Help choosing a resistor value

Discussion in 'General Electronics Chat' started by gte, Jun 20, 2010.

  1. gte

    Thread Starter Active Member

    Sep 18, 2009
    347
    4
    Attached is a picture of the solenoid, with a resistance measurement across the terminals.

    I'm getting 8.2 ohms of resistance. The solenoid gets ~10vDC square wave pulses @ 312hz, so 10vDC/8.2Ω = ~1.22 amps, but since this is wired directly to the cars ECU, I'm a little nervous about my math and want to make sure my thought process is correct so that I don't overdrive something and kill the ecu. The wiring from the ecu to the solenoid looks to be 18 gauge.

    If my thought process is correct, I need an 8 ohm 12 watt resistor ... is anyone in disagreement?




    [​IMG]



    And the square wave signal that the solenoid is controlled by


    [​IMG]
     
  2. ifixit

    Distinguished Member

    Nov 20, 2008
    639
    108
    Hi gte,


    18 guage is fine for 1.22A. Need more info to determine safety, however...
    1. Is the ECU designed to drive this solenoid?
    2. Is the solenoid designed to operate at 312Hz?
    3. The solenoid should be driven with a solenoid driver circuit. Does the MCU have one of those? If not, don't connect.
    4. What were you going to do with the resistor? Put it in series with the solenoid? The current would be reduced by half... it may not operate like that.
    Regards,
    Ifixit
     
  3. gte

    Thread Starter Active Member

    Sep 18, 2009
    347
    4
    Hi

    The solenoid is a factory piece, and the ecu driving it is all OEM. The oscope screen captures are also OEM signal samples, so yes to #1 and #2, I believe yes because the ecu will drive the solenoid in stock configuration to #3.

    As for #4, I want to replace the solenoid entirely with something (a resistor), so that the ecu still thinks it is present. This way I can drive the solenoid with an aftermarket controller on my own, without the ecu being the wiser. The ecu has some detection and can tell when the solenoid is plugged in.

    What do you think?




     
  4. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    You may want to observe the return wire to the ECU with the solenoid connected as per OEM.

    See what you are scoping there. That is what you will have to match to "fool" the ECU.
     
  5. gte

    Thread Starter Active Member

    Sep 18, 2009
    347
    4
    Here is what the oscope signal looks like on the - side of the sensor ... what's the best way to test for amperage? Does a solenoid work by having a dc ground and a positive signal (with mechanical spring pressure to force it back to its resting position), or does it reverse polarity to bring the solenoid back to its resting position?

    [​IMG]




    Also, when I have it disconnected I get a check engine light code of P 2645
    And when I have a 150ohm 1/4 watt resistor (that got semi hot) I get a check engine light code of P 2647
    ... so based on the description it appears they are opposites. One it detects no connection at all and the other it detects too much of something? What do you think?





     
  6. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    Odd. You may want to connect a potentiometer to the ecu and see where it is finally happy. Start with it all the way down and it should say disconnected. keep turning up until you find the spot where it is happy but not calling for "stuck" position.
     
  7. gte

    Thread Starter Active Member

    Sep 18, 2009
    347
    4
    All the way down would be almost no resistance, correct?

    I'm worried about damaging the ecu, which is why I wanted to monitor amperage in the stock configuration, before I did anything else, ideas on the best way to do this? If the amp measurement matches the calculated amp estimate, I'll feel alot better about experimenting a little.




     
  8. sashijoseph

    New Member

    Jun 29, 2006
    5
    0
    nope...start with max resistance and then go down bit by bit.

    Alternatively how about sticking up a 8-ohm speaker voice coil instead of the solenoid?
    Let the experts chip in though...I'm a noob myself.
     
  9. ifixit

    Distinguished Member

    Nov 20, 2008
    639
    108
    Hi,
    1. Most solenoids return to there original position with a spring built into them, or by an external spring, or other external mechanism.
    2. If the solenoid is replaced with an 8Ω resistor and is power 100% of the time, then use at least a 15W wire-wound power resistor to allow for the higher-than-usual automotive ambient temperatures.
    3. Is it possible that the MCU is monitoring some other engine function that is suppose to change when the solenoid is actuated, and then when it doesn't change, the MCU complains with a code?
    4. Try a test with a real solenoid in the circuit connected electrically, but not connected mechanically. You can connect your manually operated solenoid mechanically if necessary. If the MCU doesn't complain then (3) is not likely the case.
    5. The MCU may be quite critical about the current the solenoid draws when activated, so an 8Ω resistor may be required instead of a value like 150Ω or less. You can experiment as suggested earlier, but do not use a speaker voice coil:eek:, its DC impedance is much less than 8Ω.
    Regards, Ifixit
     
    gte likes this.
  10. gte

    Thread Starter Active Member

    Sep 18, 2009
    347
    4
    Hi!

    Believe it or not, Radio Shack actually sells an 8Ω 20watt resistor that I happened to purchase the other day. It says it is NON Inductive though, which leads me to believe it is not "wire-wound" ... what do you think?

    As for your bullet points

    1. It appears this one also has a small spring at the end of the beveled tip of the actuator, which probably is what returns it to its resting position

    2. Let me now if I can use what I have

    3. The ecu does monitor it externally with another sensor, but I have built a circuit to bypass this already. I did however, for the purpose of due diligence, do the test that you suggested this evening, with my circuit in place. The check engine light did not go off, and I drove it all around like that, in may different conditions

    4. It appears it only monitors current somehow. Should I get an ammeter?

    5. Before I try this, I just want to make sure I'm not going to overdrive the ecu, what do you think about the resistor I have?



    Thanks for you responses!




     
  11. BillB3857

    Senior Member

    Feb 28, 2009
    2,400
    348
    Don't forget that the solenoid will have inductive reactance! That will limit the amount of current the ECU must provide. A resistor, even a wire wound resistor, will not have anything near the same inductive reactance the coil in the solenoid has. Hate to see you blow up your ECU. Also, won't your plan violate the emission control requirements of your vehicle?
     
    gte likes this.
  12. gte

    Thread Starter Active Member

    Sep 18, 2009
    347
    4
    Hi Bill,

    I would never violate any emission control requirements.

    I had not considered the inductive reactance, any idea how to measure that?

    inductive reactance in Ω = 2∏fL or 2 * (22/7) * Frequency * L

    I have the frequency measured, it's 312hz, and the other givens are there, how do you measure inductance?




     
  13. gte

    Thread Starter Active Member

    Sep 18, 2009
    347
    4
  14. BillB3857

    Senior Member

    Feb 28, 2009
    2,400
    348
    Any test equipment is always handy to have around. However, you can do a quick and dirty test to come up with what you are looking for in a resistor value. Since you have a scope, place a very low ohm resistor (something like two one ohm resistors in parallel) in series with the ground lead of your solenoid. The voltage developed across the resistor will represent the current pulses that the solenoid is actually seeing. From the current seen and the voltage you have already seen, you can calculate the effective resistor value you were looking for in the first place. Don't be too surprised if it is much higher than the DC resistance of your solenoid.
     
    gte likes this.
  15. ifixit

    Distinguished Member

    Nov 20, 2008
    639
    108
    Hi,

    1. That resistor will be find. 20W is good. WW resistors can be non-inductive by using special winding techniques (half one, half the other way). In your case it really doesn't matter. Ensure the resistor has good convection air cooling.
    2. A solenoids inductance has to fully saturate for it to pullin properly and pull with full force. After it has pulled in the driver will "see" only the DC resistance of 8Ω.
    3. On the scope picture it looks like the on-time of the solenoid is approximately 0.75mS. Assuming the solenoid can actuate in that time, the inductance must be approximately 1.2mH or less.
    4. An LCR meter is handy to have and you can certainly use this as an excuse to get one:).
    5. In may opinion you can sub the solenoid with the 8Ω resistor as is.
    6. I doubt the ECU looks for anything else except an expected current draw of approximately 1.25 Amps.
    Regards,
    Ifixit
     
    gte likes this.
  16. gte

    Thread Starter Active Member

    Sep 18, 2009
    347
    4
    Sweet, I will try it tonight.

    I just bought an LCR meter :) to add to my tool box
     
  17. brodders

    New Member

    Jun 22, 2010
    3
    1
    Hi GTE,

    I hope you don't mind me butting in.
    The Resistor from RS will be no probs, don't worry about it not being inductive.

    A pot would have to be a large pot to handle the possible current.

    I have my own question to ask;

    What's the original purpose of the solenoid you're trying to bypass?

    Yes you can fool the ECU into thinking it's still there but, depending on what feedback the ECU expects, a fixed resistor may not work.

    Example: If the solenoid is an 'Air Bypass' / idle Control, Fuel Quantity Control on a pump, the ECU might try and increase the PWM to the solenoid and expect a rise in Idle Speed; this isn't going to happen if you use an additional controller which, may bring on a warning light or just log a fault code.

    Regards bb
     
    Last edited: Jun 22, 2010
    gte likes this.
  18. gte

    Thread Starter Active Member

    Sep 18, 2009
    347
    4
    Hey BB,

    I don't mind you joining the conversation at all!

    The ecu does have closed loop feedback, but it's very primitive and is either a 0 or 1 at a certain rpm. By fooling that and fooling the load, I can then control it as I want, when I want, without a check engine light.

    Couple this with an AEM FIC and you have the recipe for a little more power than stock :)

    I'm going to try and drive around with the resistor in tonight, I have decided I will wire it into the cabin so that I can keep and eye on it, and put a 2amp inline fuse in place, to protect the ecu.
     
  19. gte

    Thread Starter Active Member

    Sep 18, 2009
    347
    4
    Alright gentlemen thank you for all of your input, I felt pretty safe with the inline fuse of 3 amps and so I went out testing a few minutes ago.

    The 8 ohm 20 watt resistor got a little warm to the touch, it was 85F out, so I guess the resistor was about 95F because I could feel it was a little warmer then the ambient temperature of my cars cabin (I did not have it in the engine bay)



    Here is a video of what the voltage drop is across it, you'll see it change at a certain rpm from 1.13vDC to 7.55vDC (DC average of a 0 to 10v square wave) of a drop across the 8 ohm resistor.


    On a side note, another funny thing I noticed is that it's always at a .77vDC drop across the resistor when I first start the car, until the first time I eclipse the square wave changeover point where the time on vs time off basically inverts and then the lowest resting averaged voltage it will fall to after that will be 1.13vDC until next time I cycle power with the key, then it will be .77vDC until the square wave time on vs time off changes and then the lowest it will rest at will be 1.13vDC ... any ideas as to why?

    http://www.youtube.com/watch?v=Ou2YxRqkRtU



    Any concerns with the temperature? Even if it got reasonably hot, that'd be ok correct?



     
  20. ifixit

    Distinguished Member

    Nov 20, 2008
    639
    108
    Hi gte,

    The 0.77V drop is likely as a result of a small current put through the solonoid buy the ECU to detect whether it is connected or not.

    I don't have the spec for your RS 20W resistor, but I can speak in general terms.



    Design considerations...
    1. Power resistors are designed to operate at very hot case temperatures, which can be in the 300 to 400 °C range. That doesn't mean you should, however, operate like that.
    2. The increased temperature means the temperature coefficient of the resistor will cause the resistance value to change some percentage per °C. This may be important to a design or not.
    3. The materials in the vicinity of the resistor usually cannot stand that high temperature unless its steel and the joints are welded or crimped. E.g. FR4 PCB material should not be heated above 105°C, but 85 °C is better for good long term reliability.
    4. De-rate the power rating (20W in your case) down to an exceptable level of dissipation where the case temperature does not exceed 85 °C.
    5. Item (4) is an excuse to buy a temperature probe:).
    6. With a case temperature of only 95°F, that resistor will last for 1000 years (estimate).
    7. This is likely more info than you need.
    Good luck, drive safely and remember to buckle-up.

    Ifixit
     
    gte likes this.
Loading...