January 24...Displaying Monitor Spectra in the Interactor
Revised January 26

A user recently requested the ability to compare theoretical and measured optical monitor curves (%R/%T vs. layer thickness) in the Interactor. While seemingly unavailable, the capability is implemented through 'DATA' macro commands. DESIGN 2.61.4690 makes the procedure somewhat more convenient.

 

As the Interactor screen capture shows, BLUE indicates calculated H layers, RED calculated L layers, and GREEN monitor trace (simulated here by Copy/Paste Spectrum). As the design is modified, the green trace moves (confusingly?) as the horizontal scale follows layer thicknesses changes. In addition there is the issue of physical thickness accuracy in the deposited film. We await reports as to the method's effectiveness.

Care is required to ensure that the optical monitor trace has the DESIGN data format. Column A is total thickness (nm), Column B is reflectance (0-1) and Column C is transmittance (0-1). Note that the end of Layer 1 and beginning of Layer 2 have the same values because the monitor wavelengths for the adjacent layers are the same. Deleting Row 21 or Row 22 gives the same monitor trace in DESIGN. (Unequal wavelengths give vertical jumps at boundaries. Check Specify evaluation wavelengths... in the Graph Axes dialog.) Users needing help with Macro Commands or requiring additional explanation should contact FTG Software.

---

February 18...MEASURE Scan Methods with the FilmStar Database

Recently a user expressed interest in utilizing the FilmStar Database for MEASURE Scan Methods. Typically, FilmStar BASIC exports FileMaker script commands (F*S >> FM). Another possibility is a FileMaker front end (FM >> F*S). You might, for example, include photos showing system setup as well as instructions in local languages. How does it work? FileMaker's Send DDE Execute command triggers FilmStar Macro Commands, for example sending SCANPLOT to MPE983|Main to scan and plot the spectrum. Ultimately, BASRUN or BASEXEC returns macro commands for saving spectra in a separate database. DDE commands can also be sent to Excel for pass/fail analysis. While there are a large number of MEASURE DDE macro commands, more can be added as required. Contact FTG for an online explanation or see for yourself how the  SCAN  button works: Scan Methods: Methods.zip (Test Mode) Unzip to C:\Winfilm\Measure (or ..Scantraq) FileMaker Database: Methods.fmp12 FileMaker 45 Day FREE TRIAL Win 10/11

Cary UMA: Mcary5U 2.51.2062 required

---

March 11...Tooling Factor global search
Minor Revision August 1

A user reported that Tooling Factor determination failed. As did not provide details enabling us to reproduce the issue, we assumed that initial default tooling factors (1.0) were too far from the solution, therefore giving a local minimum with Cannot invert singular matrix error.

Gedankenspektrum analysis: Using design ToolTest.faw we simulated a 'Measured' spectrum by assigning (random) values L:1.138 & H:1.356 in Setup...Tooling Factors. The spectrum was converted to optimization targets. We next reset default initial factors L:1.0 and H:1.0 and solved for tooling factors via the dialog's <Compute> button. The solution is given in the DLS column. (Tooling factor determination always uses DLS).

Noting that DLS Tooling Factors 1.2551 and 1.0445 are erroneous local minima, we implement Plan B: a simple BASIC code searching for closest initial factors.

' ToolSearch.bas for FilmStar DESIGN
' Tooling Factor global search
' Two film materials: L (Film 1, Row 3) in Film Indices and H (Film 2, Row 4)

DefDbl A-Z
Sub Main
    vBest = 10^8
    Tooling$ = "1 1" ' Default starting values
    AxesDraw
    Calculate
    Replot
    For fL = 1 To 1.5 Step 0.05   ' Modify ranges as required
        For fH = 1 To 1.5 Step 0.05
            Tooling$ = CStr(fL) & " " & CStr(fH)
            v = Merit  ' FilmStar BASIC function Merit
            If v < vBest Then vBest = v: fLbest = fL: fHbest = fH
        Next fH
    Next fL
    Tooling$ = CStr(fLbest) & " " & CStr(fHbest) ' New starting values
    AxesDraw
    Calculate
    Replot
End Sub

New factors are displayed above in the Search column. We now return to Setup...Tooling Factors and once again click <Compute>. Results in the Search + DLS column now match the 'Measured' column.

Users requiring help implementing BASIC code should contact FTG Software.

---

March 28...CIE + %T targets Design Exercise

A user asked about a design with T=10% CIE response (D65 illumination) and maximum %T from 870 to 890 nm. Starting with a standard highpass stack we set up a Workbook with Objective F6+G6+0.02*H6+(1-AVERAGE(D499:D519)) and Workbook Macro CALCULATE;BasExec Dim x!, y!, yy!|GetCie x, y, yy|WbSetNum "B6", x|WbSetNum "C6", y|WbSetNum "D6", yy;. BasExec inserts BASIC code directly into the Workbook. Sub Main and End Sub are added automatically and "|" is equivalent to CRLF. BasExec is not intended for code containing multiple subroutines. More about BasExec: here & here.

Sub Main       Dim x!, y!, yy!       GetCie x, y, yy       WbSetNum "B6", x       WbSetNum "C6", y       WbSetNum "D6", yy   End Sub

CIE targets are X=.3127, Y = .3290, YY= 10. While there is a great difference between the blue (Workbook only) and green (DLS+Workbook) spectra, that difference might not be apparent when viewed at 0°, as indicated by the coordinates given below. Ripple in the Workbook  solution can be reduced by constraining the standard deviation (STDEV).

Workbook Optimization

DLS + Workbook Optimization

Will two radically different designs look the same?

Referring to the four Comparator images above, we see no color difference between the designs at normal incidence. The Workbook design, however, shifts to red at 30° while the DLS design remains neutral. In the image at the right, the rectangle corresponds to WB 0° while the circle corresponds to WB 15°. We expect that an observer would have little difficulty in distinguishing the two designs. Considering only normal incidence, is the DLS design more stable than the  Workbook design? As shown below, the DLS CIE scatter plot (S.D. = 2%) is essentially the same as Workbook only.

DLS + Workbook 75% Yield

Workbook 74% Yield

Optimization Targets in a Workbook Objective: Transferring 380 to 890 x 1 nm spectra is wasteful if we only utilize 870-890 nm in the Objective. We speed-up calculations with Optimization Targets 870 to 890 x 1 nm and Workbook Macro BasExec Dim x!, y!, yy!|WbSetNum "G3", Merit|GetCie x, y, yy|WbSetNum "B6", x|WbSetNum "C6", y|WbSetNum "D6", yy; and Objective Function F6+G6+0.02*H6+0.2*G3.

The Workbook (tutorial) is a complex component enabling optimization scenarios not otherwise possible; users requiring further explanation or assistance should contact FTG Software to arrange a ZOOM meeting.

---

June 14...Target Generator Update

Updated Target Generator adds Reciprocal Wavelength Spacing

Inverse spacing works as follows: a 2 nm interval at 500 nm becomes a 4 nm interval at 1000 nm and an 8 nm interval at 2000 nm. This was originally added to speed calculations and has been in DESIGN for many years. Inverse spacing, however, adds relatively too many IR targets and too few UV targets for very wide band coatings. The new Reciprocal Wavelength Spacing option gives equal frequency intervals for wavelengths in nm. This better obeys the 11th Commandment "Thou shalt have more targets than wiggles." (Note: Columns A-C computed by DESIGN, Column D computed with Excel formula D2=10^7/C2, etc.) This simplifies target spacing. Previously, a designer requiring frequency-spaced wavelengths would need to create targets in 1/cm and then convert to nm or µm.

Note: Lines overlap at 500 and 1000 nm

---

July 6..Wavelength List Generator Update

The Wavelength List Generator (Setup...Wavelength List...File..Generate) has been updated in DESIGN 2.61.4731 and MEASURE 2.51.2080 with spacing options similar to the Target Generator (above).

Inverse and Reciprocal (shown below) spacing are especially useful for speeding up Workbook optimization. If reflectance between 800 and 1200 nm is unspecified it can be ignored. Note that <Ctrl+D> deletes a single cell or range of highlighted cells in the Wavelength List dialog.

---

August 12..Lorentzian Dispersion Functions Revisited

Lorentzian functions are very useful in characterizing dielectric materials. Built-in User Index functions LOREN and LORENK have now been supplemented with LORENM and LORENKM (DESIGN 2.61.4743, INDEX 2.51.0688) wherein coefficients C and D (LORENKM) are µm-based.

DESIGN/INDEX n,k calculations can be verified in Excel. Click here to download. The FilmStar BASIC code below can be used to check DESIGN calculations. Don't forget to add LORENK etc. to the list of available built-in functions. Otherwise it will not appear in the pull-down Dispersion list.

Sub Main
    GetIndex "Z", 550, xn!, xk!
    Debug.Print xn
    Debug.Print xk
End Sub

---

August 29...Optical Path Length

A user needs to compute optical path length (OPL) in a film stack. OPL is the sum of n*d where n is index and d is path length, typically in nm. We thought it best to implement as an instructive BASIC macro (shown below) thereby giving users output format flexibility. Click here to download the design illustrated here. Noting that path length increase with angle (Snell's law)... tPath = n2/Cos(Arcsin(Sin(Angle)*n1/n2)) or lacking an Arcsin function tPath = n2/Sqr(1-(Sin(Angle)*n1/n2)^2) where n1 is entrance medium index (usually 1) and n2 film index.

In the code below there are two Sub Main versions. Comment-out the version not being used.

' OpticalPathLength.bas for FilmStar DESIGN
' Copyright 2022 FTG Software Associates
' Computes total path length Σ n*d
Option Base 1
DefInt I-N
DefSng A-H, O-Z
Option Explicit
Const aRad = 57.2957795 ' Converts angle to radians

'Sub Main ' simple single wavelength case uses current calculation angle
'    Calculate  ' Recommended (loads n,k tables)
'    Debug.Print Format$(tOptical(550), "0.00") & " nm"
'End Sub

Sub Main 'Print in Immediate window & copy to Clipboard
    Dim Ang, Wave, Clip$
    Clip$ = "Ang (°)" & vbTab & "W (nm)" & vbTab & "OPL (nm)" & vbCrLf ' Header row
    Calculate  ' Recommended (loads n,k tables)
    For Wave = 400 To 800 Step 50 ' Set same min and max if no dispersion
        For Ang = 0 To 75 Step 15
            Angle = Ang ' Set calculation angle
            Debug.Print "Ang " & CStr(Ang) & "°, W (nm) " & CStr(Wave) & _
              " nm: OPL=" & Format$(tOptical(Wave), 2) & " nm"
            Clip$ = Clip$ & CStr(Ang) & vbTab & CStr(Wave) & vbTab & _
              Format$(tOptical(Wave), "0.00") & vbCrLf
        Next Ang
    Next Wave
    Clipboard Clip$ ' Paste into Excel
End Sub

Function tOptical(ByVal Wave)
    Dim symb$(), nTyp$(), nFile$(), nDesc$(), tSymb$
    Dim i, k, tPath(), t, qd(), qt$(), qdesc$(), nMatl, xn, xn1, xk, nLay
    DesignConvert 1, "N" ' Physical thickness in nm
    GetIndexList symb$(), nTyp$(), nFile$(), nDesc$()
    nMatl = UBound(symb$)
    For i = 3 To nMatl
        tSymb$ = tSymb$ & symb$(i)
    Next i
    ReDim tPath(3 To nMatl)
    ' If ".itw" error in GetIndex, Calculate spectrum and try again
    GetIndex 1, Wave, xn1, xk ' xn1 is entrance medium index
    For i = 3 To nMatl
        GetIndex i, Wave, xn, xk
        tPath(i) = xn/Sqr(1-(Sin(Angle/aRad)*xn1/xn)^2) ' d 1/cos increase
    Next i
    GetLayers nLay, qd(), qt(), qdesc() ' Converts Groups; Stack Mode not supported
    For i = 1 To nLay
        k = InStr(tSymb$, qt(i)) + 2
        t = t + qd(i) * tPath(k)
    Next i
    tOptical = t
End Function

If you've read this far you might be interested in improving FilmStar BASIC skills. Please contact FTG Software with any issues.

---

September 2...Verifying THK Variable(s) in DESIGN

DESIGN 2.61.4744 (Sep 1) erroneously disabled
BASIC Sub IndexGet in Stack Mode. Fixed in 2.61.4745.

Users need to verify dispersion functions in DESIGN. While verifying (n,k) vs. wavelength and vs. thickness is (slowly) possible in the Index Formulator, n,k arrays can be plotted in Excel or INDEX.
 
THK variables in User-Defined Index Functions are the basis of Double Dispersion. Users should verify calculations with a DESIGN BASIC macro based on the listings below. Since layers of the same material can have different refractive indices, Stack Mode is required when THK variables are specified.

Option Explicit ' Optional but always a good idea
DefSng A-H, O-Z
Sub Main ' Versus layer thickness (THK variables)
    Dim t, s$, xn, xk
    s$ = "t (nm)" & vbTab & "n" & vbTab & "k" & vbCrLf ' Row header
    Calculate ' Loads dispersion
    For t = 0 To 200 Step 20
        SetLayerThk 2, t         ' Layer 2 is in Row 4
        GetIndex 4, 550, xn, xk  ' 550 nm
        ' Debug.Print CStr(t) & ": " & CStr(xn) & ", " & CStr(xk)
        s$ = s$ & CStr(t) &vbTab & CStr(xn) & vbTab & CStr(xk) & vbCrLf
    Next t
    Clipboard s$ ' Paste into Excel
End Sub

Calculate loads n,k tables. Alternatively, click Evaluate...Recalculate <F9> before running the macro the first time. Repeat if you modify index tables. Hint: Calculating before running the macro reveals any errors. When learning about THK variables, start with a very simple function such as -(THK<=100)*1.5-(THK>100)*2.5 wherein n is 1.5 when t<=100 nm and 2.5 otherwise.

With no THK variables, calculations are simplified and GetIndex refers to either the Film Indices row number OR Stack Mode editor row number. Here we evaluate as a function of wavelength and eliminate the header line:

Option Explicit
DefSng A-H, O-Z
Sub Main  ' Versus wavelength
    Dim s$, xn, xk, wave
    Calculate ' Loads dispersion
    For wave = 400 To 800 Step 20
        GetIndex 4, wave, xn, xk ' Row 4 in either Film Indices or Stack editor
        ' Debug.Print CStr(wave) & ": " & CStr(xn) & ", " & CStr(xk)
        s$ = s$ & CStr(wave) & vbTab & CStr(xn) & vbTab & CStr(xk) & vbCrLf
    Next wave
    Clipboard s$ ' Paste into INDEX (or Excel)
End Sub

Please contact FTG Software with issues and comments. No BASIC experience? Review our BASIC Tutorials and contact us for a ZOOM meeing.
 

---

September 9...MEASURE BASIC FileSave Bug

A user asked about adding a description line to CSV spectra. In testing BASIC Property FileDescription, we found that BASIC Function FileSave failed. The bug, fixed in 2.51.2084, is in revisions updated earlier this year.

How to test FileSave:

Sub Main
    Scan ' Use Test Mode or open/paste spectrum
    FileDescription(1) = "Test A (FileSave)"
    FileSave "C:\Winfilm\Spectra\TestA.csv"
End Sub

If FileSave is defective TestA.csv is not created. In that case, update if possible or use the equivalent alternative BASIC code listed here:

Sub Main ' Plan B
    Scan ' Use Test Mode or open/paste spectrum
    FileDescription(1) = "Test B (DATASAVE)"
    Macro "DATASAVE C:\Winfilm\Spectra\TestB.csv"
End Sub

---

September 19...Font Size Multiplier (DESIGN Interactor & INDEX)

FilmStar has numerous graphics options for publications, PowerPoints web pages, etc. A user, planning to utilize DESIGN Interactor plots and INDEX n,k plots in publications, found font sizes to be too small.

The Font Multiplier setting in File..Configuration..Misc applied to both Interactor and INDEX plots. Reviewing this capability, we recognized the need for separate multipliers, as introduced in DESIGN 2.61.4748/INDEX 2.51.0691. Font Multiplier settings have been relocated as shown below.

---

October 28...Copying INDEX n,k Tables to Excel

Index tables saved in separate sheets

A user requested means to copy numerous INDEX tables to Excel in order to compare multiple files and determine which are redundant or too low resolution as in AG above. Instructive (VBA code not hidden) multi-sheet examples: IndexTablesEarly.xlsm (early binding) IndexTablesLate.xlsm (late binding) Note the format options in Sub FileToSheet. Header rows (not shown above) require INDEX 2.51.0692 (Nov 1) with updated BASIC Sub IndexCopy. Early binding is useful when writing code as IntelliSense is supported. Late binding lacks Intellisense but is less fussy regarding its connection with, in this case, FilmStar INDEX. Sheets are loaded in alphanumeric order. Should you manually add sheets later, reorder with this VBA code. Try this version to save all data in one worksheet. Users requiring help should contact FTG Software. Those learning VBA will find books and courses by searching "learn Excel VBA".

VBA IntelliSense pull-down for INDEX object

---

November 4...Modifying Plot Module Legends

A user requested options in Collector-generated Plot Module graphics. DESIGN 2.61.4758 and MEASURE (Scantraq) 2.51.2092 extend BASIC Property PlotLegend with the ability to read [t$=PlotLegend(i)] as well as assign [PlotLegend(i)=t$] labels. This modification plus new BASIC Sub PlotRedraw allows legends to be automatically modified. This capability should also be useful when dealing with Photon RT formats. Hint: If there are both %R and %T columns in the Collector, set the y-axis label to '% Response' by setting Auto Y-Axis Label to False in the Main Menu Graph Details dialog (<Ctrl+H>).

Sub Main     Dim i%     For i = 1 To 12         s$ = PlotLegend(i)         If s$ <> "" Then             s$ = Replace(s$, "RP", "%R P-pol")             s$ = Replace(s$, "RS", "%R S-pol")             s$ = Replace(s$, "RR", "%R R-pol")             s$ = Replace(s$, "TP", "%T P-pol")             s$ = Replace(s$, "TS", "%T S-pol")             s$ = Replace(s$, "TR", "%T R-pol")             PlotLegend(i) = " " & s$ & "°"         End If     Next i     PlotRedraw End Sub

Legends created by the Collector module

Legends modified by the BASIC code listed above
 

---

November 29...DESIGN Tolerancing Algorithm

A user had questiones about DESIGN's tolerancing function and its mathematics. We thought the best way to show how it works is in an Excel workbook. Pseudo-random numbers are multiplied by either Abs Error or % Error limited to ±Max Error. Basing selling price on tolerancing is better than guessing, but not as useful as pricing based on previous production. Yield estimation for complex requirements is possible in the DESIGN  Workbook

---

December 12...Importing OLIS 983 Spectra

The OLIS 983 modernizes the venerable PE 983 IR grating spectrophotometer. While FilmStar MEASURE does not directly control the OLIS 983, the Scan button (or a Macro button) can be set to import spectra from OLIS Excel files. In this case the 'instrument' is actually the Mpe983 Raw Data Server in the Scan Method dialog.

MEASURE Mpe983.exe version 2.51.2094 adds optional BASIC FileOpen parameters enabling users to open OLIS and other Excel files wherein column 1 is wavelength and additional columns are response. FileOpen also allows for multiple header lines (default 1). When triggered by the Scan button via attached Scan Macro, reflectance is normalized (reduced) according to the specified Standard file. Alternatively, a separate BASIC macro containing DataReduce is possible.

The Scan Macro is given by the BASIC code shown at the top right, specified in Scan Method 4/7 (Control Macros). Highlight the Sample Scan cell and click Edit... Edit Row <F5> to bring up a larger code window. The code at bottom right applies to a BASIC macro not connected to the Scan button. Default configuration: two columns: Column 1: Wavelength, Column 2: %R/%T. Additional data columns n: %R/%T where column number n≠2 is specified as follows: FileOpen  ,,,,"N","%R"   ' 1 Header line (default), Column 2 (default) FileOpen  ,,4,3,"N","%R"  '  4 Header lines, Column 3 FileOpen  ,,0,3,"N","%R"  ' No header lines, Column 3 Note that the optional parameters apply only to Excel files (not Collections). Typically the user imports an Olis spectrum with a File Open dialog. To display the file name in graph titles, check the following in File..Configuration..Preferences. As an alternative to the File Open dialog, a spectrum can be specified as follows:  FileOpen "C:\Spc\My983.xlsx",,,,"N", "%R". A FilmStar Excel spectrum typically has W (nm) in cell A1 (W (µm) etc.) and %R or %T or OD in cell B1. In order to retain the original OLIS spectrum, Mpe983 2.51.2094 adds variables xUnit$ (wavelength N {nm}, M {µm}, A {Ångs), C {1/cm}) and Typ$ (%R/%T) to the FileOpen function. Users requiring additional explanation or compatibility with other spectral files should contact FTG Software. For OLIS tab-separated text files, click here.

'Scan Macro  Sub Main  FileOpen ,,,,"N","%R"  DataSmooth  AxesDraw  End Sub      'Standalone Macro  Sub Main  FileOpen ,,,,"N","%R"  DataReduce  DataSmooth  AxesDraw  Replot  End Sub

Back to Technical Issues