The restoration of a vintage tube amplifier like the LEAK Stereo 50 is not just a technical repair, but a true historical and emotional recovery. In this article we document step by step the complete overhaul of this power amplifier paired with the LEAK Varislope II preamplifier, bringing it back to its original performance and, where necessary, improving safety, reliability, and noise levels. From the replacement of out-of-spec resistors and capacitors, to the cleaning of the chassis, the construction of a proper bottom cover, and the final measurements: a philological restoration without technical compromises. If you want to understand what it really means to bring a vintage amplifier back to life, avoiding improvised repairs and preserving its sonic and collectible value, here you will find a real case explained in detail. Related article: Restoration of the LEAK Varislope 2 preamplifier, which is paired with this power amp.
Aside from the four box capacitors that had been installed to replace the original coupling capacitors to the output stage – and which most likely caused the plate glow and total malfunction of the amplifier – all the other components were still original. Dried-out electrolytics, leaking paper-in-oil capacitors, and all the carbon composition resistors completely out of spec, with values wildly different from what they should have been. On top of that, all the tube sockets were smeared with sticky residue because someone had sprayed the usual lubricant spray on them, which is absolutely unsuitable for an electronic circuit. I therefore began the work by removing the transformers from the chassis to proceed with a thorough cleaning of the metal base.
Once the circuit was properly cleaned, I started by replacing the speaker terminals with slightly more practical ones.
The restoration then continued by replacing, one by one, ALL the resistors and ALL the capacitors present on the board. It may sound extreme, but there was literally nothing worth keeping. I replaced the carbon composite resistors with standard film resistors. Sorry for those who consider them “musical components”, but I measured drift far beyond acceptable values: 200k resistors reading 245k, 180k reading 190k, 2k2 reading 2k5, and so on. So much for precision and stability. Not surprisingly, many of the ECC83 tubes in this unit were completely worn out, with deformed curves and unbalanced sections, a clear sign they had been operating at improper or excessive bias points.
The non-polarized capacitors were replaced with polypropylene models from Mundorf or other good-quality NOS brands. The only ones that survived were the two main 16+16µF electrolytics, which, after all measurements and tests, proved to be in perfect condition, so I left them in place. I instead replaced the third electrolytic for the preamp decoupling stage with one supplied by the customer, mounted in an octal socket, as it fit perfectly in the hole and could be properly secured, unlike the one installed by the previous technician, which was held in place only by the pressure of foam stuffed into the oversized opening.
Another important step was building the bottom cover for the amplifier chassis. Apparently, LEAK sold them without a bottom panel, or maybe they were meant to be installed inside furniture. If anyone has information about this, feel free to comment. The fact is that around 400 volts run through this circuit, and people use it open like that. Crazy! So I built a proper metal panel to close the unit as it should be.
And please, the first person who comments saying that closing it makes it heat up inside and “it won’t sound good”, I’ll strangle them. The components that heat up (tubes and the three power resistors) are all outside; inside there is nothing that suffers significant passive heat. And in any case, an electrical appliance must be closed. Besides safety, closing it improves shielding from any kind of electrical interference. I also went through the box of tubes that was brought to me to select a complete and working set to install; the uTracer was a great help.
A couple of measurements. The amplifier is not 50 watts, but 25+25 RMS. The frequency response is about 10Hz – 16kHz at -1dB @1 watt. As power increases, a wavy trend above 10kHz appears, probably due to interactions with the strong negative feedback. Harmonic distortion is 0.12% @1 watt and 0.22% @15 watts. Let’s look at the graphs…
Spectrum at 1 watt
Spectrum at 15 watts
Frequency response at 15 watts
Frequency response on a reactive load (maximum deviation about 1dB)
With the power amp finished, it was time for its preamplifier: you can find the related article here. In the video below, the power amplifier is playing connected to the Varislope II. Obviously, the video does not demonstrate the sound quality, which must be heard in person, it only shows that the unit is working.
I would also like to point out that if you plan to use this power amplifier with a modern CD player or DAC, the Varislope 2 preamp might be unnecessary. The output signal of such devices is already strong enough to drive the power amp directly (the fewer components in the chain, the better). To do so, just disconnect the Varislope 2 and use the two RCA inputs located above the umbilical connector, adding only a passive attenuator to control the volume, as I did in this test.
