Datasheet

LM1949
www.ti.com SNLS349C –FEB 1995–REVISED MARCH 2013
PEAK AND HOLD CURRENTS
The peak and hold currents are determined by the value of the sense resistor R
S
. The driver IC, when initiated by
a logic 1 signal at Pin 1, initially drives Darlington transistor Q
1
into saturation. The injector current will rise
exponentially from zero at a rate dependent upon L
1
, R
1
, the battery voltage and the saturation voltage of Q
1
.
The drop across the sense resistor is created by the solenoid current, and when this drop reaches the peak
threshold level, typically 385 mV, the IC is tripped from the peak state into the hold state. The IC now behaves
more as an op amp and drives Q
1
within a closed loop system to maintain the hold reference voltage, typically 94
mV, across R
S
. Once the injector current drops from the peak level to the hold level, it remains there for the
duration of the input signal at Pin 1. This mode of operation is preferable when working with solenoids, since the
current required to overcome kinetic and constriction forces is often a factor of four or more times the current
necessary to hold the injector open. By holding the injector current at one fourth of the peak current, power
dissipation in the solenoids and Q
1
is reduced by at least the same factor.
In the circuit of Figure 1, it was known that the type of injector shown opens when the current exceeds 1.3 amps
and closes when the current then falls below 0.3 amps. In order to guarantee injector operation over the life and
temperature range of the system, a peak current of approximately 4 amps was chosen. This led to a value of R
S
of 0.1Ω. Dividing the peak and hold thresholds by this factor gives peak and hold currents through the solenoid of
3.85 amps and 0.94 amps respectively.
Different types of solenoids may require different values of current. The sense resistor R
S
may be changed
accordingly. An 8-amp peak injector would use R
S
equal to .05Ω, etc. Note that for large currents above one
amp, IR drops within the component leads or printed circuit board may create substantial errors unless
appropriate care is taken. The sense input and sense ground leads (Pins 4 and 5 respectively), should be Kelvin
connected to R
S
. High current should not be allowed to flow through any part of these traces or connections. An
easy solution to this problem on double-sided PC boards (without plated-through holes) is to have the high
current trace and sense trace attach to the R
S
lead from opposite sides of the board.
TIMER FUNCTION
The purpose of the timer function is to limit the power dissipated by the injector or solenoid under certain
conditions. Specifically, when the battery voltage is low due to engine cranking, or just undercharged, there may
not be sufficient voltage available for the injector to achieve the peak current. In the Figure 18 waveforms under
the low battery condition, the injector current can be seen to be leveling out at 3 amps, or 1 amp below the
normal threshold. Since continuous operation at 3 amps may overheat the injectors, the timer function on the IC
will force the transition into the hold state after one time constant (the time constant is equal to R
T
x C
T
), or when
the voltage on the TIMER pin (Pin 8) is greater than typically V
SUPPLY
x 63%. The timer is reset at the end of
each input pulse. For systems where the timer function is not needed, it can be disabled by grounding the TIMER
Pin (Pin 8). For systems where the initial peak state is not required, (i.e., where the solenoid current rises
immediately to the hold level), the timer can be used to disable the peak function. This is done by setting the time
constant equal to zero, (i.e., C
T
= 0). Leaving R
T
in place is recommended. The timer will then complete its time-
out and disable the peak condition before the solenoid current has had a chance to rise above the hold level.
The actual range of the timer in injection systems will probably never vary much from the 3.9 milliseconds shown
in Figure 1. However, the actual useful range of the timer extends from microseconds to seconds, depending on
the component values chosen. The useful range of R
T
is approximately 1k to 240k. The capacitor C
T
is limited
only by stray capacitances for low values and by leakages for large values.
The timing capacitor is reset (discharged) when the IN pin (Pin 1) is below the V
OL(MIN)
threshold. The capacitor
reset time at the end of each controller pulse is determined by the supply voltage and the timing capacitor value.
The IC resets the capacitor to an initial voltage (V
BE
) by discharging it with a current of approximately 15 mA.
Thus, a 0.1 µF cap is reset in approximately 25 µs.
COMPENSATION
Compensation of the error amplifier provides stability for the circuit during the hold state. External compensation
(from Pin 2 to Pin 3) allows each design to be tailored for the characteristics of the system and/or type of
Darlington power device used. In the vast majority of designs, the value or type of the compensation capacitor is
not critical. Values of 100 pF to 0.1 µF work well with the circuit of Figure 1. The value shown of 0.1 µF (disc)
provides a close optimum in choice between economy, speed, and noise immunity. In some systems, increased
phase and gain margin may be acquired by bypassing the collector of Q
1
to ground with an appropriately rated
0.1 µF capacitor. This is, however, rarely necessary.
Copyright © 1995–2013, Texas Instruments Incorporated Submit Documentation Feedback 7
Product Folder Links: LM1949