Laser locking (control loops with two feedback paths.)

[George Herold]

That should be a lot more zippy than 20kHz if connected to a somewhat
reasonable TIA. Or did you give it a hefty dose of Ambien?

Grin... I posted numbers for Tim. I can get a factor of 2 or so by
reverse biasing the PD, and maybe a factor of 5-10 with a faster
opamp.

George H.

 

[George Herold]

Now you'll have lots of folks banging on your door at Otis Street,
wanting to see her






Sometimes it's best not to have all the gain in the first (TIA) stage.
Opamps are fairly cheap these days and the following ones don't have to
be very fancy. Maybe something George could look at.

Hi Joerg, thanks for the advice. As Phil says later on, I've always
tried to keep signals in the Volt range for noise reasons. (But
perhaps I was too focused on noise?) Anyway I can sure reduce the TIA
gain and dial in more later. (I start to worry about the bandwidth of
the rest of the signal chain though...)

George H.

 

[Joerg]

[...]

That should be a lot more zippy than 20kHz if connected to a somewhat
reasonable TIA. Or did you give it a hefty dose of Ambien?

Grin... I posted numbers for Tim. I can get a factor of 2 or so by
reverse biasing the PD, and maybe a factor of 5-10 with a faster
opamp.

Then it's time to flick the little switch on the Weller and do that

Reverse bias is easy if there is no compelling reason why you need to be
photovoltaic with the PD. Just keep the abs max limit in view. With the
opamp also keep in mind its slew rate. GBW means nothing if it can't
swing its tail around fast enough.

 

[George Herold]

I'm not sure what you're showing me the gain _of_ here.

I hope it's the closed loop gain of the piezo feedback circuit.

It's a really good idea to have the open-loop gain be descending at
around 20dB/decade with 60 to 120 degrees of lag at the loop closure
point.  Too little slope on the gain curve gives you these weird broad
low peaks (or troughs, depending on the phase).

Going back to my original suggestion, you want a block that you can call
"laser" which _internally_ has the high & lowpass filters, and the feed
to the diode and piezo, and the feedback from the PD.  Whatever the
response of that block is, it should be well-behaved.  I've been assuming
that its frequency response would be fairly constant up to some cut-off
frequency, at which point it would start dropping by 20 or 40 dB/decade.
Then you'd wrap that with your controller to get a nice descending
characteristic, with a gain=1 point as high as you can push it.









You're not bothering to bias your photodiode?!?!?!  In a world of easy-to-
get 10MHz or better op-amps you're using 1.5MHz?!?!?!?!

Um -- I think there's some circuit improvements that you could make,
there.

Hangs head in shame, kicks dirt. "Gee Tim, there's no need to rub it
in".

But seriously, I did this in 2001, when it seems like, I harldy knew
anything electronics-wise. There are several hundred of these "slow
as molasses" detectors out there. I can't write up some side locking
technique that uses modified detectors. But maybe I can sell them all
newer faster detectors?

Anyway again thanks,

George H.

 

[George Herold]

George Herold wrote:
Tim Wescott wrote:

[...]

20kHz? That's like molasses. Why so low? And it should not cause it to
oscillate.
If he's already got substantial phase shift elsewhere, then the photodiode
rolling off would cause oscillation somewhere around 20kHz.
The fact that it happens at _exactly_ 20kHz just means that, sans
photodiode, he's got about 45 degrees of margin at 20kHz.
But where does all that phase margin fall through the cracks? Unless
everything rolls off fast, of course. 20kHz BW for the photodiode sounds
really low, unless it is one the size of a dinner plate.
--
Regards, Joerg
http://www.analogconsultants.com/-Hidequoted text -
- Show quoted text -
Grin, well not quite dinner plate size. ~6-7mm diam.
That should be a lot more zippy than 20kHz if connected to a somewhat
reasonable TIA. Or did you give it a hefty dose of Ambien?

Grin... I posted numbers for Tim.  I can get a factor of 2 or so by
reverse biasing the PD, and maybe a factor of 5-10 with a faster
opamp.

Then it's time to flick the little switch on the Weller and do that

Reverse bias is easy if there is no compelling reason why you need to be
photovoltaic with the PD. Just keep the abs max limit in view. With the
opamp also keep in mind its slew rate. GBW means nothing if it can't
swing its tail around fast enough.

Grin, yup, that's fine for me. Not so easy for most customers.

I figure I should put a cap multiplier on the bias supply too. I
don't want all the power rail noise getting into the photodiode.

I've got a 'screaming' fast FET opamp in a drawer somewhere... I can
try that too. (ADA4817.. looking at the spec sheet this may be more
than I can handle!)

George H.

 

[Joerg]

George Herold wrote:
Tim Wescott wrote:
On Fri, 18 Jan 2013 13:36:40 -0800, Joerg wrote: [...]





20kHz? That's like molasses. Why so low? And it should not cause it to
oscillate.
If he's already got substantial phase shift elsewhere, then the photodiode
rolling off would cause oscillation somewhere around 20kHz.
The fact that it happens at _exactly_ 20kHz just means that, sans
photodiode, he's got about 45 degrees of margin at 20kHz.
But where does all that phase margin fall through the cracks? Unless
everything rolls off fast, of course. 20kHz BW for the photodiode sounds
really low, unless it is one the size of a dinner plate.
--
Regards, Joerg
http://www.analogconsultants.com/-Hidequoted text -
- Show quoted text -
Grin, well not quite dinner plate size. ~6-7mm diam.
That should be a lot more zippy than 20kHz if connected to a somewhat
reasonable TIA. Or did you give it a hefty dose of Ambien?
Grin... I posted numbers for Tim. I can get a factor of 2 or so by
reverse biasing the PD, and maybe a factor of 5-10 with a faster
opamp.

Then it's time to flick the little switch on the Weller and do that

Reverse bias is easy if there is no compelling reason why you need to be
photovoltaic with the PD. Just keep the abs max limit in view. With the
opamp also keep in mind its slew rate. GBW means nothing if it can't
swing its tail around fast enough.

Grin, yup, that's fine for me. Not so easy for most customers.

Oh, I thought you were the R&D guy for this project.

I figure I should put a cap multiplier on the bias supply too. I
don't want all the power rail noise getting into the photodiode.

No worries, usually an RC filter is fine.

I've got a 'screaming' fast FET opamp in a drawer somewhere... I can
try that too. (ADA4817.. looking at the spec sheet this may be more
than I can handle!)

Usually just a fast enough low-noise amp would work. IIRC I used an
LM833 on a similar design, on account of it's nicely low 1/f knee. Good
stability around DC was important to the client and that one hit the
spot quite well.

 

[George Herold]

George Herold wrote:
George Herold wrote:
Tim Wescott wrote:
]
20kHz? That's like molasses. Why so low? And it should not causeit to
oscillate.
If he's already got substantial phase shift elsewhere, then the photodiode
rolling off would cause oscillation somewhere around 20kHz.
The fact that it happens at _exactly_ 20kHz just means that, sans
photodiode, he's got about 45 degrees of margin at 20kHz.
But where does all that phase margin fall through the cracks? Unless
everything rolls off fast, of course. 20kHz BW for the photodiode sounds
really low, unless it is one the size of a dinner plate.
--
Regards, Joerg
http://www.analogconsultants.com/-Hidequotedtext -
- Show quoted text -
Grin, well not quite dinner plate size. ~6-7mm diam.
That should be a lot more zippy than 20kHz if connected to a somewhat
reasonable TIA. Or did you give it a hefty dose of Ambien?
Grin... I posted numbers for Tim.  I can get a factor of 2 or so by
reverse biasing the PD, and maybe a factor of 5-10 with a faster
opamp.
Then it's time to flick the little switch on the Weller and do that
Reverse bias is easy if there is no compelling reason why you need to be
photovoltaic with the PD. Just keep the abs max limit in view. With the
opamp also keep in mind its slew rate. GBW means nothing if it can't
swing its tail around fast enough.

Grin,  yup, that's fine for me.  Not so easy for most customers.

Oh, I thought you were the R&D guy for this project.

I figure I should put a cap multiplier on the bias supply too.  I
don't want all the power rail noise getting into the photodiode.

No worries, usually an RC filter is fine.

I've got a 'screaming' fast FET opamp in a drawer somewhere... I can
try that too.  (ADA4817.. looking at the spec sheet this may be more
than I can handle!)

Usually just a fast enough low-noise amp would work. IIRC I used an
LM833 on a similar design, on account of it's nicely low 1/f knee. Good
stability around DC was important to the client and that one hit the
spot quite well.

--
Regards, Joerg

http://www.analogconsultants.com/- Hide quoted text -

- Show quoted text -

I don't know the lm833. 300nA 0f bias current doesn't look good.
.... 0.5pA/rtHz of current noise times 1 meg ohm.... 500nV!.. ouch.

George H.

 

[Joerg]

George Herold wrote:
George Herold wrote:
Tim Wescott wrote:
]
20kHz? That's like molasses. Why so low? And it should not cause it to
oscillate.
If he's already got substantial phase shift elsewhere, then the photodiode
rolling off would cause oscillation somewhere around 20kHz.
The fact that it happens at _exactly_ 20kHz just means that, sans
photodiode, he's got about 45 degrees of margin at 20kHz.
But where does all that phase margin fall through the cracks? Unless
everything rolls off fast, of course. 20kHz BW for the photodiode sounds
really low, unless it is one the size of a dinner plate.
--
Regards, Joerg
http://www.analogconsultants.com/-Hidequotedtext -
- Show quoted text -
Grin, well not quite dinner plate size. ~6-7mm diam.
That should be a lot more zippy than 20kHz if connected to a somewhat
reasonable TIA. Or did you give it a hefty dose of Ambien?
Grin... I posted numbers for Tim. I can get a factor of 2 or so by
reverse biasing the PD, and maybe a factor of 5-10 with a faster
opamp.
Then it's time to flick the little switch on the Weller and do that
Reverse bias is easy if there is no compelling reason why you need to be
photovoltaic with the PD. Just keep the abs max limit in view. With the
opamp also keep in mind its slew rate. GBW means nothing if it can't
swing its tail around fast enough.
Grin, yup, that's fine for me. Not so easy for most customers.

Oh, I thought you were the R&D guy for this project.

I figure I should put a cap multiplier on the bias supply too. I
don't want all the power rail noise getting into the photodiode.

No worries, usually an RC filter is fine.

I've got a 'screaming' fast FET opamp in a drawer somewhere... I can
try that too. (ADA4817.. looking at the spec sheet this may be more
than I can handle!)

Usually just a fast enough low-noise amp would work. IIRC I used an
LM833 on a similar design, on account of it's nicely low 1/f knee. Good
stability around DC was important to the client and that one hit the
spot quite well.

--
Regards, Joerg

http://www.analogconsultants.com/- Hide quoted text -

- Show quoted text -

I don't know the lm833. 300nA 0f bias current doesn't look good.
... 0.5pA/rtHz of current noise times 1 meg ohm.... 500nV!.. ouch.

Ok, if you absolutely must have tons of gain in that one stage then
you'll have to get something more expensive. I used much lower gain and
had that one in there for the low frequency portion, because of the 1/f
knee.

Old rule for me is that if the 1/f knee isn't at least visible in a
graph then chances are the marketeers didn't like it and banned it from
the datasheet