Tromso Heater log of important technical notes

This is an attempt to log faults which may have slightly affected the radiation pattern during experimnts in the past.

The notation for antenna row used here is:
[Array][Tx_number][(for A1 only) north or south row][number counting from west]
e.g. A1T1n11 =array-1, transmitter 1, north row, 11th from west end.

Overview of Array-1 antenna feeds as of 20160719.
Spreadsheet of faulty Array-1 antenna feeds and repairs.

A1T5 fault in coax

2012-09-24
On 13 July 2012 after noticing that A1T5 seemed to cause arcing on MORRO, I (MTR) examined that row of coaxial feeds. I found that the two 75-Ohm CATV cables were cut at the base of the ~3m vertical wooden post, west-most, north. It seems that this, and several similar faults found in the previous two years, was caused by rotting of the posts and their sinking causing stress on the coax. The fault was fixed by GM in mid-August 2012 by joining the cables again inside a plastic protective box. Incidentally, it is not certain whether the arcing noise seen on MORRO has anything to do with this fault.
How long has this fault existed ?
Regular SWR measurements of the antennas did not show high SWR or a particularly unusual SWR vs. freq. shape, including a measurement made on 13-Jul-2012. A new SWR measurement made on 22-Sep-2012 after the repair, did however show a SWR curve which differed in detailed shape from the previous one. Examining the history of the SWR curves for A1T5 revealed that the last one to have the same shape as the repaired coax was made on 26 May 2008. The next curve on 12 Jun 2009 was different and essentially the same as all the curves up to 13-Jul-2012. So the fault appeared between 26 May 2008 and 12 June 2009 and was fixed in mid-August 2012. Perhaps detailed examination of the SWR during experiments in that interval would give a more exact date, but this is by no means certain.
What are the implications for the antenna beam ?
A1T5 feeds two east-west aligned rows each of 12 antennas, each antenna oriented 45 deg east of north. The cut coaxes would have disconnected the two west-most antennas of the northern row. Since the SWR measured at the transmitter was not high, presumably the total measured power was correct, it just being distributed (evenly ?) among the remaining 22 antennas. This is my best estimate of the implication. Some modelling with NEC might be useful.

A1T5n3 faulty feed

2014-11-26
On 22 Nov 2014 a broken twin feed wire was found to the 3rd antenna from west, north row. This was probably the cause of 'arcing' and interference in VHF experiments on 20 Nov and earlier. On 26 Nov a water-cooled dummy load of 336 Ohm (three 1kOhm 225 Watt resitors in parallel) was connected instead of that antenna, until 23 June 2015 when it was repaired. Over the winter this dummy load was frozen in the water of a 44 gallon drum. Both A1T5n3 and A1T6n3 were burn't again on 20151022, and both repaired on 20151029.

A1T6s12 faulty feed

2015-10-22
On 22 Oct 2015 a broken twin feed wire was found to the easternmost antenna, south row. On 29 Oct a water-cooled 'waffle-mix' dummy load of 336 Ohm (three 1kOhm 225 Watt resitors in parallel) was connected instead of that antenna. It remained in use until 27 Nov 2015. It was removed 2015 for the cold winter temperatures. The antenna was fixed on 2 June 2016.

A1T6s another faulty feed

2015-11-26
On 26 Nov 2015 a broken twin feed wire was found ...

A2T10 ant nr 4 faulty elbow

20141010 ? to 2016-11-25
On 25 Nov 2016 a short circuit due to a burnt connector to the south elbow of tower nr 4 (from west) of A2T10 was repaired. This fault appeared as an asymmetry in SWR plots between 20141010 (OK) and 20150129 (poor).

A1T3/4 antenna feeds burnt.

20170710
Nr 12s burnt, two water cooled 'waffle-mix' dummy loads were put in place.

A1T3/4 antenna feeds replaced

20171004
Nr 12s burnt, two water cooled 'waffle-mix' dummy loads were removed and the rebuilt antenna feed installed.

A1T7/8 two antenna feeds burnt.

Nr 3s burnt either on 18 or 19 October 2017 and Nr 11n burnt on 19 October 2017
On 20 October 2017 two water cooled 'waffle-mix' dummy loads were placed in tx7 (Nr.3s and Nr.11n), and two new experimental panel heater resistor dummy loads were placed in tx 8 (Nr.3s and Nr.11n). An SWR test showed a slightly high SWR for Tx 8 suggesting that they are not so sood compared to the 'waffle-mix' loads.