We think the beam in the ESR is not bunched. There is a clear difference in PMT signal if ions are in the ESR compared to no ions (see screenshot of scaler PMT mid, high level with ions, low level without) but we see no structure of an ion bunch (see screenshot of accumulated bins). 
We also increased the accumulation to 10 injections, still no bunch visible.
We double checked the frequency value of the HF cavity mentioned in the last ELOG entrance. 

In contrast to PMT Mid, the difference between ions and no ions in the ESR (see screenshot scaler) is not visible. Signal gets higher with injection and shows some kind of exponential decay afterwards. 
Beam was dumped at 500 and 1200s. Even after the dump, the signal steadily drops.
We tried to adjust the threshold such that a difference between ions in the ring and not becomes visible, but even with a minimal threshold no effect was visible.

As the beam in the ESR is most probably not bunched, there was not much we could do. During the night shift we:
- adjusted the threshold of the PMTs mid and north with beam in the ring
- found that the delay generator for PMT mid is broken, so the timing is off and needs to be remeasured
- tried to see the beam with PMT north, but with no success. Maybe with bunched beam this will be possible.

We turned off the PMTs, stopped mbs and the ESR pattern, notified HKR and left at 5am.


===> 10.05.12:30 (Carsten)
I checked the bunching in the HKR - the beam is bunche - see attached Foto from scope in HKR.
The blue line shows (the "zig-zag") that the ion beam is very well bunched, and that bunching frequency and cooling energy are "well matched".

I tried scraping with pressurized-air devices in order to verify the beam poistion. All 4 devices that we use do not work reliable (at least when driven "manually" via Device Control).
1) Sometimes they do not reach the set position
2) Sometimes they overshoot the set position but "oscillate" back to finally reach the set value. EVEN with small step like 5mm the drive moves then 15m in and 10 mm back (or even further?)
3) It seems (but difficult to verify to due 2) that position calibration is quite off.

In particular the latter behaviour makes it very difficult to locate the ion beam.  It seems to be possible to set up the particle detector for the primary beam but will make it very difficult if not impossible to set up 208Bi with the secondary beam.

The following value for "scraping" are thus to be taken with a grain of caution.
GE02DD2_G (Part Det behind cooler): beam is destroyd at -25 (at -30 ?)
GE02DS3HG (Scaper in south arc): beam is destrod at -4 
GE01DD1AG (Capture North arc): beam is destroyed at about -30 
GE01DD2IG (Ion North arc): even at "endlage innen" ion beam cannot be touched.

For further tests I try the following automated settings:

IN in in SC7, BP3
GE02DD2_G, GE02DS3HG, GE01DD1AG : -45
GE01DD2IG : 20

OUT in in SC10
GE02DD2_G, GE02DS3HG, GE01DD1AG : -120 
GE01DD2IG : 120
(not fully out)

The cooling ith a tentative DRALS pattern works well. 
Cooling voltage in SC8 is now 905 V higher:
ESR manipulations -> Cooler maniuplations -> GECEBG1E is changed from +150V to +1055V. Please note the sign conventions. This means that the electrons are now 905V faster.
This is compensated by a DRALS pattern of -900 V (COOL) / 3300V (MEAS0/MEAS1). The -900V at the drifttubes decelerates the electrons.

The voltage is verified via the bunching pattern in the scope. The bunching signals look best in a flattop of about +1050V to 1060V. 1050V is the design value (1050V -900V = 150V).

• ssh esr_cooler@140.181.95.195
  
• screen -r cooler_signals (Note: If this command ends with an error message, use "-R" instead of "-r". In this case the readout was not running and must be restarted (see below). Otherwise you should see something similar to the attachment. As long as warnings are printed everything works fine The program can be stopped by typing "stop" and pressing return.)
  
• (*) cd Documents/electroncoolerelectricalsignals/
  
• (*) conda activate cooler_signals
  
• (*) python main.py -f <file name> -i <interval> (argument filename is the name of the log file and interval is the sending interval of the HV-divider)
  

At ~17:26:30 the beam process was accidentally set to not wait on breeding, affecting the data in file 0012.LMD

We have 4 valid DR files for DR with unbunched beam. Iel = 500mA
run_0011.lmd
run_0012.lmd
run_0014.lmd
run_0015.lmd
corresponding ADC data for the DT voltages are available (ESR NUC).

NTCAP is stopped at ~22:30
(Plan: go back to bunched beam)

Below is a first offline analysis of the DR data of and of the intermediate cooling step (-900V) of unbunched beam.
The DR spectrum goes from Ecm = 24eV (channel 1) to 4 eV at channel (channel 1001)  with a stepwidth of 0.02 eV per channel (equidistant in the CM-frame). Thus 100 channels are 2 eV.
The other spectrum shows the intermediate cooling rate after each measurement step. 
The maximum of he peak is at about channel 520, i.e 24 eV - 5.2*2eV = 13.6 eV (expected ~17.8 eV). The peak is very broad. FWHM about 100 channels (2 eV). For comparison, the old data from 1998 had FWHM ~0.6 eV but limited due to density of the resonances - not experimental resolution.

The rising slope is the real RR (radiative recombination at non-zero energies) slope. There is no background from residual gas (verified in the bunching spectrum).
The intermediate cooling rate is about 2.5 to 3 times the rates in the DR spectrum. Expected was AT LEAST a factor 10 but rather a factor of 100 (one hundred). This finding is inline with the long beam lifetimes and the moderate breading efficiency.

Comment: In the analysis, I carefully checked all relevant raw-data spectra, in particular the timing pattern and the measured voltages. The raw data look extremely good (from a technical point-of-view)and agree with the settings. 
=> The new setup works very well.

At about 22:30 we switched back to the bunched beam.
The new NTCAP directory is IQ_2025-05-11_22-37-04, started NTCAP Schottky data.
Same ramp as before 24 eV to 4 eV. 1001 steps dE = 0.02 eV,
Measurement step 40 ms / cooling step 40 ms.
The ramp is repeated 15 times.
I_el = 500 mA.
Opened file run_0016.lmd
Jonas checked that the bunching works using the scope in the main control room. 

The DR peak appears after very few scans.

During this night we recorded several DR measurements. Before the DR, we charged bread for about 10 minutes.

1. We took over at 0:00 with file no.16 (started 22:15) 
2. Started new file (no.17) at 4:34
3. Recorded file no.18 starting at 6:30

We stopped the measurement at 7:28 and started the shut down procedure.