PWM solenoid driver

Discussion in 'The Projects Forum' started by dmaxben, Sep 22, 2008.

  1. dmaxben

    Thread Starter Member

    Oct 27, 2006
    11
    0
    hi everyone...need some advice that will probably be pretty simple to answer, but Im new at all of this stuff. :)

    What I am looking to do is build a manual torque converter lockup switch for the transmission in my truck. Its PWM, so I cant just ground the solenoid control wire and get lockup.

    Its pulsed at 100hz and the solenoid draws about 10 amps @ 12vdc. It constantly gets 12vdc and then the TCM (transmission control module) has a control/ground wire that it pulses on to lock up the torque converter. 0% duty cycle is unlocked, and max duty cycle that the solenoid sees is 97%. Low side driver.

    However the duty cycle must be ramped up in a "soft start" fashion. I tried just grounding the control wire (effectively 100% DC I guess?) for grins and it made a bang and locked up VERY hard. Ive datalogged the TCM and depending on throttle position it generally spreads the ramp up from 0% to 97% over 1.0-1.4 seconds.

    So basically Id like to make a circuit that, when you flip a toggle, it takes over control of the solenoid and manually ramps it up to 97% DC over the period of 1.0 second.

    Ive done a lot of research and found many different methods of making a small PWM driver circuit, but Im just not sure which one is the simplest, and which one would do exactly what I want...


    this one seems simple but it has no ramp-up function...could I add a capacitor in there to give it ramp-up? Also, it shows a dc motor there, im not sure if I would need a different FET thats specifically for controlling a transmission solenoid?
    http://www.dprg.org/tutorials/2005-11a/index.html

    here is another one I found, but it looks way more complicated and I feel like what I want could be accomplished with a similar circuit..
    http://www.elecfree.com/electronic/wp-content/uploads/2007/07/ee2382390-1.gif

    and finally, I found this FDMS2380 IC that seems to be perfect...its designed for automotive apps, specifically for controlling transmission solenoids!! But im not sure how I would wire it up. The fact that its surface mount is also kind of a PITA. And im not sure how I would get it to output 100hz PWM and do a specific ramp-up. Would I add a separate 555 astable circuit? Not sure how I would set that up either. In looking at the data sheet for this, im not sure why there are two control inputs on each channel, a and b?? Do both of them have to go high and then it will actuate the solenoid?? And why are there 4 separate outputs on each channel?
    http://www.fairchildsemi.com/ds/FD/FDMS2380.pdf


    thanks again for any help with this project..

    ben
     
  2. scubasteve_911

    Senior Member

    Dec 27, 2007
    1,202
    1
    I would stick to the fairchild part, eventhough the fact that this is SMT is a pain. There are two inputs for different modes of operation. The datasheet speaks of these in the applications section. They mainly dictate the 'recirculation modes'.

    There is only one output per channel, but many pads on the device with the same signal. This is to increase current handling/minimize inductance.

    I am not sure about the ramping up that you speak of. Perhaps this is happening because of the non-linear force versus current of the solenoid. For example, the solenoid can cause more force with little current at the start position, then as the plunger moves, it requires more current for the same amount of force.

    To properly diagnose this, you should be using a current probe on an oscilloscope. This should give you a signature of what you are looking for. Secondly, I would measure the inductance of the solenoid. These two things will enable you to create a proper startup profile. I have no clue about a simple way to do this, others can help. I only work with microcontrollers for these kinds of things, since they're dynamic and more technologically-relevant.

    Goodluck,

    Steve
     
  3. mik3

    Senior Member

    Feb 4, 2008
    4,846
    63
    Well, if you want to spread the ramp up from 0% to 97% over 1.0-1.4 seconds depending on throttle position then the best solution is to use a microcontroller or a PLC with transistor outputs otherwise the circuit will be complicated.
     
  4. vetterick

    Active Member

    Aug 11, 2008
    35
    0
    Engineers, please don't read the following, I can't be held responsible for heart attacks.

    Couldn't you just skip the PWM entirely and just use a simple RC timer and a couple of transistors, might not be as fun but would be lot easier to do.
     
  5. dmaxben

    Thread Starter Member

    Oct 27, 2006
    11
    0
    thanks for all the replies!

    The ramp-up can be fixed at 1.0 seconds. I dont care about varying it based on throttle position.

    The solenoid is ramped up because the torque converter lockup clutch needs to be applied smoothly and gradually. Current remains constant. Basically the equivilent of driving a manual transmission car...if you're at a stoplight and side step the clutch, its going to lurch and start moving with a harsh bang. But if you slowly let the clutch out, its smooth. Same thing as the pulsed apply on this automatic transmission.

    The solenoid is not PPC (pressure proportionate to current) solenoids.

    Here is a description from the service manual:
    TCC (PWM) Solenoid F - This solenoid a normally
    closed, pulse width modulated, and operates at a
    frequency of 100 Hz (cycles per second). The percentage of time the voltage is ON
    during each 100th of a second is called the solenoid
    duty cycle.
    A 100 percent duty cycle indicates a maximum signal
    to the solenoid. A zero percent duty signal indicates a
    minimum or no signal to the solenoid. The TCM,
    using pulse width modulation programming, varies
    the percentage of voltage ON time during a cycle. As
    the pulse width, or duty cycle is increased, the
    solenoid is ON longer.


    the solenoid has a resistance of 10 ohms

    ben
     
  6. scubasteve_911

    Senior Member

    Dec 27, 2007
    1,202
    1
    I don't really understand what is special about this solenoid. Are you saying electronics are built in? And the 0-100% 100Hz PWM is only a control signal? If this is true, then you do not need a high power bridge.

    I didn't know a solenoid existed that doesn't have a non-linear current to force characteristic. Maybe a differential arrangement or something be used for linearity?

    So, the solution would probably be, as suggested, to use a microcontroller for this. Otherwise, you will probably have a pretty complex circuit to do. A microcontroller can do that sort of PWM with ease. Maybe buy a cheap starter kit and run the board as the controller when you're done.

    Along the lines of, http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=428-1585-ND

    Steve
     
  7. dmaxben

    Thread Starter Member

    Oct 27, 2006
    11
    0
    Steve- Im sorry I was incorrect about it not being a PPC solenoid. Its just a standard "basic" solenoid that needs a 100hz PWM control signal to ground the solenoid. I dont really know much about programming microcontrollers or PLC's. :(

    thanks
    Ben
     
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    If I remember correctly, this solenoid is actually a hydraulic valve. Changing the PWM on the solenoid affects the pressure behind the valve; thus the actuation of the components that the valve controls.

    Actually, you would want some input from the throttle position; if the driver had the accelerator pressed down more than about 1/3 of the way, the torque converter should come out of lockup. Otherwise, it may be damaged by a "hard shift" of the transmission; ie a kickdown to 1st or 2nd gear during WOT operation. The disengangement and reengagement of converter lockup must have a fair amount of hysterisis, or the cycling will quickly wear out the lockup components.

    Automotive environments are among the harshest around. You must use components that are rated for such environments; ie: -20°C to 150°C operation. Commercial grade components will fail very quickly.

    It seems this would be a decent application for an industrial temp rated 555 timer set for 100Hz operation and an RC time (or better a constant current charge/discharge) on pin 5, with an opening throttle condition or ECU override rapidly dropping the PWM to 0.

    Thoughts?
     
  9. dmaxben

    Thread Starter Member

    Oct 27, 2006
    11
    0

    This is for an Allison transmission in a chevy pickup truck. It works completely differently than all other autos; ie, the converter is supposed to stay locked through the shifts, thats how the transmission works, it doesnt need to unlock between shifts. It also is NOT designed to unlock under higher throttle percentages; its designed to haul 17,000lb trailers with the efficency and control of a manual transmission (ie, the Allison automatic physically cant "freewheel" like other autos, its always locked directly in one gear). Its a pretty interesting but VERY complex automatic transmission.

    The problem is that the TCM (trans computer) is too smart for its own good, its so advanced and over-engineered that in high hp drag racing and competition apps (yes haha I do drag race my 7600lb truck) the TCM gets confused and starts tripping over its own feet.

    Its ALWAYS learning and adjusting things by itself, so therefore its rather unpredictable as to how its going to react to something. If you drive around really easilly for an hour and then suddenly go hammer on it with a tune thats 400 horsepower over stock power level, the TCM freaks out and sometimes puts the "im supposed to lock the converter at full throttle when I get to 25mph in second gear" on the back burner so to speak, and once it gets a grip and finally says "ok now I know whats going on...lets lock the converter now" the truck might be at 65mph, and all that power was wasted because the converter didnt lock when it was supposed to.

    So basically I just want to have an option to manually lock it when I say so, for more consistant performance in drag racing or sled pulling... :)

    ben
     
    Last edited: Sep 23, 2008
  10. scubasteve_911

    Senior Member

    Dec 27, 2007
    1,202
    1
    Hi Ben,

    It's too bad, programming a micro like the one suggested doesn't take much time to learn. I actually started when I was 18 and it took me about a week to write my first medium-complexity program. So, it shouldn't be that hard to pick up if you give it a shot. The particular microcontroller I suggested actually has a psoc express environment that doesn't require any code to make an application, but I haven't played with it yet. Or, you can get the standard free software with example source code.

    I'm completely clueless when it comes to how vehicles function, other than the complete basics.. I'd rather be good at few things than not that great at many, ces't la vie!

    Steve
     
  11. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    Question , first thread lists current as 10 A. , later coil resistance is shown as 10 Ω ? One end of coil tied to + 12 V. ?
     
  12. dmaxben

    Thread Starter Member

    Oct 27, 2006
    11
    0
    yes, one end of the solenoid is tied to 12vdc
     
  13. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    What is the coil resistance?; is current 1.2 or 10 A ?
     
Loading...