As I worked on the design a bit more I found that I don't need the signal to be clamped at that point. It needs to be clamped later on, after filtering. Sorry about that. So the new requirement is that the gain stage just be a gain stage - by "isn't clipped" I mean that the output of the gain stage isn't slammed against the op amp's output limits because of bias current errors.
The next stage is going to be a bandpass filter of some type, using the rest of the opamp. I understand that you and others would much prefer it for me to simply specify the application; while that would probably get me to an optimum solution faster (along with suggestions of better methods) it removes quite a bit of the DIY fun of the project. I'm just looking for crossword hints here, not the solution to the whole puzzle, if that makes sense!
Didn't find your answer? Ask the community — no account required.
J
John Larkin
amplify
with
an NPN
clamp
the
amp.
shows up
can
10
The classic way to do this would be to generate a pretty stiff Vcc/2 rail, and use that to bias all the amp stages halfway up. The sigs are all AC coupled.
John
B
Bitrex
amplify
Also, with
at
an NPN
clamp
the
happens
amp.
shows up
can
below 10
to
because
it
Yes, I think now that the design has evolved that will be the best approach, instead of just trying to amplify one half of the input signal with everything referenced to ground. I'll just clamp off the whole negative-going portion later.
J
Jim Thompson
amplify
Also, with
at
an NPN
clamp
at the
happens
amp.
shows up
A
can
below 10
to
because
it
the
Marginally safe ----^ with an LM324
"Clamp" is not the appropriate word. "Clip" or "rectify" would be more appropriate. ...Jim Thompson
[On the Road, in New York]
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
J
John Fields
Run the file and you'll get the schematic.
M
meg
JF, read what I said, I don=92t want to fart with spice. Yeah, I could, but I don=92t want to and he=92s asking for help. All the op has do is describe the prob and provide a schemo.
His last comment re JL=92s response just doesn=92t make any sense.
I think he just wants to DC offset his AC sig off ground.
JL gave him a solution that he doesn=92t seem to understand. At least that=92s what I get from his comment.
I=92m not sure he=92s going to get there in one gain stage. But its hard to tell without any more info.
You don=92t need spice to figure this out, the op just needs to share more info.
G
George Herold
to amplify
the
=A0Also, with
main at
h uses an NPN
the
the clamp
on at the
t happens
y op amp.
t shows up
ion. A
RF
hink I can
below 10
trying to
input
nput
because
low, it
too
the
ith
I do
only
ions,
o do
hing
he
,
in
ke
a
tage
t
e of
ure is
olytic
r
be
le
th the
a
hind the
=A0 =A0 =A0...Jim Thompson
t
etc
uple
n
as
ot
=94.
e
a
-
What's your white noise source? Do you worry at all about the 1/f noise of the amplifier? It might be an issue in your frequency range.
George H.
J
Jamie
amplify
Also, with
at
an NPN
clamp
the
happens
amp.
shows up
can
below 10
to
because
it
maybe the link will help.
formatting link
Jamie
F
Fred Bloggs
to amplify
the
=A0Also, with
main at
h uses an NPN
the
the clamp
on at the
t happens
y op amp.
t shows up
ion. A
RF
hink I can
below 10
trying to
input
nput
because
low, it
too
the
ith
I do
only
ions,
o do
hing
he
,
in
ke
a
tage
t
e of
ure is
olytic
r
be
le
th the
a
hind the
=A0 =A0 =A0...Jim Thompson
t
etc
uple
he
n
as
ot
t
Okay- another hint is that noise bandwidth differs from signal bandwidth...
B
Bitrex
I found in working a bit with the LM324 Spice model I have that with a voltage rail of just 2.4 volts, the op amp needs more like 0.7 volts bias to operate correctly in that configuration. With that supply voltage the positive swing is limited to about 1.4 volts. I don't know if that's what you meant by marginally safe, or something different.
B
Bitrex
amplify
Also, with
at
an NPN
clamp
the
happens
amp.
shows up
can
below 10
to
because
it
Hi Fred, thanks for the reply. I did some Googling to refresh my memory on noise bandwidth. So the noise bandwidth is the bandwidth of a brickwall filter that contains equivalent noise power as the total power contained under the curve of the original transfer function of the circuit. I can see how this could be useful, but I'm not sure how it directly relates to my application where I am concerned only with relative signal levels.
Say I have a source resistor connected to an ideal amplifier with a certain transfer function - I can calculate the noise bandwidth of the amplifier and plug that into the noise voltage formula for the source resistance and get a figure on the output noise of the amplifier in V/sqrtHz due to the source resistance over the bandwidth in question.
Let's say I have some other noise source, but I don't know how its noise power or voltage varies with bandwidth. All I know is that over some narrow bandwidth I see an approximately constant level of noise voltage. I then amplify this voltage by 60dB with a system that has a - 3dB point of 100 Hz. The RMS level of noise voltage at the output of the amplifier at 100 Hz will still be -3dB down compared with the midpoint, yes?
If I'm misunderstanding something (which is probably the case!) please help clarify. Thanks!
M
meg
d to amplify
of the
. =A0Also, with
remain at
ith uses an NPN
ed the
n the clamp
tion at the
it happens
ply op amp.
it shows up
ation. A
nt RF
think I can
nt below 10
m trying to
e input
input
y because
o low, it
't too
, the
with
if I do
y only
ctions,
to do
thing
the
nt,
gain
like
s a
stage
ent
nce of
ature is
trolytic
her
l be
ngle
with the
e
1,
e
t
s
to a
behind the
=A0 =A0 =A0...Jim Thompson
of
est
, etc
I
couple
l
,
t the
he
ion
r as
not
e
er
just
,
y
Where you place the analysis filter matters in noise measurements.
Don=92t let the amp set the filter limits.
The Vn of the amp circuit will be a function of your source impedance, amp noise products and your feedback circuit. The noise at the output will be Vn*G. Take note, those noise density numbs are frequency dependent, they are a description of thermal noise behavior of the part. You need to verify that your frequency of operation is in that region and not in the flicker region for your chosen part.
Assuming your into the thermal noise, then your analysis bandwidth will be set by the final filter behind all this crap. Think IF filters in radios, now days implemented in dsp.
Just out of curiosity, how are you measuring this stuff? Are you still struggling with the DC component?
J
Jim Thompson
[snip]
An LM324 has a PNP Darlington style of input differential pair, which limits your positive-going common-mode range to around 1.3V below rail. Then the output is an NPN Darlington follower, so you lose on both ends :-(
Your bandwidth requirements seem low. Can't you use a rail-to-rail CMOS device? ...Jim Thompson
[On the Road, in New York]
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
J
John Fields
amplify
with
NPN
up
Why is that bad?
When the dichotomy of the source's position is reflected back to the
source, using irony, shouldn't the effect of his own words be used
against him?
J
John Fields
And, why not, since vanquishing an enemy is sweeter using his own
sword.
J
John Larkin
Vanquishing? Hiding behind a killfile, lurking and jumping out to peck at my ankles? With you, his faithful sidekick, holding up the rear? What a joke.
John
J
John Fields
What makes me laugh is the spectacle of you perpetually trying to
wriggle out of situations you get yourself into.
J
John Larkin
You have virtually nothing to say about electronics, but you like to lurk in the bushes with your fellow coop-mate and jump on people who you feel have made mistakes, spelling or syntax or definitions or whatever fuzzy stuff you think you understand.
This is a *discussion* group, Bozo, not a spelling bee.
Design any interesting electronics lately? Tell us.
John
J
John Fields
An "electronics" discussion group, asshole, so why don't you try to
remember that the next time you feel inclined to post any of your
inane, off-topic crap?
J
josephkk
a
behind the
Appearently you do not understand the idea and insult of death by
1,000,000 paper cuts.
?-(
Join the Discussion
Have something to add? Share your thoughts — no account required.
Didn't find your answer?
Ask the community — no account required
Report Content
You are reporting this content to the moderators. They will look at it
ASAP.