How to modify a 12v PWM fan control to drive a solenoid?

Discussion in 'The Projects Forum' started by defcon5, Mar 13, 2009.

Thread Status:
Not open for further replies.
  1. defcon5

    Thread Starter New Member

    Feb 17, 2009
    I'd like to apologize in advance for the length of this question - I tried to get all the details into one post, and it got very long.

    Hello everyone - this is my first post here and honestly, I'm a bit apprehensive. As I've searched through similar threads trying to find my answer, I come across folks claiming to be "noobs", who then launch into a litany of tech speak that quickly has my head ringing. Among my circle of friends, I'm "the smart tech guy", and these so-called noobs, make me look like a pre-schooler. I'm stunned at the vast array of topics that people on this board are experts on - truly impressive.

    But enough of that - here's my issue: I have a solenoid in my car that I would like to control manually via a PWM controller. This solenoid controls hydraulic pressure to a clutch that controls the rear wheel drive in my AWD transmission. Under normal operation, the car has a 70%-80% front wheel drive bias, but if you really get on the throttle, the transmission computer (TCM) will send some of the power to the rear wheels to prevent wheelspin. This all works well in normal everyday driving, but the car sees some amateur competition on weekends where I ask it to work it's hardest on gravel, mud, snow and all manner of slippery surface. When pushed this hard, the AWD system exhibits some troublesome lag. The front wheels can sometimes spin for as much as 1.5 seconds before the rear lockup is called for - usually very abruptly. This hurts performance in two ways - first, by the time the rear wheels join in, I'm well past the place where I needed them, and second, when they lock up abruptly, it upsets the balance of the car.

    Here's what I know about the solenoid/TCM:
    The solenoid is driven by a 12vDC PWM signal operating at a fixed 50hz
    The solenoid's impedence is 10 - 17Ω according to the factory service manual
    I have not been able to determine solenoid current, but believe it to be around 700mA at 95% duty cycle.

    0% duty cycle = full rear wheel lockup (50/50 front-rear torque split)
    100% duty cycle = no rear wheel lockup (front wheel drive)
    The TCM will vary the duty cycle to the solenoid anywhere from 5% to 95% depending on throttle position and vehicle speed.

    I've already put a switch and a 17Ω "dummy load" in line to fool the TCM - basically killing power to the solenoid and locking the car at a 50/50 torque split. This works, but in an effort to improve performance, and learn some new tricks along the way, I'd like to build a PWM controller that will allow me to dial in as much rear wheel traction as necessary.

    I stumbled across a schematic and build plans for a 12vDC PWM controller aimed at quieting down PC fans at this site.
    Using the stripboard layout, and some of his suggested alternate parts to change frequencies and run higher loads, I built a working 60hz PWM controller. It controls a PC fan quite nicely, but when I tried to hook it up to the car, I ran into a problem. (apologies - this is where my utter lack of electronics training is going to reveal itself) In cpemma's schematic, the MOSFET is is controlling the fan via the path to ground. I understand this is perfectly reasonable, but I don't have (easy) physical access to the solenoid's ground wire. It's wrapped up in an aluminum transfer case and surrounded by 12qts of transmission fluid.
    What I think needs to happen, is to use the MOSFET to control the current TO the solenoid instead. However, looking at schematics that do switch the 12v line to the solenoid, like this 555 version, I see that the FET symbol is different. Is this the difference between N-channel, and P-channel MOSFETS?
    I'm reading through the fantastic electronics classes at the top of this forum, and understand the concept of field effect transistors, but implemetation and actual operation are still a bit fuzzy.

    The MOSFET I chose was the IRF630B which is an N-channel version. If I swap to a P-channel MOSFET and reroute the signal to the positive side of the solenoid, will this fix my problem?
    Could it possibly be this simple? I doubt it.
    Have I drastically underestimated the complexity of what I'm attempting? Very likely.
    School me - I'll listen!
  2. thingmaker3

    Retired Moderator

    May 16, 2005

    Not as much of a "pre-schooler" as you thought you were, huh?;)

    You also have the option of keeping the N-channel MOSFET and using a technique called "bootstrapping." You can do this with single-chip MOSFET drivers.
  3. thatoneguy

    AAC Fanatic!

    Feb 19, 2009

    You asked and answered your question in your first post. You just didn't know the term you were looking for, like "High Side Driver", although you explained it well. :)

    It isn't as simple as switching in a P channel instead of the N channel, but it is possible, there are a few topics on this with various schematics in the forum search.

    Don't be afraid of topics with microcontrollers in subject, many PWM circuits are in those areas. The MOSFET Driver "parts" of those circuits are the same, only you are creating the PWM with a 555 instead of a uC. You do not need to switch to those get your project going!
  4. defcon5

    Thread Starter New Member

    Feb 17, 2009
    thingmaker3 and thatoneguy - thanks for the replies (and positive reinforcement).
    Who knew I'd actually be on the right track?? :)

    So...High Side Driver and Bootstrapping eh? I've got some more googling and reading to do! :)

    The existence of a High Side Driver leads me to believe that there must be a Low Side Driver method as well, and if I'm following you, High Side is when you're switching on the positive line and Low Side would be switching the negative, yes?

    So, would it be oversimplifying too much to think of MOSFETS as acting in the same way as a discrete relay? That is, using a lower voltage signal as a means to switch a higher voltage? In effect offering delicate IC's and microcontrollers some protection? I understand there is much more to them than that, but for my particular purpose, that's their role, correct?

    Along those same lines, could it be reasoned that the P-channel is most effectively used (in my case) as a high-side driver, and that the N-channel would be a more appropriate low-side driver? Again, I'm breaking this down into 'black & white', so tell me if I've gone too far astray.

    Well, I'm off to search, read up, and try to get this straightened out in my head.
    Thanks again for the pointers and proper terminology - I'll try my forum searches again.
  5. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    That is pretty much it, except the high side and low side terms come from H Bridge motor drivers.

    The high side goes between the power source to motor.

    The low side goes between the motor and ground.

    The voltage potential is usually very large between Source and Drain in those two placements.

    Or think of one as disconnecting/connecting the positive wire (high side)
    the other as disconnecting/connecting the negative wire (low side)

    Due to their properties, for an N channel MOSFET to be fully on, the gate needs to be around 5V (roughly, this varies widely!) more positive than the Drain. This is easy on the low side, since the Drain is ground. On the high side, it isn't very easy due to the voltage drop of the load, which has the source voltage 'floating', rather than ground potential.

    A mechanical relay is too slow for effective PWM, but many Solid State Relays are fast enough, and could go on the high side.

    On solid state relays: Yes, MOSFETS are like relays, using a low voltage/current to control a much higher voltage and/or current (to hundreds of amps from a 5V processor!)
    Last edited: Mar 14, 2009
  6. defcon5

    Thread Starter New Member

    Feb 17, 2009
    Ok, between reading up on FET theory and application (ow, my head hurts),
    and asking some electronics savvy friends and co-workers, I've got myself completely confused. It was my understanding that changing my current configuration from a "low side" driver, to a high side driver was not as simple as either:
    A) Using my existing N-channel MOSFET to switch the high side without modification. This was later reinforced by thingmaker3 in his reply.
    B) Simply using a P-channel MOSFET to switch the high side

    I was initially told that while yes, using a p-channel is the proper concept, it's not as easy as just plugging it in to the high side of my existing circuit without modification.
    Then, just yesterday, I was told by our resident "transistor guy", that I'd be fine moving my N-channel up to the high side and it would work - no mention of bootstrapping or modification.

    I know you guys are a smart bunch - think you can straighten me out?

    I've cobbled together some modified circuits to visualize what I think my options are. One disclaimer - I simply chopped up the diagram of the original circuit in photoshop (sorry - graphic designer by trade) to reflect my changes. When I built this project, I opted for a MOSFET (IRF630B) in place of the transistor, so you'll have to pretend that the transistor symbol is actually a MOSFET. Oh and the fan is actually a 17ohm solenoid:rolleyes:

    My basic question is - Is it really this simple? Can one of these possibly fix my issue without having to rework the entire circuit? Any explanation of the 'whys' and 'why nots' is always appreciated. Feel free to get up on a soapbox if you want - I'll read it.

    Here's the original circuit - Low side/N channel (thanks to cpemma)

    Option 1 - High side/P channel (apologies to cpemma)

    Option 2 - High side/N channel (ditto)

    Hopefully those poorly modified diagrams don't confuse the issue further.
    Thanks in advance - you guys have already helped a lot.
  7. defcon5

    Thread Starter New Member

    Feb 17, 2009
    Late weekend bump!
    Any chance I got any of the the above diagrams right?
  8. blinky


    May 25, 2009
    Hello folks,

    I am trying to get hold of a copy of the circuit diagram and more importantly the strip board layout for the circuit. I have made a couple of these for various things but the cpemma website does not seem to be available anymore. I was wondering if anybody has a copy of the circuit and layout they could post here or alternativly a similer circuit based on a 555 timer that they have a strip board layout for?
    This would be a great help as I am no good at creating the strip board layouts!!!

    Last edited: May 25, 2009
  9. Bernard


    Aug 7, 2008
    Option 1 looks OK, I'd replace 1kΩ gate resistor with 330Ω, adn add 3.3kΩ from gate to +12V, so that gate does'nt float.Output may be inverted??
    Option 2 suffers from low drive V.
  10. defcon5

    Thread Starter New Member

    Feb 17, 2009
    Thanks Bernard. I'll have some extra time in about a month, so I'll give your suggestions a try.

    Blinky - you've got PM.
  11. nippet

    New Member

    Aug 10, 2011
    Apologies to dig up an old thread.
    Although I am trying to find out the completed circuit design that defcon5 used (if that ever happened).

    I have the exact same issue he had.
    Need to drive a 12V solenoid via PWM, preferrably adjustable with a potentiometer from 0-100.
    If anyone was wondering the vehicle in question is a Subaru with a 4EAT automatic transmission.

    The only exception being that (his) pre 2004 model operated:
    0% duty cycle = full rear wheel lockup (50/50 front-rear torque split)
    100% duty cycle = no rear wheel lockup (front wheel drive)

    My model 2004 and newer is the opposite:
    0% duty cycle = no rear wheel lockup (front wheel drive)
    100% duty cycle = full rear wheel lockup (50/50 front-rear torque split)

    Subaru did this to force the vehicle into a safer limp home mode if a component were to fail in the circuit. Otherwise the vehicle would be wheel hopping around corners being locked into 50/50 split.

    I have some additional details on the circuit operation I am trying to replicate.
    Today I pulled as much information as I could with my Fluke DVOM and ammeter.

    Actual Measurements:
    12V Circuit @ fixed 50 Hz
    Solenoid Resistance - 12.6 Ohms
    The Transmission Module commands the solenoid differently depending on what gear is selected. Not necessarily important for designing this circuit, just interesting.
    V, A & PWM on time %
    P - 0V 0A 0%
    R - 3.6V 0.25A 42%
    N - 0V 0A 0%
    D - 3.6V 0.25A 42%
    3 - 3.6V 0.25A 42%
    2 - 3.6V 0.25A 42%
    1 - 5.3V 0.38A 55%

    I do not have an oscilloscope so I would expect the voltages to actually be 12V, since my multimeter is just averaging the pulses.

    Full lock up would draw 0.95A @ 100%

    Any help steering me in the right direction would be helpful, component sourcing, layout ect...
    I also only have access to the positive side of the solenoid, the ground is buried inside the transmission.

    Are there any off the shelf units that would work for this? ? ?

  12. SgtWookie


    Jul 17, 2007
    Sorry, we can't discuss this topic.

    "Automotive modifications" were placed on the "Restricted Topics" list about 7 months ago. Please see paragraph 6 of the Terms of Service. Links to the ToS is at the bottom of every forum page.
  13. nippet

    New Member

    Aug 10, 2011
    Hmm, well as unfortunate as that is potential liability is understandable.
    Last edited: Aug 10, 2011
  14. defcon5

    Thread Starter New Member

    Feb 17, 2009
    Apologies to keep this going after SgtWookie's moderation, but I can't seem to PM nippet directly.

    Nippet - I'll assume you've followed a paper trail from to arrive here. Feel free to email me (defcon5atgmaildotcom), or PM me on RS25.

    I didn't realize that this forum's rules had changed recently. I fully respect the reasoning behind the change, and feel quite lucky to have gotten so much help before the topic became restricted. Humble thanks to all.
    I've got plenty of goofy projects in mind, so I'll be back when my kids return some of free time....say, 18 years from now?
  15. SgtWookie


    Jul 17, 2007
    I'm not a moderator; I just post a lot. I felt obliged to mention the rules though.
  16. bertus


    Apr 5, 2008

    Thanks SgtWookie for pointing this thread.

    I am closing this thread as it violates AAC policy and/or safety issues.

    6. Restricted topics. The following topics are regularly raised however are considered “off-topic” at all times and will results in Your thread being closed without question:

    • Any kind of over-unity devices and systems
    • Automotive modifications
    • Devices designed to electrocute or shock another person
    • LEDs to mains
    • Phone jammers
    • Rail guns and high-energy projectile devices
    • Transformer-less power supplies
    This comes from our Tos:
    Terms of Service
    There will be enough sites where automotive questions can be discussed :
    Member selected automotive forums

Thread Status:
Not open for further replies.