Section Half Multisensor Logger ( SHMSL )

X402

• 22 Mar 2024 Ocean optics Bulb has been changed

X401

• Scientists were concerned about the MS point probe not working after observing discrepancies in magnetic susceptibility peaks present in the WRMSL data, and not in the SHMSL data for several sections. The sections targeted to verify the MS probe were archive and working halves of Core U1611A-56R-1. Both 56R-1A and 56R-1W were measured with the SHMSL and the WRMSL and compared with the WRMSL data measured on the whole-round. The magnetic susceptibility peaks were not seen on the SHMSL, but present on the WRMSL.

Section U1611A-56R-1 A measured with the SHMSL and the WRMSL. Note that the section shows magnetic susceptibility peaks only when measured with the WRMSL. Data are compared with the WRND measurement on the WRMSL.

The WRMSL and SHMSL measure different volume and have different sensitivity:

• WRMSL: The magnetic susceptibility sensor is a loop. It measures all around the whole-round over a certain interval (a few cm)
• SHMSL: The sensor is a point-probe and measures around the sensor (a couple of cm) and what is on the surface (does not reach what is deep within the sediment).

If some material has a contrasting magnetic susceptibility from the sediment matrix, the WRMSL is likely able to detect it because it measures everything.
The SHMSL would detect this material either because it is strong enough or close enough to be detected by the probe. For instance, if it is too deeply buried in the sediment (i.e., not at the surface) and not strong enough, the probe would not see it because the signal of this material would be hidden by the sediment matrix signature. If the magnetic susceptibility of this material is relatively higher, it would likely be seen by the probe (even if it is not at the surface).

In order to support this statement, we secured Section U1611A-56R-1W and measured it flipped over in addition to the normal way of measuring sections.

Sediment of the working half secured with paper towel and tightly shrink wrapped

Working half flipped over and measured with the SHMSL

The plot below shows the magnetic susceptibility for the different measurements carried out on the working half and compared with the results of the whole-round core.

Section U1611A-56R-1 W measured with the SHMSL and the WRMSL. Note that the section shows magnetic susceptibility peaks when measured flipped over with the SHMSL. Data are compared with the WRND measurement on the WRMSL.

• Scientists reported to ALO that the Y-axis of the SHMSL stopped moving at the end of a measurement. No details were provided to the ALO on the circumstances. Restarting IMS reinitialized the communication with the Y-axis.
• Bulb changed on 12/17/2023 before arrival on Site U1609.

X400T

2023 Dec 10. Ship-wide power out: 1030-1130 hr utc+1. Restart of SHMSL. AR700 laser defaulted back to 9600 baud while the Windows host retained the 230400 baud rate setting.

• AR700. Re-applied the 230400 baud rate.
  • P(arameter) and S(etting) lights are to be as shown to match computer-side baud rate 230400. To observe this setting, press the Function key 7 times--quick 1/4 second presses.
    
    After ~10 seconds the indicators will turn off.
  • Method to modify the setting.
    The method is outlined in the AR700 manual (pp 40-43).
    • Starting with all indicator lights off.
    • 7 quick presses of the Function key.
    • Press the Function key for ~2 sec. The S(ettings) light bank will blink.
    • 9 quick presses of the Function key will choose the 230400 baud rate setting (corner light by 10 lit).
    • One long press (~2 sec) of the Function key to cause the setting to be stored. Blinks to indicate acknowledgement.
    • After 10 seconds the indicator lights turn off.
• Used the NI-MAX > AR700 > VISA Test Panel > I/O tab to send the command "W1234" to the Laser device to save its current configuration to firmware (preserved during short power outages).

2023 Nov 28. SHMSL first power-up. Following 2 OS update cyces and 3 shipwide power outages during dry-dock. All devices lost their LabVIEW identifiers. Reapplied as indicated by (current) SHMSL MS, Laser, and QEPro Motion Communication and Control Setup.

• Matched cable serial numbers. Applied names. Note that serial number to name is a constant association. The COM# change over extended power outs and some types of OS updates.
  • cable s/n ...DCE1E0 = AR700 now COM5
  • cable s/n ...DCF414 = MS3 now COM6
  • cable s/n ...DCDFDC = X-AXIS now COM4
  • cable s/n ...DCE00F = Y-AXIS now COM3
• QE-Pro
  • USB hub-connector at QEPro body showed two amber lights. NI-Max showed QE-PRO blink in, then out again every few seconds. Unplugged USB hub-connector. 5 seconds. Replugged. Green lights for both.
  • USB device with no name in NI-MAX shows like this.
    
    Renamed to QEPro-2 in NI-MAX per current docs.
  • Task definition lost also. Reconstructed per above doc; section NI-DAQmx Task.
    
    
• For normal operation (computer control of lights), the toggle switches should be as indicated
  • Toggled toward ON
    
    
  • The Power switch is toggled to ON.
    The Shutter switch is toggled to TTL.
    
    

X400

• The white standard for the QEPro spectrophotometer needed to be cleaned several times during the expedition because it got dirty very frequently. The use of sandpaper was not very efficient because it created grooves on the surface. Therefore, we used a soft polisher to make the surface clean and flat.
• New bulb installed by Core Description Technician on September 7. New integration time: 0.24

X395

• New bulb installed on Aug 8th very close to the end of expedition. (calibration value was 0.2)
  
• after IODP software update, MUT does not upload some files  and copy that to the error folder but the files has no problem, we move it to the in folder again and MUT upload with no issues

X399

• New IMS software (version 14) was released during X398P, and was tested again on X399 after a few changes. Sample information window required debugging. When the user enters section information with the Manual Tab, the information disappears if the user changes tab. This was fixed.
  
• After installing the Windows Updates and restarting the computer, when IMS is opened for the first time, there will be a message regarding a loss of communication for the Y-axis. Close the window and IMS, and restart the computer. This will clear the error. When opening IMS again, there won't be any error message and the SHMSL will work properly. The reason for this issue is unknown.
• Changes in the calibration file to report the correct value of the dark pixel for the white calibration. The Dark Pixels for white and dark calibrations should be similar.
  

• Positions excluded during a measurement was not reported correctly in the data file (value=false, whereas it was true) - this issue was corrected. In addition, a new indicator on the sample information window was added to inform the user that intervals are excluded from the analysis.

• Bulb change on April 23, 2023 (old integration time=0.5, new integration time after bulb change: 0.2)
• For information/reminder, when testing the SHMSL, it is necessary to put a play core to test and check that the length is correctly reported. An empty track will give wrong length information.

X398P

• IMS software updated and tested

X398

• Some times during the first part of the expedition, at the end of the measurements the color reflectance sphere dragged the MS point standard, the lid was opened in one occasion.

A new standard with softer borders and with the lid on the bottom has been done by the ETs. It was installed on January 8th, 2023 at 17:01 UTC time.

Another spare standard has been done and kept in drawer PPTRKF10. The old standard is also in that drawer.

Software Suggestions

• In C:/ > DATA > Archive three folders are generated (MS point, Profile and RSC) but no data were saved on them because they are saved in the AUX_DATA folder. If possible, it will be useful no remove them.
• After the laser profile is done a window pops up indicating the length of the core, this length could be accepted or the measurement could be canceled. When the measurement is cancelled a window pops up asking if you want to Return or if you want to Don't Move, in both cases the measurement of the core has to be restarted in order to measured it. Scientists suggested to add to this window also the option of 'continue with the measurement' in case the cancel button was selected by mistake. 

X397

• bulb has replaced because it does not turned on.
• Changed out the Ocean Optics Bulb on Nov 18, 2022.

X397T

• There were no major issues with the SHMSL.
• At one point the RSC sensor suspension system came loose on one side when the orange rubber inserts slipped out of place. Repaired and good to go.
• Early in the expedition, the cap came off the magnetic susceptibility point sensor calibration disk because the adhesive wore out and the RSC sensor can clip the edge of the disc depending on the measurement interval. It was resealed with Permatex silicone RTV.

X397P

• The raw data output was not actually the raw data.  We found an issue in the QEPro data writer vi that had the normalized data being output to the raw file.  The issue was corrected so the raw file now contains the raw data.
• Work is on going to get the white and dark calibration data uploaded.  During Kick the Tires, Light the fires testing it was noted these have been missing from our uploads
• We ran the SHMSL over an empty liner to reset the laser limit. After this we did not have issues with the laser measuring beyond the sections as was reported from 393.
• When SHMSL computer is restarted, the first time IMS is launched, the Y-axis does not respond properly.  If you relaunch the IMS software, the Y-axis will work.
• After IMS SHMSL was installed on the PC, we noticed the lights did not turn on.  After tracing the wires, we also noticed the system was not getting power because the CRIO was not closing the relays.  We checked the NI Max set up more closely and noted an issue with NI Max Task for the QE Pro lights.  We removed the task and did a fresh set up.  Afterwards the lights functioned normally.  See the NI Max set up documentation in the physical properties notebook.

X393

• Bulb was changed on June 13th, July 9th and 28th.
• On bulb installed on June 13th the integration time increased fast to 0.440, after that it was stable until July 9th, when the light source was changed.
• IMS is not reporting properly the length of some of the sections. This happens when the section is close to one of the brackets (aprox, 5 cm).
  • Possible cause is the filter applied at the end of the profile. The profile measured by the laser has the correct length before the filter is applied. The length reported is the length at the beginning of the bracket, (CHECK)
  • Change the laser speed from 5 to 2.5 cm/s, solve the problem but not for the case of sections shorter than aprox. 45 cm.
  • Changing the setup of the laser from Quality to Rate didn't make any change in the behavior.
• RS 232 light source connector cover was changed in order to allow to attach the screws. It was found not completely plugged ones, the halogen light was off.
  
• Offset problem was detected in the Color Reflectance data. It was an intermittent problem, related with the correct operation of the halogen light source shutter.

By looking at the Dark Calibration was also possible to see that the shutter was not working properly because all the information was concentrated under higher wavelength values. Light source sn 034990124 was changed and sn 034990296 installed. Measurements on several sections were repeated.

• Halogen Light source sn 034990124 was checked, it had a dry joint. It was welded again. Test it before install it in the SHMSL.

• On hole U1560B in the last 20 cm of the a section measurement. The Y Axis stopped to identify the limit switches and it kicked the physical stopper on the top. Limit switches and connections were checked and Y-Axis board was cleaned from dust. Next measurement was fine. This issue happened two more times. Laser profile was review in order to see a spike or something that could cause the issue but nothing was found. Origin of the problem unknown. 
• Several times during the measurement of hard rocks the color reflectance sensor dragged pieces of rock, the origin of the problem is that the sensors started to move in the X-Axis before they finish the movement on the Y-Axis.

The safety clearance was increased to 5 cm and speed and acceleration of X and Y axis were modified in order to allow the Y axis to go higher before the X-axis started to move. Also scientist were asked not to set hard rock pieces higher than the core liner. The problem was minimized but not solve.

The origin of the problem could be in IMS or the setup parameters.

X390

• No issues reported.

X392

• Bulb was changed on February 8th and March 17th.
• No issues reported.

X391

• 10 Dec 2021  Light bulb was changed in halogen source , the integration time was 0.77 and with the new bulb is 0.22

X396T

• Light bulb was not changed.
• New ruler was installed in the track.
• Spectrometers inventory was completed with the other instruments/items installed on the track.
• After upgrade to LabVIEW 21, blue light will not turn OFF causing IMS to crash each time that black calibration was performed. Trying to solve it using NI MAX, or resetting power didn't work. IMS code was improved to better accommodate the orders to lights and shutter.
• Each time that SHMSL computer restarts IMS will not recognize the Y-Axis, to solve this problem, user should restart IMS. 

X396

• Lightbulb changed in the RF52052 light source on SHMSL, but the calibration time worsened (.590 from .320).  We noted the light was not being emitted from this light source.  After changing the bulb again the integration time remained the same. When the TTL/ Shutter switch was flipped to TTL the light source made many clicking noises as if the shutter was not functioning properly.  We swapped light sources and calibration time immediately was lowered.  Adjusted shutter brass knob on Sept. 6 and put back on the SHMSL.  Testing indicates that RF52052 is working with good calibrations and no issues.  Left RF52052 on SHMSL and Put RF52053 back in the drawer.
• Experienced some issues with the SHIL laser not recognizing the end of the core material, specifically with hard rock sections with empty liner below the core material.  This did slow the measurements down, but scientists could enter the correct length to avoid measuring the empty spaces.

X395C

• 12 Jun 2021 Light bulb was changed.
• On final measurements of U155E4 the SHMSL starts to make squeak while doing the up-down movement. 28 Jun 2021 gas strut was replaced, but the noise continue. 29 Jun 2021 Jurie Kotze and Zenon Mateo serviced and grease the Y-Axys Exlar actuator. Squeak stopped.
• Measurement interval on U1554 and U1555 was 2 cm to match the SRM measurements. Measurement interval on U1562 and 1563 was 2.5 cm for APC and XCB and 2 com for RCB.
• QE Pro 91127 and 90900 were sent to IODP. QE Pro 90900 will be sent again to the ship after recalibration.
• QE Pro 90899 will be installed on the track at the end of the expedition in order to replace QE Pro 90900 that will be send to shore for recalibration. They will be tested to assure that measurements are taken correctly.
• Few times during measurements, after the doing the laser profile the instrument started automatically the measurement of the section, without waiting for the user to verify its length. During last hole the touch of the monitor will be turn off to check if there is a relation between the touch screen and this issue. This behavior has also been observed in the NGR (couple times this expedition).
• Every time Windows updates are made in the station, IMS stops recognizing the Y-axys. This problem can be solved by restarting IMS.
• At the end of the expedition was noticed that Y-axis Exlar actuator continue to make noise. The actuator should be checked before start to measure samples on next expedition.

X395E

Drivers updated

No Issues

• The X and Y axes movements were adjusted for both sensors to ensure proper positions and simultaneous compression during both measuring and calibration motion.
• The "white" calibration puck was resurfaced and sanded smooth with D.I. water and 220 grit sandpaper on a glass surface, rinsed clean with D.I. water, and dried thoroughly before replacing.
• The SHMSL laser was not registering the ‘end of section’ properly and would scan to the end of the track. This was corrected by updating the IMS code. The AR700 Single Distance.vi had unwired cases for the error codes from the laser (Not seen, too far, etc.).  The system continued to use the default value, which was an “in-range” value, allowing the laser to scan to the end of the track.  By wiring these cases with the error codes, the track correctly measures and recognizes the end of the section.

X395P

No Issues

X393R

Not used

Spare parts arrived: 3 m of high-temperature silicon o-ring cord stock for the SHMSL sensors (PN 96505K24). Stored at SHMSL drawer PPTRKF3.

X390C

• 30 Nov 2020 New calibration data sheets and linearity tests of QEpro 90899 and 91126 provided by the vendor were added to the QEPro Inventory page.
• 28 Nov 2020 Two tests were performed in order to compare the measurements between the QE Pro that it is already installed on the SHMSL (90900) and the two QE Pro that were sent to the vendor (90899, 91126). They all behave similarly. 
  
  • Test 1: Perform the QEPro calibration, continuously, seven times. On QE Pro 90899 and 91126, the calibration was performed again after 1 hour of work. Dark calibration values were collected. Data is saved on T:\IODP_Share\PhysProps\Testing and Calibration Notes\SHMSL\QEPro_Tests\SHMSL_X390C_QEProCalibrationTest. A summary of the results is on the rights part of this page under "QE Pro (90900, 90899, 91126) Calibrations Comparison Data X390C (Nov 2020)".

• • Test 2: Perform seven continuous Color Standards measurements per each QE Pro. Data is saved on T:\IODP_Share\PhysProps\Testing and Calibration Notes\SHMSL\QEPro_Tests\SHMSL_X390C_QEProColorStandardTest. A summary of the results is on the rights part of this page under "QE Pro (90900, 90899, 91126) Color Standards Comparison Data X390C (Nov 2020)".

• 25 Oct 2020 QSG and instrument checklist were reviewed and updated in Confluence.
• 16 Oct 2020 Halogen bulb replaced.
• 12 Oct 2020 On X384 it was reported that the laser profile keeps measuring far past the end of the section, been most prominent on short sections and core catchers where the profile records the length to be 125 cm and shows a sloping end to the section.

After performing several tests we realized that this issue was divided into three different situations:

• • The laser profile keeps measuring far past the end of the section: Once the laser detects that core liner is not present it will continue for another ± 10 cm to make sure that it is the end of the section. If the laser is set to the incorrect height could see the black supports and continue the measurement.
  • The profile shows a sloping end to the section: This feature is due to a process apply to the raw data when the laser finish to take the profile. The processing applies edge corrections at the top and bottom of the section, values there are smoothed. This also implies that gaps at top and bottom of the section are removed. Another question is that the laser can also reject a measurement for quality reasons (ambient light sources, rough surface, etc.), if the rejection happens over several cm the plot will look like a ramp. For solving that it is possible to measure at a slower speed.
  • The profile records a 125 cm length for short sections and core catchers: 125 cm is the position of the end of the last bracket of the track. If the laser is taking this measurement means that it is measuring bellow the range that it is suppose to measure, and is taking the bracket as part of the section. To solve this problem it is necessary to move the laser up, in this sense we are looking for a range where we can see the core liner but not the bracket.

The measurement range of the laser is, starting from the laser, from 78.2 mm to 179.9 mm. Greater accuracy is at 129 mm.

At the beginning of the expedition the distance from the laser to the benchmark was ± 118.5 mm, at that distance the laser could see the track and because of that it could confuse it with the core. After lifting the laser to 134 mm from the benchmark it starts to measure correctly the end of the section. (Note: Now the benchmark distance measured on the Y-Axys Setup is 55.53 mm, that it is close to the default value of 56.99 mm reported on the User Guide).

X390P II

Tech report notes:

• Connected the two QE Pro devices returned from the vendor Ocean Optics after repairs.
  • Both units calibrated properly.

• Suggested tests for 390C: Run white standard calibrations with each unit to ensure consistent measurements.

X384/390P I

• 31 Jul 2020 Reflectance measurement comparisons being carried out between the QEPro mounted on the SHMSL and the handheld Minolta. 

Tech report notes: Problems were initially reported in getting the instrument setup. Issues were resolved and discussed in the new Minolta User Guide. Software CDs and all needed cables to setup the Minolta are now kept in the cases. 

Testing: A study was performed comparing reflectance data between the Minolta and QEPro. The spectral standards and a subset of sections were measured on both spectrometers and then compared to known values of the standards. The data was sent to David Houpt on shore for analysis. Results from Tristimulus X,Y,Z and CIELAB L*, a*, b* confirm that both spectrometers closely, though not perfectly, align with known values. The Minolta performs slightly better for some colors with the QEPro performing more accurately for other colors, in all cases the differences are small. Below is a graphical overview comparing L*,a*,b* of the Minolta, QEPro, and the color standard values provided. The results of David Houpt's analysis is found in the Physical Properties Lab Notebook and as a link here: Minolta vs QE Pro Diffuse Standards.

• 31 Jul 2020 Control Set offsets were edited in IMS, all were off by approximately 1 cm. Will look into adding the black spectral standard to the control set.
• MS Point Study was conducted to compare the standard deviations and standard errors at varying integration times for a lower magnetic susceptibility core (concrete core) and for a higher magnetic susceptibility core (core from X384). The results are located in the side panel Instrument Resources > In-House Tests . Based on the results we can measure cores with high or low susceptibility at 1000ms (optimal integration time) with an observed standard error of 0.05 and observed standard deviation of 0.55. Results can be found in IODP_Share\Physprops\Testing and Calibration Notes\SHMSL\MS Integration Time tests.

Tech report notes: MS-point/MS3 integration time testing was conducted to experiment with how different integration times affect the quality of the measurement. For these tests two core sections were used, one with very low susceptibility and the other with a higher more variable susceptibility. For each section we made measurements with four different integration times: 100ms, 500ms, 1000ms, and 3000ms. Each section was run 10 times at each time setting. In order to compare each time setting tested we can use the mean standard deviation to calculate the standard error, STDERR = STDEV/SQRT(N). For one run at 100ms the STDERR = STDEV, thus for 10 runs at 100ms we can calculate the STDERR = STDEV/SQRT(10). Plotted below is the predicted vs. observed STDERR as a function of measurements(N) taken for each section. As expected the measurement becomes more certain as the integration time increases, however the difference between standard errors at higher integration times is less than expected.  The standard deviation increases systematically for measurement times greater than 1000ms.  While it may still be beneficial to use the standard integration time of 3000ms for certain measurements, 1000ms is the new recommended measurement integration time. 

Excel files used to collect this data and create these plots can be found in: IODP_Share\PhysProps\Testing and Calibration Notes\SHMSL\MS Integration Time tests and as a link here: Concrete SHLF MS Point Study Data and Play Core SHLF MS Point Study Data.

• The laser profile keeps measuring far past the end of the section. This is most prominent on short sections and core catchers where the profile records the length to be ~125 cm and shows a sloping end to the section. The MS and QEPro measurements still start where expected. PROBLEM SOLVED DURING X390C.

• Air cylinder showed signs of deteriorating and was replaced with the same XXXX . Currently looking into if we should be using the 50lb or 25 lb cylinders.

Tech report notes: The gas strut started showing signs of deteriorating and was replaced with the 50lb gas strut. The 50lb gas strut has been added to CPP Inventory under PP5006. All the gas struts are labeled in the Track Hardware Drawer.

• 30 Jul 2020 MS2K Probe kept sitting unevenly on the core. The probe holder was removed and the orange rubber pieces were adjusted until the unit looked even and tied off with the black nylon thread, and the the red stretchy bands on the top of the unit were also replaced. The unit is sitting more evenly now. There is a consistent problem of the cage tilting down to the left, possibly owing to the nylon screws being unable to screw into the white rod and blue plastic and thus unable to support the weight of the unit.
• 24 Jul 2020 Halogen Bulb replaced.

X387P/387T

• MS3 meter was installed, replacing MS2 meter. Appropriate IMS code changes were made.
• Laser code was adjusted to perform measurements that are more accurate.
• This station is measuring the MSP drift of the data before starting the laser profile and at the end of the measurements. Air measurement is taken for this purpose. Drift corrections are not applied to the raw data, they were only applied on the auxiliary files. Further explanation about this topic will be found on the manuals.
• New AR700 laser clamp was installed on the station. IMS software code and instruments offset were modified to adapt to this change.

• CCW, CW and Home switches have been moved to adapt to the changes made on the Geometry of the SHMSL.
• Zenon Mateo designed a frame for the MS3 meters.

• White MSP standard with a hole for the sensor was placed on the station. It permits less variation in the standard measurements than the flat one.
• Acrylic standards base was fixed to the rails to allow the MSP sensor to descend in the correct point every time.

Technical Service

• Bulb was change March 18 and April 10. Bulb install March 18 shows more erratic measurements than the install on April 10, making us suspect that it was defective.
• New power supply box was installed, in order to take out most of the connectors.
• M-Drive board was moved and all the instruments were connected to UPSs. During the QEPro testing, it was discovered that the ship current affect the measurements. See PP Lab Notebook report for more information.

Testing

• 24 Mar 2020 to 05 May 2020 several tests were perform in order to find the problem on the QEPro spectrometers. See PP Lab Notebook report for more information:
  • QEPro 02181 was sent to the vendor. It was not possible to perform the white calibration, had low counts and did not follow the same pattern as QEP00674.

• QEPro 00732 was sent to the vendor. On the graph bellow it is possible to see that QEPro 00674 data are more regular than QEPro 00732 data; QEPro 00732 follows the variation of the temperature but in the opposite direction and Total counts on both QEPro went down each time that the old light source was installed (defective bulb). Also was not possible to perform the white calibration on QEPro 00732 (See image below).

Suggestions / notes

• Remind the scientist to use the Exclude Intervals feature to not measure the foam spacer in the core.

Tests

• Summary of Test 19 results (we are only using the halogen light).

The experiment is divided in 4 steps:

1.- New light source installed

2.- Old light source installed

3.- New light source installed. Light source not connected to external power supply

4.- Old light source with new bulb. Connected to external power supply.

Looking at the graph we can see that:

- QEPro 00674 data are more regular than QEPro 00732 data.

- Total counts on both QEPro go down each time that the old light source is installed.

- Detector temperature on both QEPro behaves the same. QEPro 00732 has approx. a degree less.

- QEPro 00732, follows the variation of the temperature but in the opposite direction.

- On step 2, both QEPro behave more erratic.

- We can not see a change in behavior between step 1 and 3, for QEPro 00674.

- On step 4, the behave of QEPro 00674 is better that on step 2 (maybe the bulb used on step 2 was damaged).

- On step 4, both QEPro have and step and total counts go up. That continue to happen during latter on the experiment but the results are not represented in this graph. More experiments need to be done.

- On step 3, we can see a big jump down on the temperature, we don't know what is the origin of it. (Light source was not connected to the external power supply).

- Locking at the picture is possible to see that both spectrum are different.

Experiment

TEST19 040820 QEP00674 & QEP00732
Same light source for both QEP.

HALOGEN LIGHT IS CONNECTED TO A 24V EXTERNAL POWER SUPPLY, ONLY FOR LIGHT, TO CHECK IF THE PROBLEM WITH INSTABILITY CAME FROM A ISSUE ON THE ACTUAL REGULATOR.
All electrical current goes through UPS and the UPS was connected to a low pas filter.
Calibration at the beginning of the experiment (Not possible to do white calibration on QEP00732).
Reset to default values each time before calibration.
Data measurements: Total Counts, Detector Temperature, TEC Temperature and TEC Stable.
Light source was change, using a new one (sn= 034990128, pn=RF52053)
1254 040920, CHANGE LIGHT SOURCE FOR THE ORIGINAL ONE (SN 034990124, PN=RF52052).
1249 041020, CHANGE LIGHT SOURCE SN 034990128 (PN=RF52053) WAS INSTALLED. IN THIS CASE THE 24V EXTERNAL POWER SUPPLY HAS NOT BEEN USED.
1330 041120, CHANGE LIGHT SOURCE FOR THE ORIGINAL ONE (SN 034990124, PN=RF52052).NEW BULB WAS INSTALLED (10 APRIL 2020). IT IS PLUG TO THE 24V EXTERNAL POWER SOURCE.
From 1315 040820 to 1226 041220

• Experiments performed to QEPro 02181. Total Counts data and a graphic of Test12, comparing QEP00674 and QEP02181. In the results it is possible to see that QEP02181 has always very low counts and doesn't follow the same patron as QEP00674.

TEST11 032120 QEP00674 & QEP02181
Same light source for both QEP.
Calibrated.
White + Blue light on.
All current goes through UPS and the UPS was connected to a low pass filter.
Reset to default values each time before calibration is added to the code.
AT THE BEGINNING BOTH SPECTROMETERS WORKS WELL BUT AFTER A LITTLE BIT QEP02181 FALL DOWN AND MEASUREMENTS WHERE CLOSE TO 1200 (NORMALLY 180000), BUT THE MAINTAIN THE SHAPE.
AFTER THIS EXPERIMENT WAS NOT POSSIBLE TO CALIBRATE QEP02181.
SWAP WIRES OR POSITION OF THE QEPro ON THE SOFTWARE MADE NO CHANGE. NOT POSSIBLE TO PERFORM CALIBRATION FOR QEP02181.
MAYBE THE PROBLEM IS THE TEMPERATURE?
(I don't have the date or pictures from Test 11)

TEST12 032220 QEP00674 & QEP02181
In this experiment we will test if the warming of the QEP02181 is the cause of the falling of the values.
Both QEPros and both lights will be switch off at least 11 hours.
There are three types of the temperature that the instrument record, we will record the three of them to check if the change in any of then could be the cause of the problem.
Same light source for both QEP.
All current goes through UPS and the UPS was connected to a low pas filter.
NO Calibrated.
White + Blue light on.
Reset to default values each time before calibration is added to the code.
QEP02181 PRESENT LOW COUNT VALUES SINCE THE BEGINNING (1000 approx.).
TEMPERATURE AND OTHER VALUES IN BOTH INSTRUMENTS ARE APPROX THE SAME.
THE SPECTRUM ON BOTH SPECTROMETERS HAVE THE SAME SHAPE.
We will let the experiment run for some hours to see if there is a change. NO CHANGES.
From 1950 032220 to 1223 032320

X378/378S

• Change made to file writing to stop standard measurement from being added to the section data.
• 17 Jan 2020 to 19 Jan 2020
  • SHMSL has crashed into the benchmark/ the MS standard 3 times and pulled the rubber strips out of the ring.  Adjustments were made to the y-axis height to get the system to avoid crashing into the bench mark.  Note that the safe height is based off the core surface and not the bench mark. 
  • An offset was noted in the core data for L* (more towards white) between sections 3 and 4 of core 16 U1553B.  The offset was determined to be  caused by the QEPro not touching down fully on the white calibration standard and applying an incorrect correction factor.  Note that the integration time had increased suddenly which helped diagnose the issue.
• 04 Jan 2020 to 12 Jan 2020  During Exp 385 the QEPro had many issues and the technical staff had to switch the QEPro out multiple times.  During the transit at the start of 378 testing of all three QEPros was completed in an effort to determine if a) we have any working spares and b) to test the reset procedure provided by Bill Mills during Exp 385.
  • QEPro #90900 was installed on the SHMSL at the start of Exp 378 and appeared to be functioning properly.  It was removed and replaced with QEPro #91126. 
  • QEPro #91126 initially calibrated properly, but when running measurements the spectral values were all near zero.  We attempted to do a reset using the procedure provided by Bill Mills (see below).  This reset did not correct the low spectral counts on this QEPro and we were unable to preform a successful calibration.  The QEPro was removed from the system.
  • Reset procedure: Make sure IMS itself is closed before running this procedure.  Open the I_PI_QEPro.vi and the QEPro Driver.vi.  On the front panel of the I_PI_QEPro.vi select Initialize Communications from the drop down menu and run the vi.  Then in the QEPro Driver.vi select  reset factory defaults from the drop down menu and run the vi.  The QEPro should be restored to normal operations at this point.  Re-launch IMS and test.
  • QEPro #90899 placed on SHMSL.  This device was removed during 385 because it was periodically giving poor data.  When it was reinstalled on Jan 4 2020, it appears to be functioning properly.  It was left installed and testing continues to see if the inconsistent behavior returns.  No reset has been preformed on this device as of 07 Jan 2020.
  • 09 Jan 2020 QEPro #90899: Attempted calibration but the white calibration would not stop collecting data regardless of counts collected. Aborted calibration and attempted running a section. The results for the section returned results 'NaN' for L*, a*, b*.  Attempted calibration again (figures below) with same white calibration behavior. At 0.7sec integration time clicked 'Stop' and 'Accept Calibration'. Moved to the dark calibration and result was looked like noise rather than clear spectra. Aborted calibration and ran section again. This time QEPro generated numbers for L* a* b* with L* values varying from 0 - 80. Ran section again with L* values varying between 4-26.

• 
  
  • Attempted a Restore to Factory in the QEPro driver.vi, no change.
  • Restarted computer, no change.
  • Changed Halogen light (09 Jan 2020 ), no change.
  • Noted the % saturation in QEPro set up was 95% and the manual showed 90%.  Changed to 90%.  Run calibration, and white calibration was successful.  Dark was still noisy.  The red horizontal line on the white calibration plot has a max value that appears to be set by the saturation percentage. The line was approximately 180000 when saturation was at 90%.  The red line moved to 190000 when set to 95%.  When set to 85% the red line stayed at 170000. We ran multiple successful white calibrations with no working dark calibrations.  Then the white calibration stopped working.  The counts could not reach the desired target count value.  After switching back and forth between the saturation percentage values the white calibration began working again.
  • 10 Jan 2020 to12 Jan 2020: QEPro 90899 and 91126 each underwent 'Restore to Factory' command in the QEProDriver.vi. Then each QEPro was connected to OceanOptics Software on DESClogik computer and hooked up to spare halogen light source.
    • QEPro 91126: Counts in Ocean Optics software were approximately 1200. This was in line with the low counts observed during the 'White Calibration' in the IMS software.
    • QEPro 90899: Counts in Ocean Optics software were approximately 170,000. QEPro and light source were kept on and running over a couple of days. The counts dropped down to approximately 141,000 over 90 minutes. The counts would then then drop down to approximately 137,000 and then rise back up to 142,000 over a period of several hours and repeated that pattern over about 2 days.

X378T

• 27 Nov 20191x large 75 V capacitor in the Power Supply box was replaced (ETs preventive maintenance).

• The fan on the Power Supply box was replaced for a new one (ETs preventive maintenance).

X385

• 03 Nov 2019  SHMSL still not working consistently. The data issued look different though. The l* data decreases and the b* data increases. And at times the b* increases and then decreases down section. 
  1. NEED INFO HERE ABOUT GARRICK'S CHANGES.  Garrick monitored system power and saw no dips in power when the data became inconsistent
  2. The 3rd QEPro, 0674 was placed on the SHMSL.  Issues persisted.  Note that the white calibration was coming up with very low (no) counts whenever the data became inconsistent.
  3. Bill Mills sent Labview code instructions to do a reset on the QEPro. System is working currently. Continue monitoring system.
     1. Are you see low spectral counts?  If so the last two times this happened it was caused by a loose electrical connection on the power cable.  If this happening,  you may hear Window's OS connecting and disconnecting from the device also the power light on the QEpro will flash orange.  Again if this what you are seeing it can be corrected by first fixing the loose power connection (inspect the entire length of the cable from power supply to QEPro).  Quit IMS but leave it open, next open the I_PI_QEPro.vi and then open the QEPro Driver.vi.. On the front panel you will see a drop down box  select Initialize Communications and run the vi.  Next select Reset Factory Defaults and run the VI.  This should restore the QEPro to normal operations.  Re-launch IMS.Doing this from memory so the names in the drop list may not match what I wrote above but you should be able to figure it out.  I believe there is a write up in the tech reports.

• 02 Nov 2019: SHMSL data inconsistent. As seen in below figure, the L*a*b* values dropped to very low values and then would return as noisy data. This was a duplicate hole and nothing of this type was observed for the original data sets.We attempted to do a QEPro calibration and found the white counts to be low and the dark calibration to be very noisy (see image below for dark calibration). We systematically changed out hardware to try and identify the issue. QEPro00732 was on the system when the problems were first noted.
  1. Switched out the single fiber optic cable.  No change in behavior.  The original cable was placed back on the SHMSL.
  2. Switched out the split fiber optic cable.  No change in behavior.  The original cable was placed back on the SHMSL.
  3. Checked power connections.  Garrick powered Startech from external power instead of through the USB.  This did not correct the issue.
  4. Switched from QEPro00732 to QEPro2181.  Calibration data looked normal after switching.  We switched back to 732 and the calibration was still bad.  We left QEPro 2181 on the SHMSL.  But data issues persisted even with a good calibration.
  5. Switched QEPro cable to a different USB on the Startech.  We were able to collect some data for a few sections and then the problems reappeared.  We then wired a USB from the QEPro to the computer.  This worked for a while. 

The problems persisted through multiple shifts. For 12 hours the only fix that seemed to work was switching where the USB was plugged in (Startech vs Computer).

• 31 Oct 2019 : L*a*b* values dropped to ~0. Attempted to calibrate QEPro but white calibration showed very few counts. Halogen bulb was replaced and Calibration was attempted again and saw same low counts. Following steps were taken:
  1. Restarted IMS. No change in behavior
  2. Checked for dirt inside sphere, saw none. Confirmed shutter was open
  3. Cycled power to both light sources. No change in behavior
  4. Unplugged end of fiber optic cables, confirmed light coming through
  5. Cycled power to Computer. No change in behavior
  6. Cycled power to QEPro. No change in behavior
  7. Tried different port in Startech. No change in behavior
  8. Swapped QEPro2181 for QEPro0732. Calibration was successful and section half data consistent with previous cores and sites.

• 07 Oct 2019 : RSC data acquisition stopped and was not noticed for a few cores.  One of the wires of the power cable to the QEPro broke. The cable was repaired and the system is working properly.  A few cores will have to be SHMSLed again.

• 24 Sep 2019Calibration times were slower than expected and Preventative Care and Maintenance's (PCMs) were conducted: After PCMs were completed the calibration results were improved from 0.34 to 0.21. The following PCMs were conducted:
  1. The X and Y axes movements were adjusted for both sensors to ensure proper positions and simultaneous compression during both measuring and calibration motion.
  2. 24 Sep 2019The halogen bulb was changed .
  3. The "white" calibration puck was resurfaced and sanded smooth with D.I. water and 220 grit sandpaper on a glass surface, rinsed clean with D.I. water and dried thoroughly before replacing.

X379T/385T

• Sometimes if the CAPS LOCK key is ON the label will not scan correctly, try again a couple times, fix the scanner or switch off the Caps Lock key are the possible solutions.
• Even if the laser profiles are correct, if there is a high on the section, the sensor track will not go up in order to measure or avoid that high, it stays at the Safety Clearance. Changes were made on the LabVIEW code in order to solve this problem.
• The key LIGHTS ON is not working, but if the lights are OFF the turn on when the START key is pushed. It was a confusion on the LabVIEW code that has the switch off order on it.
• Near the the end of 385T expedition, even if the label ID was reading correctly, the laser didn’t begin to measure the profile. This required a re-scanning of the label. The origin of the problem was that the program was still updating files while the section was scan, for solving that the program is tell to ignore that issue.
• Etienne Claassen made a teflon holder for the MSP standard with a chamfered cut-out hole in the middle to hold the sensor in place. The new container is 19 mm high, 54 mm in diameter, with a 5 mm deep and 26 mm diameter central hole. The space on the plexiglas support for this standard was deepened so the contact surface is the same as that of the RSC white standard.

• 
  

  

• After the implementation of the new software, even if the laser profile was done correctly the MS and QEPro sensors will begin to measure 2.5 cm before the bottom of the section. This was solved by updating the instrument offset for: MS, MS2 and MS3, X=11.09 and for QEPro, X= 21.09.
• The new sensor/instrument offsets for the integrating sphere and MSPOINT sensor are: X= 3.1 cm and Y= 4.2 cm
• Since 24 Aug 2019, offsets recorded in the RSC files and MSPOINT output files are rounded up to two decimal places, instead of three.
• Bill Mills updated the IMS-SHMSL program to incorporate the use of the new MS3 meter, which is orders of magnitude faster than the previous MS2 sensor.
• Even if same parameters for measure the same core are used, sometimes the amount of measurements is different, because the MSP sensor stops measuring in different points of the end of the section. This issue was noticed with a 5 mm sample interval, measuring a 150 cm section, the difference was not more than 1 cm (2 measurements).
• The logger occasionally makes a “clacking” sound when it is moving on the X axis, near the home position. This is despite the cleaning and lubrication of the NSK actuator.
• The QEPro High Performance Spectrometer was changed for a new one (ns QEP00674) because it was unable to perform an RSC calibration (counts were only in the order of 10³ instead of 10⁴, similar to what happened during Exp. 382). This is despite a new bulb and the fiber optic cable swapped with the old spare. Note that the QEPro should be installed by the MCSs using the appropriate driver. 
• The power plug for the QEPro High Performance Spectrometer was found to be loose and was replaced with the original. This prevented the QEPro from being intermittently detected by the Device Manager or NI-MAX.
• The RSC and MSP gimbals were slightly modified by replacing the button-head screws with counter sunk head screws.

• Bill revised the IMS routine for measuring the laser profile of the split face and for calculating the sample volume. This new approach requires measuring an empty section-half liner as a “background”.
• The wire that connected the computer with the monitor was defective and was change for a new one.
• After the re-assembling the SHMSL, the AR700 laser that was taken out during Exp 382 (sn 001918) was found to be working fine and was re-installed. It was initially thought that the cause of the problem was the power chord or the unit itself.
• The gimbal for the RSC integrating sphere was repaired twice by the ETs because the rubber supports came off.
• Once the sensors stayed pressed on the split face, the abort button became unresponsive and the computer was sluggish. Bill Mills replaced the entire PC with SHIL-2. SHMSL PC was serviced by the MCSs and is now put back on track.
• The white standard for the RSC was cleaned very often, almost every two or three days during JR100.
• The gas strut for the Y-axis was replaced because the old one was not strong enough.

• Sections 999-J1008D-1H-1A, 2A and 3W were measured with different intervals, using two MS meters and point sensors. The data are on IODP_Share/PhysProps/CORE_999_J1008D.

• For testing intents there were created three different software folders (I_PI_MS, I_PI_MS2 and I_PI_MS3) inside IMS-10/PLUG-INS/don’t use. Depending of the sensor that you want to use it is necessary to cut and paste the selected folder under IMS-10/PLUG-INS

  • MS: Typical measurements with MS2.

  • MS2: Measurements with MS2 for testing purpose. This file is the same on the WRMSL and SHMSL, in order to compare them. USE THIS ONE FOR GUAYMAS EXPEDITION.

  • MS3: Measurements with MS3 for testing purpose. This file is the same on the WRMSL and SHMSL, in order to compare them.

Testing

• Test measurement of the standards with MS2 and MS3, using the sensors MS2E and MS2K and the IMS and Bartington software were performed. On the figure it is possible to observe that the difference between the measurements it is not dependent of the meter (MS2 or MS3) or the software, but is dependent of the sensor (MS2K or MS2E). It is important to notice that the output file of the IMS only reports the base of the actual reading, which is in 10E-4.

• Test measurements of 999-J1008D-1H3 were performed with a sample interval of 5 mm. Meters MS2 and MS3, and MS2K and MS2E sensors were used. No significant differences between meters or sensors were found. For comparison, this section was also measured with the MS susceptibility loop on the WRMSL.

• The MS2K/MS2E factor for standard measurements is 1.94, but on the core measurement the factor is 1.17, the reason of this difference could be because the 1 mm thick of the teflon standard cap.

Suggestions

There are intermittent communication faults on the QEpro High Performance Spectrometer, halogen light sources and the Acuity AR700 Laser due to the size of the universal round plugs (5.43mm outside diameter, 2.56mm inside diameter). ETs recommend that it could be good to have the original pluggers recommended by the supplier. The size for the AR700 and halogen light source is 5.45 mm outside diameter and 2.11 mm inside diameter, and 5.41 mm outside diameter and 2.48 mm inside diameter for the QEPro Spectrometer.

X383

• 24 May 2019Halogen bulb changed.

X382

• 05 May 2019: QEPro began having issues talking with IMS during a calibration attempt.  It had flashing red and orange lights by its USB port.  After a brief look we discovered that we are still using an old Star Tech USB hub for the instruments on the SHMSL.  We switched to another open port on this box and everything started working again.  Weren't all the old Star Techs removed from the tracks because of issues like this?

  • In the process of troubleshooting the laser profiling process, we learned a lot about the features of the AR700, and provided a chance to closely look at the profile data uploaded to LIMS. The laser profile calculations for the uncorrected section half thickness and area should be reviewed (and revised as needed; see image below for references and equations. Vector file for future editing.)

• 14 Apr 2019 to 17 Apr 2019:  Much more troubleshooting on the AR700 issues afflicting the end-of-section measurements. Ultimately we wound up replacing the laser with a new spare and replacing its power supply with a factory-supplied power brick. It definitely improved things but every once in a while the laser "sees" and includes the bottoms of the delrin saddles holding up the rails in the total section length, if they are close to the end of a section. (Anyone else please add details/corrections to this brief narrative!)
• 13 Apr 2019Return laser length is variable cms longer than actual section. User has to manually enter length.  This has been an on and off issue.
• 23 Mar 2019to24 Mar 2019:  During routine testing we noticed that the integration time on the QEPro was extremely long, especially in the calibration process.  Per standard practice the halogen bulb was replaced.  Integration times remained long however, and the white spectrum during calibration was clipped at about 120000 counts although the spectrum below the clipped limit continued to grow towards full saturation.  Other troubleshooting steps included re-calibration after the bulb has warmed up or an hour, rebooting the PC, re-seating the USB cable on the QEPro, and renaming the QEPro config file in IMS so it would recreate a fresh version. No difference in behavior was observed.  The SVN logs also show no recent changes to the IMS-SHMSL software. Finally, the QEPro spectrometer itself was replaced with the newest spare, a new device driver was installed and and a trial version of OceanView was installed. NIMAX was used in adding the new QEPro, but we ran into the problem of what the correct name that is hard coded in IMS. For this, we briefly re-connected the old QEPro spectrometer to find out the correct name. The lugs for the X2g power wire was also re-crimped to provide firm contact and prevent communication loss with the PC. The power plug for the cDAQ USB-serial module was also reset as it came loose during the course of troubleshooting. In a nutshell this provided an updated Ocean Optics driver and the QEPro resumed normal behavior.  The suspect older QEPro is in the ET shop for more evaluation. A full version of OceanView has been installed on the DESClogik computer in the logging office for offline evaluation of our QEPros.
• 24 Mar 2019: gas strut replaced to remedy sagging chassis after quitting IMS.

X379

No Issues

X368T

No Issues

X378P/368X

• 26 Nov 2018SHMSL-Move to safe height command in the Y-axis setup does not move correctly.  Motion looks to be only half a centimeter rather than the 3 cm entered.
• SHMSL-The software does not appear to use the Laser Profile to determine correct Y-axis motions.  This caused instances where the sensors did not touch a core.  The core surface was a few millimeters above the benchmark and the instruments were remaining 5mm above the core.  The hardware was adjusted to ensure the instruments would touch the core.
• 18 Oct 2018SHMSL - changed out QEPRO holder bands, previously the holder bands had been epoxied in place and with the changes in temperature recently they had become extremely stiff. New ones were installed and the instrument reconfigured, the downward motion is now more flexible removing the chance of a gap between the sample and the collector.