An equation of the `real` solar cell

[ptc-1793698]

Hello!

I would like to have a fit for the current-voltage curve of one Schottky junction (probably this is, or may be).
The problem is that the equation for real solar cell that may be used here, has not have an explicit form.
How to fit the data using such an equation?

KL

[alnstevens]

Like so? Compressed M14 file attached as M11 didn't like the Minerr solve block (probably just my poor initial guesses!).

stv

[ptc-1793698]

I have MC 13 and 14 available, but none of them opens the worksheet.

KL

[alnstevens]

Strange! Here are uncompressed 13 and 14 files.

stv

[ptc-1793698]

Thank You!

Now the version for MC 13 opens, the more new does not. Maybe there are updates of MC14 that I do not have? This may be.

KL

[alnstevens]

My M14 is version M020.

stv

[ptc-1793698]

How would it be possible to obtain a derivative of this implicit function: dJ/dV (V)? Or inverse of it dV/dJ (V)?

KL

[alnstevens]

It's fairly simple to differentiate and rearrange to get what you want - see attached (however, you should double check that I've done the differentiation correctly, as I did it rather quickly!). Also, the function could be implemented more efficiently - the exponential terms are calculated twice in the attached.



Oops! Error. This is the corrected file.





stv

[ptc-1793698]

Thank You!

I compared with my solution. Yes, it was somewhat different.

KL

[ptc-1368288]

On 10/20/2009 10:22:41 AM, KristjanLaes wrote:
>Thank You!
>
>I compared with my solution.
>Yes, it was somewhat
>different.
>
>KL
___________________________

All red IV20modeling .
I have already replied.

"There is a function that I would like to use for fitting the data in Data".
The model is wrong for the data set. Some terms are missing to model the physics appropriately. Why don't you use the model that was done for the other collab. The model you are proposing is not "modelable", not shapable so it won't Minerr.

jmG

[alnstevens]

On 10/20/2009 10:22:41 AM, KristjanLaes wrote:
>Thank You!
>
>I compared with my solution.
>Yes, it was somewhat
>different.
>
>KL

Yes, I missed aterm from the numerator - a case of more haste less speed! I've included my derivation in the attached, and corrected the error, which makes both forms now give the same result.

To jmg: The reason for all the red that you see is that Mathcad 11 can't cope with the Minerr solve block. Mathcads 13 and 14 manage it ok.

stv

[ptc-1368288]

On 10/21/2009 12:08:22 PM, stv wrote:
>On 10/20/2009 10:22:41 AM, KristjanLaes
>wrote:
>>Thank You!
>>
>>I compared with my solution.
>>Yes, it was somewhat
>>different.
>>
>>KL
>
>Yes, I missed a term from the numerator -
>a case of more haste less speed! I've
>included my derivation in the attached,
>and corrected the error, which makes
>both forms now give the same result.
>
>To jmg: The reason for all the red that
>you see is that Mathcad 11 can't cope
>with the Minerr solve block. Mathcads
>13 and 14 manage it ok.
>
>stv
____________________________

I don't understand who missed a term and where, an what's different. I don't understand either why k, q that are physical constants should be so much different. Mathcad 11.2a has no problem solving that peanut gallery Minerr. I stopped reading at Differentiation ... no idea what you were attempting.

jmG

[ptc-1368288]

On 10/20/2009 4:12:02 AM, KristjanLaes wrote:
>Hello!
>
>I would like to have a fit for
>the current-voltage curve of
>one Schottky junction
>(probably this is, or may be).
>The problem is that the
>equation for real solar cell
>that may be used here, has not
>have an explicit form.
>How to fit the data using such
>an equation?
>
>KL_______________________________

Your model is wrong, you don't have a circuit to look at it ... J0 is Bessel built-in, rather use Jsubscript. A very similar work was done for other collab, please use same symbology as much as you can ... NOT k*T but kT where kT is defined , what about q ? k ? That there is a solution (W) to a wrong model will not correct the model. It does not Minerr, does not Given/Find.

Start from the model that more collab can look at. I will not attempt to fit a function, and the rational model will glitch ! DeadLock !

jmG

[TomGutman]

Have you looked at the thread at http://collab.mathsoft.com/read?125801,16e#125801 ? It seems to cover your case. Modelling using the diode equation starts at message 27.
__________________
� � � � Tom Gutman