Prev: Migrate from SPICE MOSFET Level 6 to BSIM4
Next: [?] Looking for a small Bluetooth radio-microphone.
From: George Herold on 6 May 2010 11:06
On May 6, 10:25 am, Rafael Palacio <rafael.pala...(a)gmail.com> wrote:
> Hi,
>
> I've been searching for a while on the  Web but all
> manufacturers(Analog,Texas,Maxim,...) offer only digital solutions or
> low freq(tens of Khz) analog modulation.
> Does  anyone know a LD Driver IC with analog modulation capability up
> to 2Ghz ?
> I want to assembly an small  demo board on FR4 with some TO-can Laser
> diodes but I dont find any solution to provide a high freq analog
> modulation.
>
> Thanks a lot.
> Regards,
> Rafael

The only way I've seen this done is by directly injecting the RF into
the diode. (I've only done it up to several hundred MHz.... the limit
of my frequency generator.) AC couple into the DC current line, then
a 50 ohm 'termniation' resistor in parallel with an inductor (to allow
the DC current to flow) and this feeds the diode.

George H.
From: Phil Hobbs on 6 May 2010 17:05
On 5/6/2010 11:06 AM, George Herold wrote:
> On May 6, 10:25 am, Rafael Palacio<rafael.pala...(a)gmail.com> wrote:
>> Hi,
>>
>> I've been searching for a while on the Web but all
>> manufacturers(Analog,Texas,Maxim,...) offer only digital solutions or
>> low freq(tens of Khz) analog modulation.
>> Does anyone know a LD Driver IC with analog modulation capability up
>> to 2Ghz ?
>> I want to assembly an small demo board on FR4 with some TO-can Laser
>> diodes but I dont find any solution to provide a high freq analog
>> modulation.
>>
>> Thanks a lot.
>> Regards,
>> Rafael
>
> The only way I've seen this done is by directly injecting the RF into
> the diode. (I've only done it up to several hundred MHz.... the limit
> of my frequency generator.) AC couple into the DC current line, then
> a 50 ohm 'termniation' resistor in parallel with an inductor (to allow
> the DC current to flow) and this feeds the diode.
>
> George H.

Doing 2 GHz with a TO-can is no joke, even if the diode chip itself is
fast enough, which it probably isn't. The problems are both electrical
and optical. The main optical reason is that the level populations in
the upper and lower laser states often don't respond fast enough.
Ordinary cleaved-cavity Fabry-Perot lasers IME have relaxation peaks
near 1 GHz, so they don't respond terribly well at 2 GHz. You'll get
some response, for sure, but it's very hard to get nice looking edges or
nice open eye diagrams at 2 GHz with those lasers. (Even 1.25 Gb/s is
a bit of a trick.)

There are VCSELs and drivers available commercially that work well up to
10 Gb/s, with higher frequencies in the works.

On the other hand, if you just want to get 10% sinusoidal modulation at
2 GHz, you can probably do that by wiring a transformer secondary in
series with the LD, with a nice big bypass cap on the other end of the
secondary. That has the additional advantage of providing excellent ESD
protection.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net
From: George Herold on 6 May 2010 21:57
On May 6, 5:05 pm, Phil Hobbs <pcdhSpamMeSensel...(a)electrooptical.net>
wrote:
> On 5/6/2010 11:06 AM, George Herold wrote:
>
>
>
>
>
> > On May 6, 10:25 am, Rafael Palacio<rafael.pala...(a)gmail.com>  wrote:
> >> Hi,
>
> >> I've been searching for a while on the  Web but all
> >> manufacturers(Analog,Texas,Maxim,...) offer only digital solutions or
> >> low freq(tens of Khz) analog modulation.
> >> Does  anyone know a LD Driver IC with analog modulation capability up
> >> to 2Ghz ?
> >> I want to assembly an small  demo board on FR4 with some TO-can Laser
> >> diodes but I dont find any solution to provide a high freq analog
> >> modulation.
>
> >> Thanks a lot.
> >> Regards,
> >> Rafael
>
> > The only way I've seen this done is by directly injecting the RF into
> > the diode.  (I've only done it up to several hundred MHz.... the limit
> > of my frequency generator.)  AC couple into the DC current line, then
> > a 50 ohm 'termniation' resistor in parallel with an inductor (to allow
> > the DC current to flow) and this feeds the diode.
>
> > George H.
>
> Doing 2 GHz with a TO-can is no joke, even if the diode chip itself is
> fast enough, which it probably isn't.   The problems are both electrical
> and optical.   The main optical reason is that the level populations in
> the upper and lower laser states often don't respond fast enough.
> Ordinary cleaved-cavity Fabry-Perot lasers IME have relaxation peaks
> near 1 GHz, so they don't respond terribly well at 2 GHz.  You'll get
> some response, for sure, but it's very hard to get nice looking edges or
> nice open eye diagrams at 2 GHz with those lasers.   (Even 1.25 Gb/s is
> a bit of a trick.)

I've always wanted to try this up at higher frequencies. We put
sidebands on a Sanyo 74DL201 (?) and observe it with a confocal Fabry-
Perot. (It's an easy way to 'put a yard stick' on the spectrum as I'm
sure you know.) The diodes have relaxation oscillations right about
6.8X GHz. You might be able to pull it into the Rb 87 hyperfine
transition.

George H.

>
> There are VCSELs and drivers available commercially that work well up to
> 10 Gb/s, with higher frequencies in the works.
>
> On the other hand, if you just want to get 10% sinusoidal modulation at
> 2 GHz, you can probably do that by wiring a transformer secondary in
> series with the LD, with a nice big bypass cap on the other end of the
> secondary.  That has the additional advantage of providing excellent ESD
> protection.
>
> Cheers
>
> Phil Hobbs
>
> --
> Dr Philip C D Hobbs
> Principal
> ElectroOptical Innovations
> 55 Orchard Rd
> Briarcliff Manor NY 10510
> 845-480-2058
> hobbs at electrooptical dot nethttp://electrooptical.net- Hide quoted text -
>
> - Show quoted text -

From: miso on 6 May 2010 22:30
On May 6, 6:57 pm, George Herold <ggher...(a)gmail.com> wrote:
> On May 6, 5:05 pm, Phil Hobbs <pcdhSpamMeSensel...(a)electrooptical.net>
> wrote:
>
>
>
> > On 5/6/2010 11:06 AM, George Herold wrote:
>
> > > On May 6, 10:25 am, Rafael Palacio<rafael.pala...(a)gmail.com>  wrote:
> > >> Hi,
>
> > >> I've been searching for a while on the  Web but all
> > >> manufacturers(Analog,Texas,Maxim,...) offer only digital solutions or
> > >> low freq(tens of Khz) analog modulation.
> > >> Does  anyone know a LD Driver IC with analog modulation capability up
> > >> to 2Ghz ?
> > >> I want to assembly an small  demo board on FR4 with some TO-can Laser
> > >> diodes but I dont find any solution to provide a high freq analog
> > >> modulation.
>
> > >> Thanks a lot.
> > >> Regards,
> > >> Rafael
>
> > > The only way I've seen this done is by directly injecting the RF into
> > > the diode.  (I've only done it up to several hundred MHz.... the limit
> > > of my frequency generator.)  AC couple into the DC current line, then
> > > a 50 ohm 'termniation' resistor in parallel with an inductor (to allow
> > > the DC current to flow) and this feeds the diode.
>
> > > George H.
>
> > Doing 2 GHz with a TO-can is no joke, even if the diode chip itself is
> > fast enough, which it probably isn't.   The problems are both electrical
> > and optical.   The main optical reason is that the level populations in
> > the upper and lower laser states often don't respond fast enough.
> > Ordinary cleaved-cavity Fabry-Perot lasers IME have relaxation peaks
> > near 1 GHz, so they don't respond terribly well at 2 GHz.  You'll get
> > some response, for sure, but it's very hard to get nice looking edges or
> > nice open eye diagrams at 2 GHz with those lasers.   (Even 1.25 Gb/s is
> > a bit of a trick.)
>
> I've always wanted to try this up at higher frequencies.  We put
> sidebands on a Sanyo 74DL201 (?) and observe it with a confocal Fabry-
> Perot.  (It's an easy way to 'put a yard stick' on the spectrum as I'm
> sure you know.)  The diodes have relaxation oscillations right about
> 6.8X GHz.  You might be able to pull it into the Rb 87 hyperfine
> transition.
>
> George H.
>
>
>
> > There are VCSELs and drivers available commercially that work well up to
> > 10 Gb/s, with higher frequencies in the works.
>
> > On the other hand, if you just want to get 10% sinusoidal modulation at
> > 2 GHz, you can probably do that by wiring a transformer secondary in
> > series with the LD, with a nice big bypass cap on the other end of the
> > secondary.  That has the additional advantage of providing excellent ESD
> > protection.
>
> > Cheers
>
> > Phil Hobbs
>
> > --
> > Dr Philip C D Hobbs
> > Principal
> > ElectroOptical Innovations
> > 55 Orchard Rd
> > Briarcliff Manor NY 10510
> > 845-480-2058
> > hobbs at electrooptical dot nethttp://electrooptical.net-Hide quoted text -
>
> > - Show quoted text -

Using a transformer to inject a little AC on a DC signal is an old
trick for doing high frequency PSRR testing. Often RF generators can't
supple the DC to run the DUT, so you roll your own transfer.
minicircuits has nothing above 1.4GHz.
From: RafaP on 7 May 2010 04:31

.. What you suggest looks like a bias tee. i will try that solution,
the only problem comes from the impedance matching because I dont know
the input impedance of the photodiode.
Thanks again.
Rafael

On 6 mayo, 17:06, George Herold <ggher...(a)gmail.com> wrote:
> On May 6, 10:25 am, Rafael Palacio <rafael.pala...(a)gmail.com> wrote:
>
> > Hi,
>
> > I've been searching for a while on the  Web but all
> > manufacturers(Analog,Texas,Maxim,...) offer only digital solutions or
> > low freq(tens of Khz) analog modulation.
> > Does  anyone know a LD Driver IC with analog modulation capability up
> > to 2Ghz ?
> > I want to assembly an small  demo board on FR4 with some TO-can Laser
> > diodes but I dont find any solution to provide a high freq analog
> > modulation.
>
> > Thanks a lot.
> > Regards,
> > Rafael
>
> The only way I've seen this done is by directly injecting the RF into
> the diode.  (I've only done it up to several hundred MHz.... the limit
> of my frequency generator.)  AC couple into the DC current line, then
> a 50 ohm 'termniation' resistor in parallel with an inductor (to allow
> the DC current to flow) and this feeds the diode.
>
> George H.

 |  Next  |  Last
Pages: 1 2
Prev: Migrate from SPICE MOSFET Level 6 to BSIM4
Next: [?] Looking for a small Bluetooth radio-microphone.