added report
This commit is contained in:
Binary file not shown.
|
After Width: | Height: | Size: 3.4 MiB |
Binary file not shown.
|
After Width: | Height: | Size: 3.4 MiB |
Binary file not shown.
|
After Width: | Height: | Size: 3.6 MiB |
Binary file not shown.
|
After Width: | Height: | Size: 78 KiB |
Binary file not shown.
@@ -0,0 +1,326 @@
|
||||
\documentclass[11pt]{article}
|
||||
\usepackage[a4paper,margin=22mm]{geometry}
|
||||
\usepackage[T1]{fontenc}
|
||||
\usepackage{lmodern}
|
||||
\usepackage{microtype}
|
||||
\usepackage{graphicx}
|
||||
\usepackage{booktabs}
|
||||
\usepackage{array}
|
||||
\usepackage{tabularx}
|
||||
\usepackage{longtable}
|
||||
\usepackage{xcolor}
|
||||
\usepackage[hidelinks]{hyperref}
|
||||
\usepackage{caption}
|
||||
\captionsetup{font=small,labelfont=bf}
|
||||
\Urlmuskip=0mu plus 1mu
|
||||
\pagestyle{plain}
|
||||
\setlength{\parskip}{0.55em}
|
||||
\setlength{\parindent}{0pt}
|
||||
\newcommand{\um}{\ensuremath{\,\mu\mathrm{m}}}
|
||||
\newcommand{\degrC}{\ensuremath{\,^{\circ}\mathrm{C}}}
|
||||
\newcommand{\sourcefile}[1]{\texttt{\detokenize{#1}}}
|
||||
\definecolor{reportblue}{RGB}{26,74,115}
|
||||
|
||||
\begin{document}
|
||||
\hypersetup{pageanchor=false}
|
||||
\begin{titlepage}
|
||||
\centering
|
||||
\vspace*{28mm}
|
||||
{\color{reportblue}\rule{\textwidth}{1.2pt}}\\[12mm]
|
||||
{\Huge\bfseries OWIS Linear Stage\\Factory Acceptance Test Report\par}
|
||||
\vspace{7mm}
|
||||
{\Large Engineering Version\par}
|
||||
\vspace{12mm}
|
||||
{\color{reportblue}\rule{\textwidth}{1.2pt}}\\[20mm]
|
||||
\begin{tabular}{@{}ll@{}}
|
||||
Project: & \texttt{FAT\_OwisStageTREncoder}\\
|
||||
Prepared by: & Henrique Garcia\\
|
||||
Organization: & Paul Scherrer Institut, Mechatronic Systems\\
|
||||
Report date: & 7 September 2026\\
|
||||
Test dates represented: & 27 August--3 September 2026\\
|
||||
\end{tabular}
|
||||
\vfill
|
||||
\textbf{Document status: evidence-based engineering assessment}\\[3mm]
|
||||
No formal acceptance threshold was supplied. Results are ranked, but no pass/fail decision is made.
|
||||
\end{titlepage}
|
||||
|
||||
\hypersetup{pageanchor=true}
|
||||
\pagenumbering{roman}
|
||||
\tableofcontents
|
||||
\clearpage
|
||||
\pagenumbering{arabic}
|
||||
|
||||
\section{Executive Summary}
|
||||
Sixteen independent positioning datasets were reviewed together with the ECMC source folders, nine setup photographs, two ALMEMO temperature logs, and the available Oriental Motor PK-series manual. Fifteen datasets have supplied Keysight ISO 230-2:2014 (modified) result PDFs; the additional ED7062 dataset is available as a \sourcefile{.Lin} file and was evaluated with the same numerical definitions.
|
||||
|
||||
The best comparable full-travel test is:
|
||||
\begin{center}
|
||||
\renewcommand{\arraystretch}{1.25}
|
||||
\begin{tabular}{@{}ll@{}}
|
||||
\toprule
|
||||
Configuration & EL7062 DC, closed loop, S-curve\\
|
||||
Approach method & 0.2 mm preliminary point below each target\\
|
||||
Test travel & 1--90 mm, ten targets, five forward/reverse cycles\\
|
||||
Positioning accuracy, $A$ & \textbf{7.3211 $\mu$m}\\
|
||||
Repeatability, $R$ & \textbf{1.6524 $\mu$m}\\
|
||||
Maximum reversal error, $B$ & \textbf{1.1254 $\mu$m}\\
|
||||
\bottomrule
|
||||
\end{tabular}
|
||||
\end{center}
|
||||
|
||||
The controlled final approach improved the otherwise similar EL7062 closed-loop S-curve result: accuracy decreased from 12.2173 to 7.3211\,$\mu$m and repeatability decreased from 5.7420 to 1.6524\,$\mu$m. Lower values are better. This is strong evidence that approach direction and settling dominate a material part of the observed error.
|
||||
|
||||
A separate fine-step test from 10 to 11\,mm used 20\,$\mu$m target increments and returned $A=4.1456\um$. It is useful for local incremental behavior but must not be presented as 90\,mm full-travel accuracy.
|
||||
|
||||
The newly added ED7062 CSV closed-loop S-curve run with a 0.1\,mm preliminary approach returned $A=9.2284\um$, $R=3.6663\um$, and maximum $B=0.8462\um$. It is competitive but does not replace the EL7062 result above as the best full-travel dataset.
|
||||
|
||||
\section{Scope, Sources, and Limitations}
|
||||
Only the following source areas were used:
|
||||
\begin{itemize}
|
||||
\item \sourcefile{Measurements}: 18 \sourcefile{.Lin} files, 15 one-page result PDFs, and two temperature CSV files. Two \sourcefile{.Lin} files are duplicate-labelled copies, leaving 16 independent positioning datasets.
|
||||
\item \sourcefile{ECMC}: EL7041, EL7062 CSP, EL7062 CSP-PC, EL7062 CSV, and ED7062 CSV configurations.
|
||||
\item \sourcefile{Media}: nine setup photographs dated 3 and 7 September 2026.
|
||||
\item \sourcefile{Motor}: \sourcefile{HM-7471E.pdf}, a generic Oriental Motor PK-series operating manual.
|
||||
\end{itemize}
|
||||
|
||||
The report does not assume that every configuration folder was loaded during every historical measurement. File labels are used to associate a result with a configuration. No live IOC, EtherCAT bus, drive, encoder voltage, limit input, controller tuning, or physical motion was tested while preparing this document.
|
||||
|
||||
The result PDFs state ISO 230-2:2014 (Modified), algebraic sign convention, and a coverage factor of 2. No acceptance criterion, interferometer calibration certificate, alignment worksheet, uncertainty budget, or signed test checklist was present in the requested source folders. Consequently, the report does not declare pass/fail or a complete measurement uncertainty.
|
||||
|
||||
\section{Physical Setup}
|
||||
The photographs show an OWIS linear stage mounted on an optical table. A stepper motor drives a screw through a coupling. A TR-Electronic linear encoder is mounted along the stage, and a laser interferometer measures carriage displacement along the motion direction. The moving retroreflector is supported by an orange fixture on the carriage. Air/material sensing and cable restraint are visible.
|
||||
|
||||
The separate ALMEMO logger uses three temperature measurements. ALMEMO channel 0 is a PT100 attached to the stepper motor. Channel 01 is a second PT100 mounted on the stage, directly above the ball bearing at the point where the spindle begins. The remaining sensor measures room temperature. This placement distinguishes motor heating, temperature at the spindle-support region of the stage, and ambient room conditions.
|
||||
|
||||
\begin{figure}[htbp]
|
||||
\centering
|
||||
\includegraphics[width=0.96\textwidth]{../Media/20260903_163042.jpg}
|
||||
\caption{Overall FAT setup with laser interferometer on the left and OWIS stage on the right.}
|
||||
\end{figure}
|
||||
|
||||
\begin{figure}[htbp]
|
||||
\centering
|
||||
\includegraphics[width=0.96\textwidth]{../Media/20260907_110051.jpg}
|
||||
\caption{Stage, carriage, encoder, motor coupling, and interferometer arrangement during the later setup.}
|
||||
\end{figure}
|
||||
|
||||
The images establish component placement, not metrological alignment. A future signed setup sheet should record beam parallelism, dead path, Abbe offset, reflector fixing, sensor positions, cable forces, stage mounting torque, and warm-up time.
|
||||
|
||||
\section{Motor and Mechanical Drive}
|
||||
The ECMC files describe an OWIS two-phase stepper arrangement with 1\,mm travel per revolution, 400 full steps per revolution, 0.8\,A maximum current, 0.5\,A standby current, 48\,V nominal supply, 7.5\,$\Omega$ phase resistance, and 3080\,$\mu$H phase inductance.
|
||||
|
||||
The supplied HM-7471E document is an Oriental Motor PK-series operating manual. It provides general installation and safety information but not the exact installed motor's electrical data. Relevant manual requirements include:
|
||||
\begin{itemize}
|
||||
\item qualified electrical/mechanical personnel;
|
||||
\item a guarded rotating shaft/coupling and an external emergency-stop circuit;
|
||||
\item use of a specified compatible motor/driver combination;
|
||||
\item secure mounting to a flat, thermally conductive plate;
|
||||
\item motor-case surface temperature not exceeding 100$\degrC$;
|
||||
\item inspection for loose fasteners, bearing noise, stressed wiring, loose connectors, and shaft misalignment.
|
||||
\end{itemize}
|
||||
|
||||
The exact motor model/nameplate is not legible or stated in the reviewed source set. Also, the ECMC source contains a resistance inconsistency: 7500\,m$\Omega$ in the motor component and 2471\,m$\Omega$ in a controller component. This must be resolved by identifying and measuring the installed motor before final tuning.
|
||||
|
||||
\section{ECMC and EtherCAT Configuration}
|
||||
\subsection{Common architecture}
|
||||
The reviewed configurations use ECMC 11.0.8 in engineering mode at a 500\,Hz EtherCAT rate. Axis 1 is exposed as \texttt{M1}. The primary feedback is a TR-Electronic LT140-S absolute SSI linear encoder through EL5042 slave 2, channel 1.
|
||||
|
||||
The source settings for the SSI channel are:
|
||||
\begin{itemize}
|
||||
\item 24 position bits, binary coding, no multiturn or offset bits;
|
||||
\item 1\,MHz clock;
|
||||
\item scale $1/10000$ mm/count = 0.1\,$\mu$m/count;
|
||||
\item absolute offset $-11.826$\,mm;
|
||||
\item 9.0\,V encoder supply requested through EL5042 configuration;
|
||||
\item counter wraparound disabled.
|
||||
\end{itemize}
|
||||
The 9.0\,V setting is not a voltage measurement. Confirm the live SDO and connector voltage with the encoder disconnected, and verify it against the exact encoder order code before connection.
|
||||
|
||||
\subsection{Compared drive variants}
|
||||
\begin{tabularx}{\textwidth}{@{}p{28mm}X@{}}
|
||||
\toprule
|
||||
Variant & Static source description\\
|
||||
\midrule
|
||||
EL7041 & Stepper terminal with a secondary 16-bit open-loop counter and SSI as primary feedback.\\
|
||||
EL7062 CSP & CSP position setpoint on \texttt{positionSetpoint01}; internal 32-bit position counter at 1,048,576 counts per 1 mm revolution.\\
|
||||
EL7062 CSP-PC & Inner EL7062 position loop plus outer ECMC position loop closed on the SSI encoder.\\
|
||||
EL7062 CSV & ECMC closes position on SSI and commands \texttt{velocitySetpoint01}; source warns of a firmware-related counter jump on disable and recommends CSP unless fixed firmware is confirmed.\\
|
||||
ED7062 CSV & Separate drive variant with local tuning values; no EL7062 firmware behavior is assumed for ED7062.\\
|
||||
\bottomrule
|
||||
\end{tabularx}
|
||||
|
||||
In CSV, the configured conversion is $n_{raw}=v\,2^{31}/500$ for $v$ in mm/s. In CSP, the drive position scale is $1/1{,}048{,}576$ mm/count for the stated 1 mm/revolution mechanics.
|
||||
|
||||
\subsection{Motion sequence}
|
||||
The later automatic FAT sequence uses ten targets from 1 to 90\,mm, five forward/backward cycles, 2\,mm/s motion velocity, and 3\,mm/s$^2$ acceleration/deceleration. The gap sequence first approaches a point 0.2\,mm below each target, dwells for approximately 1\,s, then moves to the measurement target and dwells for approximately 6\,s. This enforces a common final approach direction.
|
||||
|
||||
Software limits are $-1$ and 95\,mm. The reviewed EL7062 CSP configuration uses 0.05\,mm following-error tolerance, 0.0015\,mm at-target tolerance, and hardware-limit inputs on the drive's channel-1 digital inputs. EL2008 channel 4 supplies the limit switches.
|
||||
|
||||
The PLC documentation has an execution-handshake weakness: busy-low plus at-target can be accepted before busy was ever observed high. For unattended motion, keep execute asserted while active, latch busy-high, and declare completion only after busy returns low while at-target remains true.
|
||||
|
||||
\section{Measurement Method}
|
||||
Most full-travel files use ten targets, five bidirectional cycles, 120\,mm/min feed rate, 3\,s measurement dwell, and a 4\,s trigger dwell. Earlier trigger windows were 0.2 or 0.3\,mm; later gap tests use 0.025\,mm. Target patterns also changed. These differences matter when comparing results.
|
||||
|
||||
The Keysight sheets report:
|
||||
\begin{itemize}
|
||||
\item $A$: bidirectional positioning accuracy;
|
||||
\item $E$: systematic position deviation;
|
||||
\item $M$: mean bidirectional position deviation;
|
||||
\item $R$: bidirectional repeatability;
|
||||
\item max $B$: maximum reversal error.
|
||||
\end{itemize}
|
||||
Values below are copied from the supplied PDFs and converted from millimetres to micrometres. The ED7062 row was calculated from its 100 stored linear deviations: five forward and five reverse runs at ten targets. The calculation uses the same two-sigma ISO 230-2 (modified) definitions as the supplied result sheets and was cross-checked by reproducing the published metrics of an existing EL7062 result from its \sourcefile{.Lin} data.
|
||||
|
||||
\section{Results}
|
||||
\small
|
||||
\begin{longtable}{@{}p{48mm}p{35mm}rrrr@{}}
|
||||
\caption{Reported positioning results. Lower values are better.}\label{tab:basicresults}\\
|
||||
\toprule
|
||||
Drive/control & Profile/scope & $A$ & $E$ & $R$ & max $B$\\
|
||||
& & \multicolumn{4}{c}{$\mu$m}\\
|
||||
\midrule
|
||||
\endfirsthead
|
||||
\toprule
|
||||
Drive/control & Profile/scope & $A$ & $E$ & $R$ & max $B$\\
|
||||
& & \multicolumn{4}{c}{$\mu$m}\\
|
||||
\midrule
|
||||
\endhead
|
||||
EL7041 closed & S-curve, 89.312 mm & 12.9155 & 9.9679 & 6.0430 & 1.5967\\
|
||||
EL7041 closed & Trapezoidal, 90 mm & 7.5989 & 6.0386 & 2.6739 & 1.5143\\
|
||||
EL7041 open & Trapezoidal, 90 mm & 24.8280 & 20.8842 & 8.4932 & 5.6624\\
|
||||
EL7041 DC closed & Trapezoidal, 90 mm & 7.7985 & 6.3297 & 2.2667 & 1.1797\\
|
||||
EL7041 DC open & Trapezoidal, 90 mm & 13.9195 & 10.2204 & 10.7227 & 7.5308\\
|
||||
EL7041 DC closed & S-curve, 90 mm & 11.9568 & 9.0849 & 4.9041 & 2.7315\\
|
||||
EL7062 closed & S-curve, 90 mm & 10.4298 & 8.9517 & 4.1907 & 2.5880\\
|
||||
EL7062 open & S-curve, 90 mm & 22.7101 & 18.6696 & 8.1442 & 4.2550\\
|
||||
EL7062 open & Trapezoidal, 90 mm & 21.4864 & 18.8733 & 5.6532 & 3.9325\\
|
||||
EL7062 DC open & Trapezoidal, 90 mm & 8.5891 & 6.1477 & 6.1303 & 3.7104\\
|
||||
EL7062 DC closed & Trapezoidal, 90 mm & 12.2126 & 9.8056 & 4.9802 & 2.7991\\
|
||||
EL7062 DC closed & S-curve, 90 mm & 12.2173 & 9.8419 & 5.7420 & 3.0251\\
|
||||
\textbf{EL7062 DC closed} & \textbf{S-curve + 0.2 mm approach, 90 mm} & \textbf{7.3211} & 6.9837 & \textbf{1.6524} & \textbf{1.1254}\\
|
||||
EL7062 CSV DC closed & S-curve + gap, 90 mm & 9.4819 & 6.9473 & 2.8791 & 0.7820\\
|
||||
ED7062 CSV DC closed & S-curve + 0.1 mm approach, 90 mm & 9.2284 & 7.0082 & 3.6663 & 0.8462\\
|
||||
EL7062 DC closed & S-curve, 20 $\mu$m steps, 10--11 mm & 4.1456 & 1.3173 & 4.1456 & 1.1760\\
|
||||
\bottomrule
|
||||
\end{longtable}
|
||||
\normalsize
|
||||
|
||||
Two duplicate-label cases are excluded from Table~\ref{tab:basicresults}. The EL7062 CSV trapezoidal-gap file duplicates the corresponding S-curve dataset. The two newly added ED7062 files labelled CSP and CSV also contain identical measurement and environment arrays; both have the same 7 September 2026 10:02:27 test timestamp, and both internally describe \texttt{ED7062\_CSV}. They are therefore treated as one ED7062 CSV run, not separate CSP and CSV evidence.
|
||||
|
||||
\begin{figure}[htbp]
|
||||
\centering
|
||||
\includegraphics[width=0.92\textwidth,page=1]{../Measurements/OWIS_stageTR_Electronics_FAT_ISO230_Stepper_EL7062_EL5042_DC_ClosedLoop_SCurve_gap.pdf}
|
||||
\caption{Best supplied full-travel result: EL7062 closed-loop S-curve with a 0.2 mm controlled final approach.}
|
||||
\end{figure}
|
||||
|
||||
\section{Engineering Interpretation}
|
||||
\subsection{Effect of closed-loop control}
|
||||
Closed-loop configurations generally reduce the combined positioning error relative to their open-loop counterparts. For the non-DC EL7041 trapezoidal pair, accuracy improves from 24.8280 to 7.5989\,$\mu$m. For the corresponding EL7062 S-curve pair, it improves from 22.7101 to 10.4298\,$\mu$m.
|
||||
|
||||
The EL7062 DC open-loop trapezoidal run has relatively good $A=8.5891\um$, but repeatability and reversal error are worse than the best gap-controlled closed-loop test. It should not be selected on accuracy alone.
|
||||
|
||||
\subsection{Effect of controlled final approach}
|
||||
The clearest improvement is produced by the 0.2\,mm preliminary approach. Relative to the EL7062 DC closed-loop S-curve run without the documented gap:
|
||||
\begin{itemize}
|
||||
\item accuracy improves by 4.8962\,$\mu$m, or 40.1\%;
|
||||
\item repeatability improves by 4.0896\,$\mu$m, or 71.2\%;
|
||||
\item maximum reversal error improves by 1.8997\,$\mu$m, or 62.8\%.
|
||||
\end{itemize}
|
||||
This supports retaining a unidirectional final approach in the FAT sequence.
|
||||
|
||||
\subsection{CSV versus CSP risk}
|
||||
The CSV gap result is competitive, but the EL7062 CSV source contains a firmware warning about open-loop counter discontinuity during disable/re-enable. The measured table alone does not resolve that safety/continuity risk. Use CSP as the default unless the installed EL7062 firmware is identified and a reduced-speed enable/disable continuity test passes.
|
||||
|
||||
\subsection{New ED7062 result}
|
||||
The ED7062 CSV result improves accuracy slightly relative to the earlier EL7062 CSV gap result ($9.2284\um$ versus $9.4819\um$), while repeatability is worse ($3.6663\um$ versus $2.8791\um$). Maximum reversal error is similar ($0.8462\um$ versus $0.7820\um$). The ED7062 result used a stated 0.1\,mm preliminary approach rather than the 0.2\,mm approach of the best EL7062 CSP-oriented sequence. Because drive type, approach gap, thermal state, and test time differ, this is not a controlled terminal-only comparison.
|
||||
|
||||
\section{Environmental Data}
|
||||
The best 90\,mm result sheet reports:
|
||||
\begin{center}
|
||||
\begin{tabular}{@{}lccc@{}}
|
||||
\toprule
|
||||
Quantity & Minimum & Maximum & Mean\\
|
||||
\midrule
|
||||
Air temperature ($^\circ$C) & 22.38 & 22.41 & 22.40\\
|
||||
Pressure (mmHg) & 736.1 & 736.2 & 736.2\\
|
||||
Relative humidity (\%) & 42 & 43 & 43\\
|
||||
Material sensor 1 ($^\circ$C) & 22.56 & 22.59 & 22.58\\
|
||||
Material sensor 2 ($^\circ$C) & 22.77 & 23.07 & 22.91\\
|
||||
\bottomrule
|
||||
\end{tabular}
|
||||
\end{center}
|
||||
|
||||
Material sensor 2 rose by 0.30$\degrC$ during the best EL7062 positioning run. This is relevant because thermal expansion can create micrometre-scale change over the 90\,mm travel depending on the represented material and sensor contact.
|
||||
|
||||
The expanded ALMEMO file \path{Measurements/Temperature/FromRest_almemo_20260907_083344.csv} contains 1,206 samples from 09:40:32.918 to 10:20:42.719, a stored interval of 40.16\,min. Its column names now directly match the supplied sensor map: \texttt{ch0} is the motor PT100, \texttt{ch1} is the stage PT100 above the spindle-start bearing, and \texttt{ch3} is room temperature.
|
||||
|
||||
\begin{center}
|
||||
\small
|
||||
\begin{tabular}{@{}llrrrrr@{}}
|
||||
\toprule
|
||||
Record & Sensor & Start & Minimum & Maximum & End & Mean\\
|
||||
\midrule
|
||||
ED7062 CSV & Motor, ch. 0 & 22.538 & 22.529 & 30.741 & 24.374 & 27.135\\
|
||||
& Stage, ch. 01 & 22.585 & 22.583 & 24.807 & 23.651 & 23.919\\
|
||||
& Room, ch. 3 & 23.208 & 22.551 & 23.853 & 23.034 & 23.103\\
|
||||
\midrule
|
||||
ED7062 CSP & Motor, ch. 0 & 23.333 & 23.318 & 23.968 & 23.760 & 23.736\\
|
||||
& Stage, ch. 01 & 23.201 & 23.188 & 23.264 & 23.264 & 23.222\\
|
||||
& Room, ch. 3 & 23.164 & 22.935 & 23.276 & 23.179 & 23.143\\
|
||||
\bottomrule
|
||||
\end{tabular}
|
||||
\normalsize
|
||||
\end{center}
|
||||
|
||||
The ED7062 CSV positioning file is timestamped 10:02:27, which falls within the longer logger record. The nearest stored sample gives 30.721$\degrC$ at the motor, 24.732$\degrC$ at the stage, and 23.520$\degrC$ in the room. The motor and stage reached maxima of 30.741$\degrC$ and 24.807$\degrC$ approximately 22.1 and 23.3 minutes after the first stored sample, respectively. Thus the positioning measurement occurred close to peak motor temperature and while the stage was still warm; it was not a room-temperature equilibrium condition.
|
||||
|
||||
The second file, \path{Measurements/Temperature/FromRest_ED7062_CSP_almemo_20260907_083344.csv}, contains 116 samples over 3.83\,min. During it the motor rose from 23.333$\degrC$ to 23.760$\degrC$ and peaked at 23.968$\degrC$; the stage changed from 23.201$\degrC$ to 23.264$\degrC$. No independent CSP positioning dataset with a matching 10:47--10:51 timestamp is present, so this log is treated as a thermal-only record.
|
||||
|
||||
\begin{figure}[htbp]
|
||||
\centering
|
||||
\includegraphics[width=0.98\textwidth]{ALMEMO_Temperature.png}
|
||||
\caption{ALMEMO PT100 records. Top: the 40.16-minute record containing the ED7062 CSV positioning timestamp. Bottom: the later 3.83-minute ED7062 CSP thermal-only record.}
|
||||
\end{figure}
|
||||
|
||||
The logger data establish substantial transient heating, but they do not by themselves quantify thermal positioning error. PT100 tolerance, logger-channel uncertainty, attachment/contact quality, sensor time constants, and the temperature distribution between the bearing, spindle, frame, encoder, and interferometer reference remain uncharacterized.
|
||||
|
||||
\section{Risks and Open Items}
|
||||
\begin{enumerate}
|
||||
\item \textbf{No acceptance limit:} numerical ranking is possible, but formal FAT disposition is not.
|
||||
\item \textbf{Motor identity and resistance:} record the nameplate and resolve the 7.5\,$\Omega$ versus 2.471\,$\Omega$ source inconsistency before tuning.
|
||||
\item \textbf{Encoder supply:} the 9.0\,V configuration is not proof of terminal voltage or compatibility. Verify live SDO and connector voltage with the encoder disconnected.
|
||||
\item \textbf{PLC handshake:} require busy-high before accepting busy-low plus at-target as completion.
|
||||
\item \textbf{Limits and signs:} test hardware limits, raw inputs, software limits, motor direction, and SSI direction at reduced speed.
|
||||
\item \textbf{CSV firmware:} identify firmware and test counter continuity before using EL7062 CSV.
|
||||
\item \textbf{Thermal stability:} the longer logger record captures warm-up and cooldown, but the ED7062 positioning run occurred close to peak motor temperature. Repeat the positioning test after a defined equilibrium criterion is met and retain synchronized temperature data.
|
||||
\item \textbf{Comparability:} repeat the preferred configuration with unchanged targets, dwell, trigger window, approach, mounting, and environment.
|
||||
\item \textbf{Uncertainty:} add calibration certificates, alignment/dead-path checks, Abbe offset, expansion model, and mounting-repeatability data.
|
||||
\end{enumerate}
|
||||
|
||||
\section{Recommended FAT Configuration and Next Test}
|
||||
Use the EL7062 CSP closed-loop configuration with the SSI encoder as primary feedback and the 0.2\,mm positive-direction final approach. Before the next automatic run:
|
||||
\begin{enumerate}
|
||||
\item confirm the active IOC path and exact configuration loaded;
|
||||
\item verify motor wiring, nameplate, resistance, inductance, current settings, and drive tuning;
|
||||
\item verify disconnected EL5042 supply voltage and SSI status;
|
||||
\item compare motor-counter and SSI direction/scale using a very small move;
|
||||
\item exercise both limits and stop behavior at low speed;
|
||||
\item validate the execute/busy/at-target sequence;
|
||||
\item stabilize thermally and acquire a complete environmental log;
|
||||
\item perform at least three complete 1--90\,mm campaigns with identical settings;
|
||||
\item define the acceptance criterion before signing the FAT.
|
||||
\end{enumerate}
|
||||
|
||||
\section{Conclusion}
|
||||
The available evidence shows that the OWIS stage can achieve a best reported 90\,mm bidirectional accuracy of 7.3211\,$\mu$m and repeatability of 1.6524\,$\mu$m when closed-loop control is combined with a controlled one-direction final approach. This is the preferred configuration for confirmation testing. Formal acceptance remains open because no requirement or complete uncertainty budget was supplied, and the static ECMC safety/tuning issues require live verification.
|
||||
|
||||
\subsection*{Source Register}
|
||||
\scriptsize
|
||||
\noindent\textbf{Position data:} \sourcefile{Measurements/*.Lin} and corresponding Keysight PDFs.\\
|
||||
\textbf{Temperature data:} both CSV files in \sourcefile{Measurements/Temperature}.\\
|
||||
\textbf{ECMC sources:} \sourcefile{EL7041}, \sourcefile{EL7062_CSP}, \sourcefile{EL7062_CSP_PC}, \sourcefile{EL7062_CSV}, and \sourcefile{ED7062_CSV}.\\
|
||||
\textbf{Photographs:} \sourcefile{Media/*.jpg}.\\
|
||||
\textbf{Motor manual:} \sourcefile{Motor/HM-7471E.pdf}, Oriental Motor PK-series operating manual HM-7471-4, December 2025.
|
||||
\normalsize
|
||||
|
||||
\end{document}
|
||||
Binary file not shown.
@@ -0,0 +1,133 @@
|
||||
\documentclass[10pt,twocolumn]{article}
|
||||
\usepackage[a4paper,top=18mm,bottom=20mm,left=16mm,right=16mm,columnsep=6mm]{geometry}
|
||||
\usepackage[T1]{fontenc}
|
||||
\usepackage{lmodern}
|
||||
\usepackage{microtype}
|
||||
\usepackage{graphicx}
|
||||
\usepackage{booktabs}
|
||||
\usepackage{array}
|
||||
\usepackage{tabularx}
|
||||
\usepackage{xcolor}
|
||||
\usepackage[hidelinks]{hyperref}
|
||||
\usepackage{caption}
|
||||
\captionsetup{font=small,labelfont=bf}
|
||||
\Urlmuskip=0mu plus 1mu
|
||||
\setlength{\parindent}{1em}
|
||||
\setlength{\parskip}{0pt}
|
||||
\newcommand{\um}{\ensuremath{\,\mu\mathrm{m}}}
|
||||
\newcommand{\mm}{\ensuremath{\,\mathrm{mm}}}
|
||||
\newcommand{\degrC}{\ensuremath{\,^{\circ}\mathrm{C}}}
|
||||
\newcommand{\sourcefile}[1]{\texttt{\detokenize{#1}}}
|
||||
|
||||
\title{\vspace{-8mm}\textbf{Positioning Performance of an OWIS Linear Stage Under ECMC Control}\\
|
||||
\large Factory Acceptance Test Report in IEEE-Style Two-Column Format}
|
||||
\author{Henrique Garcia\\Paul Scherrer Institut, Mechatronic Systems\\\texttt{FAT\_OwisStageTREncoder}}
|
||||
\date{7 September 2026}
|
||||
|
||||
\begin{document}
|
||||
\maketitle
|
||||
|
||||
\begin{abstract}
|
||||
This report consolidates 15 ISO 230-2:2014 (modified) linear-positioning result sheets for an OWIS stepper-driven linear stage, the associated ECMC configurations, setup photographs, a short temperature acquisition, and the available Oriental Motor PK-series operating manual. Tests compare EL7041 and EL7062 drive terminals, open- and closed-loop control, trapezoidal and S-curve trajectories, direct-current (DC) distributed-clock configurations, and controlled positive-direction approach gaps. Across the comparable 90\,mm tests, the lowest reported bidirectional positioning accuracy was $A=7.3211\um$ for EL7062 closed-loop S-curve control with a 0.2\,mm preliminary approach. Its repeatability was $R=1.6524\um$ and maximum reversal error was $B=1.1254\um$. A separate 10--11\,mm fine-step test returned $A=4.1456\um$, but it is not a full-travel result. No acceptance limit was supplied, so the report ranks configurations without declaring a formal pass or fail.
|
||||
\end{abstract}
|
||||
|
||||
\textbf{Keywords---}ECMC, EPICS, EtherCAT, EL7041, EL7062, EL5042, ISO 230-2, laser interferometry, positioning accuracy, repeatability, stepper motor.
|
||||
|
||||
\section{Scope and Evidence Boundary}
|
||||
The analysis uses only files in \sourcefile{Measurements}, \sourcefile{ECMC}, \sourcefile{Media}, and \sourcefile{Motor}. The numerical values in Table~\ref{tab:results} are transcribed from the one-page Keysight result PDFs, whose numerical analysis identifies ``ISO 230-2 2014 (Modified)'', algebraic error sign, and coverage factor 2. Configuration statements are static source findings; they do not prove which IOC configuration was active for every historic test. Photographs document the physical arrangement but do not establish alignment uncertainty or calibration status. The motor manual is a generic Oriental Motor PK-series operating manual and does not identify the exact installed motor model.
|
||||
|
||||
\section{Test Article and Measurement Arrangement}
|
||||
The photographs show an OWIS screw-driven linear stage fixed to an optical table, a stepper motor coupled to the screw, a carriage-mounted linear encoder head, and a laser interferometer aligned approximately parallel to the motion axis. The moving retroreflector is attached to the carriage through an orange fixture. Air and material-temperature sensors are visible around the apparatus. Figure~\ref{fig:setup} shows the representative arrangement.
|
||||
|
||||
\begin{figure*}[t]
|
||||
\centering
|
||||
\includegraphics[width=0.49\textwidth]{../Media/20260903_163042.jpg}\hfill
|
||||
\includegraphics[width=0.49\textwidth]{../Media/20260907_110051.jpg}
|
||||
\caption{FAT arrangement. Left: interferometer, stage, encoder, and restraints. Right: later close view of the stage, motor coupling, carriage, and interferometer alignment. The photographs are configuration evidence only; no dimensional inference is made from image scale.}
|
||||
\label{fig:setup}
|
||||
\end{figure*}
|
||||
|
||||
The measurement files generally use ten targets over 1--90\,mm, five forward/reverse cycles, 120\,mm/min feed rate, and 3--4\,s dwell. Later ``gap'' sequences use targets 1, 10.89, 20.7775, 30.6675, 40.555, 50.445, 60.3325, 70.2225, 80.11, and 90\,mm with a 25\,$\mu$m trigger window. A preliminary target below each measurement point forces the final approach in the positive direction. One fine-increment test covers only 10--11\,mm at 20\,$\mu$m target increments with a 6\,$\mu$m trigger window.
|
||||
|
||||
\section{ECMC Architecture}
|
||||
The reviewed configurations use ECMC/EPICS with EtherCAT at 500\,Hz and axis 1 (EPICS axis name \texttt{M1}). The TR-Electronic LT140-S absolute SSI encoder is connected through EL5042 slave 2, channel 1, and is the primary position feedback. Source configuration specifies 24-bit binary SSI, 1\,MHz clock, 0.1\,$\mu$m/count scale, and an absolute offset of $-11.826$\,mm. It also requests a 9.0\,V EL5042 encoder supply. This voltage and compatibility with the exact encoder order code require physical verification before connection.
|
||||
|
||||
The stage model in the ECMC files assumes 1\,mm of linear travel per motor revolution, 400 full steps/revolution, 0.8\,A maximum current, 0.5\,A standby current, 48\,V nominal supply, 7.5\,$\Omega$/phase, and 3080\,$\mu$H. The generic PK-series manual confirms applicable handling constraints such as qualified installation, external emergency stopping, guarding rotating parts, secure motor/driver combinations, and a maximum motor-case surface temperature of 100$\degrC$. It does not independently confirm the local winding values or exact motor type.
|
||||
|
||||
\subsection{Drive variants}
|
||||
EL7041 configurations use a secondary 16-bit open-loop counter. EL7062 CSP configurations use \texttt{positionSetpoint01} and an internal 32-bit position counter scaled at $2^{20}=1{,}048{,}576$ counts per 1\,mm revolution. The CSP-PC folder describes an inner EL7062 position loop and an outer ECMC position loop closed on the SSI encoder. CSV variants instead map \texttt{velocitySetpoint01}; the configured velocity conversion is
|
||||
\begin{equation}
|
||||
n_{\mathrm{raw}}=v\frac{2^{31}}{500},
|
||||
\end{equation}
|
||||
where $v$ is in mm/s for the stated 1\,mm/revolution mechanics and 500\,Hz cycle rate. The source README warns that an EL7062 CSV firmware defect can cause its open-loop counter to jump to a whole-turn boundary on disable. CSP is therefore the stated preferred operating mode unless the installed firmware is confirmed fixed.
|
||||
|
||||
\subsection{Motion sequence and safeguards}
|
||||
The later FAT PLC files command five forward/backward cycles at 2\,mm/s and 3\,mm/s$^2$. The 0.2\,mm approach-gap sequence dwells nominally 1\,s at the preliminary point and 6\,s at the measurement point. Software limits are $-1$ and 95\,mm. Hardware limit inputs use channel-1 digital inputs, with an EL2008 output supplying the switches. The source sets a 0.05\,mm following-error tolerance and 0.0015\,mm at-target tolerance in the reviewed EL7062 CSP configuration.
|
||||
|
||||
Two configuration risks remain. First, the motor component declares 7500\,m$\Omega$ while a controller component declares 2471\,m$\Omega$; the installed motor resistance and tuning basis must be resolved. Second, the active PLC descriptions accept busy-low plus at-target as completion without first proving busy became high. For unattended execution, execute should remain asserted during motion and completion should require a busy-high latch followed by busy-low and at-target.
|
||||
|
||||
\section{Positioning Results}
|
||||
Table~\ref{tab:results} reports the values printed by the supplied Keysight result sheets. All values are converted from millimetres to micrometres. ``DC'' and control/profile labels reproduce file metadata and are not independently reconstructed from runtime logs.
|
||||
|
||||
\begin{table*}[t]
|
||||
\centering
|
||||
\caption{ISO 230-2:2014 (Modified) Numerical Results}
|
||||
\label{tab:results}
|
||||
\scriptsize
|
||||
\begin{tabular}{@{}llrrrrr@{}}
|
||||
\toprule
|
||||
Drive / configuration & Profile and scope & $A$ ($\mu$m) & $E$ ($\mu$m) & $M$ ($\mu$m) & $R$ ($\mu$m) & max $B$ ($\mu$m)\\
|
||||
\midrule
|
||||
EL7041, closed loop & S-curve, 89.312 mm & 12.9155 & 9.9679 & 9.3136 & 6.0430 & 1.5967\\
|
||||
EL7041, closed loop & Trapezoidal, 90 mm & 7.5989 & 6.0386 & 5.9269 & 2.6739 & 1.5143\\
|
||||
EL7041, open loop & Trapezoidal, 90 mm & 24.8280 & 20.8842 & 19.0523 & 8.4932 & 5.6624\\
|
||||
EL7041 DC, closed loop & Trapezoidal, 90 mm & 7.7985 & 6.3297 & 6.1324 & 2.2667 & 1.1797\\
|
||||
EL7041 DC, open loop & Trapezoidal, 90 mm & 13.9195 & 10.2204 & 4.3243 & 10.7227 & 7.5308\\
|
||||
EL7041 DC, closed loop & S-curve, 90 mm & 11.9568 & 9.0849 & 7.9591 & 4.9041 & 2.7315\\
|
||||
EL7062, closed loop & S-curve, 90 mm & 10.4298 & 8.9517 & 6.9815 & 4.1907 & 2.5880\\
|
||||
EL7062, open loop & S-curve, 90 mm & 22.7101 & 18.6696 & 17.5019 & 8.1442 & 4.2550\\
|
||||
EL7062, open loop & Trapezoidal, 90 mm & 21.4864 & 18.8733 & 17.9853 & 5.6532 & 3.9325\\
|
||||
EL7062 DC, open loop & Trapezoidal, 90 mm & 8.5891 & 6.1477 & 4.1465 & 6.1303 & 3.7104\\
|
||||
EL7062 DC, closed loop & Trapezoidal, 90 mm & 12.2126 & 9.8056 & 8.2630 & 4.9802 & 2.7991\\
|
||||
EL7062 DC, closed loop & S-curve, 90 mm & 12.2173 & 9.8419 & 8.2988 & 5.7420 & 3.0251\\
|
||||
EL7062 DC, closed loop & S-curve + 0.2 mm approach, 90 mm & \textbf{7.3211} & 6.9837 & 6.7060 & \textbf{1.6524} & \textbf{1.1254}\\
|
||||
EL7062 CSV DC, closed loop & S-curve + gap, 90 mm & 9.4819 & 6.9473 & 6.7269 & 2.8791 & 0.7820\\
|
||||
EL7062 DC, closed loop & S-curve, 20 $\mu$m steps, 10--11 mm & 4.1456 & 1.3173 & 0.9411 & 4.1456 & 1.1760\\
|
||||
\bottomrule
|
||||
\end{tabular}
|
||||
\end{table*}
|
||||
|
||||
\subsection{Comparative interpretation}
|
||||
The strongest full-travel result is the EL7062 DC closed-loop S-curve run with a 0.2\,mm preliminary approach. Compared with the otherwise similar EL7062 DC closed-loop S-curve run without a documented approach gap, $A$ improves from 12.2173 to 7.3211\,$\mu$m (40.1\% lower), and $R$ improves from 5.7420 to 1.6524\,$\mu$m (71.2\% lower). This indicates that controlling final approach direction materially suppresses reversal/settling variability in this dataset.
|
||||
|
||||
For EL7041, closed-loop trapezoidal control also performs well: $A=7.5989\um$ without the DC label and $7.7985\um$ with it. Open-loop cases generally show larger error or poorer repeatability. The exception is the EL7062 DC open-loop trapezoidal result at $A=8.5891\um$, but its $R=6.1303\um$ and max $B=3.7104\um$ remain worse than the best gap-controlled closed-loop run. Accuracy alone therefore does not capture robustness.
|
||||
|
||||
The CSV gap run has lower systematic deviation than the best CSP-labelled run ($E=6.9473\um$ versus $6.9837\um$) and lower max reversal error ($0.7820\um$ versus $1.1254\um$), but worse overall accuracy and repeatability. It is not recommended solely on those two metrics because the configuration itself records a firmware-risk caveat. The file named \sourcefile{...ClosedLoop_Trapezoidal_gap.Lin} is data-identical to the CSV S-curve file except for filename/comment metadata and has no separate result PDF; it is excluded as a duplicate, not treated as an additional experiment.
|
||||
|
||||
\begin{figure*}[t]
|
||||
\centering
|
||||
\includegraphics[width=0.48\textwidth,page=1]{../Measurements/OWIS_stageTR_Electronics_FAT_ISO230_Stepper_EL7062_EL5042_DC_ClosedLoop_SCurve_gap.pdf}\hfill
|
||||
\includegraphics[width=0.48\textwidth,page=1]{../Measurements/OWIS_stageTR_Electronics_FAT_ISO230_Stepper_EL7062_EL5042_DC_ClosedLoop_SCurve_SmallSteps.pdf}
|
||||
\caption{Original Keysight analysis sheets. Left: best comparable 90\,mm result with 0.2\,mm positive-direction final approach. Right: 20\,$\mu$m increment test over 10--11\,mm. The latter characterizes local incremental behavior only.}
|
||||
\label{fig:keyresults}
|
||||
\end{figure*}
|
||||
|
||||
\section{Environmental Observations}
|
||||
The best 90\,mm sheet reports air temperature 22.38--22.41$\degrC$, pressure 736.1--736.2\,mmHg, relative humidity 42--43\%, material sensor 1 at 22.56--22.59$\degrC$, and material sensor 2 at 22.77--23.07$\degrC$. The 0.30$\degrC$ rise of the second material sensor during that run suggests thermal drift was present. The supplied 7 September CSV contains only 19 samples over 36.887\,s: channel 1 is 22.301--22.303$\degrC$ (mean 22.3021$\degrC$), channel 2 is 22.422--22.425$\degrC$ (mean 22.4233$\degrC$), and channels 6 and 7 contain no valid samples. This record is too short to characterize motor warm-up or long-term stage equilibrium.
|
||||
|
||||
Environmental conditions vary among runs, and target patterns and trigger windows also changed. The ranking is therefore a comparative engineering screening, not a controlled one-factor experiment. A qualified uncertainty statement would additionally require traceable interferometer calibration data, alignment/cosine and dead-path assessments, material expansion treatment, sensor placement and calibration, mounting-repeatability studies, and repeated runs under stabilized conditions.
|
||||
|
||||
\section{Conclusions and Recommendations}
|
||||
The evidence favors closed-loop operation with a controlled one-direction final approach. The best full-travel dataset is the EL7062 DC closed-loop S-curve gap run, at $A=7.3211\um$, $R=1.6524\um$, and max $B=1.1254\um$. No requirement value was supplied, so these figures cannot be labeled pass or fail.
|
||||
|
||||
For the next FAT campaign: (1) use the EL7062 CSP configuration rather than CSV unless the installed firmware is confirmed unaffected; (2) resolve the 7.5\,$\Omega$/2.471\,$\Omega$ source mismatch and document the motor nameplate; (3) verify the EL5042 voltage at the disconnected connector and compatibility with the exact SSI encoder; (4) implement and runtime-test the execute/busy/at-target handshake; (5) verify both hardware limits and direction signs at low speed; (6) stabilize the apparatus thermally and log all relevant channels for the full test; and (7) repeat at least three complete 90\,mm campaigns using identical targets, dwell, trigger window, and environment before freezing compensation values.
|
||||
|
||||
\begin{thebibliography}{9}
|
||||
\bibitem{keysight} Keysight E1733A result files, \sourcefile{Measurements/*.Lin} and corresponding PDFs, 27 August--3 September 2026.
|
||||
\bibitem{ecmc} ECMC source configurations and local READMEs, \sourcefile{ECMC/EL7041}, \sourcefile{ECMC/EL7062_CSP}, \sourcefile{ECMC/EL7062_CSP_PC}, \sourcefile{ECMC/EL7062_CSV}, and \sourcefile{ECMC/ED7062_CSV}, reviewed 7 September 2026.
|
||||
\bibitem{photos} FAT setup photographs, \sourcefile{Media/*.jpg}, 3 and 7 September 2026.
|
||||
\bibitem{motor} Oriental Motor Co., Ltd., ``2-Phase/5-Phase Stepping Motors PK Series Operating Manual,'' HM-7471-4, published December 2025, file \sourcefile{Motor/HM-7471E.pdf}.
|
||||
\bibitem{temp} Temperature acquisition, \path{Measurements/Temperature/FromRest_almemo_20260907_083344.csv}, 7 September 2026.
|
||||
\end{thebibliography}
|
||||
|
||||
\end{document}
|
||||
Reference in New Issue
Block a user