Leak Troughline Stereo vs. Quad FM3. First impressions
Introduction
This report describes the comparison I have made between a newly acquired Leak Troughline tuner and a Quad FM3 tuner that is in regular use. I had intended this comparison to be part of a larger survey of ‘classic’ tuner designs, but for reasons that will become apparent I am reporting my first impressions. Depending on the response to this report and with further experimentation I hope to expand the survey to include two valve designs and two solid-state designs.
The contestants
Leak Troughline Stereo tuner.
Encouraged by Neil’s excellent writing on this tuner I recently acquired a sample in excellent condition, s/n 56184. This came fitted with Leak’s own stereo decoder; a solid-state design using germanium transistors. Original Mullard brand valves (with the shield badge) were fitted and the vendor stated that the tuner had been recently serviced. Two flying leads fitted with cheap quality phono plugs provide stereo output. The RF aerial input is via screw connections, permitting the use of 75Ω or 300Ω designs.
The Troughline tuner has an RF sensitivity of 1.5μV for full limiting. I can’t find the output level, apart from fact that the voltage level from the cathode follower feeding the stereo decoder is 1V. It turns out that the stereo output is considerably less than this; estimated to be ~25mV. The captive 2-core mains cable is fitted with an American style plug, with two flat blades; complicating installation.
Quad FM3
This is Quad’s first solid-state tuner, replacing the FM2 valve design. This particular sample (s/n 32352) has been in regular use for 30 years and has had an estimated cumulative use of ~ 27000 hours.
The RF sensitivity is 1μV for -30dB noise level, and provides an output of 100mV for 30% modulation.
Methodology
With radio being an ephemeral medium it is difficult to perform A-B comparisons. Whilst it is easy to be misguided by A-B comparison, it is the only simple way to realistically assess the sound quality of tuners fed by the same signal. To this end the signal from a roof mounted 3-element Yagi was divided between the two tuners using a simple resistive splitter. The signal loss through the splitter is calculated to be 4dB at worse.
The output from each tuner was fed to a separate input of a Quad 44 preamp: these inputs initially having a sensitivity of 100mV and a loading of 1MΩ. With both tuners tuned to the same station, by simply switching inputs it was easy to compare their performance. The power amplifier was a modified Quad 405 feeding Quad ELS 57 speakers.
The results, Day 1
Despite the Leak having a lower sensitivity, some 13 stations could be received. Obviously not all of these were at full strength, but all were sufficiently strong that noise was not a problem. The signal from the more powerful transmitters (BBC Radio 2, 3 and 4, as well as local radio) was sufficiently strong that the gap of the EM84 tuning indicator was almost closed. The output of the Leak was considerably lower than that of the Quad. The Quad 44 uses a stepped attenuator as volume control. To obtain a similar listening level, I found that the difference in volume control setting translated to a difference of 13dB; implying the output level from the Leak was only ~ 22mV.
The listening started with speech on Radio 4. I regard the human voice, whether spoken or sung, to be an excellent test of tonal fidelity, especially in the mid range. From my memory of a previous sample of the Troughline Stereo, as well as from those users reporting on the web and in magazines, I was expecting the sound to be warmer and more open than that of the Quad. I was therefore somewhat shocked to find the sound to be not only warmer, but also warm and woolly – positively ‘pipe and slippers’! This was not simply a case of the Quad displaying a dryer and less forward bass; all voices male or female through the Leak were unnaturally ‘chesty’.
In other respects the performance of the Leak was fine: good stereo spread with focus and depth. The soundstage from the Leak came no further than the plane of the speakers, whereas from the Quad the soundstage started a little forward of the speakers. Frequency range was satisfactory: it was the balance that was troubling me.
I then retuned to Radio 3 for a concert of Chopin piano works. Again through the Leak the piano sounded soft (no pun intended) and distant. Sweeter than the Quad maybe and perhaps more relaxing at lower volume, but the Quad was more involving and engaging. Through the Quad, the piano had the correct percussive nature and to me sounded more realistic. The Leak gave a good impression of height during sung Lieder, one could easily imagine the height of the vocalist, but again the vocal registers seemed to be unnaturally depressed. It was not really possible to assess the acoustics of the venue through either tuner but this is more a comment on the recordings rather than the BBC, who are usually very good at mike placement for this.
One recording was ’live’ and through the Quad one could hear individual hand claps in the applause at the end, whereas the Leak tended to make the applause sound like ‘rustling cellophane’.
Tuning to Radio 2, I was able to listen to Dylan’s ‘Like Rolling Stone’. Through the Leak, Dylan’s nasal ‘whine’ was suppressed (don’t get me wrong I love Dylan but he doesn’t have the best of voices) and a number by Chrissie Hind was relaxing but not involving, and again her voice seemed too ‘chesty’. The Leak did make a good job of Phil Collin’s percussion, so attack and transients were all there.
Returning to Radio 4 I found both Bob Brydon and Charlotte Green’s voices to be unnaturally deep.
The Leak needs a good twenty minutes to ‘warm up’, whereupon the sound stage moves more forward a little and also becomes a little wider and deeper.
Day 2
I did a bit of rearranging; feeding the Quad tuner into a tape input set to 300mV (at 120kΩ) and moving the Leak to a radio input (100mV at 1MΩ). Now the volume levels were similar to within a click or two on the volume control.
Listening started with Radio 2. The Leak gave Dire Strait’s ‘Sultans of Swing’ a nice full sound with plenty of ‘swing’. Next up was Dionne Warwick’s ‘Message for Michael’, there was too much bass and her voice sounded distant. Again Chrissie Hind’s voice was too chesty. The only time I preferred the Leak was when playing the Righteous Brothers’ ‘You’ve lost that loving feeling’, here the bass heavy balance added to the Spector ‘wall of sound’.
Retuned to Radio 3, for today’s Chopin concert. Again this was enjoyable, but the Quad made the piano sound more realistic. Finally to Radio 4 for the afternoon play. The BBC always does a good job here, and the Leak portrayed width and depth and atmosphere well. At times the added warmth helped the intimacy, but again most of the voices seemed too deep.
Conclusions?
Well I am perplexed and a little disappointed. I was expecting the Leak to be warmer and more open than the Quad, but to be honest the balance of the Leak is such that there is bass ‘bloom’ at best and bass ‘boom’ at worse. Not what I was expecting at all. Maybe the internal decoder is letting the Leak down – I don’t know.
I shall persevere, maybe the Quad is bass weak, quite likely, but voices through the Leak are definitely wrong at the moment. Any comments will be gratefully received, as I feel the Leak is capable of performing better than it has done so far. To be continued....
Barry