A RF-robust 9V supply...

[G6Tanuki]

For a while I've tried various power supplies to feed my Roberts R606: unregulated 9V "Wall-warts" and a couple of supposedly-quiet SMPS. None were really satisfactory - either giving a whine audible in the 'silent' gaps of music, or seemingly introducing strange burbly hash [intermodulation] - presumably because the diodes were seeing RF and creating the FM-0equivalent of 'modulation hum'?.

So I decided to do the job properly.

A 13V 400mA transformer with built-in 125-degree thermal-fuse was in the junk-box along with a diecast box and some DIN-connectors. There were also a matching 1980s-era Belling-Lee fuseholder/indicator-lamp and an old ferrite toroid.

Add in a "Velleman" LM317-based variable-voltage regulator board, and an hour or so of assemblage.

I added 1000pF capacitors across each of the bridge-rectifier diodes to minimise noise.
The DC side is entirely isolated from the rest of the PSU. The "Power On" lamp is a 28V 40mA bulb wired across the transformer secondary - the idea being that it helps to reduce the peak AC voltage fed into the bridge-rectifier - and, after all, there's no point rectifying/smoothing/regulating power that's only going to light a bulb. Though a 28V bulb, it glows quite bright enough when fed with half its design-voltage.

The toroid on the output - wound with red/yellow wire - is intended to keep the PSU's end of the DC output RF-decoupled (at Band-II FM frequencies) from the rest of the circuit.

On the input side there's a 100mA fuse, and the mains-lead is rather well anchored by way of a 20mm IP67-rated compression-gland and a couple of cable-ties.

The result? Burble-free FM reception, and in tests it can supply around 300mA continuous without getting at all hot. If I wanted more, a friend has offered to machine me a brass thermal-transfer block to sit between the LM317 and the case. A R606 doesn't take that sort of continuous current even when turned up to 11.

As to the blue paint-job? I was doing some work on the engine of a friend's Ford RS2000 Mk.1 Escort, and had to respray the camshaft/cambelt-covers. Ford RS-Blue was in the spraygun, so why waste it?

[paulsherwin]

The built in PSU in the 606MB is a fairly straightforward affair, though it is regulated and obviously linear.

[G6Tanuki]

Mine's not a 606MB, alas.... hence the need for an external, linear supply.

I'm a bit of a stickler for silence-when-there-should-be-silence on FM radios; the other supplies I tried didn't achieve the sort of no-signal silence I got when using batteries. With this supply I can use an xtal-controlled oscillator on 100MHz to generate a local unmodulated carrier and I get no burblies at all.

On this little one I even arranged for the DC cable between the radio and the PSU to be a 1/2-wavelength long at 102MHz [guessing the velocity-factor of twin PVC-insulated wire] so that the point where it connected to the PSU and to the radio would be high-impedance.

[G8HQP Dave]

Half a wavelength merely ensures that each end will see the same impedance (assuming no losses), not necessarily a high impedance.

A halfwave antenna looks high impedance at one end because the other end sees infinite impedance. It is high and not infinite because of radiation loss.

[G6Tanuki]

You're right about the impedance at the ends: which is why I put a bifilar toroid on the PSU DC output (so providing what should be an infinite-impedance to common-mode currents at that point).

Reflected 1/2-wave up the power-lead that should in turn be a high impedance at the radio itself - so discouraging the power-lead from also acting as a 'radial' for the radio's own antenna.

I dislike having DC supply-leads also contributing to the RF ground....