PSPICE Diode Problem

[siliconmike]

PSpice doesn't seem to take diode recovery times into consideration.

It is showing that 1N4007 diode is turning on in 0.5ns from a reverse
bias of 12V to a forward 4A current condition.

This is surprising. Any hints?

(I'm using student version of OrCAD PSpice.)

Mike

 

[Helmut Sennewald]

PSpice doesn't seem to take diode recovery times into consideration.

It is showing that 1N4007 diode is turning on in 0.5ns from a reverse
bias of 12V to a forward 4A current condition.

This is surprising. Any hints?

(I'm using student version of OrCAD PSpice.)

Mike

Hello Mike,

Recovery time means the reverse recovery time.
Reverse recovery time is the time required to switch off.

Stimulate the diode with a forward current of 1A.
Then switch to a negative voltage. Now look how long
the current flows in the reverse direction.
It's important to know the test setup if you want
check against any number from the datasheet.

A possible test circuit:
I_forward=1A, I_backward=1A
Voltage source _/-\_/-\_.... +10V/-10V, 50us on, 100us period
Series resistor with 9 Ohm
Diode connected to resistor and GND(0V).
Now you should get a few microseconds of turn-off delay.
I checked this with a 1N4002 model.

Best regards,
Helmut


PS: I use LTspice. It's an unlimited and free of charge
SPICE simulator with GUI(schematic entry and waveform viewer).


This is the test circuit: "1N4002_test.asc"
Just copy and paste it into a file named 1N4002_test.asc .
Remove additional line feeds caused by the mail programs.
Now you can open and run it with LTspice.

Version 4
SHEET 1 1352 680
WIRE -16 64 -128 64
WIRE 144 64 64 64
WIRE 512 64 144 64
WIRE -128 112 -128 64
WIRE 512 112 512 64
WIRE -128 240 -128 192
WIRE 512 240 512 176
WIRE 512 240 -128 240
WIRE -128 272 -128 240
FLAG -128 272 0
FLAG 144 64 IND1
SYMBOL diode 496 112 R0
SYMATTR InstName D1
SYMATTR Value DI_1N4002
SYMBOL res -32 80 R270
WINDOW 0 32 56 VTop 0
WINDOW 3 0 56 VBottom 0
SYMATTR InstName R1
SYMATTR Value 9
SYMBOL voltage -128 96 R0
SYMATTR InstName V1
SYMATTR Value PULSE(-10 11 1u 10n 10n 50u 100u)
TEXT -144 384 Left 0 !.MODEL DI_1N4002 D ( IS=76.9p RS=42.0m BV=100 IBV=5.00u CJO=39.8p M=0.333 N=1.45 TT=4.32u )
TEXT -136 -24 Left 0 !.tran 0 300u 0 .01u
TEXT -136 -64 Left 0 ;Plot I(D1)
TEXT -144 344 Left 0 ;*SRC=1N4002;DI_1N4002;Diodes;Si; 100V 1.00A 3.00us Diodes, Inc. diode

 

[colin]

PSpice doesn't seem to take diode recovery times into consideration.

It is showing that 1N4007 diode is turning on in 0.5ns from a reverse
bias of 12V to a forward 4A current condition.

This is surprising. Any hints?

(I'm using student version of OrCAD PSpice.)

I think your spice diode model only uses turn off delay,
also it doesnt use the fall time wich is sometimes
more of an issue when using soft recovery diodes or SRD.

Colin =^.^=

 

[Spehro Pefhany]

PSpice doesn't seem to take diode recovery times into consideration.

It is showing that 1N4007 diode is turning on in 0.5ns from a reverse
bias of 12V to a forward 4A current condition.

This is surprising. Any hints?

Find a better set of model parameters. Some have TT (transit
time)unspecified (0) and others use 5.76 or 5.7 usec and Colin just
posted one with TT=4.32usec.

(I'm using student version of OrCAD PSpice.)

Mike

Best regards,
Spehro Pefhany

 

[Jim Thompson]

Find a better set of model parameters. Some have TT (transit
time)unspecified (0) and others use 5.76 or 5.7 usec and Colin just
posted one with TT=4.32usec.



Best regards,
Spehro Pefhany

The latest PSpice diode.lib has TT=5.7us, but some time back there was
NO TT specified, this defaults to zero.

...Jim Thompson

 

[siliconmike]

Hello Mike,

Recovery time means the reverse recovery time.

But there is forward recovery time too. I just ran LTSpice. Looks
decent. I'll take some days to check out your circuit. Thanks for the
pointer.

 

[siliconmike]

I think your spice diode model only uses turn off delay,
also it doesnt use the fall time wich is sometimes
more of an issue when using soft recovery diodes or SRD.

Colin =^.^=

Use proportinal font. These are the diode params. As seen TT is 5.7uS.

D1N4748 D1N4007 D1N4001
IS 17.490000E-15 14.110000E-09 14.110000E-09
N 1.984 1.984
ISR 2.024000E-09
IKF 94.81 94.81
BV 22 1.500000E+03 75
IBV .16996 10.000000E-06 10.000000E-06
NBV 1.315
IBVL 7.007300E-15
NBVL 1.2735
RS 7.976 .03389 .03389
TT 5.700000E-06 5.700000E-06
CJO 49.000000E-12 25.890000E-12 25.890000E-12
VJ .75 .3245 .3245
M .2829 .44 .44
TBV1 840.910000E-06

 

[siliconmike]

The latest PSpice diode.lib has TT=5.7us, but some time back there was
NO TT specified, this defaults to zero.

I've posted the diode params in another post. Btw, I'm going to have to
take some time to study your drawing regarding the high-side mosfet
switch due to my inexperience in certain transistor configurations. I
appreciate it though.

Mike

 

[Jim Thompson]

I've posted the diode params in another post. Btw, I'm going to have to
take some time to study your drawing regarding the high-side mosfet
switch due to my inexperience in certain transistor configurations. I
appreciate it though.

Mike

Feel free to ask questions.

...Jim Thompson

 

[colin]

Use proportinal font. These are the diode params. As seen TT is 5.7uS.

D1N4748 D1N4007 D1N4001
IS 17.490000E-15 14.110000E-09 14.110000E-09
N 1.984 1.984
ISR 2.024000E-09
IKF 94.81 94.81
BV 22 1.500000E+03 75
IBV .16996 10.000000E-06 10.000000E-06
NBV 1.315
IBVL 7.007300E-15
NBVL 1.2735
RS 7.976 .03389 .03389
TT 5.700000E-06 5.700000E-06
CJO 49.000000E-12 25.890000E-12 25.890000E-12
VJ .75 .3245 .3245
M .2829 .44 .44
TBV1 840.910000E-06

Yes this stands for the Transit Time I beleive, however it seems there are
different conventions for TT, it is also used as Transition Time for SRDs
wich is different again, this confused me no end when I was trying to
simulate SRDs.

Generaly it is the time taken for all the carriers that allow current flow
in the PN junction to recombine wich is a random process. This cuases the
'turn off' delay.

However when the junction is initialy forward biased carriers are injected
directly into the junction, the time this takes is very much smaller and so
cunduction starts much more quickly, thus Turn on delay is usualy so short
as to not be an issue and does not apear in the above model.

I beleive the effect of junction capacitance usualy makes turn on delay
unoticable anyway.

Colin =^.^=

 

[siliconmike]

Hello Mike,

Recovery time means the reverse recovery time.
Reverse recovery time is the time required to switch off.

Stimulate the diode with a forward current of 1A.
Then switch to a negative voltage. Now look how long
the current flows in the reverse direction.
It's important to know the test setup if you want
check against any number from the datasheet.

A possible test circuit:
I_forward=1A, I_backward=1A
Voltage source _/-\_/-\_.... +10V/-10V, 50us on, 100us period
Series resistor with 9 Ohm
Diode connected to resistor and GND(0V).
Now you should get a few microseconds of turn-off delay.
I checked this with a 1N4002 model.

Best regards,
Helmut


PS: I use LTspice. It's an unlimited and free of charge
SPICE simulator with GUI(schematic entry and waveform viewer).


This is the test circuit: "1N4002_test.asc"
Just copy and paste it into a file named 1N4002_test.asc .
Remove additional line feeds caused by the mail programs.
Now you can open and run it with LTspice.

<snip>

I understand, it does show a pronounced reverse recovery time of about
4uS. What I'm wondering is that why is there no forward recovery period
?

When the diode turns on, the voltage across diode seems to go abruptly
from -10V to 1V and the current from 0A to 11A.

In my case, the diode 1n4007 is used as a snubber parallel to a 40mH
12Ohms Inductor to which 12 volts are switched using a relay.

Therefore when the supply to the inductor is cut (assuming in a period
of 1nS) , the diode turns on absorbing the current. But I'd like to
focus my study on what happens "during the time" the diode turns on.

-Mike

 

[colin]

Therefore when the supply to the inductor is cut (assuming in a period
of 1nS) , the diode turns on absorbing the current. But I'd like to
focus my study on what happens "during the time" the diode turns on.

The turn on and turn off actions of a PN junction diode are very different
indeed.

Its like turning on a tap ... as soon as its on water flows imediatly,
but as you turn it off it takes a while for the water to soak away.

If you imagine the water is actualy the charge carriers that permit
conduction (in either direction) its a bit closer to whats realy going on.

The reason this turn off delay acts like this is that PN junction use
minority charge carriers for conduction accros the barrier, and these stay
in the junction untill an electron finds an empty hole or a hole finds a
floating electron,
they dont just simply disapear out of the terminals when you reverse the
voltage.

A schotky diode however doesnt use minority carriers, and is therefroe a lot
faster at turning off.
as soon as you reverse the voltage they simply move out the elctrodes.

This is my simplified understanding of a very mathmatically complex process
so somone correct me if im wrong please.

Colin =^.^=