OT: Detecting bounce in fuel injector

Do you have a question? Post it now! No Registration Necessary

Translate This Thread From English to

Threaded View
Hi,

A colleague is developing a new fuel injector system which may have
broad application after some more research. Basically he is pulsing
the injector and can vary the pulse width, when at lower pulsewidths
the predicted mass flow does not appear to follow what it would
assuming a trapezoidal waveform/motion as a standard type petrol
injector does.

He thinks it may be bounce causing this. He wants to measure either/or
sound and magnetic field via pickups. I said I thought that both are
difficult measurements to interpret with a CRO. He hopes to pickup the
harder sound of the injector seal hitting the metal seat. The motion
is about 0.1mm p-p. The pulse drive supply is believed to be an open
collector/drain driving the injector solenoid, with a reverse biased
diode across the coil.

Just looking for ideas on detecting bounce. Apparently electrical
detection is difficult because the whole assembly and spring are
connected. Seems to me a distance measurement might be the best way,
unless a microphone can detect the higher harmonic content of seal-
plunger collision.

Re: OT: Detecting bounce in fuel injector

:Hi,
:
:A colleague is developing a new fuel injector system which may have
:broad application after some more research. Basically he is pulsing
:the injector and can vary the pulse width, when at lower pulsewidths
:the predicted mass flow does not appear to follow what it would
:assuming a trapezoidal waveform/motion as a standard type petrol
:injector does.
:
:He thinks it may be bounce causing this. He wants to measure either/or
:sound and magnetic field via pickups. I said I thought that both are
:difficult measurements to interpret with a CRO. He hopes to pickup the
:harder sound of the injector seal hitting the metal seat. The motion
:is about 0.1mm p-p. The pulse drive supply is believed to be an open
:collector/drain driving the injector solenoid, with a reverse biased
:diode across the coil.
:
:Just looking for ideas on detecting bounce. Apparently electrical
:detection is difficult because the whole assembly and spring are
:connected. Seems to me a distance measurement might be the best way,
:unless a microphone can detect the higher harmonic content of seal-
:plunger collision.


If the injector solenoid actuates a valve it is the valve which is most likely
bouncing, not the solenoid.

In a diesel engine valve bounce can excessively lengthen the period between
pre-injection and main injection thus reducing the efficiency of fuel burn and
producing poorer CO2 results. You didn't say whether you are talking about
petrol or diesel injection but I assume that valve bounce is undesirable in
both. Modern diesel engines are now using piezo injectors in preference to
solenoid injectors. They are 3 times faster and the injection sequence can be
more precisely controlled.

A search for patents pertaining to the problem may provide some clues for you
http://www.freepatentsonline.com/EP1082535.pdf (you may have to register)
I haven't studied the text in detail but you can do that. I don't know the
equipment used to detect the valve bounce timing graphs but it is probably
commercially available.

Essentially, the scheme appears to provide an initial energise pulse to the
solenoid and then just prior to the valve reaching the fully open position,
re-energising the solenoid again. This purportedly prevents the initial and
subsequent bounces and thus reduces the lag between pre-injection and main
injection fuel burns.

Re: OT: Detecting bounce in fuel injector

Quoted text here. Click to load it
youhttp://www.freepatentsonline.com/EP1082535.pdf (you may have to register)
Quoted text here. Click to load it

Low viscosity fuel. Thats good info about the piezo and the patent,
thanks.

Re: Detecting bounce in fuel injector

Quoted text here. Click to load it

You don't say what the duty cycle is. Also all rules for automotive use have
now changed and I would suggest before going any further to get your mate to
acquaint himself with CANbus technology.  PWM Solenoids are now working on
much higher frequencies/duty cycles.

IIRC most automotive PWM solenoids use the liquid they are metering as a
"shock absorber" and so reducing the bounce/"hysteresis"

Cheers TT



Re: Detecting bounce in fuel injector

Quoted text here. Click to load it

Not sure I understand the question, but the duty cycle will be much
the same as for a petrol engine, maybe a bit higher due to a lower
calorific value of the fuel.
I know about the fuel damping the bounce. Maybe you are implying that
high PWM frequencies are being used to effectively control the
solenoid a bit more accurately so as to simulate a current source or
whatever.

Re: Detecting bounce in fuel injector

:
: > You don't say what the duty cycle is. Also all rules for
automotive use have
: > now changed and I would suggest before going any further
to get your mate to
: > acquaint himself with CANbus technology.  PWM Solenoids
are now working on
: > much higher frequencies/duty cycles.
: >
: > IIRC most automotive PWM solenoids use the liquid they
are metering as a
: > "shock absorber" and so reducing the bounce/"hysteresis"
: >
: > Cheers TT
:
: Not sure I understand the question, but the duty cycle
will be much
: the same as for a petrol engine,

Frequency = How many times per second it turns on and off.
Duty Cycle = How long it is left on or off as a percentage,
hence "pulse width"

 maybe a bit higher due to a lower
: calorific value of the fuel.
: I know about the fuel damping the bounce. Maybe you are
implying that
: high PWM frequencies are being used to effectively control
the
: solenoid a bit more accurately so as to simulate a current
source or
: whatever.

Obviously the way to stop seat bounce is to raise the
frequency as the idea is to not have the valve to continue
to slam into the seat and thus cause wear but to have it in
a controlled "float" position so it can vary an orifice size
and thus meter the liquid.  The only time you want the valve
to contact the seat is when it has stopped metering.

I should add I am in no way an expert in the field I just
fix the things when they are broken ;-)

Cheers TT



Site Timeline