OBSERVATIONS    GM1SXX          19th August 2010

A story of a sick transmitter, a Cormorant, and a lot of Jiggery-Pokery

Things have not been quiet here of late. I've been working on some 'big stuff' lately in the form of a Medium Wave Broadcast band transmitter belonging to a local radio station.  The transmitter had been providing sterling service but on Friday the 13th, that all changed when it had to be closed down after suffering some sort of failure. The transmitter is a Redifon unit and after some investigation, the master oscillator was found to be generating spurious signals over a broad range of frequencies.  The transmitter of course didn't take kindly to this and by the time we managed to reach it and disable it, two of the four channels ... the transmit path feeds 4 blocks containing modulators and PA's ... had suffered damage to their modulators and power amplifiers. Since we still had some operational kit, I took my RACAL MA1720 transmit exciter and set it to the correct frequency as a replacement for the damaged master oscillator.  While we were there we did some work on the transmitter's aerial system which brought the aerial system's VSWR down to an acceptable value. For this we used an MFJ-259B aerial analyser with a 100pf NPO capacitor across the internal tuning capacitor to load the analyser down to a suitable frequency range for the job.  The MFJ analyser doesn't normally go that low!

We put the transmitter back on the air running half power with the MA1720 driving it and checked with some station listeners for signal reports. They came back good and so with the two remaining channels working OK and the PA's being reasonably cool. we headed home.

I put the oscillator unit on the bench and it was immediately obvious that is was sick. It was custom made by HCD Research and is a colpitts oscillator in an ovened arrangement running on 1.530Mhz.   The unit is self-contained and although it plugs into the transmitter rack, it has it's own independent mains supply. The oscillator delivers 2V P/P to the transmitter circuitry. When I examined the signal from the front-panel BNC jack, it looked fine and was on frequency but when I checked the output BNC on the rear, it showed a comb of rubbish!

Oscillator circuit (apologies for the poor image quality)

Some probing turned up a few issues.   The parallel resonant circuit on the oscillator output was mistuned.   When adjusted properly using the HP Analyser, normal service was resumed with only the second and third harmonics being present in addition to the wanted signal. All other harmonics were well down in the noise.  I also noticed that the operation of the front panel power LED was intermittent.  The oscillator unit enclosure is made of anodised aluminium and the grounding arrangements were obviously inadequate. I beefed up the grounding by bonding the aluminium panels together with heavy silver plated copper wire. This is a pretty simple oscillator and it has worked 100% for almost a week now. It'll be re-fitted in the transmitter rack so I can reclaim my RACAL MA1720.

This is what the front panel jack was delivering when I first looked at the exciter Not at all bad with a nice strong fundamental on 1.530Mhz plus the second and third harmonics.  You could believe all was well..... but it was actually in fact VERY sick!

and THIS is what the output BNC to the transmitter looked like!

AAAAAAAAAAAAAAAAAAAAAAAAAAAAARRRRRRRGGGGHHHH!

Err, not very pleasant IMHO. Summat was a bit fishy in the state of Denmark!  No wonder the modulators and PA blocks popped their clogs!

 The horizontal scale is DC-11Mhz. Sproggs and nasties all over the shop!   It's a *very* good thing indeed that the transmitter is fitted with a very effective harmonic filter or OFCON would have been all over us like a bad rash!

They say a photo is worth 1K words so here are some photos.

Here's me preparing the Racal MA1720 for use as a temporary master oscillator for the big Redifon BT500 transmitter. In the background behind me is the Redifon TX and on the table is the PC used for the communications link back to the studio on 13Cms. In front of the electric radiator is my trusty old RACAL-Dana counter.

Here is a shot of the big transmitter. You'd need a crane to move it. It's very heavy, especially the power supply units.  It's a commercial Redifon BT500 transmitter. You can see the four PA blocks (black boxes with the brown coax leaving them) and above is the combiner, harmonic filter and switching/metering.  The missing section contained the faulty master oscillator which we removed for repair.

The transmitter is convection cooled and is a switching type... not a linear tx.  This provides high efficiency and allows the use of small ferrite transformers in the modulators rather than massive iron-cored transformers used in normal AM high-level modulators..

Here to the right of the TX rack is the little mixer and the Orban AM Optimod unit in a steel cabinet with my temporary replacement master oscillator, the RACAL MA1720 synth sitting on top.

 The transmitter is housed in a steel shipping container. It's ideal for the job.  To the left of the container out of view is the aerial matching system. The aerial is a short vertical monopole fed against an extensive ground and fitted with a wire capacity hat.  The matching unit contains an RF choke between the feeder and ground  plus a pair of large inductors to provide inductive loading for the electrically short aerial. I used an MJJ-259B aerial analyser with 100pf across the tuning circuit to pull the analyser oscillator low enough to sweep the aerial for 1530Mhz. We managed to get the VSWR to 1.5:1, quite well within the transmitter's 'happy operating range'.

 A shot of the transmitter cabin from outside after we had tidied things up a little.

This cormorant seems to like the vibes. This bird is a regular visitor to Celtic Music Radio. I can understand why. A nice high perch, a river to fish in, and plenty of RF to keep it's feet warm.

You can see here the top aerial insulator and the capacity hat wires spreading out from it. It's not pretty, but it's a reliable aerial that works very well. The insulator in the photo is a large ceramic power line type so Mr Cormorant isn't exactly small either!   The mast is an old shipping navigation light.

Alex MM0OIL removing one of the very sick PA modules.  The transmitter has 4 separate modulator and PA paths that feed a signal combiner.

The transmitter, looking a bit empty and bare, but operational.  The master oscillator failure damaged two of the high-level modulators and two of the PA blocks.  The harmonic filters are under the steel cover that reads BT500. I was surprised to find that I could easily hear what programme material was going out and discovered that the audio was actually coming from the harmonic filters.

The RACAL MA1720 made an excellent substitute for the faulty master oscillator. Tuned to 1530 MHz to provide the master signal while the original HCD Research oscillator was on the bench.

 I hope you enjoyed reading this one as much as I enjoyed putting it together. On Thursday 19th August, I handed the repaired transmitter master oscillator back to Alex MM0OIL to be reinstalled in the transmitter rack. Hopefully he'll bring back my superb old MA1720 soon to be reinstalled in my shack.