Pulser Tests with the Long Wires in Lab 8 --------------------------------------------- Original Rev. 23-FEB-2006 Current Rev. 26-FEB-2006 Tests done on 23-FEB-2006 ------------------------- Worked with Doug to collect some pictures of pulses on the long wires in Lab 8. The long wires are in a 3 layer array with ?? wires in each layer. The layers are horizontal. Under the bottom layer there are a few additional wires for High Voltage tests. In the array the wires are on 5mm x 5mm centers ?? The wires are ?? feet long. There is a tent of aluminum foil over the top of the wires that comes down along the sides of the wires. This shield is held up by two support wires that run, above, parallel to, and spread out horizontally from the "detector" wires. These support wires hold the shield foil a couple of inches away from the detector wires. The shield is at an approximate ground potential because its support wires are attached directly to the metal anchor at each end. Using this simple shield, the noise level at the output of the preamp is a lot lower than it was without the shield. This shield also provides some "visual protection" for the long wires. We currently have only one of the five preamps connected to a long detector wire. It is connected to the top layer West most wire, i.e. the easiest one to access. The top layer wire next to this "instrumented" wire, and the the middle layer West most wire are tied to an approximate ground at the near end. This is to approximately simulate having these two wires, that are closest to the "instrumented" wire, also connected to a low impedance preamp inputs. Pulses were put onto the wire from the battery operated "floating" pulser. The connection to the wire is made with a logic analyzer clip and is through a final 54 K Ohm resistor and a 500 pFd capacitor. The pulser output is about 5.3 Volts and falls off with a 500 nsec time constant, i.e 2.65 V usec of output. With the 54 K Ohm resistor this makes the injected signal about 4.9 E-11 Coulombs. Note that this is about 15,000 times larger than a real LArTPC signal. The Run I D-Zero preamplifier used has a 10 pFd feedback capacitor. Thus with this 4.9 E-11 Coulomb input signal one expects a preamp output signal step of about 4.9 Volts. In all cases the scope is trigger by the optical trigger signal from the floating pulser. The scope trigger is not changed in any way from one scope picture to the next. The optical trigger cable setup is such that you can move the pulser from the near end of the long detector wire to the far end without changing anything except where the pulser is clipped onto the detector wire. The scope is triggered on the rising edge of its trigger input waveform. Description of the scope pictures: - The upper #2 trace is the trigger signal. - The lower #1 trace is the output of the preamp that is connected to the top layer West most long detector wire. - Three pictures were taken at sweep speeds of: 4 usec, 1 usec, and 200 nsec with the pulser connected 3 feet from the near end of the detector wires. - One picture was taken at a sweep speed of 4 usec with the pulser connected to the middle layer next to the West most wire. The intent was to make a crosstalk measurement of diagonally adjacent wires. - The above 4 scope pictures were taken again with the pulser connected 3 ft from the far end of the detector wires. The intent was to collect data to measure attenuation of the signal and transit time of the signal when it is injected at the far end. - The scope pictures were given self describing filenames in the directory: www.pa.msu.edu/~edmunds/LArTPC/Long_Wire_Tests/ Tentative Results: - Look at the attenuation of the signal when the pulse is injected at the far end. Compare the peak of the preamp output signals in: tek_0_near_end_pulsed_4_usec.tif vs tek_4_far_end_pulsed_4_usec.tif and tek_1_near_end_pulsed_1_usec.tif vs tek_5_far_end_pulsed_1_usec.tif The near end signal is about 4.5 Volts vs 4.4 Volts when the pulse is injected at the far end. - Look at the transit time of the signal when the pulse is injected at the far end. Compare the relative times of points along the rising edge of the preamp output signals in: tek_2_near_end_pulsed_200_ns.tif vs tek_6_far_end_pulsed_200_ns.tif The initial rise from the baseline appears to happen about 70 nsec later when the pulse is injected at the far end compared to near end pulse injection. By the time that the rising edge of the preamp output signals have reached one half of their full amplitude, the signal from far end pulse injection is about 180 nsec later than the signal from near end pulse injection. Is this apparent slow down in the response of the preamp output signal when the pulse is injected at the far end due to the resistance of the long detector wire contributing to the effective preamp input impedance ? The difference is path length was about 6 ft less than the full length of the long detector wires. - Look at the diagonal crosstalk. Compare the peak of the preamp output signals in: tek_0_near_end_pulsed_4_usec.tif vs tek_3_near_end_cross_4_usec.tif The near end signal is about 4.5 Volts vs 0.5 Volts when the pulse is injected in the diagonally adjacent wire. - Reflections of the pulse along the long detector wire. Look in the scope pictures for structure in the front edge of the preamp output on the time scale of the signal transit time along the wire, i.e. about 200 nsec. I do not see any obvious structure. Concerns: Clearly all of this is just a first look at these signals. The grounds and preamp connections at the near end could be cleaned up. The pulse injection clip can be shielded (for the crosstalk measurement). We need to make the same measurements for a wire in the middle of the array. We need to try injecting the pulse at the middle of the long detector wire. We should tie up the other 4 preamps to detector wires and capture multiple outputs. We could also capture the digital output data files from the scope if any numeric analysis of this data is useful. What effects can we attribute to the resistance of the long detector wire ?