SEMICONDUCTOR DIODE GATES 



1153 



Fig. 12 illustrates both these design ideas. One of the control inputs 

 has the control pulse impressed by means of a transformer and so has a 

 very low dc impedance without making excessive demands on the 

 control pulse generator. The biases are so adjusted that, in the disabled 

 state, all the control diodes are just on the edge of conducting except the 

 one in the transformer path. Because of the low DC impedance, practi- 

 cally all the bias current flows in this path. Thus there is no possibility 

 of hole storage except in this one diode. If a positive control pulse is 

 impressed while the potentials on the other controls are at their more 

 negative value, the bias current just switches into those control paths; 

 the output diode remains non-conducting and any spurious hole storage 



D4 



Fig. 12 — Switching type gate which minimizes "hole storage" effects. 



current from the diode in the transformer control also goes into the other 

 control paths. Since there is no storage in the other control paths, posi- 

 tive pulses may be simultaneously impressed on all the controls, the 

 diodes in all but the transformer control path will immediately become 

 high impedance and any hole storage current from the one diode will 

 harmlessly add to the bias current flowing into the output. 



Junction diodes are used in all the control inputs except the one 

 with a transformer. A point contact unit is used here to minimize the 

 hole storage. A point contact unit is also used in the output position. This 

 is a critical location where good diode action is more important than a 

 very high discrimination. 



There is an additional advantage, in this configuration, that the rela- 

 tively large bias current flows in the control generators only very 

 briefly — while the disabled gate is being pulsed by the transformer 

 control. 



