We have found by measurement (at a transfer slope of 50 mV/dB) that when the value of R17 is
selected so that the +85
0C
and -55
0
C transfer curves are superimposed, both these curves lie
below the room-temperature curve, but within about 0.6 dBm of it. Thus, the common
specification of
0.5 dBm over the temperature range -55
0
C to +85
0
C is easily met simply by
selecting the proper value of R17, without having to use RT3, S2, and RT4. For a transfer slope of
50 mV/dB, the value of R17 which accomplishes this is about 2640 ohms (for a feedback
resistance value of 2210 ohms at room-temperature). The original transfer slope may change
slightly as a result of the change in R17, so that if a very precise specific value of the slope is
required, it may be necessary to slightly adjust the feedback resistance.
The optimum value of R17 in this procedure depends on the slope of the transfer curve, and thus
on the output feedback resistance. At a transfer slope of 50 mV/dB, the optimum ratio of R17 to
the room temperature feedback resistance is about 1.19. Model calculations for other transfer
slopes indicate that this ratio should be about 1.12, 1.14, and 1.22, at transfer slopes of 20, 30,
and 60 mV/dB, respectively. Our lab measurements indicate that in practice a ratio of 1.18 is good
enough to meet the above-mentioned
0.5dBm spec for transfer slopes from 25 to 50 mV/dB, and
that 1.22 works well at 70 mV/dB. In general, if the high temp and low temp transfer curves are
essentially superimposed (at medium to high power), but lie a little below the room temp curve,
you have the correct R17 value.
8.3.7
A Summary of Temperature Compensation Procedures:
There are three stages
to the temperature compensation procedure. These three stages are:
Stage I:
Make the overall slope of the transfer curve independent of the temperature.
Stage II:
Eliminate or minimize any temperature-induced displacement of the high-
power portion of the transfer curve.
Stage III:
Eliminate or minimize any temperature-induced displacement of the low-
power end of the transfer curve.
Until the user gains familiarity with the temperature behavior of a particular DLVA configuration, it
is best to acquire the data needed for conducting these three stages of compensation by running
complete transfer curves at room temperature, and at both the upper and lower temperature
extremes for which compensation is required.
Stage I consists of the selection of the values of R15 and R20 in the output amplifier feedback
sensistor network (which is shown in Fig. 2(c).) First, any change in the transfer slope over
temperature is measured, and expressed as the percentage change in slope per 60
0C
temperature interval. This number is then used to select the required value of R20 from Fig.
8.3(a). (The curves of Fig. 8.3(a) are based on the assumption that S1 is a 500-ohm sensistor
with a positive tempco of 0.007/
0
C.) The value of R15 is then calculated from the simple formula
given in the second paragraph of page 27. As discussed further below, the same values of R15
and R20 should work for succeeding units of the same configuration.
-30A-
 |
 |
 |
 |
 |