DVD laser and photodiode

[Jamie Morken]

Hi,

This article says that a typical DVD 16X writer requires 230mW of
optical power, and the wavelength of the laser is around 650nm, red.

"http://www.edn.com/article/CA485669.html"

Also for a 16X DVD the datarate is up to 418 Mbps, so the laser and
photodiode have very fast rise/fall times.

Does anyone know of a site where the DVD optics have been "experimented"
with? Is it feasible to use the laser and photodiode for other
purposes than as in a DVD, ie. long distance optical datalink or laser
rangefinding (time of flight and/or phaseshift measurement)

cheers,
Jamie

 

[Rene Tschaggelar]

This article says that a typical DVD 16X writer requires 230mW of
optical power, and the wavelength of the laser is around 650nm, red.

"http://www.edn.com/article/CA485669.html"

Also for a 16X DVD the datarate is up to 418 Mbps, so the laser and
photodiode have very fast rise/fall times.

Does anyone know of a site where the DVD optics have been "experimented"
with? Is it feasible to use the laser and photodiode for other
purposes than as in a DVD, ie. long distance optical datalink or laser
rangefinding (time of flight and/or phaseshift measurement)

Jamie,
The fast signals require significant power, a few nA won't
give 500MHz, and neither do just a few photons.
Mreaning long distance, whatever this means, requires some
expensive optics. The narrower the laser beam, the more it
widens over the distance, thus you're quickly at 10cm for
a mile or two.

Yes, I think an optical link is doable. How fast over
which distance require some experiments.

Rene

 

[Winfield Hill]

Rene Tschaggelar wrote...
I've been experimenting with a red Hitachi DVD laser obtained on eBay.
Hitachi's data sheet has a frequency-response plot saying it goes to
700MHz, peaks and drops like a rock. I found that properly wired, it
goes smoothly to 1200MHz, without peaking, and rolls off gradually.
It's possible my detector was primarily responsible for the rolloff.

 

[John Larkin]

Rene Tschaggelar wrote...

I've been experimenting with a red Hitachi DVD laser obtained on eBay.
Hitachi's data sheet has a frequency-response plot saying it goes to
700MHz, peaks and drops like a rock. I found that properly wired, it
goes smoothly to 1200MHz, without peaking, and rolls off gradually.
It's possible my detector was primarily responsible for the rolloff.

I got some samples of a plastic-packaged Osram IR laser that puts out
about 10 watts peak, and managed to get 1 ns pulses out of it driving
with a gaasfet. You can also get the lasers that have just the chip
soldered to the edge of a little square brass block with a mounting
hole in the middle, which is probably optically superior to the
plastic package. Either one goes for a few dollars in quantity.

The little 850 nm fiber-coupled VCSELS have risetimes around 100 ps...
hard to measure, since the good fast detectors are blind at 850. The
Optek parts are available stock from Newark and are pretty good.

Probably an avalanche transistor driver is easiest to make to do lidar
type stuff. ECL, maybe eclips lite, has enough swing to drive a laser.

John

 

[Adrian Jansen]

I got some samples of a plastic-packaged Osram IR laser that puts out
about 10 watts peak, and managed to get 1 ns pulses out of it driving
with a gaasfet. You can also get the lasers that have just the chip
soldered to the edge of a little square brass block with a mounting
hole in the middle, which is probably optically superior to the
plastic package. Either one goes for a few dollars in quantity.

The little 850 nm fiber-coupled VCSELS have risetimes around 100 ps...
hard to measure, since the good fast detectors are blind at 850. The
Optek parts are available stock from Newark and are pretty good.

Probably an avalanche transistor driver is easiest to make to do lidar
type stuff. ECL, maybe eclips lite, has enough swing to drive a laser.

John

I have a possible application for lidar, but it wont stand the $200,000
price tag of commercial stuff. The idea is to measure doppler shift on
particles in the air to get the local wind speed. Any info on cheap
lasers, optics and detectors capable of getting a range of the order of
200 metres backscatter in air is welcome.


--
Regards,

Adrian Jansen adrianjansen at internode dot on dot net
Design Engineer J & K Micro Systems
Microcomputer solutions for industrial control
Note reply address is invalid, convert address above to machine form.

 

[Jamie Morken]

Hi Win,

Rene Tschaggelar wrote...



I've been experimenting with a red Hitachi DVD laser obtained on eBay.
Hitachi's data sheet has a frequency-response plot saying it goes to
700MHz, peaks and drops like a rock. I found that properly wired, it
goes smoothly to 1200MHz, without peaking, and rolls off gradually.
It's possible my detector was primarily responsible for the rolloff.

The DVD reader I took apart has two lasers (one infrared for CD and one
red light for DVD) with 3pins each, I guess one pin is the modulation
signal and the other two pins are power and ground? Here is a pic from
what I was able to salvage from the DVD player:

"http://red79.net/rocketresearch/pics/DVD optics/DVD reader optics.JPG"

Sorry for the poor image quality, I am still using a sub 1-megapixel
camera with no optical zoom

bottom left is the photodetector, it is a 10pin IC with a clear top on
it to let the light in, any idea on where to find a datasheet and/or how
to hook this up to detect light? I think it has multiple photodiodes
inside it. I think the chip that interfaces to this photodetector is
the Toshiba TA1323.

the black chunk above the photodetector is the DVD laser diode I think.

Then there are 4 lenses from the DVD optics on the right of the picture.

I couldn't remove the suspected CD laser as it was glued in pretty tightly.

Any tips on hooking this stuff up? I am most interested in the
photodiode as I haven't been able to find any other high speed red light
sensitive photodiodes.

cheers,
Jamie

 

[Jamie Morken]

Hi,

I hooked up the red laser diode and it seems to work fine, from reading
on the net, the three pins on it are 1. ground, 2. photodiode cathode
for laser intensity feedback, 3. laser diode cathode.

The ground pin is easy to see as it is hooked up to the case, I used a
~4Volt (lithium battery) supply with a 2k resistor and this gave a red
light coming out of the laser diode, the light is not focused at all and
scatters quickly, without a lens. I went up to an 820ohm resistor and
the red light still looked the same. Be careful if you do this! I put
some cotton in front of the laser and watched as it glowed red. I don't
know how much of the spectrum of light is red, and if there is any
infrared, so I am wary of doing any "open air" tests with these laser
diodes, my vision is too valuable Also the red light is close to
infra red already, so it probably is brighter, and more dangerous than
it appears.

Here is the hookup I used:

4V ----> 2k resistor ---> laser cathode
0V ---> ground pin (connected to the diode case)

Now to try the photodiode IC.. (need help on this one!)

cheers,
Jamie

 

[Rich Grise]

I have a possible application for lidar, but it wont stand the $200,000
price tag of commercial stuff. The idea is to measure doppler shift on
particles in the air to get the local wind speed. Any info on cheap
lasers, optics and detectors capable of getting a range of the order of
200 metres backscatter in air is welcome.

For some reason, "Laser Gyro" springs to my brainpan. They used
interferometers with an audio output. I remember the doppler radar in the
USAF had an audio amplifier.

Suddenly, it actually sounds doable. ;-)

Optics? Check with astronomy and photography folks.

Good Luck!
Rich

 

[Rene Tschaggelar]

I have a possible application for lidar, but it wont stand the $200,000
price tag of commercial stuff. The idea is to measure doppler shift on
particles in the air to get the local wind speed. Any info on cheap
lasers, optics and detectors capable of getting a range of the order of
200 metres backscatter in air is welcome.

LIDAR means measuring backscatter from usually pulsed lasers.
The mentioned pricetag is not just because these guys are expensive.
Consider a backscatter coefficient of say 10^-4, of which by means
of a 30cm reflective mirror you again get only a fraction. Throw
in the r^2 attenuation and you quickly reach 120dB of dynamic range
extending far into the noise. The only way to get the signal out
of the noise is to start with enormous peak power at excitation.
The response is a modulation of the r^2 attenuation, giving a
distance profile of scattering.

You could make the setup far simpler by measuring forward attenuation,
eg between some highrise buildings, in case you're not interested on a
distance profile.

Rene

 

[Rene Tschaggelar]

Jamie Morken wrote:

Any tips on hooking this stuff up? I am most interested in the
photodiode as I haven't been able to find any other high speed red light
sensitive photodiodes.

There are plenty of fast photodiodes around. Their responsivity
is given by the material.
Eg the AEPX65 is recommended.

Rene

 

[Winfield Hill]

Rene Tschaggelar wrote...

There are plenty of fast photodiodes around. Their responsivity is
given by the material. Eg the AEPX65 is recommended.

Not so much the material as the size, because high capacitance is a
killer. That part's 4pF is a serious problem, up to 10 times higher
than the detector I used measuring the Hitachi's 1200MHz response.

 

[Winfield Hill]

Winfield Hill wrote...

Rene Tschaggelar wrote...

Not so much the material as the size, because high capacitance is a
killer. That part's 4pF is a serious problem, up to 10 times higher
than the detector I used measuring the Hitachi's 1200MHz response.

BTW the 664nm red DVD laser was an Hitachi HL6504FM and my detector
an Optek OPF480 PIN diode, biased at -100V, measured into a 25 ohm
load (double-end termination) with an Agilent network analyzer.

 

[Jamie Morken]

Hi Win,

Winfield Hill wrote...



BTW the 664nm red DVD laser was an Hitachi HL6504FM and my detector
an Optek OPF480 PIN diode, biased at -100V, measured into a 25 ohm
load (double-end termination) with an Agilent network analyzer.

Could you give more details of your circuit that you used? I can
visualize it several different ways Weren't you a bit scared of the
laser light too?

cheers,
Jamie

 

[Rene Tschaggelar]

Rene Tschaggelar wrote...



Not so much the material as the size, because high capacitance is a
killer. That part's 4pF is a serious problem, up to 10 times higher
than the detector I used measuring the Hitachi's 1200MHz response.

The responsivity was meant for the spectral color of the light.
You're right, the speed is done with the size. Unfortunately
that calls for some aligned optics unless plenty of light is
available.

Rene

 

[Winfield Hill]

Jamie Morken wrote...

Could you give more details of your circuit that you used? I can
visualize it several different ways

It was just a quick lash up.

.. bias-tee laser
.. ________| |______,---------,
.. ________|-||-+--|______(-50R-|>|-'
.. 50-ohm |____X__| \\ mirror PD bias
.. coax | \\| optics 100V
.. 50R //| & etc |
.. | // __|____
.. laser // ________| X |___50-ohm
.. current ,-|>|----+-)_______|--+-||-|____ coax
.. supply +-||-50R-' | coax |_______| term.
.. '----------' bias-tee

Components were 1206 SMT hand-soldered with zero-distance spacing.
Bias-tees were 12GHz-bandwidth Picosecond Pulse Labs. A first
attempt at a PCB stripline replacement only goes to 600MHz so far.

Weren't you a bit scared of the laser light too?

One should always be careful of course, but in this case the laser
assembly included optics to spread the beam to 1/4" substantially
lowering its intensity down to a veritable dull glow. :>)

 

[Adrian Jansen]

Rene Tschaggelar wrote:

LIDAR means measuring backscatter from usually pulsed lasers.
The mentioned pricetag is not just because these guys are expensive.
Consider a backscatter coefficient of say 10^-4, of which by means
of a 30cm reflective mirror you again get only a fraction. Throw
in the r^2 attenuation and you quickly reach 120dB of dynamic range
extending far into the noise. The only way to get the signal out
of the noise is to start with enormous peak power at excitation.
The response is a modulation of the r^2 attenuation, giving a
distance profile of scattering.

You could make the setup far simpler by measuring forward attenuation,
eg between some highrise buildings, in case you're not interested on a
distance profile.

Rene

Thanks for the thoughts, but the app requires the source and receiver at
the same location ( moving vehicle ). And we cant have high power, the
laser must be eye-safe. Guess its one of those expensive solutions, but
I would still like to know what sources and detectors are around.


--
Regards,

Adrian Jansen adrianjansen at internode dot on dot net
Design Engineer J & K Micro Systems
Microcomputer solutions for industrial control
Note reply address is invalid, convert address above to machine form.

 

[Rene Tschaggelar]

Rene Tschaggelar wrote:



Thanks for the thoughts, but the app requires the source and receiver at
the same location ( moving vehicle ). And we cant have high power, the
laser must be eye-safe. Guess its one of those expensive solutions, but
I would still like to know what sources and detectors are around.

AFAIK, the pulsed lasers are still considered eye save when the
rep rate is sufficiently low, say 1 Hz or lower. I've seen a
handheld Q switched YAG issuing single pulses of say 20kW with a
pulsetime of 1ns or below. At least at that time, the regulations
didn't cover it. Such a laser is still doable in the size of
two fists. The optics is much bigger. I'd suggest a 20 to 30cm
parabol mirror for the receiver. I'd don't know the angular margin of
the backscatter, but it could be sufficiently narrow, especially
on short range such that you have to use the same mirror for the
sending.
Your receiving electronics then would have to be fast enough to
get the 1ns pulse and its distance square decay. There are
wonderful logamps that make the exponential appear linear.

Rene

 

[Jamie Morken]

Hi Win,

Jamie Morken wrote...



It was just a quick lash up.

. bias-tee laser
. ________| |______,---------,
. ________|-||-+--|______(-50R-|>|-'
. 50-ohm |____X__| \\ mirror PD bias
. coax | \\| optics 100V
. 50R //| & etc |
. | // __|____
. laser // ________| X |___50-ohm
. current ,-|>|----+-)_______|--+-||-|____ coax
. supply +-||-50R-' | coax |_______| term.
. '----------' bias-tee

Components were 1206 SMT hand-soldered with zero-distance spacing.
Bias-tees were 12GHz-bandwidth Picosecond Pulse Labs. A first
attempt at a PCB stripline replacement only goes to 600MHz so far.

Cool! That looks like a neat circuit, I am buying that same photodiode
from digikey (part#: 365-1029-ND) and will try to emulate your circuit
with my antique 10MHz scope to start with as I don't have the Ghz+
network analyzer!

Are you going to feed the output of that circuit into an RF amplifer for
your application? I was thinking the max2611 or max2650 LNA amps might
work well for this, although they start to lose gain at around 500MHz.

I will make my own bias tee's, and I don't mind about getting RF in my
powersupply at this point, so they should be rather simple with maybe a
couple series inductor to the bias supply, and for the AC coupling caps
would 0.1uF be ok? I checked some of the "Picosecond Pulse Labs" bias
tee datasheets, and they seem to have a range of different capacitor
values. I guess the actual value of the AC coupling capacitor isn't
that important as long as it is big enough so that it doesn't charge up
more than ~10% from the AC input?

What are the cap and resistor for across the photodiode? I would guess
it is an RC snubber but I wouldn't think that would be desired for this
circuit unless it is a very small cap value?

One should always be careful of course, but in this case the laser
assembly included optics to spread the beam to 1/4" substantially
lowering its intensity down to a veritable dull glow. :>)

I have been reading that red light appears much less bright than
similiar power green or blue light would appear. I don't know if this
makes red light potentially more dangerous than green/blue light though?

cheers,
Jamie

 

[Jamie Morken]

Hi,

I posted a conversion of the ascii art circuit to eagle cad pdf:
"http://red79.net/rocketresearch/pics/DVD optics/Winfield Hill's circuit.pdf"

Is the laser current supply used to keep the laser always lasing, and
then a small signal modulation is fed to the laser through the coax?
Does the laser need to always be lasing to be able to achieve high
frequency operation? I was thinking of turning the laser on and off to
send 0's and 1's (but this is probably slower than just changing the
lasing brightness. I don't know what Win's application is, but the next
step seems to be getting this signal into digital logic

cheers,
Jamie

 

[Winfield Hill]

Jamie Morken wrote...

I posted a conversion of the ascii art circuit to eagle cad pdf:
"http://red79.net/rocketresearch/pics/DVD optics/Winfield Hill's circuit.pdf"

Is the laser current supply used to keep the laser always lasing, and
then a small signal modulation is fed to the laser through the coax?
Does the laser need to always be lasing to be able to achieve high
frequency operation? I was thinking of turning the laser on and off
to send 0's and 1's (but this is probably slower than just changing
the lasing brightness.

Yes, at very high frequencies like 1GHz, it's best not to attempt to
fully turn the laser on and off. At 300MHz or under, that's OK, but
may still not be the best approach.

I don't know what Win's application is, but the next step seems to
be getting this signal into digital logic

Not my next step - one of our scientists will be using it to evaluate
nanowire FETs.

The circuit I posted (and you wrote up) doesn't look much like a
typical laser driver or optical receiver circuit, certainly not like
most of the ones I design. The primary goal throughout, which you
didn't show or analyze on your little writeup, it to maintain a very
clean 50-ohm impedance at all points for the drive signal and for the
received signal, and double terminated as well. As I mentioned, when
one of us attempted to translate the lashup to a PCB, the performance
dropped from 1200MHz to 600MHz despite being very careful, indicating
the important of avoiding any impedance discontinuities.