Hamamatsu / General Chatting / Forums

Forums

4hv.org :: Forums :: General Chatting

« Previous topic | Next topic »

Hamamatsu

previous 1 2 3 next

Move Thread

LAN_403

klugesmith

Tue Jul 16 2013, 05:41PM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

Can you use a Position Sensitive Detector (PSD, sometimes called Position Sensitive Diode)?

It's a linear photosensor with no pixels.
Photocurrent divides between the two ends according to where the light strikes the detector.

You can read the position of a bright spot with micron resolution and excellent linearity.
The response time for position changes can be much less than 1 microsecond.
In some applications, it's hard to match that by digital processing of a sensor output
that is needlessly discrete in space, time, and voltage.
Here's an analog distance sensor that uses a PSD, and is based on optical parallax.

I once got some new PSD's from Hamamatsu for classroom demo use.
Ordered them on behalf of a licensed nonprofit educational organization.
They gave us 10-unit pricing for a 3-unit order, which got us to the $100 minimum order IIRC.

Linas

Tue Jul 16 2013, 06:41PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

klugesmith wrote ...

Can you use a Position Sensitive Detector (PSD, sometimes called Position Sensitive Diode)?

I have multiple peaks in spectrum ( it like Gauss modulated with sinus ) and i need to make DFT from that data to get phase, but maybe where is potential in thing like that, i need to sit down, and think.
hm, i could get rid of few peaks, but i will still be left with one or two, with near identical intensity,but damn, that could mean i could do that in analog, that would give true real time performance

Spectrum looks like this (need some imagination)
Jpb407039fig09

2Spoons

Tue Jul 16 2013, 10:48PM

Registered Member #2939 Joined: Fri Jun 25 2010, 04:25AM
Location:
Posts: 615

This might work for you. Its a CMOS linear camera, 1024 pixels, clockable up to 30MHz. It also has selectable pixel binning so you can run it as 512, 256 or 128 pixels. Active area is 8mm long.
I'll try to find out where we get them from.
]elis1024_pds0004revj.pdf[/file]

...

Wed Jul 17 2013, 04:41AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

I am curious Linas, which group are you working for, and more specifically are you using this as a way to look at the CEP of a laser pulse?

If you are in fact looking at CEP, then there are much better ways to do it than by looking at f-2f interferograms, you only need to lock a single fringe (you can of course measure the whole spectrum and if you are feeling ambitious lock it in a slow feedback loop as well...)

Linas

Wed Jul 17 2013, 06:14AM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

... wrote ...

I am curious Linas, which group are you working for, and more specifically are you using this as a way to look at the CEP of a laser pulse?

If you are in fact looking at CEP, then there are much better ways to do it than by looking at f-2f interferograms, you only need to lock a single fringe (you can of course measure the whole spectrum and if you are feeling ambitious lock it in a slow feedback loop as well...)

my system for slow feedback loop give Yb:KGW laser 130mrad stability ( world best for this type of lasers as far i we know) now i want to make fast feedback loop so no other electronic is needed to stabilize laser, i know that noise of cep phase is more than 20-30KHz, so i need double or triple that speed in order to lock cep (well, try to lock, no one knows is that is even possible)
Normal system use diode and pll to get beating note from 70MHz pulse train. I have only 200KHz from RA amplifier
Main problem is that my system can track cep only in 2pi range, original system based on pll can do that to 128pi

2Spoons wrote ...

This might work for you. Its a CMOS linear camera, 1024 pixels, clockable up to 30MHz. It also has selectable pixel binning so you can run it as 512, 256 or 128 pixels. Active area is 8mm long.
I'll try to find out where we get them from.

Hm, it looks like this one know how to average 8 pixels to get 1 pixel in video output, and with 30MHz clk this would be same readout speed as hamamatsu 1024 pixel ccd, because it has 8 readouts ports, but how can i get one ?
Interesting, when i was searching detector i came a cross this company, and i asked for best detector for my application, and they suggested DLIS2K (but that was no good), but clearly what i was needed is ELIS1024. that why i was searching elsewhere

...

Wed Jul 17 2013, 03:53PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

You might want to read papers of people doing CEP locking of Ti:S if you haven't already (people do it for regen amps as well, at much lower rep rates than your system), I do not think that you will really benefit from looking at the whole interferogram as opposed to locking just a single fringe (consider that at the end of the day you need to generate a single error signal to feed back into your modulator, the added information of the rest of the fringes is not going to help). Especially when you are already locked to within 130mrad you need not worry about the modulus 2*pi, since you already know that it is on the correct fringe.

You also mention than you have only access to the output of the RA, can't you stabilize the seed as well? I would expect that most of the high frequency noise is coming from the seed oscillator (although I am not familiar with yb:kgw, I work with nd:ylf and ti:s), so if you stabilize the seed well you may only need your slow loop after the RA.

In any case, the reason I ask is that I am currently working on the laser system that was used to generate the plots you posted

Linas

Wed Jul 17 2013, 04:14PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

That is not that simple. First i have Yb:KGW Laser (not CEP phase stable Ti:s), without any feedback it will supply random CEP. From oscilator we take each pulse, measure CEP with f-2f and with photodiode we get beating note, and lock phase that it runs at constant pi/2, so any 4n impulse will have same phase.

with pulse picker we take phase stable pulses (4n), and amplify that with RA , and from that i measure again CEP phase, which is drifted from oscillator phase. so if i turn off first feedback loop ( the very very fast one ) i get random phase, but we measured that phase noise drops very sharply after 50KHz or so. So if my RA runs at 200KHz ( max speed, but with new PP it is possible to get 1MHz pulse train) and if i can measure single shot phase (and yes, i can do that at 58KHz so far) , and generate feedback to my pump diode, i could use only second f-2f interferometer with my electronics to lock whole system, and yes, i can use only one fringe, but since i use linear detector, i get lot of dc near low frequency, so 3-10 fringes will keep me away from dc

This detector can read single laser pulse at 62.5KHz

2Spoons

Thu Jul 18 2013, 01:16AM

Registered Member #2939 Joined: Fri Jun 25 2010, 04:25AM
Location:
Posts: 615

We bought the ELIS chips direct from panavision, it seems.
Check your PM's.

2Spoons

Thu Jul 18 2013, 11:21PM

Registered Member #2939 Joined: Fri Jun 25 2010, 04:25AM
Location:
Posts: 615

Sending you some of these. Sorry about poor quality pic - my phone is about 8 years old.

Board contains camera and an amplifier. Configured on board for 1024 pixels, but should be easy enough to hack. The amp isn't high frequency, so you may want to change it.

Coming via DHL. Hope they work for you

Linas

Fri Jul 19 2013, 11:20AM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

i will make my own analog front end with low noise opamps with 16b adc.

At 30MHz, not lot of signal will be in output

previous 1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob