DPRG List  

[DPRG] Working with very small voltage margins & a op amp question

Subject: [DPRG] Working with very small voltage margins & a op amp question
From: Rick Bickle rbickle at intconsys.com
Date: Thu Aug 20 14:03:05 CDT 2009



Use the large capacitors and inductors on the power supply rails and not
on the signal to the laser circuitry.

Use a linear power supply instead of a switching power supply. Add some
filtering on the AC side. (i.e. balun or chokes)

Keep the wires or PCB tracks short between the filtered supply and the

Keep all associated op-amp components as close to the chip as possible.

For a PCB, put a ground plane underneath the op-amp chips and associated
components. Tie all component grounds to the plane locally.

For a prototype, use a solid copper board with the copper as a ground
plane. Solder the chip grounds directly to it as well as all component

Be sure to bypass all high speed switching devices with small monolithic
capacitors close to each device.

For really sensitive circuits, you can put a faraday shield (i.e. metal
housing) around the op-amp section.


Not sure I understand how the op-amps are going to set the current of
your diode, but there are op-amps available with high current open
collector outputs. These will give a very linear response.

I recommend the book "IC Op-Amp Cookbook" by Walter Jung, available from
Amazon. - and of course the bible "The art of Electronics" by Horowitz
and Hill.


Good Luck,





From: dprglist-bounces at dprg.org [mailto:dprglist-bounces at dprg.org] On
Behalf Of Ray Xu
Sent: Wednesday, August 19, 2009 6:35 PM
To: dprglist at dprg.org
Subject: [DPRG] Working with very small voltage margins & a op amp


Hi guys, 

I'm now finalizing my schematic for my laser diode controller, and I
have some concerns for the design of it.  To start, part of my design
looks like this: a digital voltage reference generator that is connected
to a op amp/MOSFET version voltage-current converter.  The laser diode
current would be digitally programmable through the voltage reference
generator by adjusting the gain (of the op amps) and adjusting a digital
potentiometer (that drives a LM4121), where the voltage from the voltage
reference generator to the laser diode current ratio is 1V:100mA.

So for example if I was going to drive a small 25mA laser diode; that
would mean the reference voltage generator's output would equal 250mV,
and a +/- 10mV change would mean the laser being over-currented by +/-

My concern here would be that, in the real world, the electronic noise
would either offset the "drive" voltage, or it would generate transient
spikes enough to blow the laser diode (they are sensitive to being
over-driven, BTW).  On paper, my design would work flawlessly so far.  

Also, I cannot use any heavy (large inductors/capacitors, etc) filtering
since the signal has to be modulated at 1KHz (part of the calibration

How do I go about this (besides the usual capacitors and inductors...)?
How should I isolate/decouple/bypass all of my supply rails and the
signals around the "sensitive" area?  How should I lay out my PCB

So impractical... :(

Also a quick question about op amps: For example I'm using the
LMC6062...Does the input or output capacitance make it unstable?  As far
as the datasheet says, I think it means both makes it unstable.  I don't
really get it; since the capacitance at the input is not within the
"feedback loop", it should be able to handle any amount of capacitance.
Can anyone please correct me here?

Thanks everyone for all the help

Ray Xu  

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://list.dprg.org/pipermail/dprglist/attachments/20090820/9e609d4e/attachment.html

More information about the DPRG mailing list