Analysis Run Lists:

I have written shell scripts that select and extract various runs with particular properties from halog and show them in the html file. This will help me select and quickly ispect runs with required settings (ie. all hydrogen runs, runs with HWP-in etc. ) that I would like to analyze. Here are examples of such lists:

My Scaler Analysis:

I am trying to read BigBite scalers. I would like to know the frequency of hits in each E/dE bar. I am also interested in the event frequency in the logically anded channels (L&R) on the output from the LC4516 module. I was able to read TDC scalers that are attached to scaler modules in slots 1-3 and I get reasonable numbers. However, I have problems reading the scalers from the outputs of the LC4516 modules. For this we used C380 modules which are attached to slots 9 and 10. The problem is that I am able to see only first 16 channels, while the channels 17 to 24 are empty. This happens for E plane. In the dE plane I can see only 8 channels. I amy trying to figure out, what causes these problems.

Photo of BB scalers used during ee'd

First I suspected that I do not read all the channels correctly. In particular, I was not sure if my channel mapping was correct. I was trying various combination but nothing paid out.

After a while a learned that there are two "types" of scaler banks: bbite and evbbite. If I understand correctly bbite scalers are read out as a special type of event (140), while evbbite is included into a data stream. I comapred results from both types of scalers and realized that in the "evbbite" scalers I can normally see channels E17-24 and dE17-24 that are missing in the "bbite" scaler However, in the evbbite scaler I am missing channels from the beginning: E01L-16L and E01R-16R, which are all attached to the first scaler.

I started suspecting, that this problem could be related to the fact that we removed scaler module from slot 0 and start with slot 1 instead. Because of this change I started my map with slot 1.

"evbbite:"

I have asked Bryan Moffit for help. He told me how to use a CODA tool to look directy into raw data files and directly check data (without analyzer inbetween) Using this code you suggested I went through several runs from different periods of the ee'd experiment. If I learned this scaler stuff correctly, all BigBite scaler information should be stored in banks (sections of data) that start with 0xbbeN0000, where N represents the scaler module slot number in crate #9. This means following:

1.) bank 0xbbe10000 contains data from the most left scaler module which reads E-L,R TDC data.

2.) bank 0xbbe20000 contains data from second scaler module which counts E,dE-L,R TDC data.

3.) bank 0xbbe30000 contains data from third scaler module which reads dE-L,R TDC scalers.

4.) bank 0xbbe40000 corresponds to ungated scaler for triggers etc.

5.) banks 0xbbe50000 to 0xbbe80000 are for gated scalers

6.) bank 0xbbe90000 is used to record E-(L+R) scalers

7.) bank 0xbbea0000 is used to record dE-(L+R) scalers

However, when I was looking at the data, I discovered following things. Bank 0xbbe10000 was empty in all considered runs. This would mean that the information from one of the 10 scalers is missing. When I inspected data into more detail, I concluded, that scalers 1-8 could be shifted for 1 to the right. So, the first scaler module has address 2 etc. Last two C380 scalers remain at the same addresses (9,10). Therefore one gated scaler is missing. This is my working hypothesis. I am not sure about it, but I have strong suspicion that this is the case. I also made several checks and they confirmed my hunch. Please see the pdf file below. There are two examples: "first example" and "second example".

Pdf file with information about "evbbite" scalers

1.) I took two runs with different target-spin orientations. If everything would be correct, two data banks in each run should be empty. In the "second case" this is indeed true. In the "first one"- when flipping the target polarization-only one bank is left empty.

2.) The two non-zero gated scalers should also not increase simultaneously. When one increases the second one should remain fixed and vice-versa. This is the case in the "second example" (channels 6 and 7). In addition, when I sum up the numbers from scalers 6+7, I get number in scaler 5. That would suggest that data in bank 5 correspond to ungated scaler. (In the attached pdf file I always plot readout (converted from HEX to DEC) from the first channel in each bank)

3.) In the "first example" things do not work that nicely. Banks 6,7 are empty as they should be and bank #8 is containing non-zero values. However, bank #8 does not have a pair. I believe that its pair should be in bank 9, which is already occupied by C380 module. Is this possible?

4.) I believe that banks 9,10 correctly contain information from C380 modules. In these modules we occupied only first 24 out of 32 channel. The remaining 8 empty channels are clearly visible in data.

5.) In my belief banks 2-4 are containing data from TDCs, because all 32 channels are occupied, while we know, that we did not used all the available channels in the normalization scalers.

Results of "evbbite" scaler analysis (with the assumed corrections):

1.)

2.)

"bbite:"

After this analysis Bryan suggested me to look also into the "bbite" scalers, which are stored into raw data file as a special type of event (140). When I checked the raw data, I found out that in this case scaler mapping is Ok. However, the last two scaler modules C(830) readback only 16 out of 32 channels. In addition in the dE scaler (last module) only 8 of them are non zero. Why this happens still needs to be investigated. I believe that setting for reading only 16 ch. was left from previous experiments, where only half of the scaler channels were being considered.

Pdf file with information about "bbite" scalers.

BigBite scaler analysis - Final Results

Here are my final results of the BigBite scaler analysis. Plots below show trigger rates and rates in all BigBite TDC channels. In plots 4.) and 5.) green line shows TDC scaler rate for right PMT from a particular scint. bar and red lines represents scaler rate in left PMT. Mean rates in E,dE planes are approx. 20kHz. Warning: All these rates depend on the set TDC thresholds!!!

3.)