From novaes@fnal.gov Mon Apr 29 11:01:11 2002 Date: Mon, 29 Apr 2002 10:58:29 -0500 From: Sergio F. Novaes To: novaes@fnal.gov Subject: FPD News 26/Apr/02 Hello FPDers, You don't need to keep or print the FPD News messages. A copy of all messages is kept at: http://www-d0.fnal.gov/fpd/FPD_News/news.html Thanks, Sergio PS: Sorry for the delay and, consequently, the long message. __________________________________________________________________________ Dmitri_Denisov Monday, April 08, 2002 "Summer" shutdown Dear Colleagues! At All Experimenters Meeting today Lab plan for "Summer" 2002 shutdown has been announced: 1. Shutdown to install stochastic cooling tanks will start as soon as tanks are ready with duration dictated by time needed to install the tanks. The earliest start date for this shutdown is May 15th, most probable is June 1st. Estimated duration of shutdown is no longer then 2 weeks. 2. 6 weeks shutdown is tentatively scheduled to begin on September 30th, 2002. __________________________________________________________________________ Michael Strang Tuesday, April 09, 2002 draft of rate watcher algorithm, please comment The only HV names I have are from the hardware database: Trigger: (these are the things we will want to monitor) FPD_HVC_PMTA1 -> voltage to A1U, A1D, A1I, A1O FPD_HVC_PMTA2 -> voltage to A2U, A2D, A2I, A2O FPD_HVC_PMTDI -> voltage to DI1, DI2 FPD_HVC_PMTP1 -> voltage to P1U, P1D, P1I, P1O FPD_HVC_PMTP2 -> voltage to P2U, P2D, P2I, P2O FPD_HVC_SPARE Veto: FPD_HVC_VETON FPD_HVC_VETOS MAPMTs: (A1I and A1O were "borrowed" for the VETON and VETOS) FPD_HVC_MAA1D FPD_HVC_MAA1U FPD_HVC_MAA2D FPD_HVC_MAA2I FPD_HVC_MAA2O FPD_HVC_MAA2U FPD_HVC_MADI1 FPD_HVC_MADI2 FPD_HVC_MAP1D FPD_HVC_MAP1I FPD_HVC_MAP1O FPD_HVC_MAP1U FPD_HVC_MAP2D FPD_HVC_MAP2I FPD_HVC_MAP2O FPD_HVC_MAP2U __________________________________________________________________________ Mike Martens Wednesday, April 10, 2002 Thursday access//vacuum work Doug Allen and I looked over the drawings for the pots and will try to leak check the internal bellows by removing the small plate over the worm gear view port. This is a simple task which does not involve dis-assembling the pot arms yet still lets helium into the region of the internal bellows. I have looked at photographs of the detectors and the view port plate appear very accessable even with the detectors in place as long as the pots are in the out position. I think that the view port plate is also accessable when the detector is in the IN position as well, but to be sure we should leave the pots in the out position until Doug has finished leak checking. Just in case there is trouble, will there be someone from FPD in the tunnel Thursday morning in case we need to remove the detectors? I will be on shift for the Tev until midnight tonight so I won't be in early in the morning. Whatever we find we plan on leaving the A2 pot in place for now. We may or may not have to remove it during the May shutdown depending on the type of leak we find. Doug has already thoroughly checked the conflat connections between the castle and BPMs and beam valves, the vacuum feedthroughs on the BPMs, the conflat flanges between the pot arm assembly and the castle, and the pot tips on the outer and upper pot arms. As candidates for leaks this leaves the pot tips on the inner or lower pot arms, the internal bellows on one of the pot arms, or the weld connecting the bellows to the flanges. __________________________________________________________________________ Andrew Brandt Thursday, April 11, 2002 Curious about what's going on in FPD operations? See Sergio's logbook instructions for how to get access to the electronic logbook: http://d0server1.fnal.gov/users/novaes/logbook.html If you are a shifter not using the logbook, see this page for instructions on how to use it. It's easy and it should be being used every shift. __________________________________________________________________________ Mike Martens Thursday, April 11, 2002 Vacuum leak in the A2 (GENTIL) pot. A vacuum leak has developed in the A2/GENTIL FPD pot region. Presently there is no urgency to remove the pot but we should consider making a repair during the 2 week shutdown. Apparently the leak begin near October 8th, 2001 which coincided with the start of a 6 week shutdown of the Tevatron. On Oct 8th the vacuum reading on the ion pump attached to the A2 pot jumped from a steady value of 8E-12 Torr to about 1E-10 Torr and has been steadly climbing to its present value of about 2E-9 Torr. (See attached plot of the ion pump vaccum reading over the past 52 weeks. The ion gauge in this region has failed so we do not have a vacuum measurement from the ion gauge.) During the period since October the titanium sublimation pumps in the adjacent separator have had to be fired in order to keep the separator vacuum at a decent level. Several weeks ago the beam valves between the separators and the A2 pot region were closed and the separator and pot vacuum were isolated. During this test the vacuum in the separator region improved and the vacuum in the pot region deteriorated confirming that the leak was somewhere in the pot region. An initial leak check of the pot region, which includes a BPM, found no leaks on any of the beampipe conflat flanges, no leaks on the flanges connecting the pot arms to the pot castle, and no leaks on the upper or inner pot tips. Today we leaked checked the internal bellows of the upper arm and found a leak. The small, worm gear view port window was removed and helium sprayed into the internal section of the upper pot arm. The leak detector showed a rise in the He pressure from 2E-8 Torr to 1E-6 Torr giving a clear indication of a leak. This means that there is a leak in the internal bellows or the welds connecting the bellows to the flanges. We will test the bellows on the other three pot arms this afternoon. It seems prudent at this point to avoid moving the A2 upper pot any further until some sort of repair/replacement can be made during the May shutdown. The fear is that a small crack has developed in the bellows or the weld which could become worse with continued flexing of the bellows. For now the vacuum in the adjacent electrostatic separators is fine, but if the vacuum deteriorates it could lead to sparking of the separators. During the May shutdown we will attempt some sort of repair to the pot. One suggestion is to remove the upper pot arm, move the inner pot arm to the upper position on the castle, and blank off the inner port of the castle. This leaves us without a inner arm but restores our vertical spectrometer. With the upper pot arm removed we could inspect, test, and hopefuly repair the upper pot arm. I will put in a work request to have this repair work done during the May shutdown so we can get the help we need (i.e. Doug Allen and company) to perform the vacuum repairs. Perhaps Helio and LNLS would have some insight or advice for us on this problem. One thing I am worried about is trying to safely remove a pot arm in the horizontal position without it "dropping" and crusshing the bellows. We will have to look into this operation very carefully to make sure it can be done safely. __________________________________________________________________________ Pierrick Hanlet Friday, April 12, 2002 access summary Today had mixed results: 1) Victor completed filling A/S blanks, and adding missing A/S cards to A1, A2, and D. 2) Christophe and Jorge completed the cable (and other stuff) removal and measurements for the MUX testing. 3) Doug Allen found the leak; it is in A2U. All of A2 was checked. 4) Tim ? (Control's Group) moved D cameras. They also fixed all lights and tested all other FPD detector cameras. However, by late morning, we were left with two problems: 1) A1 A/S LVPS was not working - no +-15V and no +-12V (I had forgotten about this problem, and so didn't have it on the list) 2) For some reason, the smoke detector was no longer working in A1; by this I mean that Baretto called me in the tunnel to announce that he was getting smoke detector errors in A1, and he couldn't reset them. There were no lights on the RMI indicating this, but we could never turn on the A1 A/S power. In the end, it was the RMI that wasn't putting out the requisit 20V for the smoke detectors (I measured this in several ways). We swapped the RMI, and all works now. At A1, I swapped an input fuse on the 15V supply (it was the wrong value). However, we could still never get 12V to work; all fuses were ok. In the end, Victor and I pulled the supply out of the tunnel for repairs. We also had a scare with the pot motion. Christophe and Jorge were testing and found A1 was not working. If I understood them correctly, the end result was that a cable was loose. When I was looking at it earlier, I noticed that there are no strain reliefs on these cables; we need to fix that in the May/June access. I had apparently knocked the cable loose when I was rerouting an extention cord. Lastly, we did have the ORC review. We did not pass. Everything was up to snuff on the documentation, and all systems in the inspection, except for using extension cords to power our racks. This is disconcerting as some of them existed before the fall tunnel cleanup, and they passed the inspection then. I had also gotten permission (I don't recall from whom) to run the extension cords from the Q2 to Q1 stations. In any case, Dan Wolff will get back to me with recommendations. Also, the Diple system needs covers on the backs of the racks. For documentation purposes, each quadrupole A/S crate has 14 A/S cards, but only 7 are completely installed for the U and D detectors. The remaining ones are there for air flow purposes. They are not fully installed so as to minimize the use of the LVPSs. However, for the sake of excersizing and testing the LVPS, Victor convinced me to put all A/S cards in A2. Also for documentation purpose, the A1D cables are jumpered from the MAPMTs to the ribbon cables which end in PW08. (This implies a swapping of channels from normal configuration). Next CH access we should test the signals with the LMB system; we can later compare them to a system with A/S's. So, here's our new to do list for future tunnel accesses: 1) repair and replace A1 LVPS 2) repair leak to A2 castle (May/June shutdown) 3) install strain reliefs for pot motion cables 4) install back panels on D racks 5) install AU and PU detectors 6) fix AC distribution problems for all 4 Q stations (May/June shutdown) 7) install MUX and Control devices (May/June shutdown) 8) After testing jumpered cables in A1D, install them in A/S crate Anything else? Thanks to Bodyagin, Molina, Baretto, and Royon for tasks well done, __________________________________________________________________________ Helio da Motta Filho Friday, April 12, 2002 internal bellows I have just talked with Marcelo Juni about the problem with Gentil. First, it is important to know if Gentil is one of the overbaked castles. If it is, this leak is probably consequence of this. (It would be sort of expected). The use of the pot is far from reach the material fatigue (unless the case of overbaking did affected the pot). The fact of the vacuum being external shall not be a problem, but it is a possibility (all this, of course , if the leak is due to a problem in the bellows). The incident you report in the other message shall not be a problem, since hitting the mechanical stop prevents the bellows from beeing overexpanded. Now, how to replace the bellows if we have to: a) no extra part was built due to the lack of money for that. b) replace the bellows means to built a new piece consisting of a new bellows attached to two flanges. The flanges are easy to built. The bellows must be purchased. The system now in the prototype castle shall not be used. c) swapping an horizontal pot and the top one is feasible. The operation must be done with care and someone who knows the system should at least supervise it. Of course this will change the survey completely. I feel a little bit unconfortable with the idea. I think that as much one mess up with things bigger the chances of running into more problems. My first instinct would be not to make changes in things that are working. However, if we can not go for some time without the upper pot, then we must go for the operation (it should really be well planned including some meetings - Gosh, I never thought I would propose meetings !!) d) Once the assembly piece is removed, test it for leaks is not a problem. Installing the new bellows, if this is the case, and all other works must be done with all the care to prevent the inner parts of the assembly to get dirty. Of course must be sure that the leak is due to bellows fatigue. If it is a flange or a weld it can be fixed (with all the care for the internal welds). It seems unlikely, though, that a weld would suddenly start leaking. The bellows would be the first suspect. __________________________________________________________________________ Pierrick Hanlet Tuesday, April 16, 2002 Today, Molina and I will set the SLP timing for the A1 and A2 MAPMTs, then run the agreed upon LMB tests. The A1 LVPS +-12V supply is dead. Victor and I will make an access into the tunnel and put the D LVPS into the A1 position. Since the LMB tests for the A1 will have been completed, we will then connect the MAPMT signal cables back to the A/S crate for A1D. Afterwards, Jorge can run the LMB tests again. As far as I understand, all of the A1 and A2 A/S cards for Up and Down are bench tested and have their pedestals correctly adjusted. The remaining 7 A/S cards in A1 and and A2 are for the In and Out, and are not presently critical. The 7 D A/S cards still need adjusting and bench testing, so since there will be no LVPS in this crate, we will remove these A/S cards for Victor. The other possible jobs in the tunnel will be: 1) re-route P side extension cord 2) install standoffs and strain relief for A1 pot motion cables 3) check all other cables for strain reliefs Before we leave the tunnel, we would like Sergio and/or Jorge Baretto to test the pot motion of all pots. We will leave the MAPMT jumper cables in PW08 on the AD spectrometer, since we will start using it with the EL_AD-PU trigger for data taking. This is assuming that our tests show that things work. If time permits, Ricardo (and I) will run the grey cables and install the AFE backplane (the latter to be done with Pat Sheehan). __________________________________________________________________________ Victor Bodyagin Thursday, April 18, 2002 morning access results Pierrick and me got into the tunnel this morning. As you remember yesterday we discovered a trouble with the 7th A/S board. The trouble was that all 16 channels of this A/S went down. It appeared as zero pedestals in all ADC's channels. We had to replace the board. Jeorge Barreto was helpful with testing the results. Seemingly one of the +/-7 volts regulators is dead. So the output current offset of the board was too big, and pedestals could move either too high either to perfect zero. __________________________________________________________________________ Pierrick Hanlet Thursday, April 18, 2002 more TeV access stuff In addition to the work that Victor did, I found new doors for A1 and A2 CTL racks. This freed the ones previously used for the D racks. All are installed, thought the one at A2 needs to have a pin removed so that it sits in its slot correctly. This leaves: 1) installing detectors 2) fixing the vacuum leak 3) replacing the D LVPS 4) fixing A2 CTL back door mounting 5) fixing the AC distribution 6) installing MUX and Control devices as projects in the tunnel for the final 10 pot configuration. We also checked out all of the system. There are appropriate strain reliefs for ALL IIB cables. All appeared to be running correctly. Barreto checked pot motion and A/S with Victor. __________________________________________________________________________ Pierrick Hanlet Sunday, April 21, 2002 trigger and daq status Today Jorge completed the MAPMT signal cable extensions. Victor fixed the daq. The problem seemed to have been a hardware problem. In cycling the power on d0olctl02, all came back up again correctly. We are still using the terminal server to connect to d0olctl02. I completed most of the timing issues. We are now using 10ns signals gated by clocks of 18ns. This entailed adding delays to all of the Q1 signals to time them in with the Q2 signals (they were all adjusted to within 1-2ns). Doing this, however, creates a larger gap between Q2 and Q1 in the early signal, so I added 54ns of delay to the Q2 signals into the early octal discriminator. Both the LM and Veto counters are correctly timed in to form a VETO for the elastic triggers. There remains a mystery as to why the EOR signals (halo particles) are 26ns out of time wrt the ECLK, so we are presently without EOR signals in the VETO. I also timed in the TDC signals such that the Q2 signal arrive 27ns after the Q1 signals. Though I haven't yet figured out why our EOR terms are out of time, we can now apply an offline TDC cut to remove any halo particles. Even without the EOR terms are presently not in the elastic triggers, the narrowing of the CLK gates has dramatically improved our rates: at D0 L=7.6E30 the rate for EL_AD-PU is 15.11Hz! The explanation for this result is that when the Q1 and Q2 signals were not in time upon entering the discriminator (they were ~27ns apart due to TOF), we used wide signals (60ns) to get a good overlap in the coincidence unit. We used 70ns CLK and ECLK gates, which corresponds to 3 RF buckets. Since when filling the buckets there is some spill, as well as DC beam (not in a bucket), we had significant out of time signals which made acceptable coincidences. Narrowing the CLK and ECLK gates to 18ns removes most of the overlapping particles. Lastly, the MAPMT signals are now timed into the ADC gate. So, we're ready to take data. There are still a few things to do, but these will be non-disruptive to the trigger setup. I will follow up in the next couple of days with updated documentation as well as fixing the EOR problem. __________________________________________________________________________ Andrew Brandt Monday, April 22, 2002 Halo task force meetings The following comments are in relation to the CDF background. Mike Albrow had a meeting on Thursday afternoon and subsequently Dean still and I had a meeting with Nikolai on Friday. 1. Nikolai pointed out that there is a scraper at D48? That has tungsten targets front and rear and is steel in between. Right now there is a .5 cm tungsten "target" that is used to start the scraping process. An experiment should be made to try the other scraper. Some general comments: 1. The present scraper system with a target at D49 and two scrapings at E0 and F17 is very clever and designed to clean up halo close to the beam. This will be important for the Roman Pots. However, the present problem is not what the scraper was designed to cope with! We are loosing beam from the buckets which then causes all kinds of troubles. The cure for this will come from fixing the looses. When that happens, the problem will become one that the scrapers were designed to solve. 2. At present, I note that F17 is almost 270 degrees in phase away from where beta max occurs at the CDF low beta quads. Thus a severely off momentum/off axis particle at F17 is sure to hit the low beta quads and cause lots of hell at CDF. There is no protection between F17 and CDF. That this is causing trouble is verified by the way the scraping is set up. Dean first runs in F17 till he he sees the LostP rate increase at CDF. This shows directly that stuff from there can reach CDF. The situation is not much better for E0 scraper. 3. The bent Si crystal target at D49 could be very important for future running with the Roman pots, but I don't think it is very pertinent to the present problem. I would suggest that we consider mounting an experiment since the equipment seems to be available. alvin Alvin Tollestrup Tel: 630-840-4331 TelCell: 630-881-4881 MS: 221 __________________________________________________________________________ Michael Strang [strang@fnal.gov] Sent: Monday, April 22, 2002 4:34 PM To: D0FPD_LOCAL@listserv.fnal.gov Subject: [D0FPD_LOCAL] information to keep track of during a shift During a shift please keep track of the following information: 1) Before pots are inserted, send an email with D0 Lum, A-Halo, P-Halo, P-Intensity, A-Intensity and store number. 2) In the log book, write down the voltages for all tubes and final pot locations for all pots inserted. 3) Make a log book entry for each run including the full name of the ped file for that run. 4) If any condition changes between runs (voltages, positions, ped files) make a note in the elog. 5) If you notice a hot channel, try to determine what channel it is (either in terms of ADC module and channel or in terms of detector plane channel number) and make a note in the elog. __________________________________________________________________________ Ricardo Ramirez Tuesday, April 23, 2002 FPD protocols - Proposal You can find the latest version of the CTT protocols in http://d0server1.fnal.gov/projects/VHDL/General/ They are interesting for us in order to agree with the D0 standard. By the moment I would like to suggest you to take a look at pages 5, 11 and 25, in which you will find the AFE-DFE L1CTT (CTOC) and the DFE-TM (CTTT) standard communication protocols. __________________________________________________________________________ Michael Strang Wednesday, April 24, 2002 ADC and TDC plots Pierrick, Victor and I were looking at the ADC and TDC plots for the data coming from the AD detectors and we noticed some strange things. First, for the ADCs, we see counts below the pedestal for many channels. Victor says this may be due to timing problems. Also, the ADC plots for A1D are different than the plots for A2D (the A2D plots look like the plots we got from the P side) Also the TDCs are odd looking. You can see the plots from Victor to get a better idea. All of the data has been unpacked if you want to look at it yourself. __________________________________________________________________________ Victor Bodyagin Thursday, April 25, 2002 FPD saga updates Right after morning meetings Jeorge Molina and me came to D0CR and SCR to check what was going on with LVPSs and other things. Yes, LVPS control was reporting a problem with A1 -12v power supply. Moreover, having started DAQ we saw that nothing resonable was coming from FERA ADCs. In SCR we've seen that CAMAC crate was as hot as an iron, because somebody has turned 9-unit fan cooling panel OFF. The crate was really hot and I got into panic. My initial guess was that due to overheating all ADCs went down. Jeorge and me immediately went to PREP and took a new set of 14 FERAs. Having returned to SCR and replaced two FERAs we discovered that it did not help. Finally we guessed that DAQ was running with pedestals being subtracted. We got back to ZERO_PEDESTALS and everything was OK. So, nothing bad has happened with ADCs despite our misbehaviour. This was not the last good news. The other good news is that A1 power supply is OK. LVPS control message is wrong, seemingly paricular this analog input does not work properly. Moreover, I'turned MAPMTS' HV on and saw normal signals from A1. But up to this moment our luck was over. I decided to take some data at home position. Having started DAQ I found with sadness that -12v A2 power supply died. It died literally on my eyes. Now our status is: 1. FERAs are OK. 2. A1 is OK excepting the false messages. 3. A2 LVPS -12v is down and the messages are correct. __________________________________________________________________________ Sergio F. Novaes Fermilab E-mail: novaes@fnal.gov P.O. Box 500 Phone: (630) 840.2454 Mail Station 352 Cell: (630) 272.1819 Batavia, IL 60510-0500 Fax: (630) 840.8886 USA __________________________________________________________________________