First thing this morning, I emailed our June 2013 McGuiReport to our email mailing lists.

When the transmitters finished the morning broadcasts at 10 am, I began tuning TX3 on higher frequencies.  The first frequency was 9500 kHz.  Next 11750 kHz was tried, but would not tune properly.  Eventually, it was found that the PA tube badly needed neutralizing.  The rest of the day was spent neutralizing both the PA and driver tubes and showing Graham how it is done.  Hopefully this might explain why this transmitter would not tune on the high frequencies.

It was ‘Grandparent’s Day’ today at the school that the mission children attend.  They invited Virginia to be their ‘grandmother’ for the day.   After school, Virginia invited all five of the mission children to our house for a party.  They made huge bubbles– thanks to Joyce for giving us the idea and tips–(check it out here), did various crafts and had ice cream cones.

Reviewed the logs and found out that TX3 was still tripping off due to RF unbalance when used on Antenna 6 at 7300 kHz, in spite of the work that we did two days ago.  I then researched logs back to January, when they started using this frequency, and found this has been an ongoing problem.  We decided that it would be beneficial to observe this antenna in the evening during the half hour time slot of this broadcast.

There has been a problem with TX3 not tuning properly on higher frequencies, resulting in excessive cooling water temperatures.  I decided to begin at 9500 kHz and work up through the bands.  I just got started on this process, when Steve asked me to leave early (about 3pm) so we could return after dark to observe Antenna 6.

At 7pm, we were back at the site and positioned under the antenna for a close up, yet safe, view.  During the next few minutes we saw small sparks in various places, which we later concluded were most likely small insects being ‘cremated’.  About 7:26, the west end of the most active element at 7300 kHz lit up like a huge torch with a loud voice emitting from the flames, which were bluish colored and about 4 feet long.  I think Steve jumped back almost 20 feet when it happened!  [Wonder if this is like the voice that Moses heard in the ‘burning’ bush?]  This flame went out almost immediately, but about 2 minutes later the other end of the same element also lit up and quickly extinguished.  Simultaneously with this second arc, there was another smaller arc about a half wavelength away on the feedline — an interesting phenomena.  After that, no other problems were seen for the rest of the broadcast.  Back at the transmitter building, we checked with the operator and the transmitter had sensed both arcs and tripped off, thus extinguishing the arcs.  Without the sensitive circuitry that shuts the transmitter down, the flames (actually corona) would have burnt the elements off.

Back at home, I successfully replaced the circuit breaker for the water heater (geyser [pronounced geezer] here).  For several days when I have tried to turn it on for hot water in the morning, it would trip multiple times before I could get it to stay on.  Wednesday morning it took about 30 attempts before it would stay on, so we stopped by Swaziland Electrical Sales on the way to staff meeting that morning.

Wednesday morning  we attended the weekly staff meeting and had the rest of the day off.  Virginia and I used the rest of the morning to walk around the east side of town to observe how things have changed.  Actually, there is not much that has not changed in the past 24 years.  We took quite a few photos, some of which we plan to include here when we get a better internet connection.

Just before lunch we met Jonathan at the TWR Manzini office and showed him the Manzini Market on the west side of town, where fruit and vegetables are sold in bulk and smaller quantities, where crafts are sold and lots of clothing is made in the small garage type stalls in the market.  We also saw galvanized sheet metal work (buckets, tubs, etc.), hair cutting, shoe repair, and lots of medical herbs being sold from these stalls.

We then stopped by the bakery and introduced Jonathan to samoosas for a late lunch.  After dropping Jonathan at the office, Virginia and I did some more shopping then headed home to be ready to attend the Wednesday evening service at the Evangelical Bible Church.

We got a chance to wish our granddaughter Audrey a happy 10th birthday using Skype.

First thing this morning, I noticed that Transmitter 2 (TX2) efficiency was back to normal, but no reason could be found!  One of the ways of comparing the efficiency of like transmitters is to check their cooling water temperatures.  They should be the same.  Using a hand-held digital thermometer, I compared all three 100,000 watt transmitters.  TX1 & TX2 readings were comparable and similar to their computer readouts, but TX3 read was about 15 degrees higher than its computer reading, which was comparable to the other two transmitters!  When TX3 was shut off, the hand-held thermometer instantly returned to a ‘normal’ reading!!  TX3 had been on 9500kHz.  After it was retuned to 7315kHz, the reading was normal.  This confirmed that the hand-held electronic thermometer is sensitive to certain RF frequencies, which can cause erroneous readings.

The copper ring shown here is the same as the ones used on the feedline

We made some adjustments to the feedline to Antenna 6 to see if the RF unbalance, when it is used on TX3, can be improved.  The 300 ohm feedline consists of two conductors.  Each conductor is made up of six wires spaced equally by 2 inch diameter rings (like the antenna element in the photo).  On one of the feedline spans, the two top wires of the six were sagging, so we fabricated and added another ring, with clamps, to correct the sagging.  When this was tested, there was no obvious improvement in the RF unbalance.  We noticed that another span had two bottom wires of the six that were sagging.  We cut two pieces of PVC pipe to temporarily pull these two lower wires into position  Again, there was no noticeable improvement in VSWR or RF unbalance.

In the afternoon, it was decided to test TX3 on Antenna 6 on 7300kHz at full power with modulation and have observers under the antenna to pinpoint any arcing problems.  After about 30 minutes we heard minor arcing and Graham spotted its exact location.  This arcing was not enough to trip the transmitter off, but if left to continue, it could possibly flare up to a much bigger arc due to heating and ionization of the air at this high voltage point.  This may be what has been happening in the evening to cause the transmitter to occasionally trip off due to RF unbalance.  At about 2:45pm it was decided to quickly let the antenna down and investigate.  We found that the arcing was between the corona ring and where it was attached to the element.  The bolt was not making a very tight connection.  The connection was taken apart and a file was used to clean up the surfaces.  The corona ring was then bolted more tightly onto its radiating element.  Everything was put back together and the antenna raised in record time — before 3:30pm.  We just had time to bring up the transmitter and test the antenna before evening broadcasts began.  I listened to the broadcast in the evening and heard no dropouts at all.  Will have to verify this by reviewing the logs when we get back to the site on Thursday.

It was a really full day!!

On Sunday, Pastor Deon’s sermon was on seals 4 to 6 from Revelation 6.  This was the first Sunday that we  ate at home, so I was able to spend the afternoon working on the blog entry for last week.

Two motion sensor on one corner of the house. There are a total of 7 such sensors that cover the perimeter of the house and are switched on shortly after sunset, then switched off just before sunup.

Since Joshua and Angie had made a quick trip back to Indiana and would need to drive back on Monday, Allan and Joyce invited everyone to their house for an early Memorial Day family get-together on Sunday.  We wanted to talk to everyone on Skype, so Virginia and I played some Rummikub to stay awake.  We got to see and talk to everyone after everyone got there about 5pm (11pm our time).  As a result, we got to bed late.  Unfortunately, it was windy during the night and the outside motion detectors sensed some moving branches and kept ‘squawking’ about every 30 minutes all through the night.  I had a good idea which branches were causing the problem, but decided not to go outside to trim them in the dark!

BEFORE — Notice the motion sensor on this corner of the house and the bushes starting to grow up in front of it. Taken about five weeks ago.

Monday, I covered the Maintenance section of the HC100 manual during the morning training.  After lunch, I took an exploratory walk down to the river crossing to get a little exercise.  In the afternoon and following up on the morning’s discussion, we confirmed that all the transmitters were using newer PA inductors with both copper and silver plating, which results in cooler operation than earlier versions with silver plating alone.

AFTER Virginia trimmed the bushes, there were no more false motion sensor ‘squawks’.

When I got home at sunset, I was ready to cut the bushes which we believed to be the source of motion sensor triggering the night before.  To my pleasant surprise, I found that Virginia had already done it and we have had no unwarranted ‘sqawks’ since.

Virginia had written the blog about a ‘typical day’, so I helped add the photos and post it.  Then we headed to bed a little early.  We needed to, after the short and interrupted night before.

WARNING — if you want to skip technical jargon, you will probably want to skip down to the last three paragraphs!  

Tuesday morning, I trained everyone how to calibrate the A-D cards in the metering systems of TX1 and TX3 (TX2 had been calibrated last week to help me remember how to do it).  Several were out of calibration because a wrong method of calibration had been used.  Five out of six cards calibrated nicely, but the RF metering card of TX3 could not be calibrated.  We had to replace the metering input board that ‘reads’ the +/- 15 supply voltage, which voltage is used as a reference for the calibration.

In the afternoon, we started testing a transmitter/antenna combination that has been causing intermittent problems, only to found out that the ‘new’ metering card that we just installed in the morning had a problem and the transmitter would not come on.  Fortunately, we did this testing, otherwise it would have caused lost air time later!   We found a faulty integrated circuit (IC) on the ‘new’ card and four diodes that were installed wrong.  After replacing the IC and correcting the diode problem, we were able to commence the test started earlier.

TX3 was finally brought up on Antenna 6 at a frequency of 7300 kHz and determined that the VSWR was 1.88 and the RF unbalance was 4.2.  Both of these are very close to the shutdown limits.  We did not have time to do additional testing before the evening transmissions.

As TX 2 was coming up we noticed that it had a error message regarding the 2.73 reference voltage used for temperature monitoring.  This was re-calibrated just before programming began.  This has proven to be a full day but worthwhile — finding unknown problems and making many corrections!

Wednesday morning was staff meeting in town, then Selema and I went to the transmitter site, while Graham picked up the 1000 foot reel of cable to replace the antenna control cable temporarily repaired last week.  When Graham arrived at the site, we resumed the testing of TX3 on Antenna 6 to determine at what frequencies the VSWR and RF unbalance problems ‘peak’ around 7300 kHz.  This proved to be excellent training for Graham on how to tune up the transmitter on new frequencies, but revealed nothing conclusive about the antenna problem.

We decided to test TX2, instead of TX3, on Antenna 6, in the afternoon.  This entailed, once again, tuning this transmitter on multiple new frequencies.  Graham was now getting even more experience.

Thursday morning, we completed the testing of TX2 on Antenna 6 and found that it had slightly higher VSWR on this transmitter,  but the RF unbalance was considerably lower.  This seems to indicate the problem is not totally with the antenna.  So what is the problem?!!

In the afternoon, Klaus wanted us to reinstall the original, now ‘repaired,’ A/D converter card and metering input card into TX3.  We ‘happened’ to discover that the water flow indication was showing very high flow without the pump running!  This was finally traced to ‘leaky’ diodes on the metering input card and corrected.  After the repair, it was found that the water flow sender also had a faulty connection (cold solder joint), which was also corrected.  The water flow limits are normally set to about 2.0 minimum and 4.5 maximum, but the maximum limit had been reset to 16 (the maximum possible reading is 10.24) to keep the transmitter running.  With the fault on the metering card and with the maximum set to 16, if water flow would have stopped, the transmitter would not have shut down!  I then checked the other two transmitters and found that the maximum limits on both had been set to 16, also.  I reset them to about 5 and explained that the maximum limit was set specifically to indicate a metering failure.  How long all of the transmitters had been operating this way is unknown!

Friday, we decided to continue investigating the Antenna 6 problem, by testing TX3 on Antenna 8, which is electrically identical to Antenna 6.  Antenna 8 had slightly higher VSWR and much lower RF unbalance.  Actually it looked much like what Antenna 6 looked like to TX2!

After analyzing this information, we decided the problem may be with the routing through the switchbay, since there is a different route to get to Antenna 6 than to Antenna 8.  Since Antenna 4 is also similar to Antenna 6 and is accessed by a common route through the switchbay, we decided to test Antenna 4 on TX3.  The afternoon was spent in testing again on multiple frequencies on Antenna 4.  This time the VSWR was lower and the RF balance was lower.  This seems to prove the switchbay routing is not the problem.  So, once again, what is the problem??!!

As this transmitter, TX3, started the evening transmission, its computer’s CRT monitor failed.  The transmitter will operate without the monitor, but its failure leaves the operator ‘in the dark’ about what is happening.  We quickly installed one of the two new LCD monitors that we shipped from Elkhart in March.  It sure looks bright compared to how the old CRT monitor looked.

Our grandson, Mark, calls these ‘sample’ bananas. They are about half the size of those normally purchased in the store.

It had been a full week, so Saturday provided a nice break.  In the morning we slept in until about 7am, then harvested bananas, nartjies and grenadillas from the garden.  From the other side of the fence, Klaus pointed out another stalk of bananas that we had not seen that were really ripe, so we cut down the whole stalk.  It looks like we will have to give away some, or eat bananas until ‘they come out our ears!

At noon, we were planning on attending the reunion of a  TWR missionary family, who has retired in Swaziland.  They had five girls at the same time we had our five boys.  At the last minute the reunion was postponed, because it was discovered that one of their grand-children’s passport had expired.  Hopefully it will happen before we leave.

This change of plans allowed us to attend the birthday party for Graham’s two daughters that was held in the afternoon.  We had a good time celebrating with them, the other TWR families and a couple from FEBA, former co-workers of Graham.