From: John Popelish
Organization: This space not available for advertising.
X-Mailer: Mozilla 4.7 [en] (Win98; U)
Subject: Re: photodetector circuit, high speed with good ambient light rejection
Date: Sat, 30 Nov 2002 20:36:57 GMT
NNTP-Posting-Date: Sat, 30 Nov 2002 15:36:57 EST
Sorry about having this post out of order, but my news server has
thrown the post I want to reply to, away. This is a repeat of an
earlier post that did not show up correctly on Google (evidently,
lines ending in some characters cause them to discard the carriage
return line feed and runs lines together), and I will attempt to
correct that, here.
Winfield Hill wrote:
> Hobbs points out that with moderate signal currents that create a high
> r_e, an external bias current can be added to the BJT's emitter and
> a like current taken from the collector. This approach relies on the
> absence of shot noise in resistor-generated currents. It would seem
> practical up to bias/signal ratios of up to say 25 or 50, but at some
> point weak signals get one into trouble attempting to balance the two
> bias currents and struggling with an unknown lost-current factor from
> the BJT's beta. Hobbs suggests using high-beta transistors like the
> MPSA18 (B>500), but watch out for their input higher capacitance (e.g.
> 8pF), further complicating issues at high frequencies. As is often
> the case, a panoply of compromises must be examined.
Win, I found Hobbs paper to be very interesting and thought
provoking. I am weak on noise calculation problems and intend to
study this at some length, to learn more. But the paper has given me
a couple of other ideas I would like you and others to criticize.
Again, referring to Philip Hobbs article:
Figure 9 shows a combination DC voltage bias source and follower added
to the diode to help reduce the effect of diode capacitance by
bootstrapping. Is there any benefit to this approach without the
cascode amplifier between the diode and the transimpedance amplifier?
Isn't the main purpose of the cascode to allow the opamp to tolerate
the diode capacitance better?
What if you replaced the follower transistor with a fast opamp
follower like this:
. +--R3-----+ .
. | | .
. +-----------+----|-\ | .
. | | | >-+-- .
. | ^ 0v-|+/ .
. +-|+\ | .
. | >---+--+ .
. +-|-/ R1 | .
. | | C1 .
. +--------+--+ .
. R2 | .
. | C2 .
. | | .
. +15 0v .
With, say, R1 and R2 1 meg each, And C2 very much smaller than C1
(and, possibly adjustable) to produce a net follower gain of slightly
more than 1 at the high frequency end of the response, to not only
reduce the diode capacitance effect, but to neutralize the diode
capacitance at the transimpedance amplifier input. Wouldn't this do
away with the need for the cascode transistor and current adding and
One might even add a series resistor to C1 to extend the bandwidth of
Even if this follower had the desired capacitance reducing effect,
what other problems would such a circuit create?