Re: [Fractint] Diode Fractal
. No PAR or FRM is seen, and no attachments. Maybe it is problem on my end? I am using comcast web client email. .
I shall try again . ----- Original Message ----- From: JackOfTradeZ@comcast.net To: fractint@mailman.xmission.com Sent: Thursday, July 16, 2009 4:21 PM Subject: Re: [Fractint] Diode Fractal . No PAR or FRM is seen, and no attachments. Maybe it is problem on my end? I am using comcast web client email. . ------------------------------------------------------------------------------ _______________________________________________ Fractint mailing list Fractint@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/fractint
In case the attachments aren't working ; here's a pure text file , within this message . LnDioGH.par AND LnDioGH.frm , together . START PARAMETER FILE======================================= Diode { reset=1821 type=formula formulafile=c:\winf1821\alndgh.frm formulaname=LnDioGH corners=-0.16588419/0.25978091/-0.15031315/0.1691023 float=y maxiter=512 logmap=yes colors=00000j0j0jj0j00j0j0jjkkkkrkdmwW000W0WW000WW0W0WWFNF<5>zNFFVF<5>zV\ FFbF<5>zbFFjF<5>zjFFrF<5>zrFFzF<5>zzFFFN<5>zFNFNN<5>zNNFVN<5>zVNFbN<5>zb\ NFjN<5>zjNFrN<5>zrNFzN<5>zzNFFV<5>zFVFNV<5>zNVFVV<5>zVVFbV<5>zbVFjV<5>zj\ VFrV<5>zrVFzV<5>zzVFFb<5>zFbFNb<5>zNbFVb<5>zVbFbb<5>zbbFjb<5>zjbFrb<5>zr\ bFzb<5>zzbFFj<5>zFjFNj<5>zNjFVj<5>zVjFbj<5>zbjFjj<5>zjjFrj<4>rrjzywccdWW\ Wz000z0zz000zz0z0zzzzz } comment { From : sciwise@ihug.co.nz July 2009 copyright c 2009 This is a model of a diode , as an electronic component . It utilizes Newton's method to solve a non-linear equation ; known as the ideal diode or Shockley equation . The very basic form of this equation is : i = Ids*(exp(Vs/(n*Vt))-1) where , Q = 1.6*(10^-19) Electronic Charge . K = 1.38*(10^-23) Boltzman's constant . T = Absolute Temperature , in Kelvins . T = 273 [K] + degreesC For degreesC = 25 , Vt ~= 0.025 [v] Ids = Saturation current . Vs = Applied , source , voltage . n = Ideality factor = ( 1 to 2 ) . Vt = K * T / Q Thermionic Voltage . A more realistic model of a diode will include the finite ohmic resistance of the leads . A parallel conductance is also modeled , as this speeds the convergence of Newton's method . Therefore the circuit we're modelling looks like this . i --> ----------------------------, ' | | -------------------, | | .---. Vs - D --- | | Gm --- /_\ |___| | |__________________| | | | .---. | | | Rs | |___| '----------------------------' Where D is represented by the ideal diode equation . Now the total current , i , flowing through this circuit is : i = Ids*(exp((Vs-i*Rs)/(n*Vt))-1) + Gm*(Vs-i*Rs) This equation needs to be solved in terms of i and be put in a form that can be readily and accurately solved through Newton's method . The steps involved are : 1) Let log(x) = (Vs-i*Rs)/(n*Vt) 2) Solve log(x) = (Vs-i*Rs)/(n*Vt) in terms of i . i = -(n*Vt*log(x)-Vs)/Rs 3) Substitute the expression for i , from (2) into i = Ids*(exp((Vs-i*Rs)/(n*Vt))-1) + Gm*(Vs-i*Rs) -(n*Vt*log(x)-Vs)/Rs = Ids*(x-1) + Gm*(Vs--((n*Vt*log(x)-Vs)/Rs)*Rs) -(n*Vt*log(x)-Vs)/Rs = Ids*(x-1) + Gm*((n*Vt*log(x)/Rs)*Rs) (Vs-n*Vt*log(x))/Rs = Ids*(x-1) + Gm*(n*Vt*log(x)) 4) Solve (3) for x , x = -((Gm*n*Rs*Vt+n*Vt)*log(x)-Vs-Is*Rs)/(Is*Rs),Is*Rs,1); 5) Find the derivative of (4) in terms of x . 1 = -(Gm*n*Rs*Vt+n*Vt)/(Is*Rs*x) 6) Now Newton's iteration formula is : x = x - f(x) / f '(x) , where f '(x) == d f(x) / dx . 7) For this example : f(x) = x = -((Gm*n*Rs*Vt+n*Vt)*log(x)-Vs-Is*Rs)/(Is*Rs),Is*Rs,1) and f '(x) = 1 = -(Gm*n*Rs*Vt+n*Vt)/(Is*Rs*x) 8) Therefore Newton's iteration formula for this example is : x = x - ( -((Gm*n*Rs*Vt+n*Vt)*log(x)-Vs-Is*Rs)/(Is*Rs),Is*Rs,1) - x )/( -(Gm*n*Rs*Vt+n*Vt)/(Is*Rs*x)-1) 9) Simplifying (8) x = x -( (-Ids*Rs-Vs)*x + Ids*Rs*x*x + (n+Gm*n*Rs)*Vt*x*log(x) ) / ( (n + Gm*n*Rs)*Vt+Ids*Rs*x ) Substitute u = Gm*n*Rs*Vt+n*Vt vr = Ids*Rs x = x-( -(Vs+vr)*x + vr*x*x + u*x*log(x) )/(vr*x+u)) In this logarithmic format overflow , and possibly underflow , is avoided . 10 ) We need to be able to convert x back to i , to do this we use the inverse of the original substitution : log(x) = (Vs-i*Rs)/(n*Vt) Solving for i , i = - ( log(x)*(n*Vt) - Vs ) / Rs Substitute Pt = n*Vt i = - ( log(x)*Pt - Vs ) / Rs Between successive iterations of Newton's formula the previous value of i is compared with the present, if the absolute difference is less than a predefined tolerance , Pd , then the iteration stops . Vs is a complex voltage source , possibly equivalent to a multi phase voltage source . This is a basic diode model and doesn't include the effects of diode junction capacitance . For small voltages a purely exponential representation of the diode model may converge quicker than the logarithmic and still remain accurate . The logarithmic model has been implemented as a BASIC language program and compared with data from a SPICE simulation , for a real voltage source , there is good agreement . For a Complex voltage source , the few samples generated from the BASIC program are in good agreement with those from MAXIMA CAS . } LnDioGH{ ; ; ; Complex Voltage - Complex Current ; Diode Characteristics . ; ; for 1N4148 silicon type . ; Level 1 model . ; ; at 25 degrees C ; ; Pd = (10^-13), Gm = 9.4*(10^-13), Ids = 2.52*(10^-09), Rs = 0.968, n = 1.752, Vt = 0.025, u = Gm*n*Rs*Vt+n*Vt, Pt = n*Vt, vr = Ids*Rs, Vs = Pixel, x = (0.1,0), x = x-((u*x*log(x)+vr*x*x-(Vs+vr)*x)/(vr*x+u)), y = -(Pt*log(x)-Vs)/Rs, tol = (1,0): u = y; x = x-((u*x*log(x)+vr*x*x-(Vs+vr)*x)/(vr*x+u)); y = -(Pt*log(x)-Vs)/Rs, tol = |u-y|, tol > Pd } comment { The floating point option needs to be selected , 512 iterations , or more , should be selected . Also a logarithmic palette . Minibrot structures aren't immediately apparent at the default parameter settings . A parameter setting of : xmin - 0.005770310598 xmax - 0.005666943901 ymin 0.000516417395 ymax 0.000594104392 shows embedded mimibrots and spirals , wether this is relevant in a real diode is unknown . You might be able to locate mimibrots elsewhere on the map . } END PARAMETER FILE========================================= ----- Original Message ----- From: JackOfTradeZ@comcast.net To: fractint@mailman.xmission.com Sent: Thursday, July 16, 2009 4:21 PM Subject: Re: [Fractint] Diode Fractal . No PAR or FRM is seen, and no attachments. Maybe it is problem on my end? I am using comcast web client email. . ------------------------------------------------------------------------------ _______________________________________________ Fractint mailing list Fractint@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/fractint
Dear Fractinters, sciwise wrote: here is the Diode .par and .frm without comments. A red error comes after opening in fractint (DOS-Version). Something is wrong in the .frm (parent´s must be set).
START PARAMETER FILE======================================= Diode { reset=1821 type=formula formulafile=c:\winf1821\alndgh.frm formulaname=LnDioGH corners=-0.16588419/0.25978091/-0.15031315/0.1691023 float=y maxiter=512 logmap=yes colors=00000j0j0jj0j00j0j0jjkkkkrkdmwW000W0WW000WW0W0WWFNF<5>zNFFVF<5>zV\ FFbF<5>zbFFjF<5>zjFFrF<5>zrFFzF<5>zzFFFN<5>zFNFNN<5>zNNFVN<5>zVNFbN<5>zb\ NFjN<5>zjNFrN<5>zrNFzN<5>zzNFFV<5>zFVFNV<5>zNVFVV<5>zVVFbV<5>zbVFjV<5>zj\ VFrV<5>zrVFzV<5>zzVFFb<5>zFbFNb<5>zNbFVb<5>zVbFbb<5>zbbFjb<5>zjbFrb<5>zr\ bFzb<5>zzbFFj<5>zFjFNj<5>zNjFVj<5>zVjFbj<5>zbjFjj<5>zjjFrj<4>rrjzywccdWW\ Wz000z0zz000zz0z0zzzzz } frm:LnDioGH{ ; ; ; Complex Voltage - Complex Current ; Diode Characteristics . ; ; for 1N4148 silicon type . ; Level 1 model . ; ; at 25 degrees C ; ; Pd = (10^-13), Gm = 9.4*(10^-13), Ids = 2.52*(10^-09), Rs = 0.968, n = 1.752, Vt = 0.025, u = Gm*n*Rs*Vt+n*Vt, Pt = n*Vt, vr = Ids*Rs, Vs = Pixel, x = (0.1,0), x = x-((u*x*log(x)+vr*x*x-(Vs+vr)*x)/(vr*x+u)), y = -(Pt*log(x)-Vs)/Rs, tol = (1,0): u = y; x = x-((u*x*log(x)+vr*x*x-(Vs+vr)*x)/(vr*x+u)); y = -(Pt*log(x)-Vs)/Rs, tol = |u-y|, tol > Pd } END PARAMETER FILE========================================= cheers, Guy ##################################################################################### Scanned by MailMarshal - Marshal's comprehensive email content security solution. #####################################################################################
Okay , for Fractint , a little editing is required to make the *.frm file compatible with both versions. This is how the new combined file should look : START PARAMETER FILE======================================= Diode { reset=1821 type=formula formulafile=c:\winf1821\alndgh.frm formulaname=LnDioGH corners=-0.16588419/0.25978091/-0.15031315/0.1691023 float=y maxiter=512 logmap=yes colors=00000j0j0jj0j00j0j0jjkkkkrkdmwW000W0WW000WW0W0WWFNF<5>zNFFVF<5>zV\ FFbF<5>zbFFjF<5>zjFFrF<5>zrFFzF<5>zzFFFN<5>zFNFNN<5>zNNFVN<5>zVNFbN<5>zb\ NFjN<5>zjNFrN<5>zrNFzN<5>zzNFFV<5>zFVFNV<5>zNVFVV<5>zVVFbV<5>zbVFjV<5>zj\ VFrV<5>zrVFzV<5>zzVFFb<5>zFbFNb<5>zNbFVb<5>zVbFbb<5>zbbFjb<5>zjbFrb<5>zr\ bFzb<5>zzbFFj<5>zFjFNj<5>zNjFVj<5>zVjFbj<5>zbjFjj<5>zjjFrj<4>rrjzywccdWW\ Wz000z0zz000zz0z0zzzzz } comment { From : sciwise@ihug.co.nz July 2009 copyright c 2009 This is a model of a diode , as an electronic component . It utilizes Newton's method to solve a non-linear equation ; known as the ideal diode or Shockley equation . The very basic form of this equation is : i = Ids*(exp(Vs/(n*Vt))-1) where , Q = 1.6*(10^-19) Electronic Charge . K = 1.38*(10^-23) Boltzman's constant . T = Absolute Temperature , in Kelvins . T = 273 [K] + degreesC For degreesC = 25 , Vt ~= 0.025 [v] Ids = Saturation current . Vs = Applied , source , voltage . n = Ideality factor = ( 1 to 2 ) . Vt = K * T / Q Thermionic Voltage . A more realistic model of a diode will include the finite ohmic resistance of the leads . A parallel conductance is also modeled , as this speeds the convergence of Newton's method . Therefore the circuit we're modelling looks like this . i --> ----------------------------, ' | | -------------------, | | .---. Vs - D --- | | Gm --- /_\ |___| | |__________________| | | | .---. | | | Rs | |___| '----------------------------' Where D is represented by the ideal diode equation . Now the total current , i , flowing through this circuit is : i = Ids*(exp((Vs-i*Rs)/(n*Vt))-1) + Gm*(Vs-i*Rs) This equation needs to be solved in terms of i and be put in a form that can be readily and accurately solved through Newton's method . The steps involved are : 1) Let log(x) = (Vs-i*Rs)/(n*Vt) 2) Solve log(x) = (Vs-i*Rs)/(n*Vt) in terms of i . i = -(n*Vt*log(x)-Vs)/Rs 3) Substitute the expression for i , from (2) into i = Ids*(exp((Vs-i*Rs)/(n*Vt))-1) + Gm*(Vs-i*Rs) -(n*Vt*log(x)-Vs)/Rs = Ids*(x-1) + Gm*(Vs--((n*Vt*log(x)-Vs)/Rs)*Rs) -(n*Vt*log(x)-Vs)/Rs = Ids*(x-1) + Gm*((n*Vt*log(x)/Rs)*Rs) (Vs-n*Vt*log(x))/Rs = Ids*(x-1) + Gm*(n*Vt*log(x)) 4) Solve (3) for x , x = -((Gm*n*Rs*Vt+n*Vt)*log(x)-Vs-Is*Rs)/(Is*Rs),Is*Rs,1); 5) Find the derivative of (4) in terms of x . 1 = -(Gm*n*Rs*Vt+n*Vt)/(Is*Rs*x) 6) Now Newton's iteration formula is : x = x - f(x) / f '(x) , where f '(x) == d f(x) / dx . 7) For this example : f(x) = x = -((Gm*n*Rs*Vt+n*Vt)*log(x)-Vs-Is*Rs)/(Is*Rs),Is*Rs,1) and f '(x) = 1 = -(Gm*n*Rs*Vt+n*Vt)/(Is*Rs*x) 8) Therefore Newton's iteration formula for this example is : x = x - ( -((Gm*n*Rs*Vt+n*Vt)*log(x)-Vs-Is*Rs)/(Is*Rs),Is*Rs,1) - x )/( -(Gm*n*Rs*Vt+n*Vt)/(Is*Rs*x)-1) 9) Simplifying (8) x = x -( (-Ids*Rs-Vs)*x + Ids*Rs*x*x + (n+Gm*n*Rs)*Vt*x*log(x) ) / ( (n + Gm*n*Rs)*Vt+Ids*Rs*x ) Substitute u = Gm*n*Rs*Vt+n*Vt vr = Ids*Rs x = x-( -(Vs+vr)*x + vr*x*x + u*x*log(x) )/(vr*x+u)) In this logarithmic format overflow , and possibly underflow , is avoided . 10 ) We need to be able to convert x back to i , to do this we use the inverse of the original substitution : log(x) = (Vs-i*Rs)/(n*Vt) Solving for i , i = - ( log(x)*(n*Vt) - Vs ) / Rs Substitute Pt = n*Vt i = - ( log(x)*Pt - Vs ) / Rs Between successive iterations of Newton's formula the previous value of i is compared with the present, if the absolute difference is less than a predefined tolerance , Pd , then the iteration stops . Vs is a complex voltage source , possibly equivalent to a multi phase voltage source . This is a basic diode model and doesn't include the effects of diode junction capacitance . For small voltages a purely exponential representation of the diode model may converge quicker than the logarithmic and still remain accurate . The logarithmic model has been implemented as a BASIC language program and compared with data from a SPICE simulation , for a real voltage source , there is good agreement . For a Complex voltage source , the few samples generated from the BASIC program are in good agreement with those from MAXIMA CAS . } frm:LnDioGH{ ; ; ; Complex Voltage - Complex Current ; Diode Characteristics . ; ; for 1N4148 silicon type . ; Level 1 model . ; ; at 25 degrees C ; ; Pd = (10^(-13)), Gm = 9.4*(10^(-13)), Ids = 2.52*(10^(-09)), Rs = 0.968, n = 1.752, Vt = 0.025, u = Gm*n*Rs*Vt+n*Vt, Pt = n*Vt, vr = Ids*Rs, Vs = Pixel, x = (0.1,0), x = x-((u*x*log(x)+vr*x*x-(Vs+vr)*x)/(vr*x+u)), y = -(Pt*log(x)-Vs)/Rs, tol = (1,0): u = y; x = x-((u*x*log(x)+vr*x*x-(Vs+vr)*x)/(vr*x+u)); y = -(Pt*log(x)-Vs)/Rs, tol = |u-y|, tol > Pd } comment { The floating point option needs to be selected , 512 iterations , or more , should be selected . Also a logarithmic palette . Minibrot structures aren't immediately apparent at the default parameter settings . A parameter setting of : xmin - 0.005770310598 xmax - 0.005666943901 ymin 0.000516417395 ymax 0.000594104392 shows embedded mimibrots and spirals , wether this is relevant in a real diode is unknown . You might be able to locate mimibrots elsewhere on the map . } END PARAMETER FILE========================================= ----- Original Message ----- From: "Guy Marson" <gmarson@mnhn.lu> To: "Fractint and General Fractals Discussion" <fractint@mailman.xmission.com> Sent: Thursday, July 16, 2009 6:21 PM Subject: Re: [Fractint] Diode Fractal
Dear Fractinters,
sciwise wrote:
here is the Diode .par and .frm without comments. A red error comes after opening in fractint (DOS-Version). Something is wrong in the .frm (parent´s must be set).
START PARAMETER FILE=======================================
Diode { reset=1821 type=formula formulafile=c:\winf1821\alndgh.frm formulaname=LnDioGH corners=-0.16588419/0.25978091/-0.15031315/0.1691023 float=y maxiter=512 logmap=yes
colors=00000j0j0jj0j00j0j0jjkkkkrkdmwW000W0WW000WW0W0WWFNF<5>zNFFVF<5>zV\
FFbF<5>zbFFjF<5>zjFFrF<5>zrFFzF<5>zzFFFN<5>zFNFNN<5>zNNFVN<5>zVNFbN<5>zb\
NFjN<5>zjNFrN<5>zrNFzN<5>zzNFFV<5>zFVFNV<5>zNVFVV<5>zVVFbV<5>zbVFjV<5>zj\
VFrV<5>zrVFzV<5>zzVFFb<5>zFbFNb<5>zNbFVb<5>zVbFbb<5>zbbFjb<5>zjbFrb<5>zr\
bFzb<5>zzbFFj<5>zFjFNj<5>zNjFVj<5>zVjFbj<5>zbjFjj<5>zjjFrj<4>rrjzywccdWW\
Wz000z0zz000zz0z0zzzzz }
frm:LnDioGH{ ; ; ; Complex Voltage - Complex Current ; Diode Characteristics . ; ; for 1N4148 silicon type . ; Level 1 model . ; ; at 25 degrees C ; ; Pd = (10^-13), Gm = 9.4*(10^-13), Ids = 2.52*(10^-09), Rs = 0.968, n = 1.752, Vt = 0.025, u = Gm*n*Rs*Vt+n*Vt, Pt = n*Vt, vr = Ids*Rs, Vs = Pixel, x = (0.1,0), x = x-((u*x*log(x)+vr*x*x-(Vs+vr)*x)/(vr*x+u)), y = -(Pt*log(x)-Vs)/Rs, tol = (1,0): u = y; x = x-((u*x*log(x)+vr*x*x-(Vs+vr)*x)/(vr*x+u)); y = -(Pt*log(x)-Vs)/Rs, tol = |u-y|, tol > Pd }
END PARAMETER FILE=========================================
cheers,
Guy
############################################################################ #########
Scanned by MailMarshal - Marshal's comprehensive email content security solution.
############################################################################ #########
_______________________________________________ Fractint mailing list Fractint@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/fractint
sciwise, do you really use Winfract 18.21? It's very old (anything before version 19.6 is more or less "stone age territory"). There's an updated version here (I can't reach www.fractint.org at the moment): http://twegner.dyndns.org/fractint/ftp/current/windows/winfract-20.04p09.zip Cut and pasted from Winfract 20.04patch9 Help: --------------------------------------------------------------------------------------------------------- This release of Winfract is essentially the same as the version 18.2 but with the following enhancements: - All fractal types from Fractint 20.4 (with the exception of the 'julia_inverse' fractal type, which requires some extra logic we haven't yet added to Winfract) have been incorporated into this version of Winfract. - The File Open, File SaveAs, and File Print options now use the new Windows-style "common dialog boxes", with the extra capabilities provided by those routines. This is particularly handy if you're using Windows-for-Workgroups or Windows NT, 2000, XP with their networking options. - The fractal history feature of FFD has been implemented and uses the h key to move back in history and the H key to move forward in history. --------------------------------------------------------------------------------------------------------- I hope you excuse me from scrambling your code (not all of it was necessary, but is the result of trying to get it to display something on my machine. Note that the inclusion of formulas in .par sets works only from version 19.6 on, if my memory serves me right. Regards, Gerald ---------------- .PAR SET BELOW ------------------------------------ Diode { reset=2004 type=formula formulafile=fractint.frm formulaname=LnDioGH corners=-0.16588419/0.25978091/-0.15031315/0.1691023 float=y maxiter=512 logmap=yes periodicity=no colors=00000j0j0jj0j00j0j0jjkkkkrkdmwW000W0WW000WW0W0WWFNF<5>zNFFVF<5>zV\ FFbF<5>zbFFjF<5>zjFFrF<5>zrFFzF<5>zzFFFN<5>zFNFNN<5>zNNFVN<5>zVNFbN<5>zb\ NFjN<5>zjNFrN<5>zrNFzN<5>zzNFFV<5>zFVFNV<5>zNVFVV<5>zVVFbV<5>zbVFjV<5>zj\ VFrV<5>zrVFzV<5>zzVFFb<5>zFbFNb<5>zNbFVb<5>zVbFbb<5>zbbFjb<5>zjbFrb<5>zr\ bFzb<5>zzbFFj<5>zFjFNj<5>zNjFVj<5>zVjFbj<5>zbjFjj<5>zjjFrj<4>rrjzywccdWW\ Wz000z0zz000zz0z0zzzzz } frm:LnDioGH{ ; ; Complex Voltage - Complex Current ; Diode Characteristics . ; ; for 1N4148 silicon type . ; Level 1 model . ; ; at 25 degrees C ; Pd = 0.0000000000001 Gm = 0.00000000000094 Ids = 0.00000000252 Rs = 0.968 n = 1.752 Vt = 0.025 u = Gm*n*Rs*Vt+n*Vt Pt = n*Vt vr = Ids*Rs x = 0.1 x = x-((u*x*log(x)+vr*x*x-(Pixel+vr)*x)/(vr*x+u)) z = 0-(Pt*log(x)-Pixel)/Rs: u = z x = x-((u*x*log(x)+vr*x*x-(Pixel+vr)*x)/(vr*x+u)) z = 0-(Pt*log(x)-Pixel)/Rs |u-z| > Pd } -------------- END OF .PAR SET ------------------------------------------
Thank you, Gerald. Your formulation runs nicely in my machine... but I find little of note in the Fractals produced. Maybe I should look deeper! John W. ----- Original Message ----- From: "Gerald K. Dobiasovsky" <gerald.dob@aon.at> To: "Fractint and General Fractals Discussion" <fractint@mailman.xmission.com> Sent: Thursday, July 16, 2009 5:36 PM Subject: Re: [Fractint] Diode Fractal
sciwise,
do you really use Winfract 18.21? It's very old (anything before version 19.6 is more or less "stone age territory"). There's an updated version here (I can't reach www.fractint.org at the moment):
http://twegner.dyndns.org/fractint/ftp/current/windows/winfract-20.04p09.zip
Cut and pasted from Winfract 20.04patch9 Help: --------------------------------------------------------------------------------------------------------- This release of Winfract is essentially the same as the version 18.2 but with the following enhancements:
- All fractal types from Fractint 20.4 (with the exception of the 'julia_inverse' fractal type, which requires some extra logic we haven't yet added to Winfract) have been incorporated into this version of Winfract.
- The File Open, File SaveAs, and File Print options now use the new Windows-style "common dialog boxes", with the extra capabilities provided by those routines. This is particularly handy if you're using Windows-for-Workgroups or Windows NT, 2000, XP with their networking options.
- The fractal history feature of FFD has been implemented and uses the h key to move back in history and the H key to move forward in history. ---------------------------------------------------------------------------------------------------------
I hope you excuse me from scrambling your code (not all of it was necessary, but is the result of trying to get it to display something on my machine.
Note that the inclusion of formulas in .par sets works only from version 19.6 on, if my memory serves me right.
Regards, Gerald
---------------- .PAR SET BELOW ------------------------------------ Diode { reset=2004 type=formula formulafile=fractint.frm formulaname=LnDioGH corners=-0.16588419/0.25978091/-0.15031315/0.1691023 float=y maxiter=512 logmap=yes periodicity=no colors=00000j0j0jj0j00j0j0jjkkkkrkdmwW000W0WW000WW0W0WWFNF<5>zNFFVF<5>zV\ FFbF<5>zbFFjF<5>zjFFrF<5>zrFFzF<5>zzFFFN<5>zFNFNN<5>zNNFVN<5>zVNFbN<5>zb\ NFjN<5>zjNFrN<5>zrNFzN<5>zzNFFV<5>zFVFNV<5>zNVFVV<5>zVVFbV<5>zbVFjV<5>zj\ VFrV<5>zrVFzV<5>zzVFFb<5>zFbFNb<5>zNbFVb<5>zVbFbb<5>zbbFjb<5>zjbFrb<5>zr\ bFzb<5>zzbFFj<5>zFjFNj<5>zNjFVj<5>zVjFbj<5>zbjFjj<5>zjjFrj<4>rrjzywccdWW\ Wz000z0zz000zz0z0zzzzz }
frm:LnDioGH{ ; ; Complex Voltage - Complex Current ; Diode Characteristics . ; ; for 1N4148 silicon type . ; Level 1 model . ; ; at 25 degrees C ; Pd = 0.0000000000001 Gm = 0.00000000000094 Ids = 0.00000000252 Rs = 0.968 n = 1.752 Vt = 0.025 u = Gm*n*Rs*Vt+n*Vt Pt = n*Vt vr = Ids*Rs x = 0.1 x = x-((u*x*log(x)+vr*x*x-(Pixel+vr)*x)/(vr*x+u)) z = 0-(Pt*log(x)-Pixel)/Rs: u = z x = x-((u*x*log(x)+vr*x*x-(Pixel+vr)*x)/(vr*x+u)) z = 0-(Pt*log(x)-Pixel)/Rs |u-z| > Pd } -------------- END OF .PAR SET ------------------------------------------
_______________________________________________ Fractint mailing list Fractint@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/fractint
Dear Fractinters, - sciwise: thanks for that interesting formula! - Gerald: thanks to make it work on my old laptop :). - JoT: ********* Well, I finally got it to work with tweax; it is one of the most interesting images I have seen, like with all the dots. However, there are no parameters to tweak so I can't do a morph animation, or as I was hoping, something that would model electricity thru a diode or something. Zooming in gets you nowhere fast; zooming out is ok for a few orders, then it STOPS at about E-005. You cant back out further - keep getting the same image. (I have seen this before. Strange) ********** Here is the same .frm with (Rs*p1). It maybe makes no sense but at least the "value" of Rs can be changed :) even on the complex plane :(. Cheers, Guy
---------------- .PAR SET BELOW ------------------------------------ Diode_&_(p1) { reset=2004 type=formula formulafile=fractint.frm formulaname=LnDioGH corners=-0.16588419/0.25978091/-0.15031315/0.1691023 float=y maxiter=512 logmap=yes periodicity=no params=1/0 colors=00000j0j0jj0j00j0j0jjkkkkrkdmwW000W0WW000WW0W0WWFNF<5>zNFFVF<5>zV\ FFbF<5>zbFFjF<5>zjFFrF<5>zrFFzF<5>zzFFFN<5>zFNFNN<5>zNNFVN<5>zVNFbN<5>zb\ NFjN<5>zjNFrN<5>zrNFzN<5>zzNFFV<5>zFVFNV<5>zNVFVV<5>zVVFbV<5>zbVFjV<5>zj\ VFrV<5>zrVFzV<5>zzVFFb<5>zFbFNb<5>zNbFVb<5>zVbFbb<5>zbbFjb<5>zjbFrb<5>zr\ bFzb<5>zzbFFj<5>zFjFNj<5>zNjFVj<5>zVjFbj<5>zbjFjj<5>zjjFrj<4>rrjzywccdWW\ Wz000z0zz000zz0z0zzzzz }
frm:LnDioGH{ ; ; Complex Voltage - Complex Current ; Diode Characteristics . ; ; for 1N4148 silicon type . ; Level 1 model . ; ; at 25 degrees C ; Pd = 0.0000000000001 Gm = 0.00000000000094 Ids = 0.00000000252 Rs = (0.968*p1) n = 1.752 Vt = 0.025 u = Gm*n*Rs*Vt+n*Vt Pt = n*Vt vr = Ids*Rs x = 0.1 x = x-((u*x*log(x)+vr*x*x-(Pixel+vr)*x)/(vr*x+u)) z = 0-(Pt*log(x)-Pixel)/Rs: u = z x = x-((u*x*log(x)+vr*x*x-(Pixel+vr)*x)/(vr*x+u)) z = 0-(Pt*log(x)-Pixel)/Rs |u-z| > Pd } -------------- END OF .PAR SET ------------------------------------------
_______________________________________________ Fractint mailing list Fractint@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/fractint
##################################################################################### Scanned by MailMarshal - Marshal's comprehensive email content security solution. #####################################################################################
Guy Marson wrote:
Here is the same .frm with (Rs*p1). It maybe makes no sense but at least the "value" of Rs can be changed :) even on the complex plane :(.
Darn, I wanted to do something like below in the formula, then plug in the original values and save a new .par, but was happy when the formula resulted in something tangible - and completely forgot the planned changes: Pd = real(p1) ;0.0000000000001 Gm = imag(p1) ;0.00000000000094 Ids = real(p2) ;0.00000000252 Rs = imag(p2) ;0.968 n = real(p3) ;1.752 Vt = imag(p3) ;0.025 ... Well, I've rummaged around my Fractint subdirectories and found a file with parameters/formulas I had planned to post - years ago! No-one would call these artistic, but the purpose of the .par sets is to give the potential user a feeling for the formula and its input variables. In case one is wondering why there are two iteration functions and what use p3i --> imag(p3) is: This is a cubic polynomial, it has two critical points and both have to be iterated (try to remove the 'zz = ...' parts and the result will be a fractal sliced in half with the other part of the screen as inside color). p3i only decides which of the two crit. points is assigned to z inc ase one wants to use some inside coloring function. I hope this is not too big, would not want to inconvenience anyone... Regards, Gerald -------------------------------- PARS and FORMULA --------------------------- BH_CubicMan_1 {;Unusual looking Mandelbrot Set, Lambda = 1 ; reset=2003 type=formula formulafile=archiv.frm formulaname=BuffHenriksenMan passes=t corners=-2.683776/2.683776/-2.012832/2.012832 params=0/0/1/0/16/0 float=y maxiter=1024 inside=0 periodicity=0 colors=@volcano.map } BH_CubicMan_2 {;Unusual looking Mandelbrot Set, Lambda = -1 ; reset=2003 type=formula formulafile=archiv.frm formulaname=BuffHenriksenMan passes=t corners=-2.517125/2.517125/3.356166/-3.356166/2.517125/-3.356166 params=0/0/-1/0/16/0 float=y maxiter=1024 inside=0 periodicity=0 colors=@volcano.map } BH_CubicMan_3 {;Unusual looking Mandelbrot Set, Lambda = 0 ; reset=2003 type=formula formulafile=archiv.frm formulaname=BuffHenriksenMan passes=t corners=-2.141123/2.141123/2.85483/-2.85483/2.141123/-2.85483 params=0/0/0/0/16/0 float=y maxiter=1024 inside=0 periodicity=0 colors=@volcano.map } BH_Original {;Unusual looking Mandelbrot Set, Siegel Disk ; reset=2003 type=formula formulafile=archiv.frm formulaname=BuffHenriksenMan passes=t corners=-0.34140806/0.34140806/-3.463212/3.463212/-3.4202782/0.6419478 params=0/0/-0.737368878078322/-0.675490294261521/16/0 float=y maxiter=1024 inside=0 periodicity=0 colors=@volcano.map } BH_CubicMan_4 {;Strong distortion ;Unusual looking Mandelbrot Set ; reset=2003 type=formula formulafile=archiv.frm formulaname=BuffHenriksenMan passes=t corners=-2.949252/2.949252/0.8251166/-0.8251166/-0.03367874/-3.062315 params=0/0/0.715/0.715/16/0 float=y maxiter=512 inside=0 periodicity=0 colors=@volcano.map } BH_CubicMan_5 {;Another bizarre form ;Unusual looking Mandelbrot Set ; reset=2003 type=formula formulafile=archiv.frm formulaname=BuffHenriksenMan passes=t corners=-3.306336/3.306336/0.9250185/-0.9250185/-0.03775637/-3.433088 params=0/0/-0.9949/0.1/16/0 float=y maxiter=512 inside=0 periodicity=0 colors=@volcano.map } BH_Detail {;Detail from BH_CubicMan_3 ;Unusual looking Mandelbrot Set ; reset=2003 type=formula formulafile=archiv.frm formulaname=BuffHenriksenMan passes=t corners=1.7079034/1.2211/-1.0240096/-0.79045796/1.4205494/-0.64087092 params=0/0/0/0/16/0 float=y maxiter=1024 inside=0 periodicity=0 colors=@volcano.map } BH_MainAntenna {;Detail from BH_CubicMan_4 ;Unusual looking Mandelbrot Set ; reset=2003 type=formula formulafile=archiv.frm formulaname=BuffHenriksenMan passes=t corners=0.432704149/0.648391824/2.17659659/1.99153324/0.599182127/1.9546\ 2595 params=0/0/0.715/0.715/16/0 float=y maxiter=1024 inside=0 periodicity=0 colors=@volcano.map } frm:BuffHenriksenMan {;by Xavier Buff and ;Christian Henriksen, "Julia sets in parameter spaces" ;Function: z^3 + b*z^2 + lambda*z, lambda fixed ;periodicity=no ;------------------------------------------------------------------------- ;p1 : Perturb z(0) ;p2 : Lambda ;p3r: Bailout ;p3i: Orbittracking(0|1) - maps either first or second Critical Point to z ;------------------------------------------------------------------------- ; zz = sqrt(sqr(pixel)-3*p2) IF (imag(p3)) z = p1 - (pixel-zz)/3 zz = p1 - (pixel+zz)/3 ELSE z = p1 - (pixel+zz)/3 zz = p1 - (pixel-zz)/3 ENDIF: z = ((z+pixel)*z+p2)*z zz = ((zz+pixel)*zz+p2)*zz p3 >= |z| && p3 >= |zz| } ------------------------------------- END -----------------------------------
I don't mind if you scrabble the code a little , I included all of the comments to inform those interested , about how I derived the appropriate formula . At this stage I haven't investigated what the iteration represents , except for the purely real axis . It is Winfract 18.21 , for now. Because of a few limitations with this I also installed the most recent DOS version of Fracint. ----- Original Message ----- From: "Gerald K. Dobiasovsky" <gerald.dob@aon.at> To: "Fractint and General Fractals Discussion" <fractint@mailman.xmission.com> Sent: Friday, July 17, 2009 12:36 PM Subject: Re: [Fractint] Diode Fractal
sciwise,
do you really use Winfract 18.21? It's very old (anything before version 19.6 is more or less "stone age territory"). There's an updated version here (I can't reach www.fractint.org at the moment):
http://twegner.dyndns.org/fractint/ftp/current/windows/winfract-20.04p09.zip
Cut and pasted from Winfract 20.04patch9 Help: --------------------------------------------------------------------------
-------------------------------
This release of Winfract is essentially the same as the version 18.2 but with the following enhancements:
- All fractal types from Fractint 20.4 (with the exception of the 'julia_inverse' fractal type, which requires some extra logic we haven't yet added to Winfract) have been incorporated into this version of Winfract.
- The File Open, File SaveAs, and File Print options now use the new Windows-style "common dialog boxes", with the extra capabilities provided by those routines. This is particularly handy if you're using Windows-for-Workgroups or Windows NT, 2000, XP with their networking options.
- The fractal history feature of FFD has been implemented and uses the h key to move back in history and the H key to move forward in history. --------------------------------------------------------------------------
I hope you excuse me from scrambling your code (not all of it was
necessary, but is the result of trying
to get it to display something on my machine.
Note that the inclusion of formulas in .par sets works only from version 19.6 on, if my memory serves me right.
Regards, Gerald
---------------- .PAR SET BELOW ------------------------------------ Diode { reset=2004 type=formula formulafile=fractint.frm formulaname=LnDioGH corners=-0.16588419/0.25978091/-0.15031315/0.1691023 float=y maxiter=512 logmap=yes periodicity=no colors=00000j0j0jj0j00j0j0jjkkkkrkdmwW000W0WW000WW0W0WWFNF<5>zNFFVF<5>zV\ FFbF<5>zbFFjF<5>zjFFrF<5>zrFFzF<5>zzFFFN<5>zFNFNN<5>zNNFVN<5>zVNFbN<5>zb\ NFjN<5>zjNFrN<5>zrNFzN<5>zzNFFV<5>zFVFNV<5>zNVFVV<5>zVVFbV<5>zbVFjV<5>zj\ VFrV<5>zrVFzV<5>zzVFFb<5>zFbFNb<5>zNbFVb<5>zVbFbb<5>zbbFjb<5>zjbFrb<5>zr\ bFzb<5>zzbFFj<5>zFjFNj<5>zNjFVj<5>zVjFbj<5>zbjFjj<5>zjjFrj<4>rrjzywccdWW\ Wz000z0zz000zz0z0zzzzz }
frm:LnDioGH{ ; ; Complex Voltage - Complex Current ; Diode Characteristics . ; ; for 1N4148 silicon type . ; Level 1 model . ; ; at 25 degrees C ; Pd = 0.0000000000001 Gm = 0.00000000000094 Ids = 0.00000000252 Rs = 0.968 n = 1.752 Vt = 0.025 u = Gm*n*Rs*Vt+n*Vt Pt = n*Vt vr = Ids*Rs x = 0.1 x = x-((u*x*log(x)+vr*x*x-(Pixel+vr)*x)/(vr*x+u)) z = 0-(Pt*log(x)-Pixel)/Rs: u = z x = x-((u*x*log(x)+vr*x*x-(Pixel+vr)*x)/(vr*x+u)) z = 0-(Pt*log(x)-Pixel)/Rs |u-z| > Pd } -------------- END OF .PAR SET ------------------------------------------
_______________________________________________ Fractint mailing list Fractint@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/fractint
JackOfTradeZ@comcast.net wrote:
No PAR or FRM is seen, and no attachments. Maybe it is problem on my end? I am using comcast web client email.
Must be Comcast web mail client. I saw them in Thunderbird. -- David gnome@hawaii.rr.com authenticity, honesty, community
participants (6)
-
david -
Gerald K. Dobiasovsky -
Guy Marson -
JackOfTradeZ＠comcast.net -
John W. -
sciwise