Help wanted, circuit for regular pulse variablephase shift.

Discussion in 'The Projects Forum' started by cork_ie, Sep 14, 2012.

  1. cork_ie

    Thread Starter Member

    Oct 8, 2011
    348
    58
    Hi Folks, I need some help.I have checked the archives and can only find time interval pulse delay circuits.

    I have two wheels driven together at 2:1 ratio. Wheel 1 spins at twice the speed of wheel 2.

    Wheel 1 is toothed and generates regular pulses with a missing tooth to indicate position. Sensor is either a magnetic coil or Hall effect type.

    Wheel 2 turns at half the speed of wheel 1 - rate of rotation is approx 10 times/sec to 60 times/sec. It has a marker tooth which generates a pulse by switching a Hall effect sensor.

    I need to advance & delay the pulse from wheel 2 by -10 Deg of angular rotation to + 10 Deg of angular rotation. The angular shift to remain steady irrespective of speed of rotation.

    I would prefer to avoid using a software based solution but if I have no option I will accept all suggestions.
     
  2. crutschow

    Expert

    Mar 14, 2008
    12,986
    3,226
    I can understand a -10 Deg delay, but how could you do a +10 Deg delay (advance?):confused: That requires predicting the future.

    Of course, you could delay the pulse by 350 degrees which would be an effective +10 degree advance, if that's acceptable.

    In any case, you need to calculate the phase angle, if the speed varies. That is most easily done with a microprocessor and software. A hardware solution would likely be complex.
     
  3. cork_ie

    Thread Starter Member

    Oct 8, 2011
    348
    58
    What I am actually trying to do is make a simulator for a stretched timing chain in an engine. By varying the phase angle of the camshaft pulse I would be able to determine whether it is the chain or the chain tensioner that is in trouble.

    I know that an oscilloscope would allow me to see what is happening but without a good waveform to start with it, it would be next to impossible to make a confirmed diagnosis. This way I can adjust the "apparent valve timing" ie what the ECU sees, electronically and increase or decrease the angular offset until such time as a diagnostic trouble code is set. That way I can exactly determine the true mid point.
    I was hoping I would be able to achieve a - 10 deg delay by shifting the pulse from wheel 1 and the + by shifting the pulse on wheel 2 - effectively two circuits, only one operating at any given time.

    I have thought a lot about this - it is a bit of a holy grail and I wonder why no one has ever come up with it before, probably uneconomic to do so.

    The phase angle as adjusted must remain fixed for a given setting irrespective of frequency ie rotational speed - that is the whole point.

    I was wondering if there was any way I could modify the attached circuit to suit my needs
     
  4. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    It's a standard enough application used all the time in automotive performance devices by modifying the phase timing of ignitions, fuel injector pulses etc etc. There are a number of devices you can buy that will do similar things with the crank and cam pulse trains.

    Making one of these will likely go against the forum rules as the #1 reason for wanting one is as a performance improving device to fool the car ECM/ECU timing in order to hot up a car's performance.
     
  5. cork_ie

    Thread Starter Member

    Oct 8, 2011
    348
    58
    What I am looking for is completely different - I need to temporarily vary the angle between cam and crank signal by a FIXED a amount irrespective of speed of rotation ,
    The item mentioned in the previous post is to vary the pulse timing of both signals simultaneously as a function of speed of rotation and yes, would definitely be against the terms of service.
     
  6. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    I don't want to get into an argument with you, but I design auto electronics professionally, and that is exactly what they do. The racer can "dial in" +/- X degrees in timing (referenced to cam timing) and the difference in degrees makes the spark or fuel injection etc happen at a different cam angle.

    That enables stock ECUs to make more power by deviating from the factory set timings which are usually set to give good fuel economy, low emissions, safety, etc.

    As for violating the forum rules that's not up to my opinion or yours either, it's entirely the moderators decision (so please don't take my comments personally).
     
  7. cork_ie

    Thread Starter Member

    Oct 8, 2011
    348
    58
    Correct me if I am wrong but any FI engine I know of has
    a crank sensor of some type - timing (piston position) is derived from this

    Most modern engines (1990 ->) also have a cam sensor for cylinder identification to facilitate sequential injection & distributorless ignition.This is why it is sometimes called Phase angle sensor.

    Cam - Crank relative position is fixed except where the engine has Variable Valve Timing. In that particular case adjusting cam angle would affect performance as valve opening times would vary.
    In all other cases it would likely just set a DTC for cam/crank range/performance P0336 or P0341.

    What I need to do is diagnose a slack timing chain if either of these previously mentioned codes are set and verify, that it is indeed the chain (or tensioner) by temporarily restoring the cam crank sync to the original factory setting and confirm that such codes are no longer set.

    I have heard of but never seen devices that will vary the crank angle signal - not a good idea as it will affect ignition timing. Anyone I know who goes to that extreme generally goes for an ECU remap.

    If you have a link to the device, you have in mind, I would be interested in having a look and seeing exactly what it does.
     
  8. Dodgydave

    Distinguished Member

    Jun 22, 2012
    4,969
    744
    Can you not alter the gears so the wheel 2 gets a pulse 10 deg earlier and then use software to delay it by -10 degrees and -20 degrees.
     
  9. cork_ie

    Thread Starter Member

    Oct 8, 2011
    348
    58
    Unfortunately the gears cannot be altered but I was hoping to introduce a similar delay in wheel 1 which would have the same effect for this particular application.
     
  10. Dodgydave

    Distinguished Member

    Jun 22, 2012
    4,969
    744
    Ok so put your sensor 10 degrees earlier on wheel 2, and use the software delays for -10 deg and -20 deg.


    So lets relate this to the face of a clock, your sensor picks up pulses at 12 o'clock position, the wheel rotates clockwise, but you want to delay it, so move the sensor to 11 o'clock (this will advance the pick-up)

    and use the software to delay the pulse, then when you want to advance the pick-up reduce the delay time on the software.
     
    Last edited: Sep 17, 2012
  11. cork_ie

    Thread Starter Member

    Oct 8, 2011
    348
    58
    Neither the sensors or the wheels are going anywhere, they are fixed and cannot be altered without major stripping and engineering.

    All I want to do is adjust the RELATIVE signal positions temporarily to facilitate refinding the mean point. This will help me diagnose whether the slack is on the tensioner side of the chain or not.

    All i want to do is introduce a frequency independent delay which is adjustable +/- 10 Deg maximum
     
  12. vk6zgo

    Active Member

    Jul 21, 2012
    677
    85
    Back in the day,drag racers used to fit a device that mechanically changed the relative phase between crankshaft & camshaft rotation.
    I seem to remember the brand of one of these things was "Cam-a-go".

    The other trick which was much cheaper & gave you a fixed offset,was to use an offset Woodruffe key to offset the crankshaft gear.
    Of course,this was an iterative process,where you started out with a bunch of keys & tried them until you got the best compromise.
    Both of these ideas only worked with gear driven camshafts.

    Back to the point at hand.
    The only way you will get a +10° degree phase change is to delay the previous pulse by 350°.

    Some analog networks can,over a limited range of frequencies, provide a constant phase delay but it is not variable.
    Equally,simple digital circuits can provide fixed phase delay..

    To achieve what you want implies to me some fairly heavy number crunching,which means the use of a microprocessor,or maybe a dedicated PC.
    As I am an analog person,I must bow out at this point!:D
     
  13. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    If your goal is to diagnose, and not to trick the ECU for changing performance, then you can diagnose using the workshop manual tests for cam chain timing.

    If you want to go more high tech than that, you simply SENSE the cam pulse and the crank pulses at a fixed RPM, which you could do on a 'scope with two 'scope leads and see the phase delay caused by a stretched cam chain.
     
  14. cork_ie

    Thread Starter Member

    Oct 8, 2011
    348
    58
    Yes I could try that, but how am I going to introduce the delay so that it is fixed (depending on my setting) and frequency independent?
    Any ideas anyone?
     
  15. cork_ie

    Thread Starter Member

    Oct 8, 2011
    348
    58
    Just to elaborate a little more. I want to build this especially to test Nissan Micra 1.0l engines. Timing chain stretch is a big issue with them, often at less than 25,000 miles. It is a $1,000 job as the engine has to be removed to do it , there is no other way.It is quite difficult to diagnose with 100% certainty and there are many cases of misdiagnosis as the symptoms could be caused by a hundred other things.

    Nissan Worshop manual says that "Just in case" you spend the day taking out the engine & stripping everything down to have a look at it. Then measure everything carefully with a micrometer and send the owner a $400 bill just to have a look.

    Worse again there is a variable valve hub on the intake camshaft and the CAM sensor is at the other end. So the timing is going to vary anyway once the engine runs.

    The main issue with this engine is that it is almost impossible to start cold if either the chain stretches or the VVT hub jams. It will start, run for 2 seconds , cut the spark signal and may eventually display a P0340 code (Which is factually incorrect and is caused by poor software).

    By temporarily adjusting the cam - crank angle back to it's original set point by either advancing or retarding the CAM pulse, it can be started easily and proper scope readings taken. Knowing whether the signal needs to be advanced or retarded and by how much tells an awful lot and is a guaranteed 100% accurate diagnosis in less than 20 minutes.

    All I want is someone to tell me how I can alter a square wave pulse within a narrow band of +/- 10 degrees so that the delay is fixed irrespective of frequency. The rest I feel adequately competent to figure out for myself.
     
    Last edited: Sep 18, 2012
  16. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    Ok, now you have outlined the problem in more detail I think you could build a neat stand alone test unit very simply, and still WITHOUT having to interrupt or interfere with the critical crank and cam timing pulses.

    I would use a PIC micro and connect to the crank and cam sensors (passively, without interfering with the signals). Then once the engine is running (at any speed) take two measurements;
    1. period of one full crank rotation.
    2. period between crank pulse and cam pulse.

    Then the PIC simply scales the period 2 into period 1 as a ratio;
    ratio = p2 / p1
    then to give the "stretched cam chain" angle;
    angle_in_degrees = ratio * 360
    which could all be done as one simple binary calc;
    angle = (p2 * 360) / p1

    Then any engine with a good cam chain would read an angle of X which is the normal angle that the cam pulse occurs after the crank pulse (which might even be 0 degrees).

    Then an engine with a faulty stretched cam chain would show an increased cam angle. Simple and no dangerous interfering with the car ECU or modifying of the cars operation.

    I'm not going to help you with that, because that's exactly what young rev heads want to do to their fuel injected cars to fool the ECU and try to make it go faster, causing significant risk of engine damage and emissions illegality. It would also (in my opinion) contravene the forum rules on car modifications, for exactly all those reasons (safety and legality).
     
  17. cork_ie

    Thread Starter Member

    Oct 8, 2011
    348
    58
    I still haven't worked out how this is going to make a car go faster as I mentioned in my previous post Crank signal = Timing, Cam = phase . I want to alter the RELATIVE angle back to where it should be so the car can be started without swinging it for a week.
    I trawled the internet for a device such as you previously mentioned and I have yet to find it. There are devices which will introduce a frequency dependant delay in both CAM & Crank to advance the overall timing without setting a fault code. That is an entirely different matter.
     
  18. cork_ie

    Thread Starter Member

    Oct 8, 2011
    348
    58
    While I didn't manage to get the help I wanted on AAC. I would like to thank those who genuinely tried to help and even those that wanted to tease.

    I eventually managed to modify another circuit posted here some time ago and it does exactly what I wanted to do.
     
Loading...