EPH KONIGSBERG 
1147 
Several comments on the above tabulation are 
in order. The drifts noted above are not random 
in nature ; except in rare cases, they tend to be 
unidirectional and hence predictable. Further, 
the researcher will usually reserve the more 
stable instruments for very long term or criti- 
cal low pressure experiments. Indeed, the very 
unidirectionality of the drifts, demonstrating 
the systematic nature of the drift phenomenon, 
has led us to the redesign of the instruments to 
minmize this effect. Such tests are now in prog- 
ress, and the short term (weeks) results are 
encouraging. Megohm leakage to ground has 
been measured on four instruments (5000 Ohm 
strain gages) which were in an animal for three 
years. The lowest value was 20 Megohms — 
which included the effect of a leaky cable. The 
sealed transducers, after being cleaned off and 
washed, had a low value of 100 Megohms, from 
strain gage to case ; all units were operable. We 
believe newer cable designs of the last two 
years have overcome this weak point, as no 
comparable leaking has been reported in this 
shorter (two year) period of time. 
We next consider the interaction of trans- 
ducers and body. It is reasonable to state that 
leaking of transducers are virtually nil for the 
period of time they have been used. Transducers 
manufactured over five years ago have been 
sent in for repair, and are still functional. Some 
of these transducers have only been in the body 
intermittently. The analysis of the four trans- 
ducers which had been implanted for three years 
corroborates the results observed on transducers 
sent in for repair. 
There have been "Megohm" failures on both 
earlier and recent instruments. (We can con- 
sider an electrical leakage to ground as evi- 
dence of a moisture leakage to the interior of 
the transducer.) Recent failure analysis sup- 
ports the view that such electrical leakage is 
the product of either poor design, improper as- 
sembly, or a combination of both, and should 
not be expected to occur. 
Wire breakage and cable leakage should be 
virtually nil. We have only two years of ex- 
perience with stainless steel cable, but in that 
period of time all failures have been attributable 
to improper installation. These almost invar- 
iably occur near the stem of the transducer, 
where ligatures have been tied around the 
cable, so as to secure it and position the trans- 
ducer in the proper place in the left ventricle. 
Such ligatures can severely bend the wire 
which, upon repeated flexing, breaks. The addi- 
tion of a reinforcement to the cable at that 
point, or greater care in tightening the ligatures 
around the cable will prevent this failure mode. 
There has been some pin breakage with the 
transducer cable connectors in common use. 
Greater care, or the use of the new polarized 
connectors now available can minimize this 
relatively minor problem. 
The influence of implantable transducers on 
the experimental animal are not well reported. 
We have not heard of great problems with clot- 
ting: the influence of surgical technique is pro- 
nounced. There are no systematic studies of the 
relative utility of coating transducers with 
heparinized colloidal graphite: both coated and 
uncoated transducers have been used with good 
results. The advantages of silicone versus PVC 
cables are equally unresolved. It has been re- 
ported that silicone seems to generate less rubor, 
but both types of cable have been used for long 
lengths of time. 
The remaining principal problem, for sys- 
tems which are not totally implanted, seems to 
be the design of the transducer cable connector. 
Where the usage pattern requires a long term 
experiment, where infection problems dictate 
that the end of the cable be buried under the 
skin, when the cable is first extracted from be- 
neath the skin, used, disinfected (attemptedly) 
and then reburied, the design of the connector 
is such that infection often takes place, prob- 
ably because there are reentrant crevices in the 
connector end of the cable. 
SUMMARY 
This paper has concerned itself with present 
performance of transducers (on bench test) and 
past performance of transducers on life test. 
On the basis of present laboratory results, we 
believe it to be realistic to predict that lifetimes 
of three years for both transducer and cable are 
currently attainable, and that drift rates (for 
both zero and sensitivity) considerably less than 
those quoted will be attainable within this year. 
