switch (on demand from the deck) fixed resistors into the device for 

 frequency calibration. This calibration will provide a determination 

 of drift and increase the accuracy of the measurement. No attempt has 

 been made thus far to linearize the output nor has any study been made 

 on utilizing matched thermistors. The unit will be calibrated as a 

 whole and subsequent units also will be calibrated individually. 



During the experimental investigation on the circuitry it was found 

 that the circuit was sensitive to a great number of things. For example, 

 the changing battery impedance affected the frequency. This effect 

 was almost entirely eliminated by inserting a 100 mf. capacitor across 

 the battery supply. The locations of individual components and the shape 

 of the finished unit appear to have considerable effect on the stability. 

 Miniature electrolytic capacitors now used in the circuitry no doubt 

 play a part in the drift. It is planned to replace these with either 

 tantalytic or silverlytic types now on order. Since this particular 

 circuit is so frequency- sensitive to changes in level, even transistors 

 must be selected which will give the steadiest performance. 



It is planned to have the first unit ready for sea tests next month. 

 The circuit is complete and is awaiting construction of a housing. 

 Bead thermistors in glass probes are on order to be used with this 

 model. Attempts at insulating on-hand thermistors (wafer type) have 

 proven unsatisfactory so far, in that the thermal time constant is 

 increased considerably. 



This frequency modulated system, if proven stable and accurate, 

 can be utilized with a variety of other transducers. In this way a certain 

 amount of compatibility in data recording and processing may be 

 realized. The experience gained in this pilot study will be extremely 

 helpful in the future development of frequency sensitive circuits and 

 systems. 



Encl. (1) to Code 5401-BEO/bah memo of 23 Jul 1958 



B-6 



