you talking about the collector voltage waveform? I would still expect to see a squarewave there; it's the CURRENT in the solenoid that should turn into a triangular-looking waveform. Have you double-checked that your frequency is correct?

I just connected probes in parrallel with the solenoid. That yield just a flat CV signal when solenoid is energized. I will check it at the collector.

Were you trying to get better behaviour by using a constant current to drive the solenoid? Is that when you saw the solenoid voltage being "smoothed out"? What advantages did you think you would get from using constant current drive to the solenoid?

I know the coil resistance changes with temperature so my first remedy to the CV response was to use a CC source.... although that did nothing but I think it's the proper way to drive a low resistance coil. correct me if I'm wrong.

 

[KrisBlueNZ]

The solenoid impedance is 27 ohm, so about 450 mA at 12 VDC. I will try what you've suggested on the oscilloscope.

I was expecting much more current than that. I thought it would be at least a couple of amps. I guess the solenoid doesn't really have to do much though.

I just connected probes in parrallel with the solenoid. That yield just a flat CV signal when solenoid is energized. I will check it at the collector.

This is with the fixed voltage drive and just a diode for back-EMF suppression? AFAIK you should have seen a squarewave across the solenoid, and a squarewave on the collector. The CURRENT waveform in the solenoid is what should become smoothed out. It might be worth double-checking with the scope.

I know the coil resistance changes with temperature so my first remedy to the CV response was to use a CC source.... although that did nothing but I think it's the proper way to drive a low resistance coil. correct me if I'm wrong.

I don't know. I hadn't heard of it, but there might be reasons for it I suppose. The coil resistance won't vary MUCH with temperature, will it? And the coil current wouldn't be too critical, would it? You would still be within tolerance to drive it at 12V (or 13.8V or whatever) over the full temperature range, wouldn't you?

 

[CDRIVE]

CDRIVE, there are proportional valves that operate > kHz. Check clippard.com for those valve

I didn't know that but the manufacturer spec on this one is 30Hz. I assume that's a max? I'm having a hard time getting my head around a solenoid traversing from fully retracted to fully extended at frequencies in that region. Do these proportional solenoid valves fully open & close or float at some point in between? That, I could get my head around.

 

[KrisBlueNZ]

I don't think there's any rotation involved. It's just a solenoid - an electromagnet that moves some component between two positions, to direct air from a "common" port to either the "normally open" or "normally closed" port. I think it responds directly to the drive frequency, so the duty cycle of the drive signal controls what proportion of the time the moving part spends in each position, which varies the airflow. That valve is specifically designed to operate at 31 Hz, it's not a maximum or a minimum. Check out the link in the original post and google "boost control solenoid" for general information.

 

[luey02]

Ok, just made the PWM flow solenoid work. I grounded the PWM and the CV supply, but just not together!

Sorry about the newbie mistake!

For some reason, I just thought the PWM generator and 12+V can be separate circuits

So to clarify, boost solenoid flow valves controls pressure by actuating at a given frequency, not by holding the valve open at different heights (though the latter is probably better for maximum flow and least flow pulsation....)

and FYI, the valve body is actually made by MAC.

http://macvalves.com/valves/3-way-valves/35-series/

 

[KrisBlueNZ]

Cool! Good job.