January 2    Collector to Targets

January 8    IndexSolve Updated with Collector

January 29    Spektrum Support for CEC Deposition Systems
(Minor revisions 13 Dec 2018, 28 Jan 2021)

February 26    Peaks & Valleys in MEASURE/Scantraq

  ' FSPlot Annotation code for FilmStar BASIC:
  s$ = vbCrLf
  PlotClose
  PlotAnnotate True
  PlotAnnotation = "VERS3" & s$ & _
  "vline 279.3,9,|t279.3 ,16711680;0" & s$ & _
  "vline 360.8,9,|t360.8 ,16711680;1" & s$ & _
  "vline 418.5,9,|t418.5 ,16711680;2" & s$ & _
  "vline 459.9,9,|t459.9 ,16711680;3" & s$ & _
  "vline 536.4,9,|t536.4 ,16711680;4" & s$ & _
  "vline 334.8,9,|t333.8 ,16711680;5"

New BASIC Sub WaveCorrect applies the Correction Function to an existing uncorrected spectrum. This should only be run ONCE!

---

A user wondered whether it was possible to display substrate and air indices within graphs of Refractive Index n vs. Thickness. As shown at the right, this is readily simulated by adding 'layers' having the same index as SUB and AIR in Setup..Film Indices, in this case 250G for glass and 250A for air. The problem is that DESIGN plots the 10 layer stack as a 12 layer construction as indicated in the FSPlot graph below. We need a way to remove labels "1" and "12" and label the actual film layers starting with "1".

 

How to tell DESIGN not to label layers 1 and 12? The following method was devised: insert an asterisk in the Film Indices Description column (<F3> for quick access), followed by optional text to be displayed. Note that this only modifies outer layers. An air layer in the middle of a stack would be counted as a layer.

We then have the following:

This works with various plot types with thickness axes in the main menu, FSPlot, and Interactor. It is especially useful when plotting electric field. The feature was introduced in DESIGN 2.61.4193.

---

A user needs to calculate absorptance in a specified layer at non-zero angle of incidence. This required an update since DESIGN was limited to angle = 0. We then needed to verify results. This was accomplished by collecting spectra in FSPlot to see whether partials add up to total %A. We tested 45° R-pol. (Design 100H 100L 100H 100L, L: 1.46 + 0.2i , H: 2.3 + 0.1i, AIR: 1.0, SUB: 1.52).

31 Mar 2020 Update...Further tests indicate that FWD Ignore Side 2 is required and that the substrate must have k=0. The substrate can also be treated as a massive layer in which k=0.

Looks good, but we need to calculate the difference. Click Data..Copy in the FSPlot Module and paste into Excel. Set $H2=$B2-$C2-$D2-$E2-$F2 {or $H2=$B2-SUM($C2:$F2)}, and copy the formula to subsequent rows. Results indicate that the algorithm works as designed.

Copy and paste? That's no solution for an advanced user, so read on!

Addendum April 23...FilmStar BASIC users can utilize enhancements in DESIGN 2.61.4211 whereby Sub XLwrite allows numeric/text formulas as well as numeric or text entries, and Sub XLnew supports R1C1 references (for our purposes far more convenient than column letters). AbsMulti.bas (updated 31 Mar 2020)calculates spectra and activates the Excel Viewer. It  works with any number of layers. XL functions are available in ALL FilmStar programs. Excel is NOT required.

Having specified R1C1 references in XLnew, the XLwrite numeric formula in column 8 evaluated at row 2 is R2C2-SUM(R2C3:R2C6). Functions such as SUM which can be utilized in formulas are listed here. A formula starting with = denotes a numeric cell; a formula starting with & denotes text. Anyone familiar with Excel notation should find this straightforward.

Finally, as shown here, formulas can be also utilized in the FilmStar Workbook (BASIC WbSetFormula) but, lacking R1C1 notation, cell addresses get messy beyond 26 layers.

---

April 30   UMA Scan Macros

MEASURE's Scan Macros provide means to display QA instructions in local languages. A macro can consist of a BASIC file name or multi-line BASIC code. Scan Macros are saved with Scan Method, thus different from Macro Commands (Tools menu) which do not change when a new Scan Method is opened. A typical macro might be MsgBox "Check that there is no optic in the sample compartment."

The updated Scan Macros dialog supports Cary UMA Sequence Mode. There are two entries (before & after UMA scan) instead of six (before & after Baseline, Reference, Sample scans). Sub CollectProc eliminates the need for technicians to select options.

Sub XLwrite does not overwrite spectral files, saving instead with appended numbers: Collection1.xlsx, Collection2.xlsx, etc. (Data file names are optional since data can also be saved in the Collector.)

DO NOT USE Scan Macros when scanning with BASIC Sub Scan.

---

May 1   CIE Tolerancing

An engineer at a filter manufacturer requires Monte Carlo calculations related to color variance. It took us about a half hour to create FilmStar BASIC code giving the screen below. Here we assumed 2% SD layer variation. Referring to the code listing, 'Tooling range' accounts for non-uniformity. Note our use of AVERAGE and STDEV. The complete list of functions can be found here; any similarity to Excel is intentional!

While there are other ways to accomplish the same in FilmStar, this method's output can be read by any Excel user. In fact Excel is not needed; here we use the free Excel viewer (replaced by built-in FTG Excel viewer June 2018.) linked to FilmStar via Sub XLview. As mentioned before, FilmStar's ability to create Excel files is standalone; Excel is not needed.

Our next step would be to define a CIE acceptance box and calculate % yield. With FilmStar that is an equally straightforward task.

---

June 14   n,k Solution Grid

The following introduces Gedankenspektrum research underway at FTG Software and Willey Optical to be reported at OSA OIC 2019. Consider a single absorbing film on a transparent substrate. What measurements are required to determine n,k (film thickness known) or n,k,d at a given wavelength? %R and %T certainly, but what about reverse %R (thanks to a FilmStar user for this suggestion) or P/S measurements at multiple angles? These issues are amenable to analysis with FilmStar's ability to directly create multi-tab Excel files. Color provides a sort of contour plot.

L-M n,k Solution Grid for 550 nm, 0° with (LM_RTRev.xls) and without (LM_RT.xls) reverse
reflectance; green: RMSD< .025; red: RMSD>1 {RMSD = Sqr((n-n0)^2 + (k-k0)^2)}

In a variation of IndexSolve.bas we determine indices from calculated (for testing) or measured spectra, the above chart (200 nm) is corroborated below where we see that 0° R/T (blue trace) is insufficient over the entire wavelength range. Adding reverse R at 0° or 45° R P/S leads to a correct solution (L-M method). (User1.itw is a user-supplied Gedankenspektrum test file.)

---

June 21   XLsave Tab-Delimited Output

NOTE: This topic is intended for advanced users familiar with FilmStar BASIC and FileMaker scripting. In summary, it demonstrates how to transfer data from FilmStar MEASURE to a FileMaker Pro database. Please contact us for an online demonstration at your convenience.

FilmStar Database example FM12_SaveSpectra.bas transfers CIE A and D65 values to FileMaker Pro by copying values to the clipboard and pasting into FileMaker. As shown below, copying is simple.

June 28   Cary UMA UV Issues

A user reported that Cary UMA spectra were noisier in FilmStar than in Agilent's WinUV. Screen-sharing revealed that WinUV had utilized default 4 nm SBW (spectral bandwidth) while FilmStar was set to 0.5 or 1.0 nm. Upon retesting MEASURE, performance was seen to be equivalent as shown in the FSPlot curves below. The traces represent two 100% baseline scans (no sample, detector 180°, no polarizer) divided by each other.

Noise will be increased if the polarizer is used and reduced by longer SAT (signal averaging time). Should long times adversely affect throughput, consider scanning UV and visible regions separately and combining afterwards.
 
This raises issues as to the performance of accessories in the UV. While AR and HR measurements should be acceptable, coatings with steep edges might require you to remove the UMA. The same considerations apply to the PerkinElmer Lambda 950-1050 URA.

---

Supplementing FilmStar support for directly opening and saving Excel files, we adapted the Microsoft Excel Viewer for verifying files before opening. Unfortunately Microsoft discontinued Office viewers in April 2018. We therefore introduce the FTG Excel Viewer (shown below) with the advantage of including <Open Workbook> and <Cancel> when accessed from Spectra.. Open or from BASIC Function XLopen.

About to open the wrong file? Click Cancel or press Escape

Advanced users:  Under BASIC control, the FTG viewer simplifies opening an XL file after viewing it. For example, in the revised version of our page discussing angular dispersion, the following code...

If Not XLopen("C:\Winfilm\Measure\WaveAngles.xls",, True) Then
    XLclose          ' close internal spreadsheet object
    End              ' quit
ElseIf MsgBox("Click OK to proceed or Cancel to quit", _ 
  vbOkCancel, "View and Verify") = vbCancel Then
    AppActivate "Microsoft Excel Viewer"
    SendKeys "%{F4}" ' close Microsoft Viewer
    XLclose          ' close internal spreadsheet object
    End              ' quit
End If

was replaced by the simpler (MsgBox not needed)...

If Not XLopen("C:\Winfilm\Measure\WaveAngles.xls",, True) Then End

---

Suppose a FilmStar computer is connected to two spectrophotometers. Two possibilities: two MEASURE executables or one (usually Mpe983.exe) running both instruments. You require a procedure that sets the startup Scan Method (*.scw) and directory. MEASURE, DESIGN and INDEX already supported startup files, but the feature was not documented. MEASURE 2.51.1923 and DESIGN 2.61.4222 set paths as well. Here are three possibilities:

1. Startup BASIC macro 1 sets method directories for multiple MEASURE EXE files:

3. Batch file (*.bat) (BASIC not required) opens the EXE and sets default Scan Method directory:

REM MEASURE Startup with Lambda 9 Start "", C:\Winfilm\Mpe983.exe "C:\Winfilm\Measure\Lambda9\PE9-AR.scw" REM DESIGN Startup with BBAR-O.faw Start "", C:\Winfilm\Design1.exe "C:\Winfilm\Designs\BBAR\BBAR-O.faw Save on your desktop. This technique eliminates having to devote a macro button to startup. (Note: Without the Start command, the Command Prompt dialog remains open and must be closed manually.)

As always, users requiring assistance with FilmStar BASIC should contact FTG Software to set up a ZOOM tutorial session.

---

We were asked about meeting L*a*b* requirements. This inspired us to test DESIGN's CIE color algorithm. We started by discovering colorizer.org and, for no particular reason, selected purple as shown below.

Colorizer RGB

We designed a filter by starting with a 22 layer SiO2/TiO2 QW stack and optimizing in the CIE dialog using the Color Optimization Targets shown below. We changed materials to SiO2/Ta2O5 and re-optimized to find a different design giving identical color values. We then obtained the same L*a*b* coordinates with a 14 layer SiO2/TiO2 stack. FTG colleague Ron Willey reported a design with only 9 layers (FILM Archive .faw). It would be interesting to repeat with different light sources, including LED.

22 Layer SiO2/TiO2 Initial (orange) and final colors

 
Three designs with identical L*a*b* values

Coincidentally (see August 2 above), a FilmStar user requires CIE L*a*b* tolerancing calculations where color difference ΔE=Sqr((a-a0)^2 + (b-b0)^2) defines acceptance. Rather than implement this directly in FilmStar DESIGN, we found it far more useful to create an Excel workbook utilizing DESIGN CIE capabilities. Users should have no problem modifying the workbook for different color criteria such as L*u*v*.

To reproduce the above chart, open Willey's 9 layer purple filter in DESIGN and then click Calculate in the Excel workbook. Referring to column A, the graph and Yield automatically update upon changing Center a*, Center b* or Radius. Press <Alt+F11> to view VBA code. Change tolerancing values in DESIGN (<Ctrl+N>) and repeat. Note how CieParams ensures that DESIGN is set to L*a*b*.

Addendum January 24, 2020...As suggested by Ron Willey, a new Excel workbook supports MacAdam ellipses. There are now two radii and a tilt angle. Should you wish to modify the workbook, unhide columns R-AD. Copy the Purple Filter design(s) to C:\Winfilm\Designs. The Full Version example includes CIE xy and u'v' in addition to L*a*b* (Purple.zip: MacAdamTolBox.bas, xyPurple.faw, LabPurple.faw, u'v'Purple.faw). The workbooks listed below were corrected and updated Jan 31.

Having implemented the Excel yield calculation, we wondered if the ellipse could serve as a Tolerance Box (Polygon) in DESIGN's built-in CIE tolerancing. As this did not work properly, the algorithm was updated in DESIGN 2.61.4445. Note that a FILM Archive file saves Tolerance Box coordinates, making it easy to use the same MacAdam ellipse with different designs.

The Excel workbook's [Copy] button (col AE) transfers ellipse coordinates (2° spacing) to DESIGN's CIE Tolerance Box. While Excel's pass/fail algorithm is based on the ellipse equation, DESIGN's point-in-polygon algorithm is more general and works with any convex polygon. Tolerance box coordinates are comma-separated. DESIGN 2.61.4452 adds tab-separated and space-separated. Scroll down to find a FilmStar BASIC program which eliminates copy and paste via Sub SetCieAxes.

  DESIGN 2.61.4445 or newer required. 

A user had questions about Wedge Mode (massive layers). In reviewing the subject we realized that the concept might not be clear. Wedge Mode returns reflectance from the first (uncoated or coated) surface, while transmittance values at each surface are simply multiplied. Coherent multiple reflections within films are included, but multiple incoherent reflections between surfaces are not. This is verified in an Excel workbook illustrated below.

Side 2 Compensation settings in Setup...Parameters are now relabeled when there are massive layers in the design. The new settings are 1-FWD Parallel (1) and 6-FWD Wedge. Consider a bandpass filter (e.g. fnum.faw) deposited on both sides of a non-absorbing substrate. While similar Parallel blocking can be achieved with absorbing substrates, peak transmittance is reduced.

Consider a 180° flip fixture for applying optical coatings, typically AR, to both sides of a substrate. Optical transmission monitoring is employed. We simulate this in FilmStar DESIGN with a macro which effectively flips the optic. Download FlipMonitor.zip for a 4 layer AR example (FlipMonitor3.zip for a 3 layer AR). Copy .faw files to c:\Winfilm\Designs and run FlipMonitor.bas (FlipMonitor3.bas).

The original design BBAR-Flip.faw is shown below at the left. 'G' is the index symbol for the massive layer. Layer 6 is the first to be deposited. After depositing layers 6-9, the optic is flipped and layer 4 deposited. FlipMonitor.bas manipulates the design into the pseudo design shown on the right. The pseudo design is NOT used to calculate, but only to draw axes and plot results.

The design contains 8 thin layers and one massive layer. The pseudo design (BBAR-Flip-t.faw) indicates the 8 layers to be monitored and includes the same index symbols (L, H ) and film materials as the actual design. Thicknesses do not matter; for example "13H 13H 13H 13H 13H 13H 13H 13H" works with the proviso that the ability to combine adjacent layers IS SUPPRESSED. Be sure to uncheck Combine Layers in the Layers Mode Design Editor Edit menu. Failure to do so leads to hard-to-diagnose errors.

As shown below, the Side 2 design is reversed in the pseudo design. This is easily done in the Groups editor with Edit.. Reverse Segment <Ctrl+S>. While both AR designs are really the same, values were altered to distinguish Side 1 and Side 2 layers.

Actual Design: 97L 130H 38L 14H 2G 13H 38L 129H 96L
Pseudo Design: 13H 38L 129H 96L 14H 38L 130H 97L

 
             Actual design                                         Pseudo design for drawing axes

The monitoring curve at 550 nm is shown below. The fixture is flipped after layer 4


Crosses indicate extrema; green circles indicate %T at layer boundaries.

FlipMonitor.bas utilizes new features in DESIGN 2.61.4282: 1. BASIC Sub SetLayersIgnore and ability to save %R/%T vs. thickness as XLS/XLSX. 2.61.4283 (19 Nov) adds the ability to display cut-off values. Click View..Peaks in the FSPlot Module. Users experiencing difficulty obtaining the above graph should contact FTG Software to arrange a ZOOM meeting.

28 Sep 2021: A user reported an error when starting from Layer Number > 1. A restriction to Layer Number 1 was added in DESIGN 2.61.4656.

As shown here, DESIGN 2.61.4524 copies values to the clipboard in Excel format. After pasting, sort by column A (physical thickness in nm) to arrange in plot order. In column C, B indicates boundaries (green circles), P peaks and V valleys. Compare the values in column B with the above plot. A user in Texas requested that values be made available in BASIC. This is shown in the code snippet below. Note that View..Peaks must be checked in FSPlot. Sub Main     PlotActivate     PlotNext     s$ = Clipboard$()     PlotClose End Sub

Back to Technical Issues