Intro: (top)

    We have a T1524/fps118V  V1 transmitter. It was part of the over the horizon radar system at moscow main.
(The xmiter part number faceplate is PN52151assy126101-1 ( -2 would be version 2. see page 854 1-10 table 1.3 of manual).   
Page to keep track of heating facility (hf)  calibration.
Tx Names: 1 through 6.

• 5.Mhz dipoles tx1-6 =  D1a,D1b,D3a,D3b,D5a,D5A
• 8 Mhz dipoles tx1-6 = D2a,D2b,D4a,D4b,D6a,D6b

• get each transmitter outputting to 100Kw on dummy load
• make sure we can protect the receivers in dome, ch when hf is transmitting.
  

Daily Measurements: (top)

• 150504:
• dummy load tx4:
• monitor cable was hooked up incorrectly on coupler.
  
•  Saw broad band glitches on dummy load
• strong harmonics around 128.125 Mhz. go away when rf drive turned off
• banging on solid state amp caused power levels to jump
• tx1,2 on antenna. 327 rcvr results:
• looked like no power output from tx2
  
• 5.125 harmonics seen at 327. tx1
• 150506
• rf monitor cable hooked up correctly (but still used spare cable)
  
• dummy load tx4
• monitor cable hooked up correctly (still used spare cable)
• 327 rcvr sees hv on, does not see 5.125 harmonics at full power into dummy load.
  
• Measured the fundamental and up to 6*F0 on the spectrum analzyer at drive=3.05 dbm
• looked at 0 to 500 Mhz with hipass filter in. 128.125 strongest harmonic.
• Tx1,Tx2 on antenna (327 rcvr)
  
• tx2: no power seen.
  
• tx1:
• 5.125 harmonics seen at -20dbm rf drive
• broad band glitches seen at -15dbm rf drive
• 1 sec spectra: 10 times Tsys
• 1 millisec spectra: 100 times Tsys
• 327 recovers after glitch in < 1 millisec
• some glitches not resolved in 1 millisecond spectra
• No N*60Hz seen in the spectra of the total power.
• 150507:
• dummy load tx1,tx4
• new monitor cables installed
  
• calibrate the spectrum analyzer and rf drive values using the dummy load.
• tx1 1 measurement (probably bad) then burned resistor in dummy load tx 1
• calorimeter measurement tx 4
• 150508:
• dummy load
• tx2 dummy load moved to tx1. Burned resistor in tx1 dummy load.
• dummy load tx3: output temp meters differed by +3C .. stop testing
• redo calorimeter measurement tx 4
• 150512
• dummy load
• prior to tests, replaced output temp sensor on tx dummy load 1 and 4.
• they had different values from the other 2 when no power applied
  
• water temp on dummy load stable 29 to 31 degC for entire run
  
• tx4 calibration
• got full curve, good fit 
  
• plots of curve and fit:   tx4 calibration curve with fit (.ps) (.pdf)
• tx3 calibration:
• only go 5 measurements then hv drip at 1,.5 dbm drive
• tx 3 Calorimeter power vs coupler power and the linear fit (.ps) (.pdf)
• tx1 calibration
• 4 measurements then powerdip, ran out of time.
• dana took 1 ghz bw timedomain samples of a burst with tek scope
  
• 150513
• tx4 on antenna:
• brought up to rf drive:-3.2, fwd meter:18kw, reflct:1kw then tripped
• tripped at this level reliably.
  
• 150514
• tx4 on antenna
  
• added 1db,2db atttn to  reflected power to forestall trips
• Got to 30kw, rfdrive:-1.3 dbm then tripped on reflected power
• measured refl pwr: 1kw (actual 2db higher)
• it is tripping on 1kw measured reflected power.
  
• tx6 on antenna:
• drive to +2dbm, rf meter:32kw, reverse power measured 0. no tripping
• saw harmonics to 40*Tsys(in 1 channel), and occasional spectra with ripples of 1 Mhz to 6*Tsys
• the harmonics did not seem to increase with output power.
  
• 150515:
• try to adjust the reverse power trip for Tx4 
• 150518
• adjust tx 3 meters
• 150519
• adjust tx 4,5,6 meters
• tx 6 AT9 attenuator bad, replaced with 3 other
• tx5: power meter bad. Replaced with spare. Readout scale not linear
• 150520:
• adjust meters tx1,tx2
• tx1 +15 volt supply card a2 has 3Volt pkToPk ripple. Dc voltage 16.4 volts.. need to fix
• tx2: 26volt supply ps1. no output. not band select. need to replace.
• 150521:
• adjust meters tx2
• the 26 volt supply was working today. So we calibrated the meters for tx2
• 150522:
• tx5 pwr meter. fixed the old one and reinstalled. so scale is now correct
• re calibrated meters for tx5
• dummy load tx6,tx5
• did dummy load calibration for tx6
• So no broad band bursts on dummy load.
  
• tx 5. did not get much power out . looks like drive stage way low.
• ran  80KW into dummy load tx6. monitored coupler 0 to 500 Mhz saw no broad band bursts.
• 150525
• 50db rf coupler. Dana remeasured for 0 to 2000 Mhz (.ps) (.pdf)
  
• 150526
• tx1:
  
• replaced bad 15 volt supply
•  calibrate tx1 meters
• tx6: on antenna. record spectra with alfa bm 0 to 20KW. saw occasional 1MHz  period spectra.
• tx3: on dummy load.
• brought up to 72 kw.
• front panel meters did not agree with calorimeter.
• 4 V pkToPk ac ripple on 15 volt line, dc value 10 volts. bad supply
• seemed to get better as things cooled off.
• 150527:
• tx3: measure loss in cable from 60db coupler to control room (1.22 db). also measured the delay
• computed xmitted pwr vs coupler output with and without cable loss (see 150527 below)
• need to refit dummy load data including thee 1.22 db cable loss.
  
• tx6 on antenna
• took data with alfa bm 0 while xmitting
• az,za 360,15.. xmitted to 100kw.occasional bursts
• stw show 1,2 usec peaks.
  
• az,za 240,15 .. xmitted 100kw . occasional bursts (stronger than az=360)
• but this may have been lightning. 1,2 usec stw peaks not obvious.
• 150602:
• tx1 on antenna. to 100KW
• take data with alfa bm0. 9.5 minutes
  
• measured rf levels at visitor center, and at rim road.
  
• tx4: won't come out of filament delay state
• tx6: on antenna,  to 100 KW
• take data with alfa bm 0 for
  
• move to 430 dome, measure harmonics, for about 5 minutes
  
• 150603:
• tx4 on antenna
• trips at 25 then 50 KW, reflected power.
• lots of bursts
  
• looked at front panel ports with scope
• had arcing at top of ant for tx4
• tx3 on antenna
• to 100 KW. no bursts
• 150610
• tx5 on dummy load
• at 70KW tripped
• saw driver anode oscillate on scope before trip. alfredo things this is a loose cable.
• tx2 on dummy load
• at 45-50kw tripped . drvr A ac flt
• saw jump on ipa output on scope at trip.
• broadband amp showed no jump
• . but needed more diagnosing...
  
• 150611
• tx4 on dummy load.
• saw no bursts. ran  60 to 90 kw for about 1 hour
• got trips drvA Ac fault at 95kw (3dbm drive, then later at 2.7 dbm drive)
• 150612: dana  used fd/tdr to measure tx3b coax after arc repair.(.pdf)
• 150616:
  
• tx5
• driver tube capacitance bad.. replaced outer sleeve
  
• capacitor on pa output replaced.. only had about 160kohms. Wouldn't hold 5KV (normal working env has 13kv across it).
• 150618: first light daytime: tx1,3,6 (dipoles 1a,3a,5a)
  
• we ran tx 1,3,6 to the tangential dipoles: 1a,3a,5a starting  around 17:07 ast
• power level was about 60KW in each tx 
• 150619:
• measure the relative phase of the signal through the 3 transmitters
• first light (nighttime 22:30 to 24:00 hrs). with 3 antennas
• 150623:tx 4 on dummy load
• dana measured S11 tx4 heliax to dipole 3b. saw broad reflection (40ns) at far end.
• 23jun15: ft/tdr measurement of tx3b after arc repair. shows change in reflection (.pdf)
  
• ran tx 4 dummy load, measured power, calorimeter
• recorded values (including pa cathode current) in table.
  
• drive power needed was about 4db higher than previous.
  
• tripped with PA Amp DC around  87KW.
• After trip, drive power needed was 3.5 db less than before trip?
• Eventually found connectors on back of broadband amp were loose. AFter a trip, the  vibration probably changed the power getting into the next tubes...
  
• 150629: tx2, tx4 on dummy load
• tx2
• trip drv AC Amps at 3dbm drv. coupler: 9.5dbm +1db
  
• adjusted biases.. then brought back up. tripped at same drive level
• Then standby -> hvOn   standby light wouldn't go off
• tx4 dummy load
• did calorimeter checks. got to 17dbm coupler value.
  
• At 17dbm coupler output tended to wander 16.6 -> 17.0
• 150820: measured heliax 3b after further repair.
• removed damaged heliax. cut it at bottom of balun and then spliced in a new piece.
• 150909: tx 5 on dummy load.
  
• couldn't get more that 15kw out. ipa filament voltage was 4.5 volts instead of 7.4.
  
• Broadband amp increased to saturation, but ipa up didn't increase much..
• 150915: 2,4,6 on antenna. arcing in heliax tx3b transmitter 4.
  
• monitor with ch and interim correlator
• bring 2,4,6 up to 60 kw .. saw glitches
• 2 up , 4 down, bring up 6, then increase 2 a little to be at 60kw
• start bringing up 4 , trips at -7 dac drive, this was arcing..
• pictures of arc:
• pictures of inside of balun showing arc was between outside of heliax and the inside of the balun.
  
• bring back up 4, 2, 6 way down
• dac drive 4 -10dbm.. specana says 8kw.. before it -10 was about 25-30 kw.
  
• bring down 4, run 2,6 at 75  kw for about 20 minutes.
• ch mon looks like 4 was arcing..
• 150918: 5,4 on dummy load
• calibrate 5 on dummy load
• calibrate 4 on dummy load
• 1/2 way through bringing 4 up (specan-10.73 22kw), gain jumped up by 4db .. and stayed there
• ran 5 for 10 minutes at 90kw
• ran 4 for 5 minutes at 90kw 
• 150922: heliax from tx 4 repaired. dana does vna measurements
• vna measurement of repaired heliax (.pdf)
• dana's email comments of plot (.txt)
• 150928: high pot test of balun section.
• we did a high pot test of a 5 foot section of balun. The setup was:
• outer conductor of heliax with spacers, pushed into section balun (see pictures)
• we used the sband high pot
• we first used a new 5 foot piece of balun, we then redid the tests with the 5 foot section that had the arc in it (see pictures). We then repeated the test with the damaged section after wetting it.
• The results were:

• section	arcing Voltage	notes
  new	20 to 25kv	arcs at spacer closest to front of section
  damaged	23 kv	heard buzzing noise start around 18kv (probably corona from damaged spot) arc did not start at the front most spacer. It probably started at the bad spot
  damaged wet	10kv	with a spray bottle, we sprayed a small amount of mist on the front most spacer. high pot current started immediately, arcing at 1st spacer started at 10kv.

•  
• We also measured about 23Mohms between the outside of the balun and the heliax outer connector (a second measurement gave overload on the fluke meter).
• Summary:
• the burned spot caused corona around 15kv
• A good section holds about 20 to 25Kv before arcing
• the arcs occur at the frontmost spacer
• Arcing is very sensitive to a small amount of water..
• dana never found the gremlins inside the balun...the spacers got in his way....
• We should compare these values with what we get when the entire helix is in place inside the balun on tower 3b (before and after connecting the dipole. We can do this from the splice.
• 150929: measure resistance between balun and outer conductor of heliax.
• The heliax was reinstalled in balun for dipole 3B.
• The short to the heliax outer conductor was removed as well as the top of the heliax was disconnected.
• We measured the resistance between the outer part of the heliax with the balun.
• The resistance was open (infinite) and then jumped to .6 ohms when we moved the bottom portion of the heliax.
• labeling the balun 5 foot sections 1.. 8 counts from the top:
• we removed section 7 and 8.
• At the top of section 7 we saw that the heliax was bowed.
  
• The spacers had slid on the heliax when the heliax was installed in the balun.
• We put more spacers at this point and then reinstalled the balun. The resistance went to infinity.. But we didn't know if the heliax was still bowed, only that it wasn't touching the balun.
• 150930: time domain measurements of balun, heliax. (also see 150930 pictures)
  
• The vector network analyzer (vna) was used to check if the heliax was well centered in the balun. If the heliax had a bow in it,  we should see it on the return echo on the time domain measurement.
  
• setup:
• vna at ground level, 70 foot cable to adaptor that connected to the outer ring of heliax at top of balun.
• bottom of balun was shorted to the outer ring of the heliax.
• The time domain measurement showed bumps along the balun. The first was about 20 feet from the top.
• Results of measurement:
  
• We decided that we needed to remove all sections of the balun and then reinstall them 1 piece at at time rather than pushing the heliax (with spacers) through the entire balun.
• The sliding of the spacers is probably what caused the arc in 15sep15. All the other baluns were installed 1 five foot section at a time about the heliax.  After tx3B burned, the spacers were shaved to make them slide easier and the heliax was pushed through the entire balun after it was assembled. We are now going back to installing one piece of balun at a time about the heliax. 
• See pictures of sma cable to heliax outer connector at top of balun.
• 151008: tx4 on antenna 3B
• no power output of PA.
• Trouble turned out to be pneumatic switch ..
  
• 151013: tx4 on antenna 3B
• transmitted 100KW for 30 minutes with no problems or noticeable glitches on ch rcvr
• at vip overlook by bowl, measured 16 millW/cm^2 .. above limit..
• 151015: try all 6 tx's on antenna
• we tried putting all 6 transmitters on antenna.
• We were aiming for about 60KW (coupler = 15dbm).
• For power monitoring
•  we plugged all 6 forward power coupler signals into the rf mux (with a 10db bad between cable an mux input).
• we setup the anritsu to have a 10.5 external attenuation offset (the mux has about .5 db loss).
  
• tx 3 and 5 didn't come up correctly
• The plots show the log from the synthesizer drive to the transmitters (.ps) (.pdf)
• Top: amplitude . scale is db below full scale drive (of synbox .. not transmitters).
• middle: 1 if rf was on, 0 --> rf is off
• bottom: phase for each transmitter.
• the * show where buttons were pushed.
• Looks like we accidentally sent maximum output to tx6 (yes that red color is tx6.. i checked) and then turned on the rf.
  
• hopefully the broad band amp was off when this happened...
• plots of synth  log with ch 1 power overplotted (.ps) (.pdf)
• same as above, but include the total power measured from ch receiver
• 10 mhz totpwr, 1 sec samples, db  above Tsys.
• Each page blows up in time a section of the original plot
• the last page shows how changed tx3 from -59.5 to -60 dbf caused the rfi to go away.
  
• 
  
• 151019: tx3,tx5 vswr on dipoles at 5.125Mhz measured by dana

• date	tx3	tx5
  19oct15	1.61: 1	1.56:1
  22apr15	1.43:1	1.42:11

•  the tdr plots looked clean
• 151020: tx3,tx5 on dummy load
• tx5 crowbarred when hv on. replaced crowbar light
• tx5 tripped when drive -25dbf
• tx3: bias was low (-220) . took data, then brought bias up to -380
• output power was 7 to 10 db less than other tx's at same drive.
• not enough drive to get to full power.
• 151021: tx3 balun measure, tx3,tx5 on dummy load, dbf to dbm synth driver
  
• vna balun measurement on tx3
• measured synth driver output dbFullscale vs dbm output
• front panel cable -> bncCable -> anritsu.
  
• Did not include the 1db loss of cable from control room to transmitter rf input
  
• tx3 on dummy load...
• got to 60KWatts then driver screen trip, and then water leak.
• tx5: on dummy load
• got to 90KWatts sat there for 30 minutes. drive levelneeded -3.6dbf
• recorded tube  stage output on scope.
• 151022:  tx5,tx1,tx6 on ant. tx2,tx6 trouble
• we put tx1, tx5, and tx4 on antenna around 85 Kwatts
• tx6 had grid current readings ipa,driver that were pegged (turns out it was a meter problem)
• tx2 came up to about 40Kw and then tripped on driver AC a. did this multiple times at the same drive level
• for some reason tx 1,2,5 had the paBias at -200 volts instead of -380.
• we adjusted them to -300.
• plots of synth driver log for the day (.ps) (.pdf)
• 151026: record meters at no rf drive, tx3 dmy ld, tx5 crowbar
  
• measured meter values with no rf drive tx1,2,3,4,6
• tx1 pa filament 8v/75amps
• all others 13.5-14V/ 175-250 amps
  
• tx5  on with no rf for about 30 min. then crowbar, then repeatedly crowbar on every  hv on
• tx3 dummy load.
• saw glitch in ipa anode output, broadband was clean.
  
• -2db only gave 40kw. broad band was > 250 Watts..
• lowering drv to -5dbf reduced the number of  glitches.. but still there.'
• 151027: tx3 on dummy load
• ipa
• Still had glitches
• 200 watts bb, but only 50kw out.
• ipa cathode meter read 0 amps and 50 KW.
• recorded meter values at -3dbf drive
• 151028: tx3 on dummy load, tx5 sky , no rf
  
• tx3:
• changed ipa tube (from 151027)
• about 30 minutes at  88KW only saw a few glitches,.
• tx5: hvon, on sky, no rf drive. recorded meter values with no rf.
• 151029: tx 1 thru 6 on antenna..for 2 hours at 80KW (approx.)
• 151102: tx 1-6 on sky for 8hrs. calibrate rcp,lcp phase, az,za swings, monitor ports.
• 151105: tx 1-6 on sky @ nite. ran with 430, but fof2 too low, looked at bursts
• 151106: tx1-6 on sky afternoon. ran with 430. got plasmalines
• see movie here...
  

General measurements/tests:  (top)

• 150512: tx4 on load. 1200 Mhz spectra of noise  burst (using tek scope) (.pdf)
  
• 150515: transmitter gain (rfCoupler/rfDrive) (.ps) (.pdf)
• all data through 150514.
  
• processing: x101/hf/process/txgain.pro
• 150518,19,20: adjusting the meters
• 150527: output power (at 5.125 Mhz)  vs 60db rf coupler (using coupling constant) (.ps) (.pdf)
• Page 1 includes cable loss (for measurements in control room)
• Page 2: is curve for measurements at the coupler (no cable loss)
• 150618: meter values when running at 100 KW
• operating values template
  

Measured rf levels  (top)

    We measured the 5.125 Mhz radiation levels using the narda monitor (get model #) while testing the hf transmitters.
These were single transmitter tests.

Table of maximum permissible exposure (mpe)

    from:Evaluating Compliance with FCC guidelines for human exposure to RadioFrequency
Electromagnetic files (OET bulletin 65 edition 97-1 (aug 1997)

Notes:

• f  = frequency in Mhz
• *  Plane-wave equivalent power density

Dummy load calibration measurements and curves (top)

    Linear fits were done to the calorimeter measurement vs rf coupler output (linear values).

• 0 coupler power should be 0 calorimeter power. this implies y-intercept should be 0
• There are 2 output temp meters. I fit computing the calorimeter power from: the average, tout1 only , and tout2 only.
• The color of the final table, corresponds to the fit i used.
  
• The ratio of PwrKw/pwrCouplerLinear  should be a constant. I threw out  outliers
• Calorimeter measurements with low output power were not very accurate (hard to read the meters with enough accuracy).
• Data prior to 150512 were ignored. They had high input temperatures (fans not working correctly).
• This data is fit to the 60db coupler as it is read in the control room with the anritsu spectrum analyzer. I used the values read off of the analyzer... no cable loss corrections were made. So you should be able to use the values straight off of the spectrum analyzer when using the plots.
  

• 150507,8: calibrate Tx4

Dana's measurements (top)

• 22apr15: final tests on 5mhz dipoles. vswr, cross coupling (.txt)
  
• 22apr15: fdtdr impulse response of hf-3a and hf-1a (.pdf)
  
• 10apr15: final shorted measurements of cables (.txt)
• 06may15: dana's blog of what we did  (.txt)
• 06may15:rf coupler 0 to 500Mhz (.ps) (.pdf)  (processing: x101/hf/dana/cplr_150505.pro)
• 150512: tx4 on load. 1200 Mhz spectra of noise  burst (using tek scope) (.pdf)
  
• 25may15: 50db rf coupler. 0 to 2Ghz (.ps) (.pdf)  (processing:x101/hf/dana/clpr_150525.pro)
  
• 17may15: discussion with steve floyd, protecting the transmitters.
• 12jun15: ft/tdr measurement  of tx3b after repair for arc damage. (.pdf)
• 01jul15: ft/tdr measurement of tx3b after arc repair. shows change in reflection (.pdf)
• 10aug15: ft/tdr measurement of tx3b after more arc repair. shows change in reflection (.pdf)
• 22sep15: ft/tdr measurements of tx3b after 2nd arc repaired (.pdf)
• tx3,tx5 on antenna, dana measures vswr
  

Jim's documents (top)

• 20apr15: Subreflector gain characteristics
• gain vs depth, height above dish.
• Dual array cassegrain syst antenna design (.pdf)
  
• kyle  labowski's thesis

miscellaneous

• 150603: Dipole 3B (tx4) arcs at end of coax  where it connects  to wires going  to dipole.
• we were xmitting about 50KW on tx4 when we started to arc at the end of the coax.
• more info shows that the arc was continuous for about 6.7 minutes
• The pictures show the teflon spacers at the end of the coax center conductor and the plastic cover that keeps the rain out.
• pict 1 (.jpeg)
• plastic cover and 1 spacer
  
• pict 2 (.jpeg)
• plastic cover and teflon spacers
• The center conductor of the coax comes out the top of the plastic connector and goes to 1 side of the dipole
• The hole blasted in the cover is where the other wire fron the other end of the dipole connected  to  the outside of the coax..
  
• pict 3 (jpeg)
• looking inside the plastic cover
• pict 4 (jpeg)    pict 5 (jpeg)
• outer sleeve and inner conductor at end of transmission line. This was cut off of the end of the line before replacing the connector. The spatial is in units of dana's hand.....
• 150915:Pictures of Dipole 3B (tx4) arcing at end of heliax.
• dipole 3B (tx4) had arcing on 150915 when we fed the dipole with about 60KW.
• The heliax from the splice up to the dipole connection was replaced (again). This is about 60 feet of heliax.
  
• I took pictures of the damaged heliax on 22sep15 as i lay on the ground close to the ao9 cement base.
• pict1 -large section of cable stretched out
• You can see the large hole toward the middle
• the cement base to the right of the heliax is about 3 feet wide.
• The large hole was 24 feet from the lower end of the heliax piece and about 30 feet from the upper end of the heliax. There was another 8 foot piece that was also part of the piece of heliax that was removed (not shown).
  
• pict2 - blowup showing the upper 1/2 of the damaged part
• pictures 3-7 show different angles of the large hole..
• pict3     
• pict4  
  
• pict5   over exposed to see inner conductor. It looks intact here, but about 2 inches to the right of the hole, the inner conductor is damaged.
  
• pict6  
  
• pict7   large hole, and smaller hole to the right.
  
• pictures 8-11 show the damage below the large hole. It includes a kink in the heliax that was probably created when a large part of that section burned away..
  
• pict8  
  
• pict9  
  
• pict10   You can see that the inner conductor is also damaged. This is true most of the large holes.
  
• pict11  
  
• there were 8 to 10 separate holes punched in the outer conductor
• the inner conductor was damaged at most of the large holes.
• 150922: pictures of the towers
  
• Tower 3 from ao9 monument stand
• This is from the center of the dish.
• pict1:
• tube on left is balun for tx3a (transmitter 3). It feeds the tangential dipoles that are parallel to the image. The helix that looks to be close to the safety  line is coming from tx3a
  
• tube on the right is balun for tx3b (transmitter 4) this is the heliax that is arcing. It feeds the dipole that comes out of the image (the other dipole is not visible).
  
• pict2
• This is a blowup of pict
• the heliax directly behind the safety line is from tx3a
• the heliax above the safety line is from tx3b (that was arcing)
• Tower 1 from rim road close to road that goes under the dish.
• pict3
• the dipoles lower left to upper right in picture are the radial dipoles.
• they are fed by the heliax in the balun on the left side of the tower.
• The short heliax feeds the half dipole coming out of the page
• the opposite dipole half is fed by the heliax that goes around the left side of the tower.
• pict4
• a blowup of picture 3.
• Tower 3 from rim road sitting on the anchor for the southeast fill area of the dish (it is the foreground tower)
  
• pict5:
• dipole lower right to upper left (radial dipole) is fed by the helix in the balun seen in the picture (the other balun is not visible). This is tx3b (transmitter 4) that was having the problems with arcing.
• pict6:
• blowup of pict5
• You can see the heliax's coming out of the cap on the top of the balun feeding the two half dipoles
• 150925: Pictures of inside of dipole 3b  balun  showing arc from heliax outer conductor to balun inner surface.
• on 150915 we had an arc in the heliax from  tx4 to dipole 3b (see above), about 24 feet from the dipole end of the heliax.
• on 25sep15 the bottom section of the balun was removed (starting from about 19 feet from the end of the balun ..check this) and brought to the hf building.
• the balun is built in sections of about 5 feet.
• We inspected the topmost section and it looked ok
• We then removed the topmost 5 foot section to look at the next one.
• The 2nd section looked ok when visually inspected from the end.. but after reaching our arm in about 2 feet, we could feel a bump.
• I then used my canon sx70 camera to see if we could get a better look at the bump...
• pict1  looking down balun from the end
• we marked the outside of the balun where we could feel the bump on the inside.
• we lowered a masking tape roll this distance (actually a little extra) into the balun to give us a distance reference.
• the bump is the black area to the lower left of the tape.
• For pictures 2,3,4 we rotated the balun so that the arc damage was on the upper right. We then slid the camera down into the balun directly below the damage to take the picture (macro mode with 10 second timer)
• pict2  arc damage is centered left,right but falls off the bottom of the image
  
• pict3  arc damage is centered up,down but off to the left
  
• pict4  entire arc damage bump in the picture.
• Looking at the blownup image, you can see narrow ridges sticking up out of the damaged area and they are bent to the right. This was probably caused when the heliax spacers passed over this area when the heliax was removed.
  
• Pictures 5 and 6 are taken with the camera looking in from the end of the balun section with the far end illuminated with the diode floodlamp.
  
• pict5 
  
• the burned section from pict2,3,4 is more than half way down from the top of the picture
• there is a white ring before the burned section. This may be from one of the spacers holding the heliax in the balun. I'm not sure whether the black area on this ring is a burn area.
  
• pict6  
• another show from the end showing the white ring and the main damaged area beyond it.
• Notes:
• We did not see the damaged area on our first visual even though it was only 2 to 3 feet from the end of the tube.
  
• We need to think of a better way to inspect the tubes.. We lucked out this time because the damaged area was only a few feet away from the end and we felt the bump with our fingers...
• 150928: pictures of spacers used to align heliax in balun
• there are two half circles next to each other spaced every few feet to hold the heliax in the balun. One edge of each spacer has been sanded down to make it easier to pass the heliax and these spacers through the balun.
• pict 1 : picture of spacer
• pict2: blowup of spacer
• The ends of each spacer is pretty sharp. They are probably making the grooves we see on the inside of the balun.
• 150928: picture of high pot setup
• pict1: balun section with heliax used for hipot test.
• pict2: dana searching for the gremlins making all the noise inside the balun.
• 150930: pictures of heliaxconnector at top of balun, adapter for vna measurement, heliax connection to dipole.
• for more info on vna measurement see here.
  
• pict1: heliax connector at top of balun. With plastic dome in place.
• The post sticking up connects to the heliax center conductor.
  
• The bolts inside the edge of the plastic dome connect to the heliax outer conductor.
  
• pict2: 2nd shot of heliax connector at top of balun (with plastic dome in place)
• You can see the cable  that goes from the heliax center conductor out to one half of the dipole.
• the two screws were added to mount the adapter board for the vna measurement.
  
• pict3:  adapter connecting 70 foot vna cable to outer conductor  of heliax.
• The 70foot sma cable was calibrated. We  used the blue 8 foot cable (seen in the picture) to push the returned echo away from the calibration 0 distance.
  
• pict4: cable connecting center of balun to dipole half.
• 151027: pictures of transmitter control cabinet
• entire rack
• top section: meters, ctrl switch
• buttons, faults, control boards (cca)
• bias knobs
• broad band amp
• bnc  outputs, breakers
  

home_~phil