Maker Pro
Maker Pro

Lock-in detectors

I

Ian Du Rieu

Jan 1, 1970
0
Hi Win,

In AoE (2nd ed) page 1034 you describe a demo for your students whereby
you detect a dim LED.
Did this demo use a commercial lock-in amplifier or did you roll your
own?

Is it practical to build a simple demo circuit using a smallish number
of parts?
I'm thinking along the lines of a "Saturday arvo project".


BTW, what's the latest on the 3rd revision?

Cheers,
Ian Du Rieu
 
R

Rene Tschaggelar

Jan 1, 1970
0
Ian said:
Hi Win,

In AoE (2nd ed) page 1034 you describe a demo for your students whereby
you detect a dim LED.
Did this demo use a commercial lock-in amplifier or did you roll your
own?

Is it practical to build a simple demo circuit using a smallish number
of parts?
I'm thinking along the lines of a "Saturday arvo project".


It is indeed a better approach for a demo to roll you own
from a few parts. Otherwise the viewer sits in front of a
bulky black box doing miracles. As starting point, have a
look at the AD630 from Analog Devices. The datasheet shows
a few examples. Add a 8038 frequency generator and a few
opamps and you're there.

Rene
 
Rene said:
It is indeed a better approach for a demo to roll you own from a few
parts. Otherwise the viewer sits in front of a bulky black box doing
miracles.

Hah, an ac-coupling capacitor at the input gets you a long way. :)
As starting point, have a look at the AD630 from Analog Devices.
The datasheet shows a few examples. Add a 8038 frequency
generator and a few opamps and you're there.

The AD630 is good for square-wave demodulation. But this chip
presents a bit of a black-box on its own. I like to use the signal
with an inverting opamp, plus a 4053 double-throw CMOS switch, to
make an inexpensive lock-in that can effectively reject room light.
For example, see AoE page 1031, and here,
http://groups.google.com/group/sci.electronics.design/msg/36bca1750cf674b2
http://groups.google.com/group/sci.electronics.design/msg/751f72b110506d67

I often place equal small-valued resistors (2.2k) in series with
the signal opamp outputs and the two HC4053 switches, to
isolate the opamps from any '4053 charge injection.

I often use an AD734 multiplier with sine-wave modulation, which
an 8038 nicely provides, but we're talking about square-wave
demodulation, which takes the signal and its harmonics. So it's
happy with square-wave modulation, which greatly simplifies the
illumination drive, just switch an LED on and off from a 555
reference oscillator.

Now, arvo = afternoon? Or http://old.arvo.org/root/index.asp
 
J

John Larkin

Jan 1, 1970
0
I often place equal small-valued resistors (2.2k) in series with
the signal opamp outputs and the two HC4053 switches, to
isolate the opamps from any '4053 charge injection.


I've also noticed that many/most opamps go bonkers for some
microseconds if their output drives a typical cmos switch. What I
usually see is a jump at the instant of switching, followed by a
linear slew back to the proper level. It can be pretty big.

John
 
P

Phil Hobbs

Jan 1, 1970
0
John said:
I've also noticed that many/most opamps go bonkers for some
microseconds if their output drives a typical cmos switch. What I
usually see is a jump at the instant of switching, followed by a
linear slew back to the proper level. It can be pretty big.

If you put a small resistor between the switch and the op amp, and a
bigger one between there and the integration cap, you can make more of
the injected charge go out the amp end, which is useful in reducing
offsets. I'm interested in the op amp problems: I try to use 12V or
+-15V supplies when I can, to preserve dynamic range, and emitter
follower outputs are pretty tough to drag around much. Does this happen
mostly with R2R bipolar or CMOS parts?

MC1496 Gilbert cells make good lock-ins, if you don't need
super-accurately calibrated gain--you can easily get a dc gain of 200,
which can reduce the need for external circuitry.

BTW does LTC still make the LT1043 switched-capacitor block? Those
things are cool, if you can afford them.

Cheers,

Phil Hobbs
 
J

Joerg

Jan 1, 1970
0
Hello Win,

Just FYI: You email address seems to be non-munged this time. Might want
to check that.

Regards, Joerg
 
W

Winfield Hill

Jan 1, 1970
0
Joerg wrote...
Hello Win,

Just FYI: You email address seems to be non-munged this
time. Might want to check that.

I saw that but had no choice. Google makes you use a real
email address to register. They hide this if one views
on Google, but send it to the rest of the world. (I had
to use Google for that one post.) They suggest you get a
fake, real email address and use that instead. But what
a crazy workaround! Anyway, I do have a few such "fake"
email addresses, but I don't use them or remember where
they are (Yahoo, CompuServe, Harvard, NewsGuy, etc?), or
the passwords. Sigh.
 
J

Joerg

Jan 1, 1970
0
Hello Win,

I saw that but had no choice. Google makes you use a real
email address to register. ...


That is one reason why I don't use them. No idea if this one would be
better but may be worth a look:
http://services.mail2web.com/FreeServices/Usenet/


NGs are probably pretty far removed from spambots but s.e.d. is being
copied into all kinds of web based forums and FAQ sites. From there to
the spambots it's just a short hop.

Regards, Joerg
 
F

Frank Bemelman

Jan 1, 1970
0
Joerg said:
NGs are probably pretty far removed from spambots but s.e.d. is being
copied into all kinds of web based forums and FAQ sites. From there to
the spambots it's just a short hop.

Very true. But spambots also make up email adresses from domain names.
Simply try to send email to [email protected] or to [email protected]
and so on. Don't you get any spam sent to your analogconsultants domain?
I'm getting email to addresses that I never created for my domains.

I'm strongly against death penalty, but for spammers I wouldn't mind if they
made an exception.
 
M

Michael A. Terrell

Jan 1, 1970
0
Frank said:
I'm strongly against death penalty, but for spammers I wouldn't mind if they
made an exception.


Only if they are televised as a warning to other spammers.

--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida
 
J

John Larkin

Jan 1, 1970
0
If you put a small resistor between the switch and the op amp, and a
bigger one between there and the integration cap, you can make more of
the injected charge go out the amp end, which is useful in reducing
offsets. I'm interested in the op amp problems: I try to use 12V or
+-15V supplies when I can, to preserve dynamic range, and emitter
follower outputs are pretty tough to drag around much. Does this happen
mostly with R2R bipolar or CMOS parts?

I've changed opamps to get me out of trouble here, bipolar to bifet as
I recall, but I don't know if there's any sort of pattern. I love the
LM8261, because you can hang any amount of capacitance from its output
to ground and swamp this problem. It's handy for all sorts of things.

John
 
J

Joerg

Jan 1, 1970
0
Hello Frank,
Very true. But spambots also make up email adresses from domain names.
Simply try to send email to [email protected] or to [email protected]
and so on. Don't you get any spam sent to your analogconsultants domain?
I'm getting email to addresses that I never created for my domains.

I don't get any to that one yet but it is also not used by friends, only
by business partners. My impression is that the spam volume has
increased once an email address was used by friends who could not resist
that dreaded temptation to send jokes, "cool stuff" or their latest cute
doggie pics to everyone in their address book. Of course some of those
recipients use free mail services and since these mails contain the
whole slew of other recipients in the source you can imagine what
happens then. Nothing is free.

I'm strongly against death penalty, but for spammers I wouldn't mind if they
made an exception.

Nah, just take away their PCs and get a court order banning them from
sending email to more than x people at the same time. Then if they
violate that lock them up.

Regards, Joerg
 
R

Rich Grise, but drunk

Jan 1, 1970
0
Joerg wrote...

I saw that but had no choice. Google makes you use a real email address
to register. They hide this if one views on Google, but send it to the
rest of the world. (I had to use Google for that one post.) They
suggest you get a fake, real email address and use that instead. But
what a crazy workaround! Anyway, I do have a few such "fake" email
addresses, but I don't use them or remember where they are (Yahoo,
CompuServe, Harvard, NewsGuy, etc?), or the passwords. Sigh.

Well, then, the answer is, use the one that's been revealed as your "fake"
address, and stay signed up on google with it, but abandon it - I have a
yahoo account that I've set to "send everything to the trash and delete it.",
and another yahoo account that's real, for real email.

I'm kinda proud of myself, because the fake google-accessing yahoo email
is richardgrise, so all I have to say to folks to have them unmunge it is
to elide ard. ;-) Thought up that little trick all by meself, I did! %-}

Cheers!
Rich
--
-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GAT(E P) dpu s: a++ C++@ P+ L++>+ !E W+ N++ o? K? w-- !O !M !V PS+++
PE Y+ PGP- t 5+++)-; X- R- tv+ b+ DI++++>+ D-? G e+$ h+ r-- z+
------END GEEK CODE BLOCK------
 
R

Rich Grise, but drunk

Jan 1, 1970
0
Only if they are televised as a warning to other spammers.

Nah. Not death. That's too merciful.

Caning, pillorying, flogging, any kind of humane torture would be much
more effective than a simple, quick, clean beheading, as far as deterrents
go.

Maybe boil them in pig fat - that'd probably cut down on Islamic
terrorists, too! ;-P

Cheers!
Rich
--
-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GAT(E P) dpu s: a++ C++@ P+ L++>+ !E W+ N++ o? K? w-- !O !M !V PS+++
PE Y+ PGP- t 5+++)-; X- R- tv+ b+ DI++++>+ D-? G e+$ h+ r-- z+
------END GEEK CODE BLOCK------
 
J

Joseph2k

Jan 1, 1970
0
Joerg said:
Hello Win,




That is one reason why I don't use them. No idea if this one would be
better but may be worth a look:
http://services.mail2web.com/FreeServices/Usenet/


NGs are probably pretty far removed from spambots but s.e.d. is being
copied into all kinds of web based forums and FAQ sites. From there to
the spambots it's just a short hop.

Regards, Joerg
The ones that make me crazy are the spammers that take the trouble to join
listservers to promulgate their trash. Given the usual method complexity
(relatively hard to automate) to subscribe, they must be really driven to
do so.
 
J

Joerg

Jan 1, 1970
0
Hello Joseph,
The ones that make me crazy are the spammers that take the trouble to join
listservers to promulgate their trash. Given the usual method complexity
(relatively hard to automate) to subscribe, they must be really driven to
do so.


It's a sign that enough people fall for spam. I was surprised when I
read how much reasonably well tested email lists sell for. My real
concern for society is phishing though. You and I are ok but imagine
older folks who have developed early stages of memory loss. They could
easily think it's a legit request from their bank.

BTW, your tag line would really go like this:
Gegen Dummheit kämpfen Götter selbst vergebens! (Schiller)

I lived a block away from Schillerstrasse a while ago but must confess
that I didn't read much from him.

Regards, Joerg
 
Top