The assignment of DAC channels in EEroom (DAQbox 40, mezzanine 26) has been changed from ch n.3 to channel n.7, so the updated DAC assignment table is the following

At WEB the "superlum" device (DET stuff?) which was connected to DAQBox 38, mezzanine 39, channel 7 has been moved to channel 6 of the same DAC mezzanine. 

We tested DAC channels in WEB and EEroom, and they proved to work fine.

The Acl configurations have  been modified in order to allow to send any photodiode signal to any noise hunting DAC channel, in order to be able to listen to the photodiodes while tapping things around.

1/ The photodiode signals are selected and decimated down to 5 kHz in LSC_ACl

I have added a new file under LSC_Acl: LSC_NoiseHunting.cfg.
This file generates a channel ENV_PD_NotStored from any LSC signal (now selected with weights in a sum for some photodiodes. A few more can be added).

I have added a button below LSC_Acl : "Reload NoiseHunting"  that reloads only this file, without reloading the whole file.
So in principle this small file can be reloaded even when the ITF is locked.

2/ This channel is read by ENVnoise, it can be filtered for better hearing, and then added to any of the noise hunting DAC output.

Here is the way to use it:
1/ select the photodiode to be listened to in LSC_NoiseHunting.cfg file
2/ reload NoiseHunting file with the VPM button.
3/ in ENVnoise, also enable this channel to the wanted DAC channel
 

We actually tested the DAC channel in the EEroom by listening to a few photodiodes (B4_6MHz, B5_6MHz,...) and it worked as advertised.

Today, we improved the setup providing DAC outputs for noise hunting. All the Acl processes to send the noise to the reserved DAC channels have been updated.

Romain G. has moved a few cables at 2 DAC outputs in order to always use a DAC channel 7 for noise hunting (moved channels: channels used to provide the Vbias to the B5 quadrants on SDB1, and channel used to provide the power/modulation of the SLED on SNEB bench). The Acl processes have been updated accordingly.

Irene has connected the outputs to ADC channels to check the paths.

There was some aliasing issue to be understood in the injection path... but right now it seems to be working fine, at least for the DAQroom noise (see figure).

• ENV_NOISE_CEB_DAQroom is the noise generated at 20 kHz in ENVnoise channel, and sent into TOLM packets at 5 kHz
• ENV_NOISE_CEB_DAQroom_NS is the same noise read by a process in another rtpc and extended up to 100 kHz after low-pass filtering at 10 kHz.
• ENV_CEB_Test is the noise read out after being sent to a DAC channel and read in an ADC channel.

The coherence is good up to 10 kHz.    To be understood why the aliasing issue found in the past days disappeared (+ to check on all the noise channels).

... the problem of missing coherence between the white noise generated by the DAC and that readout by the ADC seems to still persist (for example) with the EEroom DaqBox. See the attached plot.

Today I connected also ENV DAC channels in DetLab and DetPlatform to one ADC channel, so that we can test them and further investigate the loss of coherence found with some channels. 

This table summarizes the status:

today I was testing the ENVnoise server, injecting white noise to CEB_DACroom, while I noticed reappearing the "bad-coherence problem" between in and out signals, the problem reappeared between 9:42 UTC and 9:50 UTC and it is still there: Figure 1 shows 9:42 UTC (good coherence) and Figure 2 shows 9:5o UTC (bad coherence).

I wander if this might be related to the work on DAQ servers performed this morning, In particular channel CEB_DACroom belongs to DaqBox 24 which is configured in Acl process SIB2_SBE. I report here the extract from VPM log file during this period:

2017-03-14-09h43m19-UTC    info masserot     process:SDB2_Photodiodes stopped
2017-03-14-09h43m20-UTC    info masserot     process:SDB2_Quadrants stopped
2017-03-14-09h43m22-UTC    info masserot     process:SDB2_LC_GTW_adc stopped
2017-03-14-09h43m23-UTC    info masserot     process:SDB2_LC stopped
2017-03-14-09h43m24-UTC    info masserot     process:SDB2_LC_GTW_dac stopped
2017-03-14-09h43m26-UTC    info masserot     process:SDB2_SBE stopped
2017-03-14-09h43m36-UTC    info berni        Configuration (DMSserver) saved
2017-03-14-09h43m41-UTC    info berni        process:DMSserver stopped
2017-03-14-09h43m48-UTC    info masserot     Configuration (SDB2_LC) saved
2017-03-14-09h43m51-UTC    info masserot     process:SDB2_Photodiodes started
2017-03-14-09h43m53-UTC    info masserot     process:SDB2_Quadrants started
2017-03-14-09h43m55-UTC    info masserot     process:SDB2_LC_GTW_adc started
2017-03-14-09h43m59-UTC    info berni        Alert for process:DMSserver cleared
2017-03-14-09h44m00-UTC    info masserot     process:SDB2_LC started
2017-03-14-09h44m01-UTC    info berni        process:DMSserver started
2017-03-14-09h44m03-UTC    info masserot     process:SDB2_LC_GTW_dac started
2017-03-14-09h44m05-UTC    info masserot     process:SDB2_SBE started
2017-03-14-09h45m31-UTC    info rolland      'Reload config' sent to SIB2_dbox_rack
2017-03-14-09h45m38-UTC    info rolland      Alert for process:SIB2_dbox_rack cleared
2017-03-14-09h46m06-UTC    info mwas         Configuration (Vpm) saved
2017-03-14-09h46m10-UTC    info mwas         Configuration reloaded
2017-03-14-09h46m16-UTC    info mwas         process:SDB1_LC stopped
2017-03-14-09h46m20-UTC    info mwas         process:SDB1_LC started
2017-03-14-09h46m23-UTC    info rolland      'Reload config' sent to SIB2_dbox_rack
2017-03-14-09h46m29-UTC    info mwas         'Bench:Angular Loop [-1:Disable,1:Enable=1]' sent to SDB1_LC
2017-03-14-09h46m31-UTC    info masserot     process:EIB_SAS stopped
2017-03-14-09h46m36-UTC    info masserot     process:EIB_SAS started
2017-03-14-09h46m48-UTC    info mwas         process:SDB1_OMC stopped
2017-03-14-09h46m49-UTC    info masserot     process:ENVnoise stopped
2017-03-14-09h46m53-UTC    info mwas         process:SDB1_OMC started
2017-03-14-09h46m54-UTC    info masserot     process:ENVnoise started
2017-03-14-09h47m50-UTC    info masserot     process:EDB_PC stopped
2017-03-14-09h47m54-UTC    info masserot     process:EDB_PC started
2017-03-14-09h48m04-UTC    info masserot     process:EIB_PC stopped
2017-03-14-09h48m09-UTC    info masserot     process:EIB_PC started
2017-03-14-09h48m21-UTC    info masserot     process:ISYSnoise stopped
2017-03-14-09h48m23-UTC    info masserot     process:ISYS_moni stopped
2017-03-14-09h48m25-UTC    info masserot     process:ISYSnoise started
2017-03-14-09h48m28-UTC    info rolland      Configuration (SIB2_dbox_rack) saved
2017-03-14-09h48m29-UTC    info masserot     process:ISYS_moni started
2017-03-14-09h48m42-UTC    info masserot     Alert for process:LSC_Acl cleared
2017-03-14-09h48m47-UTC    info rolland      Reconfigure SIB2_DBOX_LC in SIB2_dbox_rack
2017-03-14-09h48m54-UTC    info rolland      Reconfigure SIB2_DBOX_SBE in SIB2_dbox_rack
2017-03-14-09h49m08-UTC    info rolland      Alert for process:SIB2_dbox_rack cleared
2017-03-14-09h49m09-UTC    info rolland      'Reload config' sent to SIB2_dbox_rack
2017-03-14-09h49m29-UTC    info masserot     process:EDB stopped
2017-03-14-09h49m35-UTC    info masserot     process:EDB started
2017-03-14-09h49m49-UTC    info masserot     process:SIB2_Quadrants stopped
2017-03-14-09h49m54-UTC    info masserot     process:SIB2_Quadrants started
2017-03-14-09h50m40-UTC    info masserot     Configuration (SIB2_LC) saved
2017-03-14-09h50m45-UTC    info masserot     process:SIB2_SBE stopped
2017-03-14-09h50m46-UTC    info masserot     process:SIB2_LC stopped
2017-03-14-09h50m51-UTC    info masserot     process:SIB2_SBE started
2017-03-14-09h50m52-UTC    info masserot     process:SIB2_LC started
 

I repeated the same test, checking the coherence between in and out signal for all other ENV DAC chennels (except for the DetPlatform and DetLab ones which need ADC to be configured on DaqBox). I found the bad coherence in all of them. I report below the single plots: 1. DetEEroom, 2. EEroom, 3. TCSroom, 4. NEB, 5.WEB

The origin of the issue has been found. ENVnoise generates 20 kHz channels and sends them into packets at 5 kHz. The packets are received on other RTPCs by processes running at 10 kHz. Acl had been validated to work in  such a configuration but with the processes on the same RTPC. When the processes are on different RTPCs, the exact time at which the 5 kHz packets are sent depends on the duration of the processing on the different real-time CPUs of the rtpc... so the packet may arrive differently within the cycle on the receiver rtpc. Hence, depending on the machine load, the 4 data of the noise channels are not always read in the correct order.

Acl is being modified to take into account the exact arrival time of the incoming packets. This new version has been put on SIB2_SBE (noise for DAQroom) and it works fine. Some more checks are being done to fix this issue in the coming days.

Today we checked again the performance of the DAC channels assigned to noise hunting after the Acl upgrade. DAC channels of DAQroom, EEroom, DET electronics room, NEB and WEB now work fine (see Figure 1). The DAC channel in TCSRoom still gives troubles: Figure 2. Alain suspects a few missing packets,  and he will check. The DAC channels in DETLab and DetPlatform DAQboxes still need to be tested.

ENV EERoom DAC channel is now: DAQbox 40, mezzanine 26, channel 4