Miscellaneous active electrical nonlinear devices – circuits – and – Signal converting – shaping – or generating – Current driver
Reexamination Certificate
1996-06-17
2002-04-23
Cunningham, Terry D. (Department: 2816)
Miscellaneous active electrical nonlinear devices, circuits, and
Signal converting, shaping, or generating
Current driver
C327S112000
Reexamination Certificate
active
06377087
ABSTRACT:
BACKGROUND OF THE INVENTION
This invention relates generally to a driving scheme for a bipolar transistor, and more particularly to a scheme for driving a push-pull pair of bipolar transistors in a ballast inverter.
Conventional ballast inverters of the push-pull type employ bipolar power transistors having relatively high switching losses when driven into deep saturation during other than maximum load conditions. Switching losses vary based on load conditions imposed on the ballast inverter and are difficult to minimize for all load conditions. The relatively high switching losses also result in the transistors operating at relatively high temperatures which are typically above 60° C. Such relatively high operating temperatures require heat sinks thereby raising the inverter manufacturing cost.
In selecting or designing a base driving scheme for a ballast inverter, the characteristics and parameters of the power transistors, such as current gain, are important and normally must be carefully specified to achieve desired ballast operating characteristics. Different types of bipolar transistors can have different characteristics and/or parameters and therefore cannot be readily substituted for one another in a particular base driving circuit. Significant voltage transients can also appear across the bipolar transistors during turn-on of the lamp ballast. These voltage transients when applied across the bipolar transistors can result in transistor failure.
Conventional ballast inverters sometimes employ Baker clamps in order to more quickly turn-off each bipolar transistor, that is, to more closely simulate the switching speed of field effect transistors. A Baker clamp, however, requires use of a relatively expensive high voltage fast recovery diode, which further increases power losses and cost. A Baker clamp is shown, for example, as diode D in FIGS. 6 or 7 of U.S. Pat. No. 4,318,011 issued to Jurgen Zeis.
A base driving scheme for a conventional ballast inverter also typically requires that the bipolar transistors have a constant, forced current gain. In other words, the base current is maintained in a fixed relationship relative to the collector or emitter current. This requirement limits the range of equivalent bipolar transistors which can be employed within the ballast inverter design, which limits standardization of components used within the ballast inverter.
Accordingly, it is desirable to provide an improved driving scheme for bipolar transistors in which relatively high switching losses are avoided resulting in relatively low operating temperatures of the bipolar transistors. The driving scheme should also permit a relatively wide range of different bipolar transistors to be used in the same circuit while optimizing (i.e. minimizing) switching losses regardless of load conditions. The driving scheme should also minimize the possibility of relatively large voltage transients appearing across the transistors and avoid the need for expensive Baker clamps to provide relatively high speed switching of the bipolar transistors.
U.S. Pat. No. 4,318,011 cited above describes circuits for maintaining what is called the storage period ts of a transistor substantially constant. The storage period ts is defined as the period of delay between the time that the base drive input current switches to its turn-off level (or blocking phase level, which is low in the described embodiment) and the time that the collector voltage begins to switch to its turn-off level (which is high in the described embodiment). This storage period ts increases as the transistor goes more deeply into saturation. Switching losses for the transistor are related to how deeply the transistor enters saturation during the switch-on (or conduction) state and thus can be reduced by reducing this storage period ts.
FIG. 1
is a block diagram of the general scheme described in this patent and corresponds to
FIG. 3
of the Zeis patent. A transistor T is provided with a base drive current Ib that is controlled in magnitude by control circuit
1
, which in turn responds to an output Ur from comparator
2
.
Comparator
2
is shown in more detail in
FIG. 2
, which corresponds to
FIG. 5
of the Zeis patent. As shown in
FIG. 2
, whenever transistor T switches off, comparator
2
obtains a measured difference value Url between the collector voltage Uc of transistor T and the inverse of base drive input voltage Ue. A memory S stores the time integral of Url for use in controlling the base current of transistor T during the next time that transistor T is turned on.
FIG. 3
corresponds to
FIG. 4
of the Zeis patent and shows voltage curves during switching of transistor T. As shown in
FIG. 3
, Url is a pulse waveform that occurs whenever transistor T is turning off. The width of this pulse corresponds to the storage period ts plus the time that it takes for the collector voltage to drop to the turn-off level, which is the total actual time taken by transistor T to turn off after the base drive input voltage Ue drops to turn off transistor T. I call this the turning off time period. It is the objective of the control circuit in the Zeis patent to maintain the time ts substantially constant.
To do this, the Url pulse is integrated each time the transistor T turns off to obtain a value Ur corresponding to the turning off time and this integrated value Ur is then stored in memory S until a new value Ur is obtained when transistor T turns off a next time. When the base drive input voltage Ue rises sufficiently to once again turn on transistor T, this stored value Ur is used by control circuit
1
to limit the base drive current so as to keep the turning off time measured the next time that the transistor T turns off substantially constant.
While the circuits shown in this prior art patent control the depth of saturation of transistor T with a feedback circuit, this is done only indirectly in that the parameter being sensed or measured (i.e., the turning off switching time period) is not the same parameter being directly controlled (i.e., the base driving current). The relationship (i.e., proportionality constant) between the base drive current and the turning off switching time period is not the same for different transistors, thereby making the gain (and therefore the performance) of any feedback circuit of this kind highly dependent upon transistor parameters. Accordingly, any such circuit apparently would need to be tuned or adjusted for different transistors if switching losses are to be minimized, which is undesirable.
SUMMARY OF THE INVENTION
In accordance with the present invention, on the other hand, saturation of the transistor is controlled instead by directly measuring and controlling the minority carrier charge stored in the transistor. When a bipolar transistor switches from an ON (or conduction) state to an OFF state, the minority carrier charge in the transistor is removed through the base of the transistor in the form of a reverse base current transient pulse that occurs each time that the transistor turns off. In accordance with this invention, saturation of a transistor is controlled by measuring the minority carrier charge itself stored during the conduction state (above a threshold) directly as it is removed in the reverse direction through the base when the transistor is tuning off. The measured value of the stored minority carrier charge (above the threshold) that is removed when the transistor is turning off is then used by a control circuit to control the base drive current when the transistor is next switched again to the conduction state. The control circuit controls the base drive current such that the minority carrier charge that becomes stored during the conduction state (and removed when the transistor is again turned off) is maintained at a substantially constant level that can be set. Since the stored minority carrier charge is being measured directly and is also being inserted and removed as a component of the base drive current (current being a flow of charge per unit of time), the sensed par
Cunningham Terry D.
U.S. Philips Corporation
LandOfFree
Driving scheme for bipolar transistors does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Driving scheme for bipolar transistors, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Driving scheme for bipolar transistors will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2923137