Transimpedance amplifier

[[email protected]]

Hi,
i've simulated a transimpedance amplifier with spice. My circuit is
composed of an opamp and a feedback resistor; in parallel to current
source a capacitor is present (i want to use a photodiode's equivalent
model). I have a question: is it possible that (with an ac sweep), V/I
transfer function is greater (at lower frequencies) than opamp open
loop gain? Why?
Another question: i wanto to use it as input of a sar adc. I've read
this interesting article
www.edn.com/article/CA6602451.html
and i want to try it. Is it more correct to use another opamp as
buffer and rc circuit as its load (as describend in the link) or use
rc load directly to transimpedance amplifier?
Thanks

 

[MooseFET]

Hi,
i've simulated a transimpedance amplifier with spice. My circuit is
composed of an opamp and a feedback resistor; in parallel to current
source a capacitor is present (i want to use a photodiode's equivalent
model). I have a question: is it possible that (with an ac sweep), V/I
transfer function is greater (at lower frequencies) than opamp open
loop gain? Why?

The open loop gain spec you are looking at may be a minimum and the
model may use the typical.

You didn't say there was a peak but if you see a peak in the curve
just before it falls off the answer may be the phase margin of the
system.

Another question: i wanto to use it as input of a sar adc. I've read
this interesting articlewww.edn.com/article/CA6602451.html
and i want to try it. Is it more correct to use another opamp as
buffer and rc circuit as its load (as describend in the link) or use
rc load directly to transimpedance amplifier?

Beware of EDN articles they are too often bogus. Does your ADC have
sample and hold in it? What sort of load does it present to its
input?

You may or may not need another op-amp in the system at all.

 

[[email protected]]

The open loop gain spec you are looking at may be a minimum and the
model may use the typical.

How can i verify it?No opamp's macromodel is present, because my opamp
is realized with transistor...

You didn't say there was a peak but if you see a peak in the curve
just before it falls off the answer may be the phase margin of the
system.

Yes, a peak is present....i've read about on graeme's book.I've used a
feedback capacitor to remove it

Beware of EDN articles they are too often bogus.  Does your ADC have
sample and hold in it?  What sort of load does it present to its
input?

It has a sample and hold with r=on switch resistor=100ohm and
c=sampling capacitor=5pF
thanks

 

[MooseFET]

How can i verify it?No opamp's macromodel is present, because my opamp
is realized with transistor...


Yes, a peak is present....i've read about on graeme's book.I've used a
feedback capacitor to remove it

Making the peak go partly away but not completely can lead to a better
system. A slight bump before the gain starts to fall can move the 3db
point up.

It has a sample and hold with r=on switch resistor=100ohm and
c=sampling capacitor=5pF

The output impedance of the driving circuit needs to be far less than
the 100 Ohms. The output will actually look inductive. Since you are
modeling the system, you can model the sample and hold action too.

 

[[email protected]]

x-no-archive:

Hi,
i've simulated a transimpedance amplifier with spice. My circuit is
composed of an opamp and a feedback resistor; in parallel to current
source a capacitor is present (i want to use a photodiode's equivalent
model). I have a question: is it possible that (with an ac sweep), V/I
transfer function is greater (at lower frequencies) than opamp open
loop gain? Why?

Well the "open loop gain" is measured in Volts out / Volts in and the
V/I transfer function is measured in Volt out / Amps in..... so you
have an apples and oranges problem I would think.

Mark