DDS Selection

ly

he

an

It is the sawtooth that has the even harmonics. You are right that the triangular wave can be considered the result of a square wave with its higher harmonics being attenuated by an integrating filter.

I can see how an array of resistor values on a shift register can be chosen to replicate the impulse response of a filter. I do not follow how a sync function can be implemented with this. A reference would be appreciated.

What advantage would a delay line filter have over a Sallen-Key filter with many poles?

. e t

Nah, they're fine. You don't want to make the Q too high, but a nice Q of ten can really help with the dynamic range. You, of course, have to adjust everything, center the filter and adjust the phase. It's tough to get down to the noise level without a bandpass somewhere in the system.

George H.

light

es

ode

h does

m is

lated

the

ies

r

s an

re

al or

tes

an

for

:

more

t one

he

inning

i-

ght in

sponse

ice

uency

y

a cut

this

nti-

if it

=A0Sounds a bit

not linear in

list

have a

mistry.

he use of DDS

th a

expect

gestion

io

ent

OP can

th 100

w order

that

ant to

he

bly

s

he

the

hen

al

e ]

Yeah seems like the same equipment can do the lockin thing and the FFT.. it'd be interesting to see the results. I did this lock-in vs amplitude detection with similar time constants and all. (to OP, the great thing about lockins is you can get a really long time constant... (narrow linewidth)) The lockin was only about square root 2 better. The phase part of lockin.

George H.

o ?

Dang optical units. The Avago part has the following unit's

50mA currrent ~2V Vf 600nm wavlength, 8 degree view angle. and the 'winners' luminous efficacy 355 lm/W luminous efficiency 27 lm/W.

So if I stuck a big photodiode in front of this at 50mA how many watts or amps of current will I measure?

George H.

So

?

Yeah for a good source ya gotta control the temp too... but that's true most optical sources to some degree* or other.

George H.

*NPI

. So

er ?

Is that a rotten tempco for wavelength or output power or both?

Or use a Hobsonian Laser Noise Canceler.

John

...

e e

the

at

er

he

I do not need any frequencies about the fundamental or carrier. Why put a notch at a harmonic when I could use a low pass to take out all the harmonics?

So

r ?

a...

The monitoring photodiode is looking at the LED sideways, not in front. The detection photodiode has the LED aimed at it but the detection chemistry is between it and the LED.

ty. So

ower ?

s e .

ted text -

More wavelength than output power. (but they don't have similar units) I don't know about the leds you posted. You've got to measure them.

George H.

y. So

wer ?

Oh I was thinking about the wavelength.

You can measure the output power lotsa ways.

George H.

om...

the

the

t the

s.

on at

rier

s.

the

h

Oh I'm sorry perhaps I misread the post. (Sorry Bill) I was thinking of a band-pass filter not a notch.

George H.

om...

the

the

t the

s.

on at

rier

s.

the

h

Opps sorry Bill, I misread your post. I was refering to a band-pass filter centered one the fundamental.

George H.

I only mention IQ is that the "null" signal is a figure of merit. If it is quiet, you have a quality signal. Quadrature isn't required, but useful.

--
With the exception of Dr. Hobbs, you can expect everyone else's reply
to be self-serving, so caveat emptor.

light

es

ode

h does

m is

lated

the

ies

r

s an

re

al or

tes

an

for

:

more

t one

he

inning

i-

ght in

sponse

ice

uency

y

a cut

this

nti-

if it

=A0Sounds a bit

not linear in

list

have a

mistry.

he use of DDS

th a

expect

gestion

io

ent

OP can

th 100

w order

that

ant to

he

bly

s

he

the

hen

al

e ]

I guess you could say an FFT is similar to lockin, is just a mix with sin/cos and an average, just done at "all" frequencies at the same time

-Lasse

ty. So

ower ?

s e . e

et

.

se

=3Da...

You're absolutely right John. I was hoping Dr. Hobbs was logged on and would give me a 'real' number. Or at least there would be a reply by the time I logged on in the morning. lumens and candela's are the worst!

George H.

Both. You can compensate for the amplitude drift with a dual-path system. What sort of wavelength stability matters to you?

John

Be wary of those intensity-angle curves. The data sheets always draw them symmetric and perfectly centered on 0 degrees. In real life, they tend to be all over the place.

John