Accelerator DivisionBLM Upgrade Proposal

Preliminary Testing of the Burr Brown ACF2101

Summary of Test Data

A. BaumbaughK. Knickerbocker

PPD/Electrical Engineering Department

03/23/2004

Fermilab

New BLM Front-end Electronics using the ACF2101 Dual Integrator.

The plan for the new BLM system is to use the Burr-Brown ACF2101 dual integrator,and to “ping-pong” (switch rapidly) between the A and B channels so as to never miss anylosses. The integration times are planned to be 20 µsec minimum. Significant features and adiagram (shown in Figure 1.) of the ACF2101 device are from the Burr-Brown datasheet andgiven below.

DESCRIPTIONThe ACF2101 is a dual switched integrator forprecision applications. Each channel can convertan input current to an output voltage byintegration, using either an internal or externalcapacitor. Included on the chip are precision 100pF integration capacitors, hold and reset switches,and output multiplexers.

FEATURESIncludes Integration Capacitor, Reset and HoldSwitches, and Output MultiplexerLow Noise: 10 µVrmsLow Charge Transfer: 0.1 pCWide Dynamic Range: 120 dBLow Bias Current: 100 fA

Figure 1. The Burr-Brown ACF2101

The ASIC Test System shown in Figure 2 was used as convenient platform forconducting the initial tests of the Burr-Brown dual integrator. The system provided all of thenecessary timing and control signals, plus power supply voltages. It also allowed easycollection and processing of test data on a standard PC.

An existing BLM system daughter card(Beam Line) was obtained from the AcceleratorDivision Instrumentation Department and used forall testing. The card had several modificationsmade to it to make it similar to the new system. Itwas modified to allow the use of both integratorsin the ACF2101 at the same time. The currentBLM system uses one integrator or the other. Thedaughter cards external integration capacitor(1000 pf) was removed and only the internalcapacitors (100 pf) of the ACF2101 were used.Other modifications were made to allow theindividual control of the following signals, ResetA, Reset B, Hold A, Hold B, Track/Hold. The final change made was to the analog powersupply voltages, they were lowered from the standard +15, -15, volts to +12, -12. The +5 voltlogic power supply voltage remained the same.

Figure 2. The ASIC Test System

The different test circuits used are shown in Figures 3, 4, and 5. The current sourceused was a Datel DVC-8500 precision voltage source and a 1.0 Meg ohm metal film leaded orsurface mount resistor. Timing signals for the ACF2101 were provided by a free runningpattern generator available in the ASIC Test System. A 16-Bit ADC available in the ASICtest system was used to make measurements of the integrator output. Figure 3 shows thecurrent source at the far end of 550 ft. of RG-58 cable. Figure 4 shows the current source atthe very input to the two integrators, and Figure 5 shows the cable in parallel with the currentsource. This last mode lets us easily switch between cable and no cable. After some effortwe were able to trigger the ADC in the test system synchronously with the Track and Holdsignal (part of the free running pattern). This allows us to make measurements in exactly thesame way as we are proposing for the new BLM system, although not quite as fast. A timingdiagram for the integrators is shown in Figure 6, we are able to free run both the A and Bchannels, and selectively digitize either A only, B only, or A and B both.

What we found was that the A and B channels have a slightly different input offsetvoltage and input resistance value which causes the two channels to be different. With nocable capacitance this difference is about .04% on a 10µA signal (see Figure 7). However,when a long cable with a large capacitance is attached, the different offset voltages at theinput either charge or discharge the cable as the system “ping-pongs” (for the effect ofdifferent cable lengths see Figure 8). The result is that the A channel reads high byapproximately 200 nA and the B channel reads low by the same amount (Figures 9 and 10).Figure 11 shows the data for reading both A and B sides, note the two separate peaks. Webelieve this to be solely due to the capacitance of the cable and the different offset voltages ofthe two inputs. If we vary the input voltage, (see Figure 12) we also see a slight effect due tothe different input resistor values. The reader should note that the noise in the system has anRMS of 15 counts which is equivalent to and RMS of 3 nA on a 10 µA signal.

In an effort to reduce or eliminate the differences between the two integrator channelstwo more circuit modification were made. At the suggestion of Craig Drennan an input offsetvoltage nulling circuit was added to the ACF2101. The “Nulling” Circuit, which is bi-polaronly needed to be added to one of the two integrators. This allowed us to adjust the offset ofthe two integrators to be equal at zero input current. This reduced the maximum A-Bdifference from 2000 counts to 859 counts (Figure 13). The remaining effect is the differenceof the “ON” resistance of the two input switches. These switches are about 1.5 kOhm in theON state. By temporarily adding a fixed resistor to A and measuring the shift in A-B, wewere able to calculate the difference of the two switches. The B switch had 43 Ohms lessresistance, by adding a 43 Ohm resistor to the B side we were able to match the twointegrators over a very large range. After adding the nulling circuit and the 43 Ohm resistor(see Fig. 14), we now have matched A-B to –10 to 35 counts (Figure 15) over the entire rangeof 0 to 10 µA. We now believe that we can “Tune” the integrators quite nicely and thuseliminate the need for a cable receiver.

We now have a very nice system for making these measurements and will continue tostudy the front end.

1.0

Meg

Ohm

PC fo

r Dat

a Co

llect

ion

and

Anal

ysis

16,5

00pf

Cabl

eCa

p.

Resi

stor

Box

550

ft. o

fRG

-58

cabl

e.

ACF2

101

Test

Set

up S

chem

atic

03/2

3/20

04

50 Ohm

pf10

0AC

F210

1

0.00

0 V

-10

.000

V

Rese

t

Sele

ct A

A

HP 5

4502

A 40

0 M

Hz D

igita

lO

scill

osco

pe

Hold

pf10

0

Rese

t

Sele

ct B

B

Hold

400K

CA31

40

100K

CA31

4010

K

HA24

25

ASIC

Tes

tSy

stem

16

- Bit

ADC

(Bur

r-Bro

wnAD

C700

KH)

4 5 36 9 8 7 10

+12V

-12V

1918

121 24 23 13 14

Trac

k &

Hold

Date

l DVC

-850

0 Pr

ecis

ion

Volta

ge S

ourc

e1.

5K

Int.

Res.

1.5K

Int.

Res.

RS-2

32 L

ink

(Lon

g Ca

ble

Conf

igur

atio

n)

Figu

re 3

.

PC fo

r Dat

a Co

llect

ion

and

Anal

ysis

31 p

fCa

ble

Cap.

1 ft.

of

RG-1

74 c

able

.

ACF2

101

Test

Set

up S

chem

atic

03/2

3/20

04

1.0

Meg

Ohm

SMR

pf10

0AC

F210

1

0.00

0 V

-10

.000

V

Rese

t

Sele

ct A

A

HP 5

4502

A 40

0 M

Hz D

igita

lO

scill

osco

pe

Hold

pf10

0

Rese

t

Sele

ct B

B

Hold

400K

CA31

40

100K

CA31

4010

K

HA24

25

ASIC

Tes

tSy

stem

16

- Bit

ADC

(Bur

r-Bro

wnAD

C700

KH)

4 5 36 9 8 7 10

+12V

-12V

1918

121 24 23 13 14

Trac

k &

Hold

Date

l DVC

-850

0 Pr

ecis

ion

Volta

ge S

ourc

e1.

5K

Int.

Res.

1.5K

Int.

Res.

RS-2

32 L

ink

(Sur

face

Mou

nt R

esis

tor (

No C

able

) Con

figur

atio

n)

Figu

re 4

.

PC fo

r Dat

a Co

llect

ion

and

Anal

ysis

31 p

fCa

ble

Cap.

1 ft.

of

RG-1

74 c

able

.

ACF2

101

Test

Set

up S

chem

atic

03/2

3/20

04

1.0

Meg

Ohm

SMR

pf10

0AC

F210

1

0.00

0 V

-10

.000

V

Rese

t

Sele

ct A

A

HP 5

4502

A 40

0 M

Hz D

igita

lO

scill

osco

pe

Hold

pf10

0

Rese

t

Sele

ct B

B

Hold

400K

CA31

40

100K

CA31

4010

K

HA24

25

ASIC

Tes

tSy

stem

16

- Bit

ADC

(Bur

r-Bro

wnAD

C700

KH)

4 5 36 9 8 7 10

+12V

-12V

1918

121 24 23 13 14

Trac

k &

Hold

Date

l DVC

-850

0 Pr

ecis

ion

Volta

ge S

ourc

e1.

5K

Int.

Res.

1.5K

Int.

Res.

RS-2

32 L

ink

(Sur

face

Mou

nt R

esis

tor (

Opt

iona

l Cab

le) C

onfig

urat

ion)

Figu

re 5

.

16,5

00pf

Cabl

eCa

p.55

0 ft.

of

RG-5

8 ca

ble.

Jum

per

Zoom

Clo

ck n

S

R1

=R

2 =

R2

- R1

=

Sel A

Bit

- 4

Sel B

Bit

- 5

Unu

sed

Hol

d A

Bit

- 0

Res

et A

Bit

- 2

T/H

A O

nly

Bit

- 6

Unu

sed

Hol

d B

Bit

- 1

Res

et B

Bit

- 3

T/H

B O

nly

Bit

- 7

Unu

sed

T/H

AB

Bit

- 8

Unu

sed

Unu

sed

AD

C B

usy

Bit

- 9

Unu

sed

C:\N

anoS

im\N

SM\B

LM_T

est.N

SM

-.42

-.42

.00

T1 =

T2 =

T2 -T

1 =

110 25

5100

051

020

5104

051

060

5108

051

100

5112

051

140

0

100

200

300

400

500

600

700

800

900

1000

Mea

n =

5107

1.86

5S

td. D

ev. =

14.

290

Num

. Sam

ples

= 4

096

Inpu

t = 1

0.00

0 V

Counts

5100

051

020

5104

051

060

5108

051

100

5112

051

140

0

100

200

300

400

500

600

700

800

900

1000

Mea

n =

5106

1.45

7S

td. D

ev. =

9.1

13N

um. S

ampl

es =

409

6In

put =

10.

000

V

Counts

5100

051

020

5104

051

060

5108

051

100

5112

051

140

0

100

200

300

400

500

600

700

800

900

1000

Figu

re 7

.

Mea

n =

5108

5.50

6S

td. D

ev. =

9.0

38N

um. S

ampl

es =

409

6In

put =

10.

000

V

Counts

Ch.

A, B

, A&

B, N

o C

able

/1 M

eg S

MR

, Inp

ut 1

0.00

0V, S

ync

AD

C, (

File

s: B

LM_X

X_SM

R_1

0000

.DA

T)

Cha

nnel

A a

nd B

plo

tted

with

diff

eren

t len

gths

of c

oaxi

al c

able

(RG

-58)

atta

ched

.

Red

, Cab

le L

engt

h =

550

ft., B

lue,

Cab

le L

engt

h =

1 ft.

, Gre

en, C

able

Len

gth

= 0

ft.

Figu

re 8

.

5212

052

140

5216

052

180

5220

052

220

5224

052

260

5228

052

300

5232

00

100

200

300

400

500

600

700

Figu

re 9

.

Mea

n =

5218

8.56

1S

td. D

ev. =

15.

405

Num

. Sam

ples

= 4

096

Inpu

t = 1

0.00

0 V

Counts

Ch.

A, 5

50 ft

. Cab

le, I

nput

10

+ .0

00V,

.010

V, .0

11V,

Syn

c A

DC

, (Fi

les:

BLM

_A_L

_100

XX.D

AT)

5212

052

140

5216

052

180

5220

052

220

5224

052

260

5228

052

300

5232

00

100

200

300

400

500

600

700

Mea

n =

5224

0.03

1S

td. D

ev. =

15.

731

Num

. Sam

ples

= 4

096

Inpu

t = 1

0.01

0 V

Counts

5212

052

140

5216

052

180

5220

052

220

5224

052

260

5228

052

300

5232

00

100

200

300

400

500

600

700

Bin

Siz

e =

5

Mea

n =

5224

6.03

5S

td. D

ev. =

15.

543

Num

. Sam

ples

= 4

096

Inpu

t = 1

0.01

1 V

Counts

5086

050

880

5090

050

920

5094

050

960

5098

051

000

5102

00

100

200

300

400

500

600

700

Mea

n =

5096

7.07

8S

td. D

ev. =

14.

928

Num

. Sam

ples

= 4

096

Inpu

t = 1

0.01

1 V

Counts

5086

050

880

5090

050

920

5094

050

960

5098

051

000

5102

00

100

200

300

400

500

600

700

Mea

n =

5097

0.99

1S

td. D

ev. =

14.

771

Num

. Sam

ples

= 4

096

Inpu

t = 1

0.01

0 V

Counts

5086

050

880

5090

050

920

5094

050

960

5098

051

000

5102

00

100

200

300

400

500

600

700

Figu

re 1

0.

Mea

n =

5091

4.74

7S

td. D

ev. =

14.

872

Num

. Sam

ples

= 4

096

Inpu

t = 1

0.00

0 V

Counts

Ch.

B, 5

50 ft

. Cab

le, I

nput

10

- .00

0V, .

010V

, .01

1V, S

ync

AD

C, (

File

: BLM

_B_L

_100

XX.D

AT)

5080

051

000

5120

051

400

5160

051

800

5200

052

200

5240

00

100

200

300

400

500

600

Figu

re 1

1.

Mea

n =

5217

8.64

0S

td. D

ev. =

15.

052

Num

. Sam

ples

= 2

059

Bin

Siz

e =

10

Mea

n =

5092

6.85

2S

td. D

ev. =

15.

200

Num

. Sam

ples

= 2

037

Counts

Ch.

A&

B, 5

50 ft

. Cab

le, I

nput

10.

000V

, Syn

c A

DC

, (Fi

le: B

LM_A

B_L

_100

00.D

AT)

1960

019

800

2000

020

200

2040

020

600

2080

021

000

2120

021

400

2160

00

100

200

300

400

500

600

Bin

Siz

e =

10

Mea

n =

2149

3.85

1S

td. D

ev. =

15.

009

Num

. Sam

ples

= 2

016

Mea

n =

1976

7.41

6S

td. D

ev. =

15.

118

Num

. Sam

ples

= 2

080

Counts

Ch.

A&

B, 5

50 ft

. Cab

le, I

nput

4.0

00V,

Syn

c AD

C, (

File

: BLM

_AB

_L_4

V.D

AT)

3000

030

200

3040

030

600

3080

031

000

3120

031

400

3160

031

800

3200

00

100

200

300

400

500

600

Bin

Siz

e =

10

Mea

n =

3171

7.32

3S

td. D

ev. =

14.

639

Num

. Sam

ples

= 1

982

Mea

n =

3015

0.46

8S

td. D

ev. =

14.

825

Num

. Sam

ples

= 2

114

Counts

Ch.

A&

B, 5

50 ft

. Cab

le, I

nput

6.0

00V,

Syn

c AD

C, (

File

: BLM

_AB

_L_6

V.D

AT)

4020

040

400

4060

040

800

4100

041

200

4140

041

600

4180

042

000

4220

00

100

200

300

400

500

600

Bin

Siz

e =

10

Mea

n =

4194

8.08

5S

td. D

ev. =

14.

997

Num

. Sam

ples

= 2

074

Mea

n =

4053

8.63

1S

td. D

ev. =

15.

332

Num

. Sam

ples

= 2

022

Counts

Ch.

A&

B, 5

50 ft

. Cab

le, I

nput

8.0

00V,

Syn

c AD

C, (

File

: BLM

_AB

_L_8

V.D

AT)

9200

9400

9600

9800

1000

010

200

1040

010

600

1080

011

000

1120

011

400

0

100

200

300

400

500

600

Figu

re 1

2.

Bin

Siz

e =

10

Mea

n =

1127

1.24

3S

td. D

ev. =

14.

524

Num

. Sam

ples

= 2

065

Mea

n =

9379

.528

Std

. Dev

. = 1

4.82

8N

um. S

ampl

es =

203

1

Counts

Ch.

A&

B, 5

50 ft

. Cab

le, I

nput

2.0

00V,

Syn

c AD

C, (

File

: BLM

_AB

_L_2

V.D

AT)

0 2 4 6 8 10

1500

1550

1600

1650

1700

1750

0 2 4 6 8 100

200

400

600

800

1000

Ch. A-B Data Difference vs. Input Current, No CompensationC

h. A

-B in

AD

C C

ount

s

Input Current in microAmps

Figure 13.

Ch. A-B Data Difference vs. Input Current, Long Cable, Nulled at 0V

Ch.

A-B

in A

DC

Cou

nts

Input Current in microAmps

PC fo

r Dat

a Co

llect

ion

and

Anal

ysis

31 p

fCa

ble

Cap.

1 ft.

of

RG-1

74 c

able

.

ACF2

101

Test

Set

up S

chem

atic

03/2

3/20

04

1.0

Meg

Ohm

SMR

pf10

0AC

F210

1

0.00

0 V

-10

.000

V

Rese

t

Sele

ct A

A

Hold

pf10

0

Rese

t

Sele

ct B

B

Hold

400K

CA31

40

100K

CA31

4010

K

HA24

25

ASIC

Tes

tSy

stem

16

- Bit

ADC

(Bur

r-Bro

wnAD

C700

KH)

4 5 36 9 8 7 10

+12V

-12V

1918

121 24 23 13 14

Trac

k &

Hold

Date

l DVC

-850

0 Pr

ecis

ion

Volta

ge S

ourc

e1.

5K

Int.

Res.

1.5K

Int.

Res.

RS-2

32 L

ink

(Offs

et N

ull a

nd O

ffset

Adj

. Con

figur

atio

n)

Figu

re 1

4.

16,5

00pf

Cabl

eCa

p.55

0 ft.

of

RG-5

8 ca

ble.

Jum

per

100K

10

+12V

-12V

10K

43.2

02

46

810

-10010203040

Figu

re 1

5.

Ch.

A-B

Dat

a D

iffer

ence

vs.

Inpu

t, Lo

ng C

able

, Nul

led

at 0

V, R

esis

tor C

ompe

nsat

edCh. A-B in ADC Counts

Inpu

t Cur

rent

in m

icro

Am

ps

Datasheets:

Texas Instruments/Burr-Brown, - ACF2101, Low Noise, Dual Switched Integrator.PDS-1078D (SBFS003) http://www.burr-brown.com/

Texas Instruments/Burr-Brown, - Improved Noise Performance of the ACF2101Switched Integrator., Application Bulletin AB-053 (SBOA032) http://www.burr-brown.com/

Intersil, - HA-2420, HA-2425, Sample and Hold Amplifier., (FN2856.4)http://www.intersil.com/

Harris Semiconductor, - CA3140, CA3140A, 4.5 MHz, BiMOS Operational Amplifierwith MOSFET Input/Bipolar Output., (File Number 957.3) http://www.harris.com/

Acknowledgements:

We would like to thank the following people for their assistance during the testing ofthe ACF2101: Craig Drennan for circuit suggestions, ideas, and advice. Marvin Olson forthe loan of a BLM daughter card, ACF2101 devices, and 550 ft. of coaxial cable. AndCharlie Nelson for circuit suggestions, ideas, and advice.