Precautions against transmitting valve failure

[high_vacuum_house]

Good afternoon,
I won the CAT 6 transmitting valve that was in the RWB auction and I was wondering about how the BBC prevented transmitting valve failure from interrupting broadcasts.

Because the valves are directly heated and the filament current is kept stable and not switched on and off if it could be helped sudden filament failure should be rare. I wonder if they kept a close eye on the individual valve heater voltage to determine if the filament was getting thinner being slowly evaporated away and its resistance increasing with a rise in valve heater voltage increasing. You could then switch over to another valve long before the heater gets to a point when it could fail.

Also constant voltage and current measurement of valve parameters to indicate the presence of gas or a valve drifting out of its specification parameter as it aged.

Christopher Capener

[G6Tanuki]

Remember that back in the Dark Ages, the BBC didn't provide a 24/7/365 broadcasting service.

They had "Closedown" in the late evening and resumed broadcasting in the morning of the following day (later on Sunday because they didn't want to provide a more attractive offer than the local church).

So they would have had plenty of time to undertake "offline" tests on the transmitter valves.

[Malcolm G6ANZ]

Not directly related to a valve per se. On the linacs I worked with they had a 'bright emitter' filament, this was a coil of tungeston heated to white heat to emit electrons for accelerating. The current was monitored and when it fell below a certain level the filament was changed. The starting current with a new filament was about 8A and when this current fell to about 7 A the filament was changed.
I would think that the filament current in the transmitting valve would be monitored in the same way so failure could be predicted.

Malcolm

[high_vacuum_house]

Good afternoon.,
That is what I thought, to monitor the filament thickness by determining the change in resistance of the filament. Presumably, the transmitters would obviously have many output valves in parallel and would be designed for ease of HV/filament power isolation to enable a replacement to be fitted without shutting down the transmitter and only lower output power being the effect.

It is interesting (or deliberate scaremongering being put out!) the information put out about the Droitwich transmitter being on its last valves and a failure of either would cause catastrophic damage! Somehow I do not believe that, as any high power installation should have ample circuit protection monitoring to stop major blow ups! The worst failure mode I can see would be if the filament snapped and shorted between the grid and anode, This should very quickly open the HT and filament overload protection relays without burning out any other components in the transmitter.

Sudden or partial loss of vacuum would alter the characteristics of the transmitter and automatic monitoring should raise an alarm and stop things running away pretty quickly!

Christopher Capener

[Ed_Dinning]

The French replaced their LW transmitter with a transistor unit of equal power. At these frequencies it is basically audio engineering. I suspect that the "last valves " story was from a manager who knew not about what he was talking !!!

Ed

[paulsherwin]

It was almost certainly PR, at a time when the BBC was testing the water as to the public reaction if 198kHz was turned off. There are obviously many possible solutions to the 'no more valves' issue (if it ever even existed) including hundreds of decommissioned modern AM transmitters available for scrap value. Arqiva could have bought the ex RTÉ Clarkestown transmitter for peanuts if they wanted to.

Despite the supposed desperate shortage of valves, Arqiva have managed to keep 198kHz operational to this day.

[russell_w_b]

Originally Posted by Ed_Dinning

The French replaced their LW transmitter with a transistor unit of equal power. At these frequencies it is basically audio engineering.

The modulator valves and PWM switching units at Droitwich were replaced with solid-state modulators mid-1990s. The PWM mod arrangement on the B6042 TXs prior to that was identical to the B6126 300kW HF transmitters at Skelton 'C' and they were trouble initially, until Marconi rolled out some modified switching boards.

But HF senders are far more accessible for fault-repairs and could generally be taken off-air by request, if cover was available at Skelton or another HF site. Not so with domestic LF!

The RF valve palarver is scaremongering. Valves of that size can be manufactured new or old-ones rebuilt: all that is needed is to throw money in the right direction. The reticence to consider surplus solid-state TXs from the likes of Nautel or ex-broadcast sets from elsewhere is corporate and/or political.

Incidentally, hypervapotron valves of that size are protected (at the insistence of the manufacturer) by 'crowbar' protection: an ignitron across the 26kV HT line to deck controlled by CTs, spark-gaps and the valve modulator switching units is fired by a thyristor, dumping the HT immediately to deck via a resistor network and shutting off the HT supplies by means of vacuum interrupter switches.

One test we had to do was connect a strand of wire, the same thickness as 15A fuse-wire, between the 26kV HT line and deck via a vacuum interrupter, bring up the sender HT, fire the vacuum interrupter and prove the transmitter tripped. The worst under normal conditions was a slight darkening of the wire. Of course, if the ignitron firing lead had been disconnected and not re-made during the ignitron pulse unit maintenance, there was an almighty bang and the wire vapourised.

Not that that ever happened, of course.

Well, maybe twice...