Config15: 23-Apr-2018 to present: This is rebuilt S/N 1 with a new PDBS and fold mirror. It was received at NIST around April 19. Mark says "This part is one of the two original test parts (Rev C, SN1 & SN2) we made as opposed to the large run of  "production" units (Rev D).(which are numbered (D001-DXXX)). These two Rev C test splitters  have a slightly different tolerance on the barrel ID to allow adequate adjustment and I consider them incompatible with the Rev D production parts."

Config14: 16-Apr-2018 to 16-Apr-2018 Fiber Bend Tests. On 16 April 2018 I used the Config 13 fiber setup for some fiber bend testing. I call this Config14. Most of the disturbance was to the output fiber. Things were not very clear, something more major happened when I moved the photodiode 2 (on the output) after E6 and before E7. Could not get it to repeat to the values for the E1 – E6. I suspect mode structure in the fibers, but it could be related to getting all the flux onto the photodiodes, or some interreflections, or something else.

Config13: The input fiber was changed to a collet, the output fiber connection to the PDBS redone, the LED/Sphere rotated 90 deg, and the gain on PD2 reduced by a factor of 10. 

Config12: Collet Fiber on one end, FC on the other; used to replace the output fiber on Config11. Collet marked "DO NOT TWIST" The input fiber is still FC/FC, from FISH, number 31This is in line PDBS with the PDBS and fold mirror removed.Fiber is longer than the FISH fibers, it is Thorlabs F1038-BK and I labeled it with Orange Tape.It was a little tricky to figure out how to modify the PDBS S/N 2 to connect to this fiber.See the drawings on page 149 of MOBY NET NB 2.

Config11: 13-Mar-2018 to 20-Mar-2018: This is the setup from Config10 with the in line PDBS S/N 2 without the PDBS or the fold mirror. There had to be a Configuration number change because the LED was removed and replaced, leading to different monitor signals, blowing the tracking using the first measurement from Config 10. The purpose of Config11 is to test for fiber bend sensitivity. The chamber door is left open and the room lights are off. Each different position of the fiber is a different variable ("experiment" on the web pages).

Config10: 13-Feb-2018 to present: We took out the PDBS and the fold mirror and are using a 9V battery on channel 1, channel 2 is the straight through. New, In Line PDBS, S/N 2. Uusing FISH fibers and SIRCUS sphere, with SEI

Dynasil - Polka Dot Beamsplitter, Fused Silica, 20:80, 12.7 mm Dia.

but how big is the collimated beam? Mimimum beam size for these things is 2mm diameter. They work better when the beam is bigger….

It is a metallic coating, which means that it should induce some elliptical polarization, but I think if the beam is unpolarized to start with, the effect is minor…and would always look the same to the rest of the system…..so I think that doesn’t matter…..

Because they are angle of incidence insensitive(up to 45 deg or so), you could get rid of the need for a right angle beam if it worked better mechanically….although I am not sure it is an advantage to have an output coming out the same was as the input….probably better to put a mirror after the 45 degree output to get it back to the two outputs going the same way.

I think you could make a pretty compact package, although in my head I can’t see how to do it in a cylinder. you are better at that….

ken

Hi Mark, I set up the polka dot BS with the input optics from FISH – using the 1” tubes w collimating leses as they were assembled. I had to take out the Vlambda filter on the LED source to get more signal.

I’m only getting 14mV on the 90 deg output channel. Did you mention these lenses had coatings?

We will try a run either Tuesday or Wednesday, Carol

Hi, that worked. Just a slight rotation to get the exit pupil filled with light, now have 77 mV on through and 71 mV on 90deg; this is without the Vlambda. We are scheduled to run Wednesday. Good sail, carol

Conversation on cfg 9

page Num 9.01 and 9.02

31 Jan 2018 - 5 Feb 2018

Carols comments: Hi, I just sent Steph the latest. My overall conclusion is the device is sensitive to what has been happening to it, but once things are on and set and in equilibrium, I don’t really see a huge dependence on either Temperature or RH. It seems to me the repeatability day to day is bigger than any correlation with temperature or RH. And this remains unexplained, giving differences in the splitting ratios of +/-0.3% for all the runs. It could be that exposure to high RH is progressive. A more advanced test would be to keep a control in a chamber at constant temp and RH (actually, the lab is not bad) and duplicate the source and detector, but we don’t have the bandwidth or parts for this now.

 My vote is a yes on these for the April deployment. We can see if the Temp and RH are maintained at a constant value, things look good. We can moderate the RH but the temperature will vary during testing vs deployment.

The chamber has an atomizer to increase the humidity, and a condensing coil to decrease it. Water must be evaporating when I turn off the chamber and leave the door shut overnight, as I have been doing. Tonight, I’m leaving the door cracked open with the chamber off. There were no films of water on the bottom plate after today’s run. The manual speaks of a drain but I’m not sure I can access it for checking.

Kens comments: From what it sounds like, we should try to have it assembled as far ahead of the calibration as possible, to allow the systems to dry out or at least equilibrate in RH before the calibration….

Carols comments: There is the possibility that the epoxy takes a very long time to really cure. I’m happy to look into a mild vacuum bake if we wThere is the possibility that the epoxy takes a very long time to really cure. I’m happy to look into a mild vacuum bake if we want to try that.ant to try that.

Marks comments: There are no vent holes in the splitter.  This would be a fairly easy modification if needed by either drilling holes in spots that won't allow light leakage or using vented set screws.  The air exchange would be slow in any case. I used methanol to clean the optics.  

Marks comments: The best place to add vents would be the Thorlabs tubes closest to the splitter. To vent the splitter body would require drilling through the mirror holders with risk of upsetting the alignment and potential mirror damage.

Conversation on cfg 9

page Num 9.02 and 9.03

5 Feb 2018 - 7 Feb 2018

Conversation on cfg 9-10

page Num 9.03 and 10.01

7 Feb 2018 - 15 Feb 2018

Everyone might want to compare this run, where the temperature changed but the RH did not, to 01 – 02 Feb in Config09, where we had the same temperature and RH profiles but with the PDBS and the fold mirror. To me, the behavior of channel 2 (the through channel) looks very similar. Eliminates the PDBS as the source of change?

Conversation on cfg 10

page Num 10.03

Carols comments:
Hi go to the Refresh page, click on Photos, scroll down to the last one
20180305_SplitterTempTests_Config10

If explanation is needed let me know. The 12” ruler is in the photo for scale only, it is not there during a run. Today I took a set of darks, then a run with the chamber turned off, then a run with the darks repeated. Seems very stable. Tomorrow I’ll bend the fiber and move the Fluke transducer around. Carol

Hi Carol,

Conversation on cfg 10

photos and setup of PDBS

Config11

13-Feb-2018 to present

This is the setup from Config10 with the in line PDBS S/N 2 without the PDBS or the fold mirror. There had to be a Configuration number change because the LED was removed and replaced, leading to different monitor signals, blowing the tracking using the first measurement from Config 10.

The purpose of Config11 is to test for fiber bend sensitivity. The chamber door is left open and the room lights are off. Each different position of the fiber is a different variable ("experiment" on the web pages).

programs

plt_output_panels_Config11_.m -- Stephs 4 subplots and zoomed versions and saves pngs as web page

data

FiberDVM_Config11.mat -- data from Fiber and DVM program

Full config 11 email

3/22/2018

Carols comments: on config 11 data: I am sorry this is hard to follow. A better way might have been to find the mean and make a table or plot saying the mean value and the adjustment I made (e.g. the height). It seems the stuff with the various blocks to bend it vertically are not that reproducible, but they are smaller in magnitude than the larger bend I put when I pushed the PDs back (20 Mar vs 19 Mar).

Config12

23-Mar-2018 to present 

• Collet Fiber on one end, FC on the other; used to replace the output fiber on Config11. Collet marked "DO NOT TWIST"
• The input fiber is still FC/FC, from FISH, number 31
• This is in line PDBS with the PDBS and fold mirror removed.
• Fiber is longer than the FISH fibers, it is Thorlabs F1038-BK and I labeled it with Orange Tape.
• It was a little tricky to figure out how to modify the PDBS S/N 2 to connect to this fiber.
• See the drawings on page 149 of MOBY NET NB 2.

The through output of the PDBS S/N 2 has lens "3 4  Vis" label. The lens is in a SM05L05 lens tube, this lens tube threads into the PDBS housing. The lens appears to be near or at the internal shoulder (i.e. all the way "in") the lens tube. A SM05L03 lens tube was threaded into the SM05L05 tube, and a second SM05L03 lens tube was threaded into it. Inside this part, again, all the way in, was the FC connector. It was spaced back from the shoulder a little bit with a group of very thin washers and a golden ring. It was held in place with a 0.5" retaining ring.

The collet fiber has external 0.535-40 threads, with a length of about 1". The issue was to get it at the correct distance from the lens. Since the FC fiber connector was near the bottom of the 0.3" long lens tube, I figured I should do the same with the collet end. I measured a distance of 12.1 mm from the back of the lens tube to the front face of the original FC fiber connector. 

I assembled the Config12 as follows:

• Removed FC fiber #21 from the PDBS and the PD2 FC connector
• Removed the lens/FC connector assembly from the PDBS body
• took apart the three SM05 lens tubes and saved the one with the FC (after I had taken out the FC to see how it was mounted and put it back.)
• threaded the SM05L03 onto the end of the collet until it bottomed out on the inner shoulder (it seemed this distance was better than using the washers from
• the FC lens tube, but I need to measure to see if it is 12.1 mm as well)
• threaded the "empty" SM05L03 onto the one on the collet
• threaded these onto the SL05L05 with the lens, turing the lens tube, not the collet/fiber
• threaded the assembled lens/collet fiber onto the PDBS body, turning the PDBS and not the fiber.
• at the other end of the fiber, I
• took the FC connector out of the SM01L05 lens tube that is holding the PD2 S2281
• attached the end of the new fiber to the FC connector
• threaded the SM01L05 lens tube onto the fiber until it bottomed out
• arranged the fibers in the port plug so there aren't any gross bends
• attached the BNC cable that goes to the preamp to the BNC on S2281.

Full config 12 email

3/26/2018

Config13

02-Apr-2018 to present

Output Fiber: Collet Fiber on one end, FC on the other; used to replace the output fiber on Config11. Collet marked "DO NOT TWIST"; fiber labeled with orange electrical tape. This is the same fiber as Config12.
Input Fiber: Collet Fiber on one end, FC on the other; used to replace the input fiber on Config12. Collet marked "DO NOT TWIST"; fiber labeled with green electrical tape.

This is in line PDBS S/N 2 with the PDBS and fold mirror removed.

I have redone how I connected the fibers to the PDBS.

On the input side to the PDBS, I removed the Thorlabs P/N F220FC-B (aspheric lens, "B" AR coating, mounted to an FC connector) and the adaptor part that got this to the 0.5" Thorlabs lens tubes. I used lens "Vis 1 2" that I had from the S/N 1 work, which is the Clarence Z selected lenses mounted with an FC connector. I took out the FC connector from its 0.5" lens tube, where the face of the FC connector was at the inner shoulder (plus some washers) of the lens tube, and I threaded this lens tube onto the external threads of the input fiber collet. This places the tip of the fiber at the same distance as it was with the FC connector.

On the output side, I undid the Config12 assembly, removed the FC connector from the "3 4 vis" lens, and threaded its lens tube (with washers) onto the external thread of the fiber collet. Following the procedure for the input fiber.

The result are the input and output lenses Clarence modeled for the splitter with the fiber collet at the PDBS and FC at the input of the green fiber and output of the orange fiber.

Since the fiber collet cannot be twisted, the output fiber was connected to PD2 as in Config12 -- by first disassembling the PD to FC, attaching the FC to the fiber, and threading the lens tube onto the FC to 1" part.  On the input side, the FC threads into the sphere. We had put some epoxy to keep it from turning and coming out. This was removed, and all the connectors to the sphere (SEI, monitor PD, LED source) were removed so the sphere could be threaded onto the fiber. The natural position of the PDBS based on its center of gravity led to a new position for the sphere -- now the input port faces up instead of south as before. So the LED source was remounted. See the photos from April 2, 2018.

The PD2 is now on gain 6. Before, it was on gain 7.

Full config 13 email

4/3/2018

Hi Steph,
On 16 April 2018 I used the Config 13 fiber setup for some fiber bend testing. I call this Config14. Most of the disturbance was to the output fiber. Things were not very clear, something more major happened when I moved the photodiode 2 (on the output) after E6 and before E7. Could not get it to repeat to the values for the E1 – E6. I suspect mode structure in the fibers, but it could be related to getting all the flux onto the photodiodes, or some interreflections, or something else. There are four photos to go with these data – the fibers with and without the “bends”.

I made the plots both ways.

Carol

Following Dark email

 Hi, I took a dark set after the fiber bend test in Config14. I ‘m not sure if you want to update these individually, because I’m going to keep putting all the darks in one program.

Full config 14 email

4/23/2018

Hi Steph, Config15 has been added (two days) and I’ve included all darks.
Enjoy, Carol

Config15
23-Apr-2018 to present

This is rebuilt S/N 1 with a new PDBS and fold mirror. It was received at NIST around April 19. Mark says "This part is one of the two original test parts (Rev C, SN1 & SN2) we made as opposed to the large run of  "production" units (Rev D).(which are numbered (D001-DXXX)). These two Rev C test splitters  have a slightly different tolerance on the barrel ID to allow adequate adjustment and I consider them incompatible with the Rev D production parts."

I had saved the lenses from S/N 1 when I sent it back to Mark for realignment. So this configuration is:

• Input: Lens "1 2" from S/N 1 with Fiber Collet at the BS body and FC at the sphere source, fiber is same as Config13;
• Through output: Lens "Vis 3 4" from S/N 1 with Fiber Collet at the BS body and FC at PD2, fiber is same as Config13;
• 90deg/fold output: Lens "NIR" from S/N 1 with FC at BS and FC at PD1, fiber is FISH #7. 

• Gains / which PD
• 6 / source sphere
• 7 / through channel (PD2)
• 8 / 90deg/fold channel (PD1)

 To assemble, everything was disconnected (output PDs and connectors on the sphere source) so not to perturb the collets, which cannot be twisted.

 B. Carol Johnson, Physicist

Full config 15 email

4/23/2018

config 15 through 30 Apr

4/30/2018