Test Lab log
Sep 16, 2015
- Peibo tested all the PMTs
- All PMTs look to be in good shape. Huan's last report indicated a few tubes broken.
Sep 18, 2015
- Measured the box for the supporting frame
Sep 21, 2015
- Opened the box to extract the mirror. It was not extracted.
- Box vacuuemed inside
The prototype open
- Pmts array.jpg
- Pmt detail 1.jpg
Detail of the PMTs windows (1)
- Pmt detail 2.jpg
Detail of the PMTs windows (2)
- Note that there is some silicon drops glued to some of the PMTs (1)
- Visually, mostly of the entrance windows seem to be in bad shape but Bogdan says it is not important (Fri, Sep 25, 2015)
- NINO card to be used, obtained from Bogdan
- HV+ from Bogdan, in test lab
Sep 25, 2015
- Dasuni explained how to use the computers and use CODA.
Sep 30, 2015
- Small meeting with Lubomir to explain us how to cable the NINO card.
- Even he refer us to John Annand test report, we cable it wrong because the scheme was different to the card we have it. Anyhow, Lubomir connected the card and it seems that the power is still working.
- Just for "fun", we connected a scintillator paddle to the NINO card, with this one unpowered, just cabled from the scintillator and the NINO's analog output with a ribbon cable through an ECL-NIM converter. I (Carlos) understand that it is due to radio frecuency coupling to the system and the lack of grounding to the board. Disconecting the cable from the scintillator causes to dissapear the signal.
- 2015-09-30 15.10.16.jpg
Signal from the NINO card UNPOWERED with a scintillator paddle connected.
Oct 1, 2015
- I tried to power the NINO card again with the right polarity (see picture). I was unsuscesful. I connected the card and raised the voltage slowly. When reaching ~3.5V the current drawing reach the maximum indicated by Lubomir and John Annand's paper. I noticed that the power source was not bridged to ground. I bridged the negative terminal to ground but the result was the same.
- Later I checked the card with Lubomir's power supply and the nominal values were the normal (5V, 1.24A) so I decided to take the card to W&M and test it there.
- 2015-10-01 15.37.17.jpg
Polarity of the NINO card
- 2015-10-01 15.37.06.jpg
Readout of the power supply of the NINO card in JLab's test lab.
Oct 2, 2015
- Todd and me tested the card and we found the same behaviour (at 3.5V we reach the 1.24A current draw by card), but Todd found that going up, the current raise just a but and drops at 5V. That's the reason I found with Lubomir's set up that the card was ok, he already has the power supply set to 5V without any ramp up.
Oct 7, 2015
- Now that we know the power proceadure for the NINO card we tried to extract information. We power the card with 5V checking that the current draw was not higher than 1.26A. We connected a scintillator paddle, power with 1400V, to the NINO, and the signal from the card through ribbon cable.
- From analog output to an ECL-NIM converter and then to an osciloscope. No signal was obtained.
- From LVDS output to the LVDS2ECL conveter, then to ECL-NIM and to the scope. No signal was obtained.
- With the use of a Probe, we tried to extract the signal directly from the card to the osciloscope.
- From the analog output, we get nothing, but according to Annand's scheme, perhaps we used wrong the probe.
- From the LVDS we got a normal signal.
- We noted that the polarity of the LVDS output changes depending of which threshold source is used. We should repeat this test.
- We increased the power to the scintillator paddle to 1800V (Mark told us that is the good value) and connected the card again to the frontend electronics. This time we got a NIM signal in the scope. We checked that the signal is correlated with scintillator signal spliting the signal. The signal in the scope was with the polarity exchanged. We noticed that the flat cable input to the ECL-NIM converter is inverted (ch 1 goes to ch 16 and so on) thus polairty is changed.
Oct 8, 2015
- Testing the NINO card with Todd, he simply cut the nail in the flat cable connector (which select orientation into the module). The signal obtained was correct polarized.
Oct 21, 2015
- The entrance window was changed wuth a plate with an LED attached. (top gallery)
- The plate was sealed with black silicon. When the silicon is cured, it will be covered with black tape.
- It is not accurate, but a simple view of the orientation of the LED to the mirror is shown in bottom gallery
- After some modifications on the set-up by Daisuni and Mark, I took some data with a scintillator paddle using itself as autotrigger.
- 2015-10-21 16.49.18.jpg
Entrance window before remove it
- 2015-10-21 16.50.41.jpg
Entrance window without tape
- 2015-10-21 16.55.44.jpg
Entrance window with LED
- 2015-10-21 17.06.15.jpg
Entrance window with LED, silicon applied
- 2015-10-21 17.06.35.jpg
Entrance window with LED, interior view
- 2015-10-21 17.07.36.jpg
Ruler aligned to the LED
- 2015-10-21 17.07.45.jpg
End part of the ruler close to the mirror
Oct 23, 2015
- Scott soldered a connector to the attached LED on the box with a 250 ohm resistor in serie.
- We connected a square function generator, with an amplitude of +5V and 1Hz frequency. In fact, the actual amplitude of the function was 2.5V but we applied an offset of 2.5V so the absolute amplitude, from 0V, was 5V.
- The LED seems to work perfectly.
- The alignment was hard to check due to the low intensity of the LED. We (Scott and Carlos) tried with a black cover and hardly we saw the spot directed to the mirror.
- We closed the box and sealed with silicon.
- In order to spread the silicon easily, we used dishwasher. It's suggested in many places, but I (Carlos) am not sure that it works with the silicon used. It didn't mess the silicon, but it didn't look normal. Anyhow it cured as normal.
- Some screws were so difficult to attach. One was impossible to screw. I guess that attaching the cover is not straight forward and there is some kind of "protocol": which screws should be placed first, which side goes first, etc.
- The same can be said with respect to the silicon. We tried as we considered to be done, but the final result is far to be optimal.
- Nevertheless, the silicon seal is not neccesary for the early test because we won't make use of gas. Light leaks is more important and it can be solved with black tape.
- The decoder macro only displays three kind of historgrams related with the number of channels and the total run. It is not useful for our purposes since in first instance we are interested in an event display, so we need a decoder to produce NTuples (to be used in ROOT) or something similar which give us data event per event.
Detail of the cable made to power the LED. A 250 Ohm resistor is connected in series at the positive cable
- 20151023 144552.jpg
Picture of the box sealed
- 20151023 144601.jpg
Detail of the silicon
- 20151023 144620.jpg
Detail of the thread which was impossible to screw
Oct 28, 2015
- Tested prototype for light leaks and covered them with black tape.
- Mounted prototype on the test stand and tested that we could see signals from the PMTs.