Repaired Roberts Rambler, now motorboating

[repairman]

Hello, I recently  revived an old Roberts Rambler that had been "badly bodged". Many of the lockfit transistors were faulty and were replaced with near equivelants. The radio appeared to work well.  

I fitted a new PP9 battery (open circuit voltage 9.7V) , you would hear an audible tick from the speaker at about 2 ticks per second when long wave was selected , long wave reception was still audible. Medium wave was not affected in this way.

By accident, i connected an older battery with a terminal voltage of about 7.5 volts. The audible ticking (motorboating ?) was not present, however, by fitting the newer battery, it returned.

I am a bit puzzled, any suggestions would be most welcome.

Regards

Repairman  

[ppppenguin]

Welcome to GVR.

Have you changed the electrolytics? They're likely to be duff on a set of this age and if any are decoupling the power rail this could easily cause your problem. Even the blue Mullard/Philips electrolytics which were the best of that era could easily be duff. Any others are simply change on sight without even asking if they might be duff.

[repairman]

Thanks for your answer, I will certainly do that and let you know. Thanks again !

[repairman]

Hello, changed all of the electroytics and it's a lot better now, thank you .

[ppppenguin]

Thanks for reporting back with your success.

For those who are a bit hazy on the function of decoupling caps here'a very brief intro. The power rails in a radio etc are at different DC voltages to the 0V rail. However they are are all meant to be at zero AC voltage. The decoupling caps act as something close to a short circuit for AC (ie signals, such as audio or RF) to ensure that that (for example) the HT rail is at the same AC potential as the chassis.

Without decouplers you can have all sorts of unintended signal paths via the HT rail etc. The effect of this can be unpredictable. Oscillation at anything from motorboating to UHF is a common effect. Odd effects as you tune across a waveband. Almost anything.

Decouplers are not perfect. They have stray inductance and are not truly zero impedance. It's the designer's job to select suitable components. Some circuits can be very critical about their decoupling. If you've ever designed with big, modern, fast digital chips you'll know that decoupling can be a serious headache. Vintage kit is a lot simpler but there's still the odd set that's too fussy for its own good.

[Terry]

Decouplers are not perfect. They have stray inductance and are not truly zero impedance. It's the designer's job to select suitable components. Some circuits can be very critical about their decoupling ...

Consider instability in a TV IF strip. There are a lot of decouplers in there and, on a wired chassis, may well have been fitted first to keep the wires as short as possible for low series inductance and, therefore, difficult to get at.

Yet it is unlikely that more than one has failed - but which one?

Easy - take another suitable cap - 1nF tubular ceramic, for example - and carefully hold it across each suspect in turn. But wait - you can't have long leads if it is going to be a guaranteed successful test and the wires are closer together than the width of your thumb and finger ...

So, what's 200V between friends?

I can't claim originality for this but here is something I made well over 50 years ago to solve the problem ...

[ppppenguin]

I've seen that device before and it's a good idea for saving on money for the swear box.

You've also been in the business for a long time