LEAK stereo 30 PLUS - Intermittent 'crash and thump' noise

[yestertech]

Really strange 'timed' fault on this one :
After being left on for anything up to 2 hours, in quiescent state, the left channel starts to crackle moderately at first and then becomes louder which builds quickly into a loud 'crashing and banging' and much movement of the test speaker cone. Due to volume of said noise, it becomes necessary to hit the off switch fairly quickly!
After a few seconds at 'off', if switched on again the noise returns at full tilt !
If the amp is left to cool, all is well again and a gentle hiss (from the disc channel ) shows the amp is still working, no smoke or blown fuses.
Clearly a thermal problem, (which takes an age to manifest itself ) I played a heatshrink gun around the output transistors , LS coupling cap, PSU caps and the LH driver board. Not a peep !
Amp is still a happy chappy ...

Measuring with the ESR meter , the LS coupling electrolytic on the LH channel all seems fine. Ditto the electrolytics on the driver board ( albeit measured when 'cool' )
I then swapped the driver boards between LH and RH and left the amp idling again. Sure enough, after a couple of hours the problem repeats but this time it's on the RH channel.
This I would take to indicate the problem is on the LH driver board ( now in the RH channel slot )
It would also seem to absolve the O/P transistors and the LS coupling cap on the LH channel.
On the driver board There is an ancient looking preset which for good measure I will replace, as well as the 4 electrolytics and 2 'tropical fish' caps for starters.
Of course it could be anything on the board !!( no dry joints or broken tracks in evidence ) so any suggestions most welcome.


Andy

[Sideband]

The 'tropical fish' caps generally don't give problems. I would be more suspicious of the ancient electrolytics breaking down. There's an outside chance it might be one of the driver transistors....I had a BC143 fail but that went short and was an easy trace. There is also the possibility of a resistor becoming noisy but that would probably show up as a rustling noise that varies in volume.

The pot sets up the quiescent current which you adjust by putting a meter in series with the collector of the 'top' output transistor....think it's 30mA.

[Nickthedentist]

I had this is a Delta 70, think a Lockfit transistor was responsible. I found it with freezer spray.

[Sideband]

Originally Posted by Nickthedentist

I had this is a Delta 70, think a Lockfit transistor was responsible. I found it with freezer spray.

Ahh yes....good thinking! If it's got BC147/8/9 lockfits in place of BC107/8/9 then they could well be the source of the trouble. Lockfits are almost 'replace on sight' if there are intermittent faults present.

[Radio Wrangler]

"Almost"....?

David

[yestertech]

Here’s a pic of what’s on the driver board - no lockfits as far as I can tell.
First stop will be to change the 220R bias preset, and the four electrolytics
Following the comment on the tropical fish caps I will leave those for now

A.

[Vintage_RC]

Have you tried tapping the skeleton pre-set pot?

[Silicon]

Andy, have you removed any heatsinks from the metal cased transistors?

These work at a high-ish voltage and power dissipation in the Leak 70.
The driver transistors are fitted with heatsinks but sometimes fail.

The Leak 30 Plus works at a lower voltage and may not need them.

[Sideband]

Those resistors fitted with end caps can go intermittent. Worth tapping around them.

Don't think the 30 has the heatsinks

[yestertech]

Originally Posted by Silicon

Andy, have you removed any heatsinks from the metal cased transistors?

These work at a high-ish voltage and power dissipation in the Leak 70.
The driver transistors are fitted with heatsinks but sometimes fail.

The Leak 30 Plus works at a lower voltage and may not need them.

No, there's no heatsinks fitted - the board is as found.

A.

[yestertech]

Originally Posted by Vintage_RC

Have you tried tapping the skeleton pre-set pot?

No, not tried that - I'm wary of doing it in case the bias voltage changes catastrophically and damages the output stage.
It will be replaced with a better quality one.
A.

[yestertech]

After much trial and error this was finally resolved. As you can imagine, it took much soak testing between component changes to prove the fault had finally been exorcised.
First tranche was the 4 electrolytics and the bias pot. A long time elapsed (coincidence) before the fault again resurfaced.
Next, were the 2 driver card input transistors, the reasoning being that the noise was extremely loud and hence probably occurring where amplification was at its greatest.
The circuit diagram shows these are BC 154/9 and BC 117. The first was actually a BC 212L and appeared to be factory fitted. Luckily both of these were to hand and in they went.
After a couple of soak tests - no nasty noises. After several mor, still all quiet.
The testing continued but all seemed well, so the covers were replaced.

A strange crackling noise was now apparent if the amplifier was moved slightly.
Removing the top cover stopped the problem occurring. This was unrelated to the original fault and after a few moments head scratching it dawned on me that the problem was due to intermittent contact between the screening foil fixed to the top cover and the top edge of the chassis. The fix was to install a couple of star washers behind the cover fixing screws.

JOB DONE !


Andy