Sunteți pe pagina 1din 8

THAT Corporation Design Note 138

Configuring Gain with the THAT1510 & THAT1512


THAT Corporation’s 1510 and 1512 are low noise, wide bandwidth microphone preamplifiers
available in several different industry-standard packages. They allow designers to upgrade
existing designs to take advantage of the superior performance of these new ICs. Both parts
include on-board laser-trimmed resistors which allow differential gain to be set with a single
external resistor. They differ from each other primarily in their internal gain structure, and
their resulting noise performance.
The 1510 is a pin-for-pin, drop-in replacement for the Analog Devices SSM2019, and the
now discontinued SSM2017, as well as the Texas Instruments INA217 audio preamplifiers.
The 1510 noise and gain characteristics are equal to or better than those of these other
parts, but the 1510 also improves on their distortion, bandwidth, and slew rate.
The 1512 is also a pin-for-pin replacement for the SSM2019/2017 and INA217, but due to
its different gain structure, produces significantly lower noise at low gains and requires
different external resistors to reach the same gain. Neither the 1510 nor the 1512 duplicates
the front-end output terminals available on the TI INA163, but in instances when these
outputs are not used, which is often true in audio applications, the 1512 can improve signifi-
cantly on the INA163 noise performance (as well as on distortion, bandwidth, and slew rate)
while producing a similar gain characteristic without other circuit modifications. It is possi-
ble to gain the noise advantages of the 1512 in substitution for the SSM2019/2017 and the
INA217, but the resulting gain characteristics may be different enough to require other
circuit modifications for some applications.
This design note is intended to offer some guidance on gain control to designers who are
considering the 1510 or 1512, first for new designs, but also for replacements in existing
circuits.
Controlling Gain
Input Stage
As shown in Figure 1, the 1510 and 1512 V+
Output Stage
use a three-amplifier instrumentation ampli- 5k
fier (IA) topology. This configuration has the 5k
(10k)
advantage of not amplifying common mode -AV -
Out
voltages as it amplifies signal, and as such, -AV
+

one can achieve excellent common mode +In


5k
(10k)
rejection ratios (CMRR) that increase with -In 5k
the gain setting. RA
5k Ref
Additionally, the input stage of these RG
RG1
RB
parts uses an unconventional arrangement. RG2 5k
The input amplifiers are configured to provide
current feedback via resistors RA and RB to V-
the emitters of the input transistors. These
nodes are brought out to pins RG1 and RG2.
This allows the differential gain of the input Figure 1. 1510/1512 Equivalent Circuit
section to be controlled by the external resis- (THAT1512 values shown in parentheses).
tor RG.
In order to reduce input voltage noise at low gains, the configuration of the 1512 differs
from that of the 1510. The 1512 output stage has 6 dB less gain than that of the 1510. This
leads to different gain equations for the two parts as follows.
10k 10k
For the 1510: A V = 1 + RG , or R G = AV − 1 .

Copyright © 2005 - 2011 Design Note 138


by THAT Corporation Page 1 of 8
All rights reserved 45 Sumner St; Milford, MA 01757 USA; www.thatcorp.com; info@thatcorp.com Doc 600064 Rev 00G
THAT Corporation Design Note 138 Configuring Gain with the THAT 1510 & 1512

RG 1510 Gain 1512 Gain Gain 1510 RG 1512 RG


∞ 0.0 dB -6.0 dB -6 dB Not possible ∞
10 kΩ +6.0 dB 0 dB 0 dB ∞ 10,000 Ω
5 kΩ +9.5 dB +3.5 dB +6 dB 10,048 Ω 3,344 Ω
1 kΩ +20.8 dB +14.8 dB +10 dB 4,625 Ω 1,878 Ω
500 Ω +26.4 dB +20.4 dB +20 dB 1,111 Ω 526 Ω
100 Ω +40.1 dB +34.1 dB +30 dB 327 Ω 161 Ω
50 Ω +46.1 dB +40.0 dB +40 dB 101 Ω 50 Ω
10 Ω +60.0 dB +54.0 dB +50 dB 32 Ω 16 Ω
5Ω +66.0 dB +60.0 dB +60 dB 10 Ω 5Ω
Table 1. Gain (rounded to the nearest 0.1dB) vs RG for Table 2. RG (rounded to the nearest 1Ω) vs Gain for
the 1510 and 1512. the 1510 and 1512.
5 k 5 k
And, for the 1512: A V = 0.5 + RG , or R G = A V − 0.5 ; where AV is the voltage gain, and RG is
in ohms.
The gain equations for the 1510 are identical to those for the SSM2017, SSM2019, and
the INA217. This makes the 1510 convenient for drop-in replacement of these parts, offering
a true second source with improved performance, and without requiring any redesign effort.
The gain equations for the 1512 are unique. Because of this, existing circuits may require
some adjustment to substitute the 1512 for a 1510 or other manufacturers’ preamp ICs.
Table 1 compares the gain of the 1510 with that of the 1512 for various values of RG, while
Table 2 shows RG for various desired gains for both parts.
Basic Potentiometer-Based Variable Gain Control
Figure 2 shows a simplified application circuit for the 1510 and 1512 using a potentiome-
ter to control gain. This circuit, and all the others shown in this design note, focuses on
gain control to the exclusion of other necessary features. For example, it does not show the
input RFI protection, phantom power and associated fault protection, input ac-coupling, or
power supply bypassing essential to implement a mic preamp with ICs available on the

Figure 2. Simplified application circuit for the 1510/1512

Copyright © 2005 - 2011 Design Note 138


by THAT Corporation Page 2 of 8
All rights reserved 45 Sumner St; Milford, MA 01757 USA; www.thatcorp.com; info@thatcorp.com Doc 600064 Rev 00G
THAT Corporation Design Note 138 Configuring Gain with the THAT 1510 & 1512

market today. For complete information on circuit configurations for these ICs, see THAT’s
1510/1512 data sheet.
Figure 3 shows how the decibel gain of the circuit of Figure 2 varies with pot rotation
using a 1510 and various 10 kΩ pots for RGV. The pots are all reverse audio (log) taper, using
a two-segment piece-wise linear approximation to the log function; the tapers vary from
2.5% to 20%. The curves assume negligible wiper- and end-resistance for RGV. RGF is 10 Ω. RGF
limits the minimum resistance between RG1 and RG2 to 10 Ω, which limits the maximum gain
to ~ 60 dB. RGV, at 10 k, sets the maximum resistance between RG1 and RG2 to ~ 10 k, which
limits minimum gain to ~ 6 dB for the 1510.
The 2.5% pot may look the most nonlinear on the plot, but in actuality, it is the best fit
for a straight line approximation between 6 dB and 60 dB. With the 20% taper pot, the last
30 degrees of rotation results in over a 30 dB change in gain from ~28 dB to 60dB,
whereas the 2.5% taper has only a 15 dB variation in gain (from ~45 dB to 60 dB) over this
same angle. This makes settings much less sensitive at high gains.

60 60
50 50
Gain (in dB)
Gain (in dB)

40 40
2.5%
5% 2.5%
30 30
10% 5%
20 10%
20 20%
20%
10 10

0 0
0 60 120 180 240 300 0 60 120 180 240 300
30 90 150 210 270 30 90 150 210 270
Rotation (in deg) Rotation (in deg)

Figure 3. Gain trajectories for the circuit of Fig. 2, Figure 4. Gain trajectories for the circuit of Fig. 2,
1510 with various pot tapers, RGV=10kΩ, RGF=10Ω. 1512 with various pot tapers, RGV=10kΩ, RGF=10Ω.

Figure 4 shows the same family of curves, but for the 1512. In this case, RGV remains 10
kΩ , but RGF is 5 Ω; for the 1512, this limits maximum gain to approximately 60 dB. Note
that with the 10 kΩ pot, the 1512 reaches a minimum of 0 dB gain. Also, the wider gain range
increases the sensitivity of gain change vs. rotation in the last 30 degrees of rotation (at
higher gains).
Note that the 1512 will produce identical curves of gain vs. pot setting as those of
Figure 3 if RGV were 5 kΩ, and RGF were 5 Ω.
SSM2017 / SSM2019 / INA217 to 1510
The 1510 is a direct replacement for the SSM2017, SSM2019, and INA217. No changes
are required in existing circuits to take advantage of the improved distortion, bandwidth,
and (in some cases) noise performance of the 1510 compared to these other parts.

Copyright © 2005 - 2011 Design Note 138


by THAT Corporation Page 3 of 8
All rights reserved 45 Sumner St; Milford, MA 01757 USA; www.thatcorp.com; info@thatcorp.com Doc 600064 Rev 00G
THAT Corporation Design Note 138 Configuring Gain with the THAT 1510 & 1512

V+ V+

In- In-
Vin- Vin- U1
RGF Vo1 RGF
R1 RG1 Vcc R1 RG1 Vcc
2k2 CW 5R Out 2k2 CW 5R Out
Sen
RGV C1 INA163 Vo RGV C1 THAT1512 Vo
2k Ref 2k Ref
RG2 Vee RG2
Vo2 U1 Vee
6800u/10V 6800u/10V
Vin+ Vin+
In+ R2 In+ R2
2k2 2k2
V- V-

Figure 5. Example application circuit for the INA163, Figure 6. Substituting a 1512 for an INA163 in
+12 ~+60 dB gain. the circuit of Figure 5.

INA163 to 1512 - Without Any Circuit Changes


While the 1512 and the INA163 have different gain equations, for many applications the
1512 may be directly substituted for the INA163. In such cases, noise at low gains can be
significantly improved, and the characteristic of gain vs. pot rotation is only slightly
changed. With this substitution, noise at
high gain will be nearly identical, but noise at
60 3 minimum gain will improve for the 1512
compared to the INA163.
50 2.5
INA163 Gain The gain equations for the INA163 are:
(dotted)
40 2 A V = 1 + 6Rk , or R G = A6Vk
dB Difference

G − 1 , which are
dB Gain

30 1.5 similar to those for the THAT1512.


Difference
Figure 5 shows an example application
20 1 circuit for the INA163. RGV is a 2 kΩ, 5%
THAT1512 Gain reverse audio taper pot (e.g., the Alps “Rev.
10 0.5
D” taper). RGF is 5 Ω. With these
0 0 resistances, gain varies from +12 to +62 dB.
0 60 120 180 240 300
C1, at 6800 uF, avoids changes in dc output
Rotation (in deg) offset with gain. Note that VO1 and VO2 (the
input stage outputs at pins 1 and 14,
respectively), are not connected.
Figure 7. Gain vs. Pot Rotation for the circuit of As shown in Figure 6, the 1512 may be
Figure 5 compared to Figure 6. substituted directly for the INA163 in this
circuit. Figure 7 shows the gain trajectories
of these two circuits along with the differ-
ence in gain that results from this direct
23 35 37 21 33 35
substitution. At low gains there is about
18 39 16 38 2.5 dB of error, and at higher gains, the
15 42 13 41 error drops to about 1.5 dB.
13 48 11 46 Figure 8 shows the theoretical front
12 62 10 60 panel scaling for these two circuits. The
primary difference between the two gain
INA163 THAT1512 trajectories is a relatively constant ~2 dB. If
this error is unacceptable, it could be
Figure 8. Theoretical front panel scales for
adjusted by changing the gain of a subse-
INA163 (Figure 5) and 1512 (Figure 6).
quent stage.

Copyright © 2005 - 2011 Design Note 138


by THAT Corporation Page 4 of 8
All rights reserved 45 Sumner St; Milford, MA 01757 USA; www.thatcorp.com; info@thatcorp.com Doc 600064 Rev 00G
THAT Corporation Design Note 138 Configuring Gain with the THAT 1510 & 1512

V+ V+

In- In-
Vin- Vin- U1
RGF Vo1 RGF
R1 RG1 Vcc R1 RG1 Vcc
2k2 CW 5R Out 2k2 CW 5R Out
Sen
RGV C1 INA163 Vo RGV
C1 THAT1512 Vo
6k Ref 6k Ref
RG2 Vee RG2
Vo2 U1 Vee
6800u/10V 6800u/10V
Vin+ Vin+
In+ R2 In+ R2
2k2 2k2
V- V-

Figure 9. INA163 circuit for +6 ~+60 dB gain. Figure 10. Substituting a 1512 for the INA163 in Figure 9.

Figure 9 shows an INA163 configured for minimum gain of +6 dB rather than +12 as
shown in Figure 5. Figure 10 shows the same circuit with a 1512 directly substituted.
Figure 11 shows the gain trajectories of these two circuits, while Figure 12 shows the front
panel scaling for each circuit.
INA163 to 1512 - Optimized Circuitry
In some cases, designers may feel that the 60 4.5
gain trajectory changes with the above circuits
are unacceptable despite the small size of the 50 4
gain differences. By changing to a more flexible
topology it is possible to keep the gain trajecto- 40 3.5

dB Difference
INA163 Gain
ries closer by changing some passive compo-
dB Gain

(dotted)
nents depending on which IC is used. The 30 3
approach shown does not require any change in THAT1512 Gain
the pot. 20 2.5
Figure 13 shows a circuit with minimum gain 10 2
of +12 dB (like Figures 5 through 8) optimized Difference
for both the INA163 and the 1512. For the 0 1.5
INA163, RGF would be a 6 Ω resistor, and for the 0 60 120 180 240 300
1512, 5 Ω. This change adjusts maximum gain to Rotation (in deg)
the same point: 60 dB, for each part. RGV is 2kΩ
for both ICs. R3 is 10 kΩ for the 1512 and
omitted for the INA163. This change adjusts Figure 11. Gain vs. Pot Rotation for the circuit of
minimum gain to the same point: +12 dB. Figure 9 compared to Figure 10.
Figure 14 shows the gain trajectories and
differences in this new circuit configuration.
Note that the maximum error in this configura-
14 26 28 12 25 26
tion is less than 1.5 dB -- about equal to the
10 30 8 29
minimum error in the previous arrangement.
Figure 15 shows theoretical front panel scalings 8 34 5 32
for this arrangement. Note that the markings 7 39 4 38
are nearly identical between the INA163 and the 6 62 2 60
1512.
INA163 THAT1512
Figure 12. Theoretical front panel scales for
INA163 (Figure 9) and 1512 (Figure 10).

Copyright © 2005 - 2011 Design Note 138


by THAT Corporation Page 5 of 8
All rights reserved 45 Sumner St; Milford, MA 01757 USA; www.thatcorp.com; info@thatcorp.com Doc 600064 Rev 00G
THAT Corporation Design Note 138 Configuring Gain with the THAT 1510 & 1512

Minimum Gain of +6dB


V+
To achieve a minimum gain of +6 dB
In-
Vin-
simply change RGF and R3. For the INA163,
U1
R1
RGF
RG1 Vcc RGF would be 6 Ω, and for the 1512, 5 Ω. RGV
2k2
RGV
CW

C3
R3
THAT1512 Vo
Out
changes to 6 kΩ for both ICs. R3 is 6 kΩ
RG2
Vee
Ref for the 1512 and omitted for the INA163.
In+ 6800u/10V Vin+ Figure 16 shows the gain trajectories and
R2
differences for this configuration. Note
2k2
V-
that the error here swings from ~ -0.5 dB
to ~ +1.5 dB. Figure 17 shows theoretical
front panel scalings for this arrangement.
Again, the markings are nearly identical
Figure 13. More versatile topology for 1512 to between the INA163 and the 1512.
better support INA163. See text for values..

60 3 60 1.5
50 2.5 50 1
INA163 Gain Difference
(dotted)
40 2 40
dB Difference

0.5

dB Difference
INA163 Gain
dB Gain

dB Gain
1512 Gain (dotted)
30 1.5 30 0
20 1 THAT1512 Gain
Difference 20 -0.5
10 0.5 10 -1
0 0 0 -1.5
0 50 100 150 200 250 300 0 50 100 150 200 250 300
Rotation (in deg) Rotation (in deg)

Figure 14. Gain vs. Pot Rotation for the (optimized) circuit Figure 16. Gain vs. Pot Rotation for the (optimized) circuit
of Figure 13 INA163 vs. 1512, 12 dB min gain. of Figure 13 INA163 vs. 1512, 6 dB min gain.

23 35 37 21 34 36 14 26 28 14 25 27
18 39 17 38 10 30 10 29
15 42 15 41 8 34 9 32
13 48 13 46 7 39 7 38
12 62 12 62 6 62 7 62

INA163 THAT1512 INA163 THAT1512

Figure 15. Theoretical front panel scales for the (optimized) Figure 17. Theoretical front panel scales for the (optimized)
circuit of Figure 13, INA163 vs. 1512, 12 dB min gain. circuit of Figure 13, INA163 vs. 1512, 6 dB min gain.

Copyright © 2005 - 2011 Design Note 138


by THAT Corporation Page 6 of 8
All rights reserved 45 Sumner St; Milford, MA 01757 USA; www.thatcorp.com; info@thatcorp.com Doc 600064 Rev 00G
THAT Corporation Design Note 138 Configuring Gain with the THAT 1510 & 1512

SSM2019/SSM2017/INA217 to 1512 - Without Any Circuit Changes


Within the available packages, the footprint of the 1512 is identical to that of the INA217,
the SSM2019/2017, and the 1510; accordingly, the 1512 generally can be put into circuits
designed for any of these devices. The advantages of replacing these devices with a 1512
include better bandwidth, higher slew rate, and lower distortion, as well as lower noise at low
gains. That last advantage is not available when replacing the SSM2019/2017 with the 1510.
Note that the 1510 noise, at low and high gain, is lower than that of the INA217.
Figure 18 shows a typical circuit based on the SSM2019 (or SSM2017, or INA217), config-
ured for ~ +12dB minimum gain. Assuming a 5% reverse audio (log) taper pot, the graph in
Figure 19 compares the gain trajectory of SSM2019 with that of the 1512 in the circuit of
figure 18. Below ~50 dB of gain, the difference in gain is 5~6 dB, which may be acceptable in
some applications.

V+ V+
In- In-

RGF Vin- U1 RGF Vin- U1


R1 RG1 Vcc R1 RG1 Vcc
2k2 CW 10R Out 2k2 CW Out
RGV C1 Vo RGV C3 R3 THAT1512 Vo
3k4 Ref 3k4 Ref
RG2 RG2
In+ Vee In+ Vee
6300u/10V Vin+ THAT1510 6800u/10V Vin+
SSM2019
R2 R2
2k2 INA217 2k2
V- V-

Figure 18. SSM2019 Circuit with Figure 20. Optimized topology for 1512, to support
minimum gain of +12dB. the SSM2019, INA217, and THAT1510.

60 6 60 6
Difference
50 5 50 SSM2019 5
Gain
40 40 Difference (dotted) 4
dB Difference
dB Difference

4
SSM2019 Gain
dB Gain
dB Gain

(dotted)
30 3 30 3

20 2 20 2
THAT1512 Gain
THAT1512 Gain
10 1 10 1

0 0 0 0
0 50 100 150 200 250 300 0 50 100 150 200 250 300
Rotation (in deg) Rotation (in deg)

Figure 19. Gain vs. Pot Rotation for the circuit of Figure 21. Gain vs. Pot Rotation for the (optimized) circuit
Figure 18, 1510/SSM2019/INA217 vs. 1512, 12 dB gain. of Figure 20, 1512 vs. SSM2019, ~ 12 dB min gain.

Copyright © 2005 - 2011 Design Note 138


by THAT Corporation Page 7 of 8
All rights reserved 45 Sumner St; Milford, MA 01757 USA; www.thatcorp.com; info@thatcorp.com Doc 600064 Rev 00G
THAT Corporation Design Note 138 Configuring Gain with the THAT 1510 & 1512

SSM2019/SSM2017/INA217 to 1512 - Optimized Circuitry


For those that find this degree of error unacceptable, there are several other
approaches, though they all entail some degree of change to the circuit.
Consider the circuit in Figure 20. This circuit uses a paralleled resistor similar to that
used in the optimized INA163 retrofit. Assuming gain range from +12 dB to +60 dB, the
1512 requires, RGF = 5 Ω and R3 = 2.5 kΩ. The SSM2019, INA217, and 1510 require RGF = 10 Ω,
while R3 is omitted (open). RGV is 3.4 kΩ in both cases. The resulting gain trajectory and
difference between the two part types is shown in Figure 21. Using RGF and R3, it is possible
to force the differences at the endpoints to zero, though this will not minimize the average
error.
Better results can be had by changing the value of the pot itself. Figure 22 shows a
circuit with RGF at 5 Ω and RGV as a 2 kΩ, 5% reverse audio taper pot. These changes result
in gain trajectory similar to that of Figure 18; +12 to +60 dB. See Figure 23 for the compari-
son. This approach is best for matching the gain trajectory of the SSM2017/2019 and
INA217 with a 1512. The result will be better
V+ bandwidth, higher slew rate, lower distor-
tion, and lower noise at low gains without
In-
Vin- U1
any compromise in gain trajectory or front
R1
RGF
RG1 Vcc
panel calibration.
2k2 CW 5R Out
THAT1512
RGV
2k
C1
RG2
Ref
Vo
Further thoughts
Vee
6800u/10V
Vin+ We reiterate that all the circuits shown
In+ R2 here are incomplete. They lack the input RFI
2k2
protection, phantom power and associated
V-
fault protection, input ac-coupling, and
power supply bypassing necessary to imple-
Figure 22. +12 dB min. gain with 1512, changing RGV to ment a mic preamp with these and other
match Figure 18. similar ICs. For complete information on
circuit configurations for these ICs, see
THAT’s 1510/1512 data sheet.

60 2 We hope that the above notes and


circuits will help point the way for designers
50 1 to take advantage of the performance
improvements available with THAT’s 1510
THAT1512 Gain 0 and 1512 audio preamplifier ICs, while
40
dB Difference

maintaining flexibility in sources of supply.


dB Gain

-1 Because of the wide range of variations


30 Difference possible with our and other makers’ pream-
-2 plifier ICs, we encourage designers to
20 SSM2019 contact us directly to discuss your specific
Gain(dotted)
-3
application and how to best configure your
10 circuits to take full advantage of our ICs’
great performance. You can reach us by
0 -4
e-mail at tech_support@thatcorp.com, by
0 50 100 150 200 250 300 fax at +1(508)478-0990, by phone at
Rotation (in deg) +1(508)478-9200, or on the web at
www.thatcorp.com. We look forward to
discussing your application directly with
Figure 23. Gain vs. Pot Rotation for the circuit of you.
Figure 22 compared to Figure 18, THAT1512 vs. SSM2019.

Copyright © 2005 - 2011 Design Note 138


by THAT Corporation Page 8 of 8
All rights reserved 45 Sumner St; Milford, MA 01757 USA; www.thatcorp.com; info@thatcorp.com Doc 600064 Rev 00G

S-ar putea să vă placă și