Roberts RM33 Desktop Radio repair. Freezer spray on the cheap.

My good friend David pops by a few times a year to bring things. My wife dreads it ....

This time he brings me a small desktop Roberts radio. I like Roberts stuff, it's nicely made and sounds reasonable.

"Goes quiet on medium wave"

OK... let's take a look....

The first battle is to get the thing apart.
First, remove the two philips screws on the edge of the back, on the left and right. Then remove the slotted screw in the centre.

Once you've done this, slide the card back upward and it'll pop out at the bottom with some gentle persuasion. Disconnect the green aerial wire from the board. Now unscrew the five silver philips screws from the bottom. Remove the power supply first, followed by the chassis.

Having removed the chassis, I set about fault finding. I previously had run the set, and after about 5 minutes MW sensitivity drops off and the volume decreases dramatically. On higher frequencies, it faded to nothing. Most likely to be a thermal fault. I left the set running until it had faded. Now what we need is some method of cooling components individually to prove which is at fault.

Now, you can buy, at quite some expense, Freezer spray, which is designed for the job. But Doz has a cheaper solution!! Enter the "Lazer" Air duster aerosol. Available for £1 a can from Poundland! Bargain. Now this produces a blast of gas, ideal for blowing the dust etc from speakers, laptops etc... here's the trick. Attach the pipe, and invert the can. Now a stream of extremely cold liquid will emerge from the end!  There is a warning on the can not to do this. It's that cold it will probably give you a cold "burn". Use caution!

Gently push the nozzle and direct the liquid at the suspect component.

Be careful when using it around higher voltages, as water will condense on the cold component. Never spray it on a hot valve, you'll crack the glass.

I started by cooling the board in areas to narrow down the search. I used a hot air gun to warm it back up again to cause the fault to reappear.

I narrowed the fault down to two caps, or maybe the
BC548 transistor. I changed the caps and the fault remained clear! The replacements are pictured.

There was no physical sign of failure of the capacitors.

Another job done!