Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Spice Model

Status
Not open for further replies.
Maybe they should give the Master's Degree to Mr. Al. Pretty cool if you can get the folks here on the forum to do you masters research for you! Where the hell is your advisor?
 
Can you disclose "the circuit" you are building and simulating? I'm curious to see how diode substitution would affect the sim results.
 
Hi

N=1.892
Is=163.464e-6
Rs=7.64e-3
 

Attachments

  • SR1660.png
    SR1660.png
    24.5 KB · Views: 141
Hello eTech,

I have calculated some parameter values by following the post #10 and #13 and some other parameters are calculated from the datasheet which are:

The first three values are calculated by following the posts and others are calculated from the datasheet.

IS = 5.67e-6
RS = 0.0113
N = 1.3330
CJO = 480pF
IBV = 0.5mA
BV = 60V
M = 0.333
EG = 0.69
XTI = 2
TT = ?

But, I am not able to calculate the parameter value of TT = ? Because I couldn't get the characteristics of the diode in order to calculate the TT value. So, Please can you provide me the value of TT.

When comparing the your's and mine values (i.e: IS,RS,N) are different so which one I have to follow. I am full confusion. So, please give me the judgement on my post...
 
Hi

N=1.892
Is=163.464e-6
Rs=7.64e-3

Hello eTech,

I have calculated some parameter values by following the post #10 and #13 and some other parameters are calculated from the datasheet which are:

The first three values are calculated by following the posts and others are calculated from the datasheet.

IS = 5.67e-6
RS = 0.0113
N = 1.3330
CJO = 480pF
IBV = 0.5mA
BV = 60V
M = 0.333
EG = 0.69
XTI = 2
TT = ?

But, I am not able to calculate the parameter value of TT = ? Because I couldn't get the characteristics of the diode in order to calculate the TT value. So, Please can you provide me the value of TT.

When comparing the your's and mine values (i.e: IS,RS,N) are different so which one I have to follow. I am full confusion. So, please give me the judgement on my post...
 
Hello again,

There may be more than one solution for a given diode such that you still get the same approximate curve. The reason for this is because i think the diode has what are called redundant parameters. This means that the parameters can be changed and still keep the same curve shape, or more mathematically, not all the parameters are orthogonal to each other. For example, i think Rs can be worked into the exponential term at least to some degree.

The method i showed you is a basic curve fitting method, that is a short cut to a full sum of squared deviates type curve fit. The reason it works is because the diode has a pretty well known curve, or at least as an approximation.
Other methods may be based on more physical structures of the diode itself, and these will yield results that are different than a curve fit, although both curves may look almost the same. In real life you cant nail this down perfectly anyway, much like you cant select a resistor with a tolerance of 1e-12 percent. There's always some tolerance anyway, the question then becomes how much deviation can you accept.

If this is for something important i would definitely void this diode and find one with spec's that are published by the manufacturer. Then you can try to curve fit that if you like and see how close you come to matching the spec sheet data.

What else you can do is run simulations using the calculated data for the diode, both ways. For example Diode1 is with one method, Diode2 is with the other method, and compare the curves in simulation.

The way to get TT i already told you. Without knowing the physics of the diode you cant do it any other way. You'd probably have to get this from the manufacturer, possibly email them.
Pictures this:
1. You have a box with a capacitor in it. What is the value of the capacitor? You cant know unless the company that put the cap in the box tells you, or they give you the dimensions of the capacitor and type of dielectric.
2. Same cap, same box, but the two leads stick out. What is the value of the cap? You can perform tests to find the value of the capacitance and the ESR too if you like.

So if you cant get TT and you really need it, then you have to switch to a different diode OR do tests on a real diode yourself.
 
Hello eTech,

But, I am not able to calculate the parameter value of TT = ? Because I couldn't get the characteristics of the diode in order to calculate the TT value. So, Please can you provide me the value of TT.

Its a schottky diode...so there is no TT value needed (its recovery is instantaneous).

When comparing the your's and mine values (i.e: IS,RS,N) are different so which one I have to follow. I am full confusion. So, please give me the judgement on my post...

The three point method is an estimation, so it depends on the points you've chosen for your calculation. They are probably different than the ones I used. But the I-V curve should still end up about the same.

eT
 
Its a schottky diode...so there is no TT value needed (its recovery is instantaneous).



The three point method is an estimation, so it depends on the points you've chosen for your calculation. They are probably different than the ones I used. But the I-V curve should still end up about the same.

eT
eT,
Can you look at reverse leakage as well?
 
When I was playing I had trouble getting this curve and the Vf curve at the same time. I'm just curious if yours gets both pretty close as it is a parameter that could effect performance.

Edit:
Cut the bottom off the picture. The x is % of reverse voltage.
 

Attachments

  • upload_2015-7-14_11-19-52.png
    upload_2015-7-14_11-19-52.png
    33.9 KB · Views: 124
Hello again,

About TT:
TT is sometimes given for a Schottky too. I cant remember exactly but i think it was for the secondary 'guard' diode in the model. There could be variations from one diode to the other also, where the modern ones have less recovery.
Any TT is usually small anyway, on the order of 1.5e-9, so you have to wonder if it is really needed anyway. Soft recovery diodes (more desirable for low EMI requirements) can be 2000e-9 for comparison.

About Reverse Characteristics:
Supposedly the reverse characteristic defines Is physically, but unfortunately i cant remember exactly how this works except that maybe Is if found from the reverse curve, for a given reverse current, and then the others are found from the forward characteristics in the usual way.
I would think there should be something about this on the web. Anybody do a good search yet?
 
When I was playing I had trouble getting this curve and the Vf curve at the same time. I'm just curious if yours gets both pretty close as it is a parameter that could effect performance.

Edit:
Cut the bottom off the picture. The x is % of reverse voltage.

Here's about the best I could do for now....it will take more playing to get it more accurate.

One image shows forward characteristics the other image reverse leakage.
The reverse leakage current is constant until breakdown since it depends on IS which is also a constant (IS doesn't change in spice like it would with reverse-bias voltage in a real diode). Anyway..for what their worth...here they are.

Oh...and I used the datasheet from Rectron Semiconductor.
 

Attachments

  • SR1660Fwd.png
    SR1660Fwd.png
    40.4 KB · Views: 121
  • SR1660Rev.png
    SR1660Rev.png
    40.7 KB · Views: 124
When I was playing I had trouble getting this curve and the Vf curve at the same time. I'm just curious if yours gets both pretty close as it is a parameter that could effect performance.

Edit:
Cut the bottom off the picture. The x is % of reverse voltage.


Hello again,

After reading eTech's most recent post i see what one of your problems is. You are trying to use the exact same exponential for the negative voltage characteristic as for the positive (forward) voltage characteristic.

A different exponential is used for the negative part of the response, if you choose to even do that. 'Is' is the 'leakage current' in the reverse mode, so to approximate you can think of it as a small current (leakage) until it hits the breakdown voltage, and then it dives very low (very high amplitude negative current).
So you have to switch equations when you go with voltages less than 0 (negative voltages), or just assume the leakage current until breakdown when the voltage goes negative. Since you probably wont need the breakdown characteristic, you can probably just use -Is as the current in the reverse mode. If you really think you need the curve then we'll have to look that up, but it will be a slightly different exponential which i think includes the BV parameter (which would make sense).

I should add that the new exponential does not kick in right away, but somewhere near the BV i think. Otherwise it is Is.
Also, some spice representations will not really take advantage of this depiction of the reverse characteristic, but will actually use a very small 'Is' and model the reverse leakage with a parallel resistor Rp instead. This value may be very high too, like 1G Ohms, or lower like 10M Ohms. For the 1N4148 diode for example it would be very high like 1G Ohms. For this kind of depiction, the leakage current parameter would be made very low like 2 picoamps, so it does not contribute much to the actual reverse response, so the parallel resistor then takes on the responsibility of modeling the reverse leakage. In cases where there is no Rp specified (with or without a very tiny 'Is"), the parallel resistance would be 1/GMIN for example, where GMIN is a global parameter that affects many component parameters as default.
Also, "Is" itself does not really kick in at exactly 0.000000 volts, but waits until the voltage gets a little bit negative, possibly around -0.25v. I think the value is taken as -5*N*Vt.
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top