Author Topic: Is it possible to model a photodiode PIN InGasAS,in MicroCAP 12 or LTspice?  (Read 8194 times)

0 Members and 1 Guest are viewing this topic.

Offline WeisesteTopic starter

  • Newbie
  • Posts: 7
  • Country: ro


I want to know if it is possible to model a photodiode InGasAs panchromatic PIN , in accord with datasheet, in simulator SPICE MicroCAP 12 or LTspice.Because the supllier where I boughit it has not a SPICE model.And if it is possible, which are those steps. Here it is data-sheet https://www.tme.eu/en/details/ig17x1000s4i/photodiodes/laser-components/ And below I will attach a electric schematic, a transimpedance amplifier with phototdiode, used like receiver of smoke.I want to monitor graphically the impulse of photodiode in function of output of transimpedance for sensibilty. There is type of photodiode,which I found it in library Micro-CAP 12, it is simpe photodiode,which does not follow in accord in datasheet.There is photodiode macro.

Thank you. :)
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 15157
  • Country: de
The photodiode macro model should be reasonably OK. One may have to adjust the capacitance. If used at different voltages one would also adjust the corresponding parameters - not sure which are the ones used. For just the circuit it should be OK to adjust at the voltage used, no need to care about the voltage dependence.

The shunt resistance may be important when it comes to noise performance, but it should not effect the pulse response.
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13218
There's not enough data in the datasheet to model it directly, but if you took a sample one, totally obscured its window then measured its Vf vs If and Ir and Vr characteristics, you could curve fit a SPICE diode model (modified Shockley equation) to it.  Its then just a matter of developing a control function for the arbitrary current source providing the photocurrent to match the fig.9 'Linearity' curve. Unfortunately trimming the whole model to match the fig.8 'Sample Pulse Response' is more problematic as junction capacitance and thus speed varies with bias, and only the zero bias response is given.
« Last Edit: June 25, 2020, 09:20:07 am by Ian.M »
 

Offline WeisesteTopic starter

  • Newbie
  • Posts: 7
  • Country: ro
Here is the datasseet.   https://www.tme.eu/Document/726cc5089c90f506f2d9a008ec28a891/ig17.pdf. Please any suggestions which help me to reach at any result, they will be great helpful me. :) :(  .
 

Offline WeisesteTopic starter

  • Newbie
  • Posts: 7
  • Country: ro
I treid, also, Do I follow the equivalent circuit, from this doecumt.Will I have any chance to get the respective results  ?
http://www.osioptoelectronics.com/application-notes/an-photodiode-parameters-characteristics.pdf
 

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2766
  • Country: ca
Hi,

I have had a quick look at this.

If I start with a very simple model for the photodiode, a current source in parallel with the junction capacitance.

I get the following model:




I am showing two test circuits, the OP amplifier and an ideal amplifier.

The result is:




The performance is dominated by the amplifier, not the PD.

I would start by reducing the value of the emitter resistor on the first transistor.

I have attached my LTspice model.

I did not model the BFP420 transistors, because they have two emitter connections, they are including package parasitic inductance etc.


Regards,
Jay_Diddy_B
* TIA Model.asc (3.69 kB - downloaded 308 times.)
« Last Edit: June 26, 2020, 02:56:13 am by Jay_Diddy_B »
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7045
  • Country: nl
I would start by reducing the value of the emitter resistor on the first transistor.

Why? Small signal doesn't see the resistor. Reducing it increases the bias current, but BFP420 will have a fair bandwidth even at 60 microamps and the low bias current will also decrease the current noise going into the base ... which might be the intention?
« Last Edit: June 25, 2020, 02:28:17 pm by Marco »
 

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2766
  • Country: ca
Hi,

I see this:




So dropping Re improves the BW.

Transient Response




This with the 2N2369 transistors, the BFP420 may be better?

Jay_Diddy_B
« Last Edit: June 25, 2020, 02:38:59 pm by Jay_Diddy_B »
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7045
  • Country: nl
Yah, these high frequency transistors will have far better bandwidths at low bias currents. SiGe even moreso.
 

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2766
  • Country: ca
Hi,



This is an LTspice model including the BPF420 models from Infineon. I had to make a symbol for an NPN BJT with two emitters. This is included in the zipfile.


This is the result of stepping Q1 emitter resistor.





The zipfile is attached.

Regards,
Jay_Diddy_B

* eevblog tia pd.zip (1.97 kB - downloaded 192 times.)
« Last Edit: June 26, 2020, 02:52:53 am by Jay_Diddy_B »
 

Offline WeisesteTopic starter

  • Newbie
  • Posts: 7
  • Country: ro
Thank you for feedback.What program do you use?And my question it is about if it possible to model a InGasAs panchromatic PIN photodiode, in accord with datasheet, in simulator SPICE MicroCAP 12 or in program LTspice.I want to monitor graphically the impulse of photodiode in function of output of transimpedance for sensibilty,the form it must be, an optical rectangular impulse, atthe input of photodiode, and that signal I wnat to monitor with the signal output ofg transimpedance amplifier.
At the third pin of photodiode, a voltage source in impuls.And this is the results graphically.
« Last Edit: June 26, 2020, 05:11:36 am by Weiseste »
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7045
  • Country: nl
The photodiode model seems fine, but the simulator seems to say your amplifier oscillates and has a weird delay (presumably it's saturating somewhere, reduce the input signal).
 

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2766
  • Country: ca
Hi,

Using the information found here:

https://www.lasercomponents.com/fileadmin/user_upload/home/Datasheets/lc/kataloge/ir_components.pdf




And modeling:
Quantum Efficiency



Non Linearity



These were modelled with polynomials obtained by curve fitting the datasheet.

LTspice Model



Voltage source V1 is the optical power that falls on the photodiode. The scaling is 1V represents 1W.

There are parameters for wavelength and Junction Capacitance.

and I get the following result:



The LTspice model is attached.


Regards,
Jay_Diddy_B

* PD Model.asc (3.48 kB - downloaded 233 times.)
« Last Edit: June 26, 2020, 01:44:10 pm by Jay_Diddy_B »
 

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2766
  • Country: ca
Hi,

These are the curves used:



and



Regards,
Jay_Diddy_B
 

Offline WeisesteTopic starter

  • Newbie
  • Posts: 7
  • Country: ro
Thank you for this feedback.Can you help me please to integreat that model of photodiode in th electrical schematic.I want to monitor graphically the impulse of photodiode in function of output of transimpedance for sensibilty,the form it must be, an optical rectangular impulse, at the input of photodiode, with a perioad of 20ns and , the interval of voltage must be between -20V and +20V, and the optical power must be between 100 and 100mW/cm^2.

Thank you.
 

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2766
  • Country: ca
Hi,

The sensor has a diameter of 1mm.

The area is Pi x d / 4 = 0.78 mm2.

You have a power density of between 100W and 100mW per cm2

So this is 1W to 1mW per mm2

The datasheet shows the photodiode is non-linear above 8mW:



It doesn't look like you have the right sensor. I suggest that you talk to the manufacturer.

Regards,
Jay_Diddy_B
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7045
  • Country: nl
Or use an attenuator.
 

Offline Weiiseste

  • Newbie
  • Posts: 4
  • Country: ro
I am sorry. I  wanted to say. between 100W and 10uW/m^2. :)

Thank you.
Dragos.
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
"between 100W and 10uW/m^2"

LOL
If that's correct you'll have to lower that range quite a bit, these transistor circuits can only keep the pulse shape over a range of about 100:1.
I don't know what the phododiode's currents are or the pulse widths from this thread.
.  That took much longer than I thought it would.
 

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2766
  • Country: ca
Hi,

"between 100W and 10uW/m^2"

LOL
If that's correct you'll have to lower that range quite a bit, these transistor circuits can only keep the pulse shape over a range of about 100:1.
I don't know what the phododiode's currents are or the pulse widths from this thread.


Yes that is what I was thinking:

10uW m-2

On the area of the sensor 0.78mm2

0.78E-6 m-2

10E-5 x 0.78E-6 = 7.8E-11

or 78 pA

Jay_Diddy_B

« Last Edit: June 28, 2020, 01:00:47 pm by Jay_Diddy_B »
 

Offline Weiiseste

  • Newbie
  • Posts: 4
  • Country: ro
Or. How can it be better for this situation?
 

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2766
  • Country: ca
Or. How can it be better for this situation?

Are you sure that the power level is 10 Micro Watts per metre2

10E-6 m2 ?

Moonlight is about 600uW m2 (Full Moon)

It is is tough to measure 78pA at DC

If you increase the area of the sensor you will get more current, but you get more capacitance.
You also have to consider dark current and noise.

Regards,
Jay_Diddy_B
 

Offline Weiiseste

  • Newbie
  • Posts: 4
  • Country: ro
Okay. Which are your advice?. What considerations must I take?. To be better. :)
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
Or. How can it be better for this situation?

You could give us some clue of the arrangement of the IR emitter and photodiode.

The 1st post says "a transimpedance amplifier with photodiode, used like receiver of smoke.".
If it's anything like these to detect the light scattered off smoke the photo currents will be at the very low end and the amp a high gain low bandwidth, probably an op amp.

www.safelincs.co.uk/smoke-alarm-types-optical-alarms-overview/
www.mavili.com.tr/en/smoke-detectors/1179-how-optical-smoke-detectors-work.html
« Last Edit: July 06, 2020, 10:25:27 am by StillTrying »
.  That took much longer than I thought it would.
 

Offline Weiiseste

  • Newbie
  • Posts: 4
  • Country: ro
Photo sensing circuits such as transimpedance amplifiers (TIAs) are useful in precision systems such as computed tomography (CT) scanners, blood analyzers, and smoke detectors.My circuit has the rooll to detect the smoke, for grenades.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf