User:Drewling
From MariachiWiki
Contents |
About Me
My name is Andrew Lingenfelter. I was born in Buffalo, NY and have lived there my whole life. Currently, I am pursuing a Physics Bachelor's degree at Worcester Polytechnic Institute. I have a full scholarship with the Naval Reserve Officer Training Corps. Being in NROTC means I have to take a few extra Navy-related classes and participate in physical training. So far I am enjoying it. At graduation I will become commisioned as a naval officer and I will have to serve for at least four years. I hope to work at Naval Reactors in Washington D.C. Here the Navy does research and engineering pertaining to nuclear reactors on naval vessels. If accepted into this program, the Navy will sponsor my master's degree in Nuclear Engineering. Eventually, I hope to get a PHD.
Summer Research at SBU and BNL
June 12, 2006
Today some of the MARIACHI participants met to watch professor Marx's presentation about what we're going to be working on. It seemed very interesting and I look forward to contributing to the project.
June 13, 2006
Today the participants of our Physics REU met many of the people involved in the different research projects. Everyone seemed very interesting. Later, my partner Ben Moeller and I met with our graduate mentor, Zejie Zhang (Jay). He went over a few things about the radio and data collection portion of the experiment. Ben nor I knows how to use MATLAB. Jay told us to look over the MATLAB tutorials and try to answer a few problems which he gave to us. Ben and I were having difficulty going from not knowing how to use MATLAB at all to creating signal processing simulations in it. The tutorials and help sections of MATLAB were not sufficiently helpful for a significant progression in the assignment. We plan to have a good grasp on MATLAB in the next few days.
June 14, 2006
Today Ben and I are working on MATLAB again. Jay could not make it to the office today. This was unfortunate because Ben and I were looking forward to asking him for help with the exercises. Although he could not show, he emailed us a better-explained version of the exercises he wanted us to do. We were able to finish a few of the problems and tomorrow we will ask Jay for assistance with the remaining ones.
June 15, 2006
Today Jay helped Ben and I answer some questions. We are starting to understand MATLAB a little better. Unfortunately, Jay was busy today so he was not able to spend as much time with us as we would have liked.
June 26, 2006
The past week we have learned a great deal and have greatly improved our ability to use MATLAB and Spectrum Laboratory. We finished another one of Jay's MATLAB assignments. We were able to finish it with much more ease than the first assignment. My favorite part of this assignment was taking an actual recorded song that we took from the radio, importing it into MATLAB, taking the fourier transform of it, applying the filter we made to it, and then inversing it back to the time domain, making it sound much different than the original file. Also, we hooked up the GPS unit, the GPS antenna, and installed the GPS software. Unfortunately, during the beginning of the week we were unable to have access to the roof so the signal strength was much worse than it could have been. In fact it was only able to pick up about 2 satellites. Later on in the week we were able to get access to the roof and we set up the antenna away from obstruction. The GPS then picked up as many as 9 satellites. Also, Ben, Jay and I went to BNL to look at the system set up there. We saw the newly installed antenna, as well as the familiar Spectrum Laboratory software that Ben and I have been using. It was interesting to see feedback from objects such as airplanes and meteors on the spectrograph. This coming week I hope to solidify my knowledge of MATLAB, Spectrum Laboratory, the GPS system, as well as the MARIACHI project as a whole.
June 28, 2006
There was a bi-conical antenna installed in a court of our building. Today we attempted to hook up both antennas and both receivers to the same sound card. We had to go to Radioshack to buy some converters. Unfortunately, we are still having trouble getting the GPS Event Time Tag to work. We have been reading the GPS manual hope to have this problem solved soon.
June 29, 2006
Today we figured out how to split the frequency scale in SpecLab so that it shows the left and right inputs. Also, we got the GPS pulse per second to work. The pulse per second created a distinct decibel jump every second on one channel. The other channel was used for the antenna. This could allow us to view the antenna output in respect to accurate time pulses. We were unable to get the Event Time Tag operational. It seems that the our new antenna reciever is not working properly. When we plug either the bi-conical or the roof antenna to the reciever there is very little response. When we plug it back into the old PCR1000 receiver we get a very significant response. We're trying to figure out if this is a software or hardware issue.
July 5, 2006
Today we connected the two antennas to two separate PCR1000 receivers and wired them to the left and right channels to one sound card. We had to go to Radioshack to buy a serial cable. We then recorded them in SpecLab simultaneously. Then we imported them into MATLAB and graphed the signal's frequencies and powers.
July 7, 2006
We took samples using the same antenna and different receivers, and using different antennas and the same receivers. After graphing the ffts and powers in the varying circumstances it seemed that the roof antenna produced much more noise than the court antenna. We plan to place the roof antenna in the courtyard with the bi-conical antenna to better compare the two antennas. The PCR1000 receiver seemed to transfer a more powerful and clear signal than the WinRadio internal receiver. We plan to determine whether this is a software or hardware error. Ben spoke with Carlos Gamboa to set up the WinRadio software to match that at BNL.
July 10, 2006
Today we went to BNL to speak with Helio Takai and to further ensure that system there matches the system at Stony Brook. We took screenshots of all the settings and took note of the hardware setup. Helio told us that something we need to do is to find out a way to determine the time delay Δt for each radio station between BNL and SBU. With this Δt we can determine the position and height of an object, and hopefully, eventually, cosmic rays. We determined a cheap way to try this. We will make all software and hardware identical at both sites. This includes the use of the GPS 10MHz output to synchronize the internal oscillator of the radio receivers and also connecting this to a counter that will connect to the sound card, synchronizing the sound cards as well. At both sites we are going to simultaneously use SpecLab and record the a radio station on one channel and the GPS PPS on the other. Because the GPS PPS will be exactly the same at both sites, we can compare a decibel spikes at the two locations with respect to the GPS PPS. We can then measure the time difference between the spikes at the two locations using MATLAB. But how will we determine the difference in each PPS? As long as the computer clocks at the two locations are synced with the GPS, Speclab will mark the PPS with respect to this time and we can find out which pulse we are comparing. At this point we need to set things up to make sure that we get very similar signals at BNL and SBU.
July 11, 2006
Today we compared the BNL screenshots to the software settings we had and made the proper adjustments. We finally discovered the problem we were having with the WinRadio receiver. The audio cable that we were using was bad. We were messing around with the cables at one point and plugged in the other wire that we were using for the other receiver. From this point on we got great signals. The signals we were receiving today looked very similar to BNL's. We saw several peaks occur that resembled those that meteors and airplanes make. Also, during the lightning storm today we could see clear spikes in the spectrograph when lightning struck.
July 14, 2006
Today we hooked up the 10 MHz output to the receiver and successfully synced it. As for syncing the sound card, we adjusted the GPS MUX1 output to 100kHz and connected it into the sound card's word clock input. We ran into a problem at this point. When we tried to adjust the sound card's options to expect this 100kHz frequency, it seemed that you can only set these to values that are available. They are 32, 44.1, 48, 88.2, 96, 176.4, and 192kHz. Unfortunately, it did not seem like there is a way to put our own value (100,000 in this case). I emailed the sound card company to whether this is the case.
July 18, 2006
M-Audio, the sound card company, as well as Spectrum Instruments, the GPS company, said that given the software and equipment that we have, we can not match the frequency of the GPS output and expected frequency input to the sound card. We are presently working on a way to put something in between the GPS unit and the sound card to split the output frequency. We considered putting a counter there. A counter would output a pulse for every specified number of input pulses. Unfortunately, it seems that the desired output frequency would have to be a multiple of the input frequency. In our situation, the closest we can get to this is inputting 10MhZ to a counter set to output every 312 pulses, producing a new 32.051kHz signal. This is close to the usable 32kHz, but it may not be close enough. I emailed WinRadio to see if there is any way to use the receiver as a intermediate device between the GPS and sound card to produce a custom pulse. We are continuing to search for other options.
Ben has been working with LabView to take the data directly from the receiver, completely bypassing the sound card and using the GPS event time tag to time the data. If we can get this working, I believe it could be much more effective than using the sound card with Spectrum Laboratory. For the time being, we are continuing to explore both methods.
July 21, 2006
WinRadio responded to me and said that I could not use it to send out a custom pulse. It turns out that we are going to get a new GPS unit that allows you to send out custom pulses. Yesterday, Ben and I went to BNL to speak with Dr. Takai and some others about a few things. Ben had some questions concerning LabView. Today, I developed an idea to get around the inaccuracy that the PC clock may pose. I split the GPS PPS to the sound card and to the Event Time Tag. Therefore, if we plug in the PPS we will know what time the first pulse is, thereby telling us what time the rest of the pulses are. I made a MATLAB program to deal with the files recorded in SpecLab. It finds where the PPSs occur and applys a time scale based on this. The user just needs to know when the first pulse occurs. To repeat what I did, do the following:
- Have your antenna signal plugged into the sound card while leaving the PPS unplugged
- Start recording in SpecLab
- Plug in the split PPS and note what time tag is assigned to the first pulse
- Stop recording
- Load file into MATLAB and run the program
This should tell you exactly what matrix index each pulse occurs at and set the time scale correctly. When I first did this, my first pulse was at 12:49:59 and I recorded 61 more pulses. My entire wav file was about 81 seconds. So I input my 22050 sampling rate and the 59 second start time and my 0 occurs right at the 2nd pulse as it should (in this case). Now you know the time that any event occurs quite precisely.
July 24, 2006
Today I have been working with MATLAB and I finished my Time Delay program. I recorded two different samples from the WinRadio receiver on different minutes. When we find the real time delay, they will of course have to be recorded at the same time, but for the sake of creating and testing the program it is ok. I then found distinct peaks in the two different samples. After entering the time frame of peaks that you want to find the time delay for, MATLAB will give you the time difference between the two peaks. Once we get the new GPS units with the custom frequency outputs, we should be able to sychronize the sound cards and take some data. If we run into no more problems and receive fairly similar signals, we should be able to find the time delay between BNL and Stony Brook.
