Difference between revisions of "DVCS3 DAQ"
From Hall A Wiki
| Line 1: | Line 1: | ||
| − | |||
| − | |||
=== ROC info === | === ROC info === | ||
Please see https://logbooks.jlab.org/entry/3292168 for information. | Please see https://logbooks.jlab.org/entry/3292168 for information. | ||
| Line 17: | Line 15: | ||
|} | |} | ||
| + | |||
| + | See also [[DVCS3 crate and channel maps]] for information about the channel layouts. | ||
| + | |||
| + | === Some ways to check the Calorimeter/spectrometer timing === | ||
| + | * The easiest check will be to look at the distribution of waves in the ARS, this lets all channels be checked simultaneously | ||
| + | * If the timing is very far off, we can compare the timing of one block with the calorimeter ADC gate. The calorimeter ADC gate is in CAEN TDC channel 9 (counting from 0), and Calorimeter block 001 with a 50mV threshold is in CAEN TDC channel 14. | ||
| + | ** The ROOT command, <code>T->Draw("tdc_val[14]-tdc_val[9]","tdc_val[14]>0 && tdc_val[9]>0","")</code>, shows the distribution of the time differences. From a DVCS cosmic run (10159), we can see that a well-timed trigger will result in an average difference value of about 400.<br /> [[Image:Pking_run10159_CB001_to_gate_timediff.png|200px]] | ||
| + | ** This command, <code>T->Draw("adcval[1]:tdc_val[14]-tdc_val[9]","tdc_val[14]>0 && tdc_val[9]>0","*")</code>, will show the correlation of the block 001 ADC value against the timing difference. Again from run 10159, we can see the strong ADC events are associated with time differences of about 400.<br /> [[Image:Pking_run10159_CB001adc_vs_CB001_to_gate_timediff.png|200px]] | ||
| + | ** If the time difference is peaked later than 400, we can increase the DVCS ADC gate width setting to try to pick up sensitivity before making an access to adjust cable delays. If the time difference is peaked substantially before 400, we will need to consider how to shift the triggering (and thus the ADC gate) earlier. | ||
=== Stress test === | === Stress test === | ||
Revision as of 13:23, 14 November 2014
ROC info
Please see https://logbooks.jlab.org/entry/3292168 for information.
| Model | Mac1 | location | Hostname | IP Adress | ROC Hardware ID | CODA ROC |
|---|---|---|---|---|---|---|
| VME | 002038046D12 | LHRS DVCS crate 1 | intelhadvcs1 (ROC27) | 129.57.192.80 | 4 | 27 |
| VME | 002038046DC1 | LHRS DVCS crate 2 | intelhadvcs2 (ROC28) | 129.57.192.86 | 6 | 28 |
| VME | 0020CEF603EA | LHRS DVCS crate 3 | intelhadvcs3 (ROC29) | 129.57.192.91 | 5 | 29 |
| VME | 002038046D0C | LHRS DVCS crate 4 | intelhadvcs4 (ROC30) | 129.57.192.92 | 0 | 30 |
| VME | Upper level crate | intelha3 | 7 |
See also DVCS3 crate and channel maps for information about the channel layouts.
Some ways to check the Calorimeter/spectrometer timing
- The easiest check will be to look at the distribution of waves in the ARS, this lets all channels be checked simultaneously
- If the timing is very far off, we can compare the timing of one block with the calorimeter ADC gate. The calorimeter ADC gate is in CAEN TDC channel 9 (counting from 0), and Calorimeter block 001 with a 50mV threshold is in CAEN TDC channel 14.
- The ROOT command,
T->Draw("tdc_val[14]-tdc_val[9]","tdc_val[14]>0 && tdc_val[9]>0",""), shows the distribution of the time differences. From a DVCS cosmic run (10159), we can see that a well-timed trigger will result in an average difference value of about 400.
- This command,
T->Draw("adcval[1]:tdc_val[14]-tdc_val[9]","tdc_val[14]>0 && tdc_val[9]>0","*"), will show the correlation of the block 001 ADC value against the timing difference. Again from run 10159, we can see the strong ADC events are associated with time differences of about 400.
- If the time difference is peaked later than 400, we can increase the DVCS ADC gate width setting to try to pick up sensitivity before making an access to adjust cable delays. If the time difference is peaked substantially before 400, we will need to consider how to shift the triggering (and thus the ADC gate) earlier.
- The ROOT command,
Stress test
- To run the analysis, on alternating aonl machines: "godvcs", "analyzer", ".L replay_DVCS.C", " replay_DVCS(run_number)".
- MSS storage: /mss/halla/dvcs/raw/dvcs14*
- to check the disk occupency: as adaq@adaq1, "df -k /adaq1/data1/*"
- Analyze deadtime on any aonl machines: "godvcs", "cd marco/deadtime/", "analyzer", ".L deadtime.C", "deadtime(run_number)".
- Please note: The tests were done on 06 November 2014. Times with no dates should be interpreted as done or observed on 06 November, otherwise the date is included in brackets.
| Run number | Buffer | DAQ start/stop time | DAQ rate TS11 (Hz) | DAQ rate ROC27 (Hz) | Analysis ran on wich aonl? | Analysis start/stop time | Analysis rate (Hz) | Time at end of copy of mss | Occupency of the data disk at end of data taking | Live time | Prescale |
| 2666 | None | 12:43/13:48 | 836.7 | N/A | 4 | 15:27(rest.)/18:14 | 342.0 | 21:21 | 0.29 |  |
4 |
| 2667 | None | 13:51/14:20 | 850.0 | N/A | 3 | 14:39(rest.)/15:15 | 684.0 | 22:45 | 0.3 |  |
4 |
| 2669 | None | 14:25/14:55 | 720.0 | 700.0 | 2 | 14:39/17:01 | 158.3 | 00:15 (Nov. 7) | 0.31 |  |
1 |
| 2670 | None | 14:56/15:26 | 752.0 | 716.0 | 1 | 15:26(rest.)/18:06 | 152.0 | 01:41 (Nov. 7) | 0.33 |  |
1 |
| 2673 | None | 15:43/16:14 | 711.0 | 690.0 | 3(rest.2 on 1) | 21:04(rest.2)/23:34 | 259.7 | 03:46 (Nov. 7) | 0.35 |  |
1 |
| 2674 | None | 16:25/16:56 | 722.0 | 718.0 | 2 | 16:43/18:23 | 152.0 | 04:45 (Nov. 7) | 0.35 |  |
1 |
| 2675 | None | 16:52/17:22 | 755.0 | 730.0 | 4 | 17:06/19:32 | 158.3 | 06:15 (Nov. 7) | 0.36 |  |
1 |
| 2676 | None | 17:25/crashed(17:45) | 870.0 | 840.0 | 1 | 17:38/19:10 | 152.0 | 07:16 (Nov. 7) | 0.38 |  |
1 |
| 2678 | ON | 17:52/18:25 | 666.0 | 648.0 | 2 | 21:06(rest.)/23:30 | 158.3 | 09:12 (Nov. 7) | 0.39 |  |
1 |
| 2679 | ON | 18:28/18:54 | 660.0 | 640.0 | 2 | 18:38/20:18 | 145.7 | 13:15 (Nov. 7) | 0.39 |  |
1 |
