BJT differential amplifer

[derek]

Hi,

If I have a differential input voltage and want to output a current to
be proportional to it, how to create a circuit using BJTs to do that?
Thanks!

 

[Tim Wescott]

Hi,

If I have a differential input voltage and want to output a current to
be proportional to it, how to create a circuit using BJTs to do that?
Thanks!

You're describing the input stage to an opamp. Search around until you
find a data sheet for something simple like a 741 or an LM353 and look
at the input stage.

If you want to do it in real life you need matched transistors, though.

 

[John Popelish]

Hi,

If I have a differential input voltage and want to output a current to
be proportional to it, how to create a circuit using BJTs to do that?
Thanks!

You have severely under specified the problem. What range of
differential voltage must the circuit respond to? What output current
per volt of input do you require? What voltage range must the current
source comply with? What load resistance can the differential
amplifier apply to the input voltage? What frequency range is
involved? Etc.

You are not ready to even start thinking about circuit designs till
you have the problem nailed down.

 

[Arie de Muynck]

"derek" ...

If I have a differential input voltage and want to output a current to
be proportional to it, how to create a circuit using BJTs to do that?

Study "Howland current source", i.e. in the AoE.

Regards,
Arie de Muynck

 

[Michele S]

Hi,

If I have a differential input voltage and want to output a current to
be proportional to it, how to create a circuit using BJTs to do that?
Thanks!

you need a differntial Gm cell....
it's possible to create a programmable Gm cell where Iout= Gm_code*Vin_diff
it's very important f you want to realize it using discrete component to use
a feedback to cancel the mismatch of the compnent...reed Pelgrom's paper on
this topic
anyway you should give some spec on your circuit...

 

[Winfield Hill]

Michele S wrote...

derek asked...
you need a differntial Gm cell....
it's possible to create a programmable Gm cell where Iout= Gm_code*Vin_diff
it's very important if you want to realize it using discrete component to
use a feedback to cancel the mismatch of the component...read Pelgrom's
paper on this topic. anyway you should give some spec on your circuit...

You're referring to this paper? M. Pelgrom, A. Duinmajer and A. Welbers,
"Matching properties of MOS transistors", IEEE J. Solid-State Circuits,
vol. 24., No. 5,pp. 1433-1439, 1989.

Bipolar transistors are more easily matched than MOFETs, and therefore
arguably are better suited for this task. An introductory discussion
of transconductance (voltage-to-current) circuits should include the
famous old RCA CA3080 chip, and mention of its relations, the CA3094,
LM13600 and LM13700. There are modern matched-transistor arrays suited
for the basis of transconductance circuits, such as Intersil's hfa3134
Gilbert cell, their hfa3102, and matched pairs, hfa3134 and hfa3135.

 

[gwhite]

Michele S wrote...

You're referring to this paper? M. Pelgrom, A. Duinmajer and A. Welbers,
"Matching properties of MOS transistors", IEEE J. Solid-State Circuits,
vol. 24., No. 5,pp. 1433-1439, 1989.

Bipolar transistors are more easily matched than MOFETs, and therefore
arguably are better suited for this task. An introductory discussion
of transconductance (voltage-to-current) circuits should include the
famous old RCA CA3080 chip, and mention of its relations, the CA3094,
LM13600 and LM13700. There are modern matched-transistor arrays suited
for the basis of transconductance circuits, such as Intersil's hfa3134
Gilbert cell, their hfa3102, and matched pairs, hfa3134 and hfa3135.

If the requirements aren't too stringent, the dual 3904's are dirt cheap in high
volume today (under 5¢ per pair. They don't claim to be matched, but I've used
them and had good luck. I think dual 2n2222's are available too.

http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?Ref=140443&Row=332568&Site=US