The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: Robert Baer
X-Mailer: Mozilla 4.75 [en] (Win98; U)
Subject: Re: Photodiodes in photovoltaic mode, speed considerations.
References: <3DACB831.F814C3D6@webaccess.net> <3DAD0481.A4004E95@NOSPAM.univ-lille1.fr> <3DAD0A0A.1F7BF2DD@webaccess.net> <3DAD2632.9E6DDA83@earthlink.net> <3DAD5AC6.642D726C@webaccess.net>
Date: Thu, 17 Oct 2002 05:38:01 GMT
NNTP-Posting-Date: Wed, 16 Oct 2002 22:38:01 PDT
Organization: EarthLink Inc. -- http://www.EarthLink.net
Chuck Simmons wrote:
> Robert Baer wrote:
> > Chuck Simmons wrote:
> > >
> > > "s.d." wrote:
> > > >
> > > > My two cents...
> > > >
> > > > If you want speed you may buffer your photodiode with a transimpedance
> > > > amplifier.
> > > > That's the usual way. But still you'll need negative supply.
> > > > At least you can try to put a low impedance load to your diode at the
> > > > expense of lower gain.
> > > > Mind that in voltage mode the diode response is logarithmic with light
> > > > power.
> > >
> > > A negative supply is outside of allowed parameters. If it costs $0.10 it
> > > is outside of allowed parameters. A load on the diode is not allowed
> > > because because I cannot specify a buffer that operates that close to
> > > the negative rail. I figure I can just live with 100mV off the rail and
> > > since I see, in devices so far, 150 to 200 mV at desired power, I'm OK.
> > > The problem is speed for clock selection for the digital loop which
> > > depends on the diode speed in photovoltaic mode. I know some clock
> > > limits for some devices but I can't predict because I only get reverse
> > > specs and I don't know how these or area relate to speed.
> > >
> > > BTW, we did use loading for a while but it did not work with a
> > > population of parts we had because the effect of the load on voltage was
> > > less predictable and caused laser damage (i.e., the things failed).
> > > Unloaded photovoltaic is safe it would seem from our experience. Some of
> > > the population of lasers tend to require close to burnout current
> > > levels. This is within spec for the lasers but a real pain for me.
> > >
> > > Chuck
> > One can bias the diode, either forward or reverse; millivolts or volts
> > (at least reverse) and have zero need for a negative supply.
> > At least with a reverse bias, the varactor-like capacitance is a lot
> > lower than forward (photo voltaic generator) mode.
> > Another advantage is that this gives a capacitance less dependent on
> > doping and other process parameters, and closer to packaging parameters.
> > The diode resulting voltage shift can drive a comparitor, or opamp
> > acting as a comparitor; or in a bridge-like circuit driving the opamp
> > (with or withot feedback).
> > And there are a lot of opamps that go rail to rail with common mode
> > input specs wider tan the supplies (many are SOIC).
> The issue is not redesigning what I have which works but rather
> predicting performance of the monitor diode in photovoltaic mode given
> only reverse specifications for the diode and possibly dimensions that I
> can measure under a microscope. In fact, the question is still one of
> predicting the speed of the diode so that I can select a clock for the
> power control device. The possible clock range is large so I have
> considerable freedom to select it for the common cathode devices.
> BTW, the reason I want to select the clock before powering a type of
> laser is because the lasers are easy to burn out. I have piles of dead
> ones some of which are victims of the clock being set too fast for them.
> Most of the other dead ones seem to be ESD failures but that's another
> ... The times have been,
> That, when the brains were out,
> the man would die. ... Macbeth
> Chuck Simmons firstname.lastname@example.org
Well, if you insist on that mode and the variations, be my guest.
There is absolutely *NO* way to determine the speed from datasheet
specs, unless that is given as a measured (ie: specified) parameter
under identical conditions that you use (an i do not think *that* will
Therefore, use the absolutely *slowest* clock speed that is useful for
the slowest device you can find, and then slow down the clock by a
factor of 10 if practical; but no less than a factor of 2 slower for
safety (since you indicate a fast clock rate is nasty).
If that does not suit the application, too bad; it is a problem you
created by refusing to use circuitry that is faster and less dependent
on diode doping and size.
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup