zo tester collin wells. original zo tester original aol circuit
TRANSCRIPT
Zo Tester
Collin Wells
Original Zo Tester
Original AOL Circuit
-
++
4
3
5
1
2
U1 OPA376
V1 2.5
V2 2.5
R2 100kR1 100k
R3
100
R4
100k
C1 100u+
Vout
Vreference
Vtest
VM
Aol = 1000 (Vtest/Vreference)
Original AOL Circuit Limitations
• #1 DUT Zo and feedback resistor form resistor divider causing errors.
• #2 DUT is not AC Coupled causing errors in the AOL curve.
• #3 Not capable of 50Ohm Drive
Example of AOL Limitations #1• DUT Zo Causing Errors
T
VM
Vreference
Vtest
Frequency (Hz)
1 10 100 1k 10k 100k 1M 10M
Ga
in (
dB
)
-140
-120
-100
-80
-60
-40
-20
0
VM
Vtest
Vreference
This region should continue to roll off at -20dB/decade
T
Computed Aol
Frequency (Hz)
1 10 100 1k 10k 100k 1M 10M
Ga
in (
dB
)
0
20
40
60
80
100
120
140
Computed Aol
Example of AOL Limitation #1• DUT Zo Causing Errors
R1 100kR
4 10
0kR
3 10
0R2 100k
Ro 200
V1 1
Vtest
VM
Vreference
Vtest limitation at high frequency due to
Ro and values of R1, R2, R3, R4
Minimum Gain at High Frequency:
Vtest/V1 = 20 log (Vout) for V1 = 1
For R1=R2=R4=100k, R3=100 --> -63.53dB
Explanation of AOL Limitation #1• DUT Zo Causing Errors
-
++
4
3
5
1
2
U1 OPA376
V1 2.5
V2 2.5
R2 100kR1 100k
R3
100
R4
100k
C1 100u+
Vout
Vreference
Vtest
VM
-
+
-
+VCVS1 1
Aol = 1000 (Vtest/Vreference)
Solution to AOL Limitation #1• DUT Zo Causing Errors
T
VM
Vreference
Vtest
Frequency (Hz)
1 10 100 1k 10k 100k 1M 10M
Vo
ltag
e (
V)
-140
-120
-100
-80
-60
-40
-20
0
Vtest
Vreference
VM
Solution to AOL Limitation #1• DUT Zo Causing Errors
T
Computed Aol
Frequency (Hz)
1 10 100 1k 10k 100k 1M 10M
Ga
in (
dB
)
-20
0
20
40
60
80
100
120
140
Computed Aol
Solution to AOL Limitation #1• DUT Zo Causing Errors
Solution to AOL Limitation #1• DUT Zo Causing Errors
– Actual Implementation
+Vbuffer
-Vbuffer
+Vbuffer -Vbuffer
-
++
4
3
5
1
2
U1 OPA376
V1 2.5
V2 2.5
R2 100kR1 100k
R3
100
R4
100k
C1 100u
+
Vout
Vreference
Vtest
VM
-
+ + U2 ths4631
R9 499
R10 499
R11 49.9
C4 8p
R5 49.9
V3 15 V4 -15
R6
49.9
Aol = 1000 (Vtest/Vreference)
+6dB
-6dB
Solution to AOL Limitation #1• DUT Zo Causing Errors
– Actual Implementation
T
Frequency (Hz)
1.00 10.00 100.00 1.00k 10.00k 100.00k 1.00M 10.00M 100.00M 1.00G
Ga
in (
dB
)
-40.00
-20.00
0.00
20.00
40.00
60.00
80.00
100.00
120.00
140.00
Example of AOL Limitation #2• DUT not AC coupled
– Even small offsets in the DUT Vcm away from (Vcc(+) - Vcc(-))/2 result in degradation of AOL at low frequencies
101
102
103
104
105
106
107
108
-40
-20
0
20
40
60
80
100
120
140
Frequency (Hz)
dB
Aol
-0.25V Offset
101
102
103
104
105
106
107
108
-40
-20
0
20
40
60
80
100
120
140
Frequency (Hz)
dB
Aol
Example of AOL Limitation #2• DUT not AC coupled
– Even small offsets in the DUT Vcm away from (Vcc(+) - Vcc(-))/2 result in degradation of AOL at low frequencies
-0.5V Offset
101
102
103
104
105
106
107
108
-40
-20
0
20
40
60
80
100
120
140
Frequency (Hz)
dB
Aol
Example of AOL Limitation #2• DUT not AC coupled
– Even small offsets in the DUT Vcm away from (Vcc(+) - Vcc(-))/2 result in degradation of AOL at low frequencies
-1V Offset
101
102
103
104
105
106
107
108
-40
-20
0
20
40
60
80
100
120
140
Frequency (Hz)
dB
Aol
Example of AOL Limitation #2• DUT not AC coupled
– Even small offsets in the DUT Vcm away from (Vcc(+) - Vcc(-))/2 result in degradation of AOL at low frequencies
-1.5V Offset
101
102
103
104
105
106
107
108
-60
-40
-20
0
20
40
60
80
100
120
140
Frequency (Hz)
dB
Aol
Example of AOL Limitation #2• DUT not AC coupled
– Even small offsets in the DUT Vcm away from (Vcc(+) - Vcc(-))/2 result in degradation of AOL at low frequencies
Railed
Example of AOL Limitation #2• DUT not AC coupled
– Although overall AC gain is “1”, DC Gain is 1+((100k+100k))/100 = ~2000V/V– Therefore any part with a Vos of >1.25mV will rail.
R2 100k
R3
100
R4
100k
Vreference
Vtest
VM
R1 100k
+
Vout
V3 1.25m
-
+
IOP1
Cap @ DC
1.25mV
2.5V
1.25V
Solution to AOL Limitation #2• DUT not AC coupled
– AC Couple the DUT with a 100uF capacitor. This will reduce the overall system offset to just the Vos of the DUT.
– It will require time when the system initially starts up for the cap to charge and cancel out the offset.
R2 100k
R3
100
R4
100k
Vreference
Vtest
VM
R1 100k
+
Vout
V3 1.25m
-
+
IOP1
Cap @ DC
C1 100u
1.25mV
1.25mV
1.25mV
Example of AOL Limitation #3• Not Capable of 50 Ohm Drive
High-Z Node
+Vbuffer
-Vbuffer
+Vbuffer -Vbuffer
-
++
4
3
5
1
2
U1 OPA376
V1 2.5
V2 2.5
R2 1MR1 1M
R3
1kR
4 10
k
C1 100u
+
Vout
Vreference
Vtest
VM
-
+ + U2 ths4631
R9 499
R10 499
R11 49.9
C4 8p
R5 49.9
V3 15 V4 -15
R6
49.9
C2
100u
Aol = R4/R3 (Vtest/Vreference)
+6dB
-6dB
Final AOL Circuit
+Vbuffer
-Vbuffer
+Vdut
-Vdut
+Vdut -Vdut
+Vbuffer -Vbuffer
+Vbuffer
-Vbuffer
-
++
4
3
5
1
2
U1 OPA376
R2 1MR1 1MC1 100u
+
Vout
VM
Vreference
R3
10k
R4
1k
C2
100u
R5 499
R6 499
-
+ + U2 ths4631
R9 499
R10 499
R11 49.9
R12 49.9
R7
49.9
V1 2.5 V2 -2.5
V3 15 V4 -15
-
+ + U3 ths4631
C4 8p
C3 8p
C5
28p
VF1
R8
49.9
C6
28p
R13 499
Aol = (R3/R4) (Vtest/Vreference)
Gain/Phase Analyzer
50R Termination +
Shunt Input Cap
+6dB
-6dB
Gain/Phase Analyzer
50R Termination +
Shunt Input Cap-6dB
+6dB
Final AOL Circuit ResultsT
Frequency (Hz)
100.00m 1.00 10.00 100.00 1.00k 10.00k 100.00k 1.00M 10.00M 100.00M 1.00G
Gai
n (d
B)
-60.00
-40.00
-20.00
0.00
20.00
40.00
60.00
80.00
100.00
120.00
140.00
160.00
Final AOL Circuit ResultsT
Frequency (Hz)
100.00m 1.00 10.00 100.00 1.00k 10.00k 100.00k 1.00M 10.00M 100.00M 1.00G
Gai
n (d
B)
-60.00
-40.00
-20.00
0.00
20.00
40.00
60.00
80.00
100.00
120.00
140.00
160.00
High-Pass Effects ofAC Coupling CapFc = 1/2πRC
Effects of DUT Cin andZo of Buffer Circuits
Final AOL Circuit Results
101
102
103
104
105
106
107
108
-40
-20
0
20
40
60
80
100
120
140
Frequency (Hz)
dB
Aol
Final AOL Circuit Limitations
Final AOL Circuit Limitations
Final AOL Circuit Limitation Partial Fix
+Vbuffer
-Vbuffer
+Vdut
-Vdut
+Vdut -Vdut
+Vbuffer -Vbuffer
+Vbuffer
-Vbuffer
-
++
4
3
5
1
2
U1 OPA376
R2 100kR1 100kC1 100u
+
Vout
VM
Vreference
R3
100k
R4
1k
C2
100u
R5 499
R6 499
-
+ + U2 ths4631
R9 499
R10 499
R11 49.9
R12 49.9
R7
49.9
V1 2.5 V2 -2.5
V3 15 V4 -15
-
+ + U3 ths4631
C4 8p
C3 8p
C5
28p
VF1
R8
49.9
C6
28p
Aol = (R3/R4) (Vtest/Vreference)
Gain/Phase Analyzer
50R Termination +
Shunt Input Cap
+6dB
-6dB
Gain/Phase Analyzer
50R Termination +
Shunt Input Cap-6dB
+6dB
False SummingJunction Gain Changed to 40dB
Change to 100k
Final AOL Circuit Limitation Partial Fix
Final AOL Circuit Results
Original Zout Circuit
V+
V-V-
V+
-
+ +U1: DUT
RG1
RF1
RS1
R1 50
V-
V+
C4 100u
+
Voutput
+-
C5 10u
+-
C6 10u
Vtest
Vreference
RNI1 0
I
I=Vref-Vtest
RS1
Zout=Vtest
I=
RS1Vtest
Vref-Vtest
Original Zout Circuit Limitations
• Not capable of 50 Ohm Drive• DUT is not AC Coupled so DC errors occur• Instrument output gets divided down by
parallel combination of 50 Ohm termination resistor and RS in series with DUT Zo.
• System noise floor corrupts low-frequency data
Example of Zout Limitation• DUT not AC Coupled
– Zout test based on the fact that both FETs in the output stage need to be biased to mid-supply with ½ Iab flowing through them.
V+
V-V-
V+
-
+ +U1: DUT
RG1
RF1
RS1
R1 50
V-
V+
C4 100u
+
Voutput
+-
C5 10u
+-
C6 10u
Vtest
Vreference
RNI1 0
I
I=Vref-Vtest
RS1
Zout=Vtest
I=
RS1Vtest
Vref-Vtest
Solution to Zout Limitation• DUT not AC Coupled
– AC Couple the Zout circuit with a 100uF Capacitor
+Vdut
-Vdut
C1 100u
+
Vout
R2 4.99k
R3 100k
R4 499 C2 100u
-
++
4
3
5
1
2
U1 OPA376
Example of Zout Limitation• Divided down Instrument Output
– Although this issue does not cause errors in the measurement system it does limit the amount of signal that we inject into the DUT therefore limiting the output of the test.
+Vdut
-Vdut
+
Vout
RS 75
R3 100k
R4 499 C2 100u
R1
49.9 Zout 10
-
++
DUT
Example of Zout Limitation• Divided down Instrument Output
– Although this issue does not cause errors in the measurement system it does limit the amount of signal that we inject into the DUT therefore limiting the output of the test.
RS = 80600
101
102
103
104
105
106
107
108
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
Frequency (Hz)
Out
put
Vol
tage
(V
)
Gain-Phase Analyzer Output
Example of Zout Limitation• Divided down Instrument Output
– Although this issue does not cause errors in the measurement system it does limit the amount of signal that we inject into the DUT therefore limiting the output of the test.
RS = 4990
101
102
103
104
105
106
107
108
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
Frequency (Hz)
Out
put
Vol
tage
(V
)
Gain-Phase Analyzer Output
Example of Zout Limitation• Divided down Instrument Output
– Although this issue does not cause errors in the measurement system it does limit the amount of signal that we inject into the DUT therefore limiting the output of the test.
RS = 698
101
102
103
104
105
106
107
108
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
Frequency (Hz)
Out
put
Vol
tage
(V
)
Gain-Phase Analyzer Output
Example of Zout Limitation• Divided down Instrument Output
– Although this issue does not cause errors in the measurement system it does limit the amount of signal that we inject into the DUT therefore limiting the output of the test.
RS = 80.6
101
102
103
104
105
106
107
108
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
Frequency (Hz)
Out
put
Vol
tage
(V
)
Gain-Phase Analyzer Output
Solution to Zout Limitation• Divided down Instrument Output
– 50 Terminate the gain/phase analyzer output and send into the input of a THS4631 high-speed amplifier.
+Vdut
-Vdut
+Vbuffer -Vbuffer
+Vbuffer
-Vbuffer
R3 100k
R4 499 C2 100u
R1 24.9
Zout 10
-
++
DUTC1 100u
+
VG1
V3 15 V4 -15
R2 499
R13 499
-
+ + U4 ths4631
C7 8p
R5 75
R6 24.9
Vtest
Vreference
+6dB
-6dB
Solution to Zout Limitation• Divided down Instrument Output
– 50Ohm Terminate the gain/phase analyzer output and send into the input of a THS4631 high-speed amplifier.
RS = 80600
101
102
103
104
105
106
107
108
0.5
1
1.5
2
2.5
3
Frequency (Hz)
Out
put
Vol
tage
(V
)
Gain-Phase Analyzer Output
Solution to Zout Limitation• Divided down Instrument Output
– 50Ohm Terminate the gain/phase analyzer output and send into the input of a THS4631 high-speed amplifier.
RS = 4990
101
102
103
104
105
106
107
108
0.5
1
1.5
2
2.5
3
Frequency (Hz)
Out
put
Vol
tage
(V
)
Gain-Phase Analyzer Output
Solution to Zout Limitation• Divided down Instrument Output
– 50Ohm Terminate the gain/phase analyzer output and send into the input of a THS4631 high-speed amplifier.
RS = 698
101
102
103
104
105
106
107
108
0.5
1
1.5
2
2.5
3
Frequency (Hz)
Out
put
Vol
tage
(V
)
Gain-Phase Analyzer Output
Solution to Zout Limitation• Divided down Instrument Output
– 50Ohm Terminate the gain/phase analyzer output and send into the input of a THS4631 high-speed amplifier.
RS = 80.6
101
102
103
104
105
106
107
108
0.5
1
1.5
2
2.5
3
Frequency (Hz)
Out
put
Vol
tage
(V
)
Gain-Phase Analyzer Output
Example of Zout Limitation• Zo Circuit not capable of 50 Ohm Drive
– 50Ohm termination on Vreference re-creates the previous problem of dividing down the input signal amplitude.
– 50Ohm on Vtest adds DC load to the DUT, which will affect Zo.
+Vdut
-Vdut
+Vbuffer -Vbuffer
+Vbuffer
-Vbuffer
R3 100k
R4 499 C2 100u
R1 24.9
Zout 10
-
++
DUTC1 100u
+
VG1
V3 15 V4 -15
R2 499
R13 499
-
+ + U4 ths4631
C7 8p
R5 75
R6 24.9
Vtest
Vreference
+6dB
-6dB
Solution to Zout Limitation• Zo Circuit not capable of 50 Ohm Drive
– Add THS4631 buffers to both Vreference and Vtest nodes.
+Vdut
-Vdut
+Vdut -Vdut
+Vbuffer -Vbuffer
+Vbuffer
-Vbuffer
+Vbuffer
-Vbuffer
+Vbuffer
-Vbuffer
C1 100u
+
Vout
VreferenceR5 499
R6 499
R12 49.9
R7
49.9
V1 2.5 V2 -2.5
V3 15 V4 -15
-
+ + U3 ths4631
C3 8p
C5
28p
VF1
R8
49.9
C6
28p
R1 499
R13 499
-
+ + U4 ths4631
C7 8p
-
+ + U2 ths4631
R9 499
R10 499
R11 49.9
C4 8p
R2 4.99k
R3 100k
R4 499 C2 100u
-
++
4
3
5
1
2
U1 OPA376
Gain/Phase Analyzer
50R Termination +
Shunt Input Cap
+6dB
-6dB
Gain/Phase Analyzer
50R Termination +
Shunt Input Cap-6dB
+6dB
+6dB
Example of Zout Limitation• Noise corrupts low-frequency data
– Zout = Zo/(1+AOLB) so at low frequencies when AOLB is still large Zout gets very small and when we measure using Vtest = Ibackdrive*Zout the signal is smaller than the system noise.
– Although this issue affects every part measured, there are two types of op-amps that cause the most problems, both cause issues due not being able to create a large enough voltage to measure (V=IR):
• 1 – Small I – Ultra-Low-Power Op-Amps. Since we have to backdrive with Iab/2 we are very limited on our backdrive current when the quiescent current of the part is in the <100uA.
• 2 – Small R - Power Op-Amps with very low Zo. After dividing down by Loop Gain the Zout is in the sub 1-Ohm range and we can still only backdrive with Iab/2.
Example of Zout Limitation• Noise corrupts low-frequency data
• OPA333
101
102
103
104
105
106
107
108
100
101
102
103
104
Frequency (Hz)
Impe
danc
e
Zout
101
102
103
104
105
106
107
108
-40
-20
0
20
40
60
80
100
Frequency (Hz)
dB
Aol
101
102
103
104
105
106
107
108
101
102
103
104
Frequency (Hz)
Impe
danc
e
Zo
Data hit noise floor on Zout Measurement
Data invalid for Zo calculation below 1kHz
Example of Zout Limitation• Noise corrupts low-frequency data
• OPA333
Zo does not go capacitiveUntil ~10Hz
101
102
103
104
105
106
107
108
10-1
100
101
102
Frequency (Hz)
Impe
danc
e
Zout
101
102
103
104
105
106
107
108
-20
0
20
40
60
80
100
120
Frequency (Hz)
dB
Aol
101
102
103
104
105
106
107
108
100
101
102
103
104
Frequency (Hz)
Impe
danc
e
Zo
Example of Zout Limitation• Noise corrupts low-frequency data
• OPA564
Data hit noise floor on Zout Measurement
Data invalid for Zo calculation
Example of Zout Limitation• Noise corrupts low-frequency data
• OPA564
Zo does not go capacitiveUntil ~60Hz
101
102
103
104
105
106
107
108
10-1
100
101
102
Frequency (Hz)
Impe
danc
e
Zout
101
102
103
104
105
106
107
108
-20
0
20
40
60
80
100
120
Frequency (Hz)dB
Aol
101
102
103
104
105
106
107
108
100
101
102
103
104
Frequency (Hz)
Impe
danc
e
Zo
Other Zout Issues• Data has a “hump” at the frequency where
the part runs out of Loop Gain.
OPA564Gain = 55dBF = 20kHz
Hump in data at 20kHz
101
102
103
104
105
106
107
108
10-1
100
101
102
Frequency (Hz)
Impe
danc
e
Zout
101
102
103
104
105
106
107
108
-20
0
20
40
60
80
100
120
Frequency (Hz)
dB
Aol
101
102
103
104
105
106
107
108
100
101
102
103
104
Frequency (Hz)
Impe
danc
e
Zo
Other Zout Issues• Data has a “hump” at the frequency where
the part runs out of Loop Gain.
OPA564Gain = 40dBF = 110kHz
Hump in data at 110kHz
Other Zout Issues• Data has a “hump” at the frequency where
the part runs out of Loop Gain.
101
102
103
104
105
106
107
108
10-2
10-1
100
101
102
Frequency (Hz)
Impe
danc
e
Zout
101
102
103
104
105
106
107
108
-20
0
20
40
60
80
100
120
Frequency (Hz)dB
Aol
101
102
103
104
105
106
107
108
100
101
102
103
Frequency (Hz)
Impe
danc
e
Zo
OPA552Gain = 42dBF = 90kHz
Hump in data at 90kHz
101
102
103
104
105
106
107
108
100
101
102
103
Frequency (Hz)
Impe
danc
e
Zout
101
102
103
104
105
106
107
108
-20
0
20
40
60
80
100
120
Frequency (Hz)
dB
Aol
101
102
103
104
105
106
107
108
101
102
103
104
Frequency (Hz)
Impe
danc
e
Zo
Other Zout Issues• Data has a “hump” at the frequency where
the part runs out of Loop Gain.
OPA344Gain = 40dBF = 10kHz
Hump in dataAt 10kHz
Other Zout Issues• Data has a “hump” at the frequency where
the part runs out of Loop Gain.
OPA364Gain = 46dBF = 40kHz
Hump in dataAt 40kHz
Areas for Future Research
• Increasing Back-Drive Current
• Measuring “DC Zo” – Marek Lis
Areas for Future Research• Increasing Back-Drive Current
– Still struggling to get enough “V” in the V=IR equation, so increase I and test the results.
OPA376RS – 4990I = Iab/2
Hi-F Zo= 200Ohms
Data Valid until500Hz
Areas for Future Research• Increasing Back-Drive Current
– Still struggling to get enough “V” in the V=IR equation, so increase I and test the results.
OPA376RS – 1150I ~ Iab/2*5
Hi-F Zo= 190Ohms
Data Valid until300Hz
Areas for Future Research• Increasing Back-Drive Current
– Still struggling to get enough “V” in the V=IR equation, so increase I and test the results.
OPA376RS – 576I ~ Iab/2*10
Hi-F Zo= 150Ohms
Data Valid until100Hz
Areas for Future Research• Increasing Back-Drive Current
– Still struggling to get enough “V” in the V=IR equation, so increase I and test the results.
OPA364RS – 3570I = Iab/2
Hi-F Zo= 200Ohms
Areas for Future Research• Increasing Back-Drive Current
– Still struggling to get enough “V” in the V=IR equation, so increase I and test the results.
Hi-F Zo= 180Ohms
OPA364RS – 825I ~ Iab/2*5
Areas for Future Research• Increasing Back-Drive Current
– Still struggling to get enough “V” in the V=IR equation, so increase I and test the results.
Hi-F Zo= 120Ohms
OPA364RS – 412I ~ Iab/2*10
Areas for Future Research• Issues with Increasing Back-Drive Current
– The reduction in the Zo values when the back-drive current is increased is not predictable. Unlike adding a DC load, the “AC Loaded” Zo results do not decrease at 1/(gm*I) for BJT or 1/sqrt(k*Id) for MOS.– There also does not appear to be any correlation between the decrease in Zo value from one part to the next.
Areas for Future Research• Measuring “DC Zo”
– Would be a method to get a low-frequency intercept point so we could interpolate between the areas where the current Zo tester runs into noise issues. – Method entails measuring AOL with and without a DC load and then calculating the Zo based on the difference between the two AOL values.