another ltspice question - laser diode current source

R

Rick

Jan 1, 1970
0
Hi again,

I have been playing around with part of a current source design Win and Tony
discussed on s.e.d. a year or two ago. I made some attempt at modeling it
in ltspice but I am really unsure of the best (cheapest, easiest,
fastest...pick two) way to modulate the supply. The problem is I really
would like to go between just under the diode lasing threshold current
(about 600mA in this case) to the operating current (which should be
programmable, but will mostly be around 1.8 to 2.2 amps for max power). If
you are at all interested, this is part of some attempt at a cheap power
supply for the very nice but very inexpensive Coherent 100mW single
longitudinal mode green laser heads showing up on ebay. Ideally you do not
want to exceed the pump diodes maximum current for any length of time (which
is about 2.4 amps for most of them).

Anyway, I experimented with several ways to servo the diode between 500mA
and some higher operating current at several khz but have not been happy
with the simulated results. Any hints? Here is one of the attempts in
ltspice format:

Version 4
SHEET 1 1160 1300
WIRE -16 304 32 304
WIRE 32 304 32 144
WIRE 32 304 64 304
WIRE 32 144 -16 144
WIRE -80 144 -128 144
WIRE -128 144 -128 288
WIRE -128 288 -80 288
WIRE -48 272 -48 256
WIRE -48 464 -48 336
WIRE 752 512 752 480
WIRE 240 80 128 80
WIRE 48 80 -128 80
WIRE -128 80 -128 144
WIRE 240 80 240 64
WIRE 240 -80 -528 -80
WIRE -704 64 -704 32
WIRE -704 -80 -704 -48
WIRE -704 -80 -704 -96
WIRE 240 -32 -208 -32
WIRE -208 -32 -208 224
WIRE -208 304 -208 320
WIRE -208 320 -80 320
WIRE -400 448 -368 448
WIRE -368 448 -368 272
WIRE -368 448 -256 448
WIRE -496 272 -496 432
WIRE -496 432 -464 432
WIRE -464 464 -624 464
WIRE -496 432 -496 576
WIRE -496 576 -336 576
WIRE -256 576 -208 576
WIRE -208 576 -208 464
WIRE -432 528 -432 480
WIRE -432 416 -432 400
WIRE -368 272 -400 272
WIRE -464 272 -496 272
WIRE -208 608 -208 576
WIRE -208 720 -208 688
WIRE -1008 496 -1008 432
WIRE -624 432 -624 464
WIRE -1008 576 -1008 608
WIRE 240 272 240 80
WIRE 144 304 160 304
WIRE 160 304 160 288
WIRE 160 288 192 288
WIRE 752 832 752 800
WIRE 752 736 752 688
WIRE 752 624 752 576
WIRE 752 368 704 368
WIRE -528 32 -528 -16
WIRE -208 368 -208 320
WIRE -1008 432 -992 432
WIRE -800 480 -800 656
WIRE -800 736 -800 784
WIRE -880 432 -912 432
WIRE -784 432 -720 432
WIRE -720 240 -720 320
WIRE -720 400 -720 432
WIRE -720 432 -624 432
WIRE -720 464 -720 432
WIRE -720 576 -720 544
WIRE -528 -80 -704 -80
WIRE 240 -32 240 -80
WIRE 240 -16 240 -32
WIRE 624 368 560 368
WIRE 480 368 240 368
WIRE 752 800 672 800
WIRE 672 800 672 672
WIRE 672 608 672 480
WIRE 672 480 752 480
WIRE 752 480 752 368
FLAG -48 256 Vcc
FLAG -704 64 0
FLAG -704 -96 Vcc
FLAG -432 528 0
FLAG -432 400 Vcc
FLAG -208 720 0
FLAG -1008 608 0
FLAG -48 464 0
FLAG 752 832 0
FLAG -528 32 0
FLAG -800 784 0
FLAG -720 576 0
FLAG -720 240 Vcc
SYMBOL pmos 192 368 M180
SYMATTR InstName M1
SYMATTR Value FDS9933A
SYMBOL res 160 288 R90
WINDOW 0 0 56 VBottom 0
WINDOW 3 32 56 VTop 0
SYMATTR InstName R1
SYMATTR Value 100
SYMBOL cap -16 128 R90
WINDOW 0 0 32 VBottom 0
WINDOW 3 32 32 VTop 0
SYMATTR InstName C1
SYMATTR Value 1n
SYMBOL res 144 64 R90
WINDOW 0 0 56 VBottom 0
WINDOW 3 32 56 VTop 0
SYMATTR InstName R3
SYMATTR Value 2.2k
SYMBOL res 224 -32 R0
SYMATTR InstName R4
SYMATTR Value 0.1
SYMATTR SpiceLine tol=1
SYMBOL voltage -704 -64 R0
WINDOW 123 0 0 Left 0
WINDOW 39 24 132 Left 0
SYMATTR SpiceLine Rser=.2
SYMATTR InstName V1
SYMATTR Value 5V
SYMBOL res -224 208 R0
SYMATTR InstName R5
SYMATTR Value 1k
SYMATTR SpiceLine tol=1
SYMBOL res -240 560 R90
WINDOW 0 0 56 VBottom 0
WINDOW 3 32 56 VTop 0
SYMATTR InstName R6
SYMATTR Value 10k
SYMATTR SpiceLine tol=1
SYMBOL res -224 592 R0
SYMATTR InstName R8
SYMATTR Value 10k
SYMBOL voltage -1008 480 R0
WINDOW 123 0 0 Left 0
WINDOW 39 0 0 Left 0
SYMATTR InstName V4
SYMATTR Value 2
SYMBOL Opamps\\LT1632 -48 240 R0
SYMATTR InstName U1
SYMBOL nmos -256 368 R0
SYMATTR InstName M2
SYMATTR Value FDS6630A
SYMBOL diode 736 512 R0
SYMATTR InstName D3
SYMATTR Value MURS320
SYMBOL diode 736 624 R0
SYMATTR InstName D4
SYMATTR Value MURS320
SYMBOL diode 736 736 R0
SYMATTR InstName D5
SYMATTR Value MURS320
SYMBOL res 720 352 R90
WINDOW 0 0 56 VBottom 0
WINDOW 3 32 56 VTop 0
SYMATTR InstName R2
SYMATTR Value 0.25
SYMBOL cap -544 -80 R0
SYMATTR InstName C3
SYMATTR Value 100µ
SYMBOL cap -400 256 R90
WINDOW 0 0 32 VBottom 0
WINDOW 3 32 32 VTop 0
SYMATTR InstName C2
SYMATTR Value 1n
SYMBOL res -896 416 R90
WINDOW 0 0 56 VBottom 0
WINDOW 3 32 56 VTop 0
SYMATTR InstName R9
SYMATTR Value 100
SYMBOL Opamps\\LT1632 -432 384 R0
SYMATTR InstName U2
SYMBOL voltage -800 640 R0
WINDOW 123 0 0 Left 0
WINDOW 39 0 0 Left 0
SYMATTR InstName V3
SYMATTR Value PULSE(0 5 0.1m 10n 10n .1m .2m 10)
SYMBOL nmos -880 480 R270
SYMATTR InstName M3
SYMATTR Value FDS6961A
SYMBOL res -736 448 R0
SYMATTR InstName R10
SYMATTR Value 10k
SYMBOL res -736 304 R0
SYMATTR InstName R11
SYMATTR Value 90k
SYMBOL ind 576 352 R90
WINDOW 0 5 56 VBottom 0
WINDOW 3 32 56 VTop 0
SYMATTR InstName L1
SYMATTR Value 2µH
SYMBOL cap 656 608 R0
SYMATTR InstName C4
SYMATTR Value .001µ
TEXT -288 896 Left 0 !.tran 0 2m 0 10n startup
TEXT 512 912 Left 0 ;Laser Diode ( Iop = 2 amps, Ithresh = 600mA )
TEXT 424 312 Left 0 ;Guess at cable inductance and resistance
TEXT -880 832 Left 0 ;Modulation Input
TEXT -1568 528 Left 0 ;Programming Voltage for Current Source
TEXT -1200 224 Left 0 ;Ideally would like to modulate between\ndiode
threshold current and programmed\ncurrent (ie 500mA <-> 2Amps)
TEXT -936 864 Left 0 ;At least 5khz would be nice
TEXT -856 -208 Left 0 ;For simplicity, source is fixed at 5V DC\n(easy
availability and minimized heating of output transistor)
LINE Normal 896 880 896 432
LINE Normal 608 880 896 880
LINE Normal 608 432 608 880
LINE Normal 896 432 608 432
 
W

Winfield Hill

Jan 1, 1970
0
Rick wrote...
I have been playing around with part of a current source design Win and Tony
discussed on s.e.d. a year or two ago. I made some attempt at modeling it
in ltspice but I am really unsure of the best (cheapest, easiest, fastest
...pick two) way to modulate the supply. ...
Click to expand...

If Tony and I were discussing it we no doubt posted viewable schematics
rather than spice listings. So why not copy and post the schematic, or
edit it to reflect the changes you're considering? No doubt there'a an
easy and approriate way to modulate the current source.

Thanks,
- Win

whill_at_picovolt-dot-com
 
R

Rick

Jan 1, 1970
0
If Tony and I were discussing it we no doubt posted viewable schematics
rather than spice listings. So why not copy and post the schematic, or
edit it to reflect the changes you're considering? No doubt there'a an
easy and approriate way to modulate the current source.

Thanks,
- Win
Click to expand...

Sure! Sorry about not doing that at first, but I am terrible at ascii
schematics :-( Give me a minute to find some sort of ascii schematic
editor.

Thanks,

Rick
 
J

Jim Thompson

Jan 1, 1970
0
Sure! Sorry about not doing that at first, but I am terrible at ascii
schematics :-( Give me a minute to find some sort of ascii schematic
editor.

Thanks,

Rick
Click to expand...

Post a binary to A.B.S.E

...Jim Thompson
 
J

John Larkin

Jan 1, 1970
0
Hi again,

I have been playing around with part of a current source design Win and Tony
discussed on s.e.d. a year or two ago. I made some attempt at modeling it
in ltspice but I am really unsure of the best (cheapest, easiest,
fastest...pick two) way to modulate the supply. The problem is I really
would like to go between just under the diode lasing threshold current
(about 600mA in this case) to the operating current (which should be
programmable, but will mostly be around 1.8 to 2.2 amps for max power). If
you are at all interested, this is part of some attempt at a cheap power
supply for the very nice but very inexpensive Coherent 100mW single
longitudinal mode green laser heads showing up on ebay. Ideally you do not
want to exceed the pump diodes maximum current for any length of time (which
is about 2.4 amps for most of them).
Click to expand...

5 volts can't be enough to drive three lasers in series.

John
 
R

Rick

Jan 1, 1970
0
John Larkin said:
5 volts can't be enough to drive three lasers in series.

John
Click to expand...

lol, no, that is just my silly excuse for a crude model of ONE high power
laser diode. It has approx. the voltage drop of 3 diodes in series or about
2.1 to 2.4V depending on current of course so I just used 3 high current
diodes in series (I didn't really examine the spice model of them
carefully).

Rick
 
J

John Larkin

Jan 1, 1970
0
lol, no, that is just my silly excuse for a crude model of ONE high power
laser diode. It has approx. the voltage drop of 3 diodes in series or about
2.1 to 2.4V depending on current of course so I just used 3 high current
diodes in series (I didn't really examine the spice model of them
carefully).

Rick
Click to expand...

Oh, OK, but 5 volts is still kinda low to give the fet and the sense
resistor and the wiring all enough to work with.

John
 
R

Rick

Jan 1, 1970
0
Oh, OK, but 5 volts is still kinda low to give the fet and the sense
resistor and the wiring all enough to work with.

John
Click to expand...

Hmmm, well that part of the circuit seems to be working OK. I am using rail
to rail op-amps and a logic FET with a low rds_on (in the tens of milliohms)
so there will not be much to worry about with FET drop. I could see with a
bjt there could be headroom problems...

As far as the wiring, I think I better keep it short anyway, less I end up
with a lot of inductance which would be bad for my little pump diode on
switching transitions.

But it doesn't really matter that much. I can use a bigger heatsink for the
FET and go with a 12V supply if it makes things work faster/better.

Rick
 
R

Rick

Jan 1, 1970
0
Winfield Hill said:
Rick wrote...

If Tony and I were discussing it we no doubt posted viewable schematics
rather than spice listings. So why not copy and post the schematic, or
edit it to reflect the changes you're considering? No doubt there'a an
easy and approriate way to modulate the current source.

Thanks,
- Win
Click to expand...

Ok, here is an attempt at the ascii schematic:

+5V
o-----------------------------+-----------------,
| |
+--------|-----------------+
| | |
| | / Rs
/ R1 | \ 0.1
\ 1.00k | / 1%
/ 1% | 2.2k | Vin R1
| ,------|--/\/\-----------+ Iout = --- ---
| | 1n | | R2 Rs
| +-||---|--, |
+5V | '---|-\| | 100 |--+S
1n | | | >--+-/\/\-----|| IRF9520 pfet
,---||-|--, +---+-|+/| |--+D
| | | | | |
+---|-\| | |--+ GND +-------L.D.---GND
Vin+ | | >--+---||
o---|---|+/| |--+ small nfet
0 to | | |
+5V '------|---/\/\---+ op-amps are rail to rail, 5V supply
GND 10K |
|
\
/ 10K
\
|
|
GND


Notes:

Laser diode has some capacitance across it and some small amount of wire
inductance and resistance in the cables. It is perfectly fine for the
compliance of the current source to be >2V. For that reason and simplicity
I removed the admittedly excellent sinks and such that would make this
a good general purpose current source. The laser diode needs to be modulated
between a low current of 500mA and the programmed current (not to exceed about
2.2 amps in this application). The modulation frequency should be as high as
reasonably possible.
 
R

Roy McCammon

Jan 1, 1970
0
Rick said:
Ok, here is an attempt at the ascii schematic:

+5V
o-----------------------------+-----------------,
| |
+--------|-----------------+
| | |
| | / Rs
/ R1 | \ 0.1
\ 1.00k | / 1%
/ 1% | 2.2k | Vin R1
| ,------|--/\/\-----------+ Iout = --- ---
| | 1n | | R2 Rs
| +-||---|--, |
+5V | '---|-\| | 100 |--+S
1n | | | >--+-/\/\-----|| IRF9520 pfet
,---||-|--, +---+-|+/| |--+D
| | | | | |
+---|-\| | |--+ GND +-------L.D.---GND
Vin+ | | >--+---||
o---|---|+/| |--+ small nfet
0 to | | |
+5V '------|---/\/\---+ op-amps are rail to rail, 5V supply
GND 10K |
|
\
/ 10K
\
|
|
GND


Notes:

Laser diode has some capacitance across it and some small amount of wire
inductance and resistance in the cables. It is perfectly fine for the
compliance of the current source to be >2V. For that reason and simplicity
I removed the admittedly excellent sinks and such that would make this
a good general purpose current source. The laser diode needs to be
modulated
between a low current of 500mA and the programmed current (not to exceed
about
2.2 amps in this application). The modulation frequency should be as
high as
reasonably possible.
Click to expand...

Easy way to get nice hi modulation freq is to shunt the LD with
a power fet. That way your source just happily runs at its
programed current and the shunt steals all or part of it.
 
R

Rick

Jan 1, 1970
0
Easy way to get nice hi modulation freq is to shunt the LD with
a power fet. That way your source just happily runs at its
programed current and the shunt steals all or part of it.
Click to expand...

Yes yes, I tried that. It was the first thought I had. Unfortunately, I could
not develop a shunt that would only steal Iop - Ithresh from the laser diode
unless I used another op-amp and sense resistor to ensure that the laser diode
always was just below threshold current even when the shunt was stealing the
rest of the current. If I just used a bjt or fet current sink for the shunt, I
would have to know the operating current of the diode in order to know what
value to sink with the shunt. This somewhat defeats the goal of having a
programmable current for the laser diode.

Having a current sensing variable shunt really did seem like the best way to go
because the voltage source for the whole thing would see a steady current draw
no matter what you did to the modulation. I am just not sure if that is the
best solution. I would rate solutions on a stability > cost > efficiency scale.

Rick
 
W

Winfield Hill

Jan 1, 1970
0
Rick wrote...
Winfield Hill wrote ...

Ok, here is an attempt at the ascii schematic:

+5V
o----------------------------+---------------,
| |
+--------|---------------+
| | |
| | / Rs
/ R1 | \ 0.1
\ 1.00k | / 1%
/ 1% | 2.2k | Vin R1
| ,------|--/\/\---------+ Iout = -------
| | 1n | | R2 Rs
| +-||---|--, |
+5V | '---|-\| | 100 |--+S Q2
1n | | | >--+-/\/\---|| IRF9520 pfet
,--||-|--, +---+-|+/| |--+D
| | | | | |
+--|-\| | |--+ GND +---L.D.---GND
Vin+ | | >--+---|| Q1
o---|--|+/| |--+ small nfet
0 to | | |
+5V '-----|---/\/\---+ op-amps are rail to rail, 5V supply
GND 10K |
\ R2
/ 10K
\
|
GND

Laser diode has some capacitance across it and some small amount of
wire inductance and resistance in the cables. It is perfectly fine
for the compliance of the current source to be >2V. For that reason
and simplicity I removed the admittedly excellent sinks and such that
would make this a good general purpose current source. The laser
diode needs to be modulated between a low current of 500mA and the
programmed current (not to exceed about 2.2 amps in this application).
The modulation frequency should be as high as reasonably possible.
Click to expand...

This looks like it's based on one of Tony's drawing, and perhaps
he'll have something good to say, but I do have a few comments.

First, considering the circuit as a high-frequency-capable current
source, it'll do poorly with the large-die-area high-capacitance
mosfets you've selected. For example, using an FDS6630 for Q1 has
Ciss = 460pF, compared to say 20pF for a 2n7000. In fact a 2n7000,
which itself is capable of conducting over 1A, badly degrades the
speed, so a small-signal BJT is really what's called for here.

Similarly using a FDS9933 for output MOSFET Q2, with Ciss = 600pF
and Coss = 300pF at 2V is a serious problem.

As for driving laser-diodes, I've always used considerably-different
circuits, tuned to the characteristics of the laser. For example
to protect the laser the voltage compliance should not much exceed
the voltage drop of the laser at its maximum current rating. Also,
most high-power laser driver circuits servo the laser's light-level.

Finally, as to really high speed, over several MHz, one easy method
is to directly drive the laser with an ac-coupled RF signal, using
the low-frequency servo simply to provide the dc bias setpoint.

Thanks,
- Win

whill_at_picovolt-dot-com
 
R

Rick

Jan 1, 1970
0
This looks like it's based on one of Tony's drawing, and perhaps
he'll have something good to say, but I do have a few comments.

First, considering the circuit as a high-frequency-capable current
source, it'll do poorly with the large-die-area high-capacitance
mosfets you've selected. For example, using an FDS6630 for Q1 has
Ciss = 460pF, compared to say 20pF for a 2n7000. In fact a 2n7000,
which itself is capable of conducting over 1A, badly degrades the
speed, so a small-signal BJT is really what's called for here.
Click to expand...

Ok. When I was playing with ltspice (new to that anyway) I was just using the
built in parts library for nfets. I guess I should have studied the datasheet
for the part first, heh heh. The same thing applies to Q2, although Tony's
original circuit had a IRF9520N with a Ciss of roughly 500pF and a Coss of about
400pF at VDS = -2V. Of course nobody specified high speed operation back then.
:)

Similarly using a FDS9933 for output MOSFET Q2, with Ciss = 600pF
and Coss = 300pF at 2V is a serious problem.

As for driving laser-diodes, I've always used considerably-different
circuits, tuned to the characteristics of the laser. For example
to protect the laser the voltage compliance should not much exceed
the voltage drop of the laser at its maximum current rating. Also,
most high-power laser driver circuits servo the laser's light-level.
Click to expand...

While bringing the light level into the loop sounds attractive, this is a very
large pump diode for a frequency doubled laser whose output (532nm) that is
sampled by a photosensor depends not only on the 808nm pump diode current, but
the temperature of the lasing crystal, non-linear crystal, and the pump diode
itself. Most of the controllers for these large pump diodes operate in constant
current mode without any type of light feedback, except via a microcontroller
for fine tuning of output power over seconds, minutes, hours.

I can always just use the small voltage compliance slow servoing current source
as depicted and do the modulation externally using a AO crystal or high speed
galvo.
Finally, as to really high speed, over several MHz, one easy method
is to directly drive the laser with an ac-coupled RF signal, using
the low-frequency servo simply to provide the dc bias setpoint.

Thanks,
- Win

whill_at_picovolt-dot-com
Click to expand...


Thanks again,

Rick
 
J

James Meyer

Jan 1, 1970
0
Sure! Sorry about not doing that at first, but I am terrible at ascii
schematics :-( Give me a minute to find some sort of ascii schematic
editor.

Thanks,

Rick
Click to expand...

I pulled the LTSpice listing into LTSpice with no problem. IMHO it's a
much better way to post schematics than one of those other schematic drawers.

Are you wedded to MOSFETs for your design? MOSFETs appear to make the
driver circuits less demanding at first glance, but when you start to take into
account their input capacitance, you'll find that bipolar transistors aren't
that much harder to drive.

Jim
 
You must log in or register to reply here.

Similar threads

G
LTSpice - How to Copy a Circuit To Another?
Replies
1
Views
27K
Helmut Sennewald
H
R
Another discrepancy (LTspice/switchcad3)
2
Replies
39
Views
5K
Robert Baer
R
R
re-wiring yet another christmas light string
Replies
1
Views
158
Harald Kapp
Harald Kapp
S
Another set of eyes to check my kicad schematic
Replies
5
Views
649
philipchambers
P
R
Superimpose Current Into Another Circuit Line?
Replies
1
Views
958
HarryA1
HarryA1
Top