Daresbury test list of issues

[sebprince]

Software

• Window too large and doesn't fit inside screen on my laptop
• Tension table and graph cut out on my laptop
• Accommodate python2.7 used for winder software on same computer (probably specifically call python3 only for DWA and use 2.7 by default system-wide)
• Have previous APA UUIDs in config page and tension page as a drop down menu maybe (possibly by reading them from the APA folder names)
• Resonance tab’s table too narrow to fit long filenames, make it wrap possibly (or remove some columns like UUID, or even submitter?)
• Scan list table width is too narrow (e.g. layer V, head board 4), possibly make wire segments wrap around in cell?
• Some visual indication when submitting resonances to DB to know when it's done
• Slow to upload resonances to DB (could be resolved by next item)
• Multithreading (1) database uploads and downloads, (2) end-of-scan relay deactivation and (3) scan abort, such as not to lock up the GUI. If multithreading DB is an issue, could reduce number of uploads and downloads by doing more steps locally.
• Database overwriting fix from James might not be sensible so Chris should check
• Use resonance structure to inform peak finding, and not use it for associating resonances to tensions, at least for X and G
• Cannot add red vertical resonance lines in raw A(f) subtab
• Aborting a run colors whichever scan is selected in the scan list
• Possibly the radio button changes to next one relative to current position selected not to current run (which could be different if clicked another scan in the table during the completion of the current one)
• Clicking abort scan before scan properly started keeps red button and no way to recover from that state
• Write in some file whether scan is submitted, so can be loaded in table afterwards instead of the question mark
• Step size does not work in table if there's a letter hexadecimal value
• Some manual config values at the bottom of the advanced tab don't output expected value (probably issue related to hexadecimal values)
• Swap side and layer in config page such as to have the same order as the folder names, or could swap the order in the folder names since there are fewer sides than layers...
• Shut off 50V 500 V led in GUI when heartbeat stops being received to show unconnectedness
• Make heartbeat more visible: something else than slightly changing shade (pink <-> red)
• Disable abort button if not receiving heartbeat (after having received it and being connected, can start scan)
• Record if run was aborted or finished successfully
• Resonance table shouldn't show aborted runs
• Not looking at A(f) at all during run make them never show up after it's completed
• Scan 9 of layer V head board 1 measures 4 wires but there are 5 plots with x-axis with fixed frequency range
• A(f) plots that have no wires should not have data displayed as it's confusing (maybe some text label that says no wire instead)
• Complete button to stop scan before end when sufficient data has been acquired (maybe quite a bit of work and limited usefulness with fast scans)
• Tension graph x-axis for X goes to 1000 while it should be 480
• Implement APA geometry plot in config tab
• Show channel number local to head board in the APA geometry plot
• Have a way to launch all configured scans one after the other automatically
• If all scans are automatically started one after the other, show in APA geometry plot the different scans by different colors, and show current scan number out of the total number to be done in sidebar
• Idea: Don’t really look at data in A(f) plots, just look at resonances tab and collect data automatically, so main focus becomes resonance tab
• Typing the Tab key on keyboard to move to next field skips fields here and there, e.g. "Measured by" -> "APA UUID"
• In A(f) plots, no way to return zoom level to default when x-axis range matches start/end frequencies
• Remove scan# from scan list table since not used anywhere?
• Could have a gradient background band that disappears when out of tolerance to indicate expected resonances frequency range
• For G and X layers, calculate tension based only on lowest-frequency resonance?
• Ability to upload a user message/note with resonances/tensions?
• In resonance table, go to more recent scan in table automatically after submission
• Record the button status in scan data somehow (not sure if would be useful when not in development). Before we were recording heartbeat but stopped, but could record deltas of button status or if error bits are active?
• In resonance table, load a selected scan just by clicking on the road, i.e. without having to click “load selected scan”
• For U and V, need to revisit if we really need to separate resonance measurement and tension measurement, since we only do subset of them (no short wires)
• Two empty rows at bottom of scan table if no recent scan loaded, which happens under some yet-to-be-understood situations (maybe don't load any recent scans and have an empty table?)
• For tension table, do not use bright yellow for cell background as can’t read white numbers on that color
• For tension table, round tension values to single digit past decimal point (or maybe 2)
• For A(f), reorder plots in increasing order of wires (which changes per layer), instead of DWA channel?
• Put layer and side of APA channels in A(f) plot titles
• Loading tensions from different layers also loads previous layers into next one

DB

• DB APA tension shows data 0.5 off along x-axis (appears in-between integers so hard to know what wire number is associated to data)
• “wireSegments” seems to affect tension table display on sietch https://dev.sietch.xyz/test/6124b5995d88df11eddb854a
• Currently impossible to edit pointer table from web interface to go back to previous resonance value (but would anybody really do that, seems simpler and faster to redo measurement)?
• Can’t easily find tension of a particular wire, say XA112, from the tension table, other than counting the values
• Creating a tension frame manually shouldn't be needed

Hardware

• HV DC creates noise
• Choosing right holes for bottom screws of probe boards: have some silkscreen indicators maybe
• Reduce G probe board height by 4mm
• Power switch not turning off
• Flaky -5V LED
• Make sure component screws into boxes are tightened during assembly (power cable one was loose)
• Pem nuts popped out: small one for thick X by too much torque on top screws, large one for thin X by too much torque on flex cable screws. Should have torque limits similar to CR board at least for top screws. Still a concern for bottom screws and X head board's pem nut
• Are holes in probe boards for the flex cable aligned between X and G probe boards?
• Shipping of probe board needs to be better since the connectors pierced the foam and were touching each other: maybe use bags, and then wrapped in foam?
• Flex cable might be too asymmetric length: hard to fold (both up for storage and down to the probe board)
• Need to mill on temporary carrier board (they can be adjusted vertically by 4 mm, usually 2mm done) 4 mm depth
• Bent pins G probe boards due to bad flex cable insertion somehow...
• Retention system for bias cables as it comes out

Specific to Daresbury lab

• UDP data not received by winder computer: open windows firewall port (allow for it in public network profile)
• Double clicking dwaDaq.py does not work so need to create “python dwaDaq.py” .bat file
• No way to tell when to click connect in the software with ethernet cable connected to PLC since PLC takes a long time (~1 minute?) to assign IP to DWA
• Screw in corner on DWA box close to PDB and flex cable is stripped

[chrisstanford]

James ​- Window too large and doesn't fit inside screen on my laptop ​- Have previous APA UUIDs in config page and tension page as a drop down menu maybe (possibly by reading them from the APA folder names) ​- Resonance tab’s table too narrow to fit long filenames, make it wrap possibly (or remove some columns like UUID, or even submitter?)​ ​- In resonance table, go to more recent scan in table automatically after submission​ ​ ​- ​In resonance table, load a selected scan just by clicking on the ro​w​, i.e. without having to click “load selected scan” ​​ ​- ​Two empty rows at bottom of scan table if no recent scan loaded, which happens under some yet-to-be-understood situations (maybe don't load any recent scans and have an empty table?)​ ​- ​(2) end-of-scan relay deactivation and (3) scan abort, such as not to lock up the GUI. If multithreading DB is an issue, could reduce number of uploads and downloads by doing more steps locally. ​- ​Use resonance structure to inform peak finding, and not use it for associating resonances to tensions, at least for X and G ​- ​Cannot add red vertical resonance lines in raw A(f) subtab ​- ​Aborting a run colors whichever scan is selected in the scan list​​​​​​ ​- ​Scan list table width is too narrow (e.g. layer V, head board 4), possibly make wire segments wrap around in cell?​​​ ​- ​Remove scan# from scan list table since not used anywhere?​​​ ​- ​Have a way to launch all configured scans one after the other automatically​ ​- ​Possibly the radio button changes to next one relative to current position selected not to current run (which could be different if clicked another scan in the table during the completion of the current one) ​- ​Clicking abort scan before scan properly started keeps red button and no way to recover from that state​.​​ ​Disable abort button if not receiving heartbeat (after having received it and being connected, can start scan)​ ​- ​Shut off 50V 500 V led in GUI when heartbeat stops being received to show unconnectedness ​- ​Make heartbeat more visible: something else than slightly changing shade (pink <-> red)​ ​- ​Record if run was aborted or finished successfully​ - record in JSON if run was aborted​ ​- ​Resonance table shouldn't show aborted runs ​- ​Not looking at A(f) at all during run make them never show up after it's completed Cannot reproduce ​- ​A(f) plots that have no wires should not have data displayed as it's confusing (maybe some text label that says no wire instead) (also suppressed V(t) plots) ​- ​Implement APA geometry plot in config tab ​- ​Show channel number local to head board in the APA geometry plot​ ​- ​If all scans are automatically started one after the other, show in APA geometry plot the different scans by different colors, and show current scan number out of the total number to be done in sidebar ​- ​Could have a gradient background band that disappears when out of tolerance to indicate expected resonances frequency range ​- ​Put layer and side of APA channels in A(f) plot titles ​- ​In A(f) plots, no way to return zoom level to default when x-axis range matches start/end frequencies (override "A" button on plot)

Chris ​- ​Tension​ table and graph cut out on my laptop (unable to reproduce) ​- ​Some visual indication when submitting resonances to DB to know when it's done ​- ​Slow to upload resonances to DB (could be resolved by next item) ​- ​Multithreading (1) database uploads and downloads,

​- ​Database overwriting fix from James might not be sensible so Chris should check ​- ​Write in some file whether scan is submitted, so can be loaded in table afterwards instead of the question mark ​- Any user-defined run parameter should be in physics units, with label. ​Table should display decimal value (physics units). Step size does not work in table if there's a letter hexadecimal value ​- ​Some manual config values at the bottom of the advanced tab don't output expected value (probably issue related to hexadecimal values) ​- ​Swap side and layer in config page such as to have the same order as the folder names, or could swap the order in the folder names since there are fewer sides than layers... ​- ​Tension graph x-axis for X goes to 1000 while it should be 480 ​- ​Typing the Tab key on keyboard to move to next field skips fields here and there, e.g. "Measured by" -> "APA UUID" ​- ​For G and X layers, calculate tension based only on lowest-frequency resonance? ​- ​Ability to upload a user message/note with resonances/tensions?​ ​- ​For tension table, do not use bright yellow for cell background as can’t read white numbers on that color ​- ​For tension table, round tension values to single digit past decimal point (or maybe 2) ​- ​For A(f), reorder plots in increasing order of wires (which changes per layer), instead of DWA channel? ​- ​Loading tensions from different layers also loads previous layers into next one

Sebastien ​- ​Record the button status in scan data somehow (not sure if would be useful when not in development). Before we were recording heartbeat but stopped, but could record deltas of button status or if error bits are active?​ ​- ​For U and V, need to revisit if we really need to separate resonance measurement and tension measurement, since we only do subset of them (no short wires)​ ​- ​Scan 9 of layer V head board 1 measures 4 wires but there are 5 plots with x-axis with fixed frequency range ​- ​Complete button to stop scan before end when sufficient data has been acquired (maybe quite a bit of work and limited usefulness with fast scans) ​- ​Idea: Don’t really look at data in A(f) plots, just look at resonances tab and collect data automatically, so main focus becomes resonance tab ​- Accommodate python2.7 used for winder software on same computer (probably specifically call python3 only for DWA and use 2.7 by default system-wide)