Leak Stereofetic FM tuner.

[North hill]

I obtained this old FM tuner the front is damaged I am getting a new metal front the thin plate with all the printed lettering on and the design will be laser etched on to the cut metal plate also I had to design a new tuning scale from a Rogers scale and shrink down on the computer to the correct size I hope the light from the neg green festoon bulb will light up the tuning scale.

I changed all the electrolytic capacitors in the PSU. I used Panasonic caps FM and FRX super low impedance long life 105C type caps. With the service manual I was able to see the values of the DC decoupling capacitors that was 10nF at 25 volts max I changed to small 5mm pitch 100nF at 100V the legs were long so I was able to bend to the correct pitch. The 100nF caps have a lower reactance this will reduce AC noise across the power lines. I also painted the dark red tuning indicator with smooth HAMMERITE paint and with a resistor a high efficiency LED (white) lights up the mono stereo meter like my Rogers Ravensbourne FM tuner. I will post pictures of the finished leak when and if I finish it!

[North hill]

More photos of my leak stereofetic FM tuner with out the thin front panel no wood sleeve or metal case. Note the new white led to illuminate the stereo mono meter I also painted the thin metal tuner indicator with silver paint. All new electrolytes on the three PCBs and two wima 1u5 at 100v film caps ac bipass caps for the silicone transtor x2 from vt20 and vt21 i am getting a laser etched plastic tuner scale I traced and designed I just hope it lights up with the green festoon bulb

[PaulR]

Nice tuners. I had to recap my Stereofetic as all the electrolytics I tested had very high ESR to the extent that stereo did not work. I also have a Delta FM which runs perfectly happily on its original caps.

[Trevor]

I remember when it was launched we were Leak retailers One of the best tuners around at the time I felt it was so advanced Much better than the amplifiers

[Jez1234]

It was very good for the time and technically advanced. Much better than the Troughline 1, 2, 3, and Stereo that everyone raves about and a fraction of the price.

The hand point to point wired front end on Stereofetic's was particularly impressive.

[North hill]

Yes I am listening to the stereofetic FM tuner now I am lucky about 1.3 miles from the transmitter here in Swansea the sound on the better radio stations with low compression and very low noise I played the tuner to a friend. His thought it was digital bicouse the background noise was so low not bad for a product design dating from 1968. Ok I did a re cap mostly electolytycs and some film caps. The question is is the LEAK STEREOFETIC or the ROGERS RAVENSBOURNE better ?
A question of taste.

[Radio Wrangler]

Which is better?

They are similar enough that there is no need to worry over the choice. Though Leak seem to have made a better range of amplifiers than Rogers did, and the Stereofetic matches the style of the Leak amps, so I think the win goes to the Stereofetishists in an indirect way.

David

(...suggested collective noun for owners of Stereofetic tuners)

[Synchrodyne]

That’s an interesting comparison.

The Rogers Ravensbourne FM tuner somewhat predated the Leak Stereofetic. It was advertised in WW 1968 April, and “announced” in WW 1968 June. The Stereofetic was announced in PW 1969 November and WW 1969 December.

The original versions of each were reviewed by “Gramophone” magazine, attached. The Leak, in cased form, cost £72.5, whilst the Rogers, also in cased form, was £66.89 so they were within 10% of each other.

Looking at the circuitry of each, the Ravensbourne had a D&W 341/AFEC/FET/2 front end, which used dual-gate mosfets in the RF and mixer positions. (See: https://www.vintage-radio.net/forum/...d.php?t=157235.) It was four-gang, but unusually as compared with typical 1970s practice, there was a bandpass input and single-tuned interstage, rather than the other way around. It could have been that this was simply copied across from the preceding bipolar 341, as used for example in the Radford FMT2. With bipolars, evidently the reduced cross-modulation benefit of a bandpass input outweighed the noise disadvantage. This had been covered in the RCA 965 December IEEE paper “Performance Analysis of 3- and 4-coil FM Tuners using RCA High-Frequency Transistors”. The noise penalty of the bandpass input was said to be 4 dB.

The IF strip, using four µA703 ICs with distributed LC selectivity appears to have been out of the Fairchild playbook. This type of IF strip was described in Electronics World 1966 October, p.32ff. The wideband (1 MHz) discrete ratio detector reflected the swing from the Foster-Seeley discriminator often preferred in the valve era. At a given bandwidth, the latter had lower distortion, but on the other hand, the ratio detector was amenable to use at very wide bandwidths, whereas the Foster-Seeley was not. The wider bandwidth reduced distortion, but it also had lower gain, and as far as I know, diminished self-limiting, so needed to be preceded by a high-gain IF strip that could provide serious limiting. This was much more easily doable with ICs than with valves.

The stereo decoder I think was more-or-less Rogers’ established discrete transistor unit. The tuning meter was a centre-zero indicator, showing the correct tuning point only, not signal strength.

The Stereofetic had a three-gang front end, unusually with a single-gate mosfet RF amplifier, although with the usual dual-gate mosfet mixer. Nonetheless, this might well have been derived from the RCA sequence, which was stepwise from all-bipolar to dual-gate mosfets in the RF and mixer positions. The first step was the use of single-gate mosfet RF amplifier with bipolar mixer, described in Application Note 3453. Next was the change to a single-gate mosfet mixer, AN-3535. Then the latter was substituted by a dual-gate mosfet mixer, as described in the 1967 July RCA paper “Application of Dual-Gate MOS Field-Effect Transistors in Practical Radio Receivers”. I don’t know if the final step, to a dual-gate mosfet RF amplifier, had its own paper or AN, but it was covered in AN-4018. It looks as if Leak stepped off the RCA bus at the last-but-one stop.

The IC-based IF strip appeared to use an earlier RCA IC line-up (a CA3011 pair, this being the lower voltage version of the CA3012) but with ceramic filters in place of the originally shown L-C type. This IC series dated from early 1966.

The stereo decoder was mostly discrete, except that it used a CA3026 IC as a transistor-tree demodulator, rather than the usual diode array. The tuning meter showed a combined signal strength and centre channel indication, essentially by subtracting the modulus of the off-tune bias (zero at the on-tune position) from the signal strength bias. That was something that had also been done in the valve era with magic eyes (e.g. the Dynatron T10A). The original Leak Troughline had obtained the same effect with a magic eye by different means, combining the NBS null-point indicator with a signal strength input.

Both tuners predated the simpler layout made possible the 1971 crop of ICs, namely the RCA CA3089 IF subsystem, and the RCA CA3090 and Motorola MC1310 PLL stereo decoders, and their various derivatives. That said a fully integrated, but non-PLL stereo decoder had been available since 1969 in the form of the Motorola MC1304/5.

The Rogers Ravensbourne was later updated to use the newer ICs, but I do not know exactly when that happened.


Cheers,

[Valvepower]

Hello,

I kind a like the Stereofetic.

Terry

[Radio Wrangler]

The Foster-Seeley and Travis discriminators are simple derivatives of the slope detector, and any variation of the amplitude of their IF input shines straight through onto their audio outputs. So they have little AM rejection. Any AM rejection wanted has to be provided by limiting amplifiers in the IF chain. Some attention to detail in those limiters is needed to minimise their AM-to-PM conversion which would translate any incoming AM to phase modulation, which any discriminator would interpret as FM with a sloped passband.

The ratio detector rose to popularity because it too has tuned resonators, but instead of simple amplitude detection, it takes into account phase between its pair of resonators. and acts as a ratiometric comparator. Hence the ratio bit in the name. This gives a degree of AM rejection in the discriminator itself. So the ratio detector surged in popularity because with one, you could dump one stage of limiting from the IF amp. So it all came down to money as it often does. Maybe the AM rejection suffered a little, and the distortion was a bit higher, but the savings were 100% bankable and few could resist.

Later on, ICs for limiting IF stages became pretty good and the ratio detector limiting advantage became worth less. Also the ICs became cheap (thereby driving the accountants into a frenzy) and the quadrature detector (which needs extra limiting stages to work well) became feasible at no extra real cost. This avoided the distortion and threshold effects of the ratio detector.

A few outlying FM tuners avoided the ceramic filters that had become popular with the IF chips because carefully designed and adjusted high-Q LC filters could be made to outperform the ceramics, especially in group delay. The designers were also more adventurous in the discriminator department. Pulse-count and delay-line monostable types for example.

AFC is a double edged sword. If it's turned on and is pulling the local oscillator much, then it means that the discriminator is being run offset from its centre frequency and into the curvier reaches of its characteristic. Distortion goes up and becomes more asymmetric. Try offsetting the tuning of an FM tuner with a centre-zero tuning meter on a clean signal and you'll hear significant distortion come in before the AFC loop gives up the fight and the tuner jumps off the station. The distortion has been curving up right the way from centred tuning and at what point you notice it depends on you and the material being listened to.

For microwave telephony links, a 0.06-18.6MHz baseband was FM'd to make a 40MHz wide IF centred on 140MHz and then it was mixed up to around 11GHz before being squirted across 40km or more of hilltop driving rain and mist. Because the baseband was something like 9600 phone channels modulated as SSB, distortion was a major problem. This can be considered to be an extreme case of FM if ever there was one. The receivers down converted to 140MHz, then came a very carefully designed and aligned limiting IF with very low AM to PM. The demodulators were usually Travis types. HP in South Queensferry made the test gear for this sort of equipment to be developed with and for service/alignment work in the field. Doing FM for stereo audio isn't exactly a walk in the park, but it does depend on which direction you come from.

THere are difficulties and compromises in an FM tuner and the articles Peter has posted are from an era when the reviewers had a grasp of the issue that you could bias a design for distant signal reception, or for surviving strong local signals but one always compromised the other.

David

[Trevor]

what a great write up Radio Wrangler I have enjoyed every word !
Re the Rogers tuner You never know what you will get with all the variations all under one model no
The Leak unit is also very good and advanced for its day another good 1 was the quad FM3 with all its variants
Trev

[Jez1234]

I preferred the Rogers amplifiers to the Leak models personally... but only when we're talking solid state! Very much the other way round with the valve ones (although I've not heard the TOTR Rogers EL34 power amp). I still have the Ravensbrook tuner and amp amongst my collection but unfortunately I let go of the more up market Ravensbourne tuner and amp I also had. The amp was basically NOS as well so I regret letting that go.

The Ravensbourne amp turned in an impressive performance subjectively. Considerably better than the Leak Stereo 70 to my ears... in spite of it using a transformer as a phase splitter!

The later Leak 2000 series amps sounded awful!!

A sweet spot in the Leak SS range for me was the Delta 75 receiver. This basically combined the Stereofetic/Delta FM with the Stereo 70/Delta 70 amp but I believe a few hindsight based improvements and tweaks were incorporated which meant it performed noticeably better than the separate tuner and amp.

[Trevor]

I also have sang the praises for the Rodgers Ravensbourne amplifier even praised the use of driver transformers The early Leak transistor amps were not as good as the valve stuff of that I am sure
Having said that We must have sold 5 leaks for every rogers back in the day
Trev

[Jez1234]

The first Leak solid state amp, the Stereo 30 (without the "Plus") which basically ripped off Toby & Dinsdale, were extremely variable from sample to sample in my experience.

[Trevor]

Yes I heard that story also having said that even that was basically a Lin circuit but with thermistor compensation I actually liked the thirty if you thought of it as an 8 watt amplifier

[Radio Wrangler]

These were amplifiers in the age where counter measures to thermal instability were being explored. Still to be handled was the transistor amplifiers more abrupt clipping than valve amplifiers and the need to counter this by having large margins of power reserve.

I think the 303 and the A48 marked the watershed into a new era where transistor amplifiers had become acceptable hifi and without any nasty habits.

The later Leak and Rodgers tuners also reached a plateau where not a lot changed for quite some time. The FM3 also fitted in here.

David

[Jez1234]

Originally Posted by Radio Wrangler

These were amplifiers in the age where counter measures to thermal instability were being explored. Still to be handled was the transistor amplifiers more abrupt clipping than valve amplifiers and the need to counter this by having large margins of power reserve.

I think the 303 and the A48 marked the watershed into a new era where transistor amplifiers had become acceptable hifi and without any nasty habits.

The later Leak and Rodgers tuners also reached a plateau where not a lot changed for quite some time. The FM3 also fitted in here.

David

Ah but the better valve amps that got to 0.1% THD and lower by using as much as 26-28+dB of NFB had reasonably abrupt clipping themselves... and early low powered transistor amps had to cope with (amongst many other issues such as you mention) tiny smoothing caps and often poor PSRR. valve amps had far greater stored energy in the high voltage of their smoothing caps at the time compared to SS amp PSU's of the day and with poor regulation on some smaller valve amps HT they could have comparatively greater peak power relative to RMS that could help in the loudness battle.
IME SS class A (and hence "proper" PSU amongst other factors) power amps of say 10WPC compare well to similarly 10WPC class A valve amps such as a Leak Stereo 20 in max volume, "effortlessness" etc etc. Something like an "Eagle" 10WPC integrated from the early 70's with maybe a 30VA mains TX and 3300uF of smoothing most certainly would not!

It'll never quite get to "comparing apples with apples" but there are subtleties enough to discuss it for far too long

Interesting that you chose the A48 as a watershed amp there! I don't recall seeing anyone have much of an opinion one way or another on it before?

[Ted Kendall]

The Sugden A21 came out at about the same time as the 303 - Sugden and Hacker went the Class A route, whilst the 303, as noted elsewhere, used Baxandall's idea of output triples. Peter Walker's answer to the assertion that Quad were late to the semiconductor party was "hang on a minute - can you think of a transistor amplifier before the 303 that was actually any good?" (His 50D, introduced a year or so earlier, was more workhorse than hi fi, but still vice-free.)

I'm with Walker on this - there was a dawn of sorts with the Dynaco 120, which Len Hulley tied himself in knots trying to account for the impressive difference in sound between it and established valve designs. What he was actually hearing was crossover distortion...

More on topic, I had a Leak Delta (dressed-up Stereofetic) tuner for a long time, and liked it.

[Trevor]

I can remember the Sugden rep coming in with an early A48 it easily outperformed a Leak stereo 70 in A-B testing
Having said that I really liked the smaller class A 21 when it was used within its power rating
Trev

[Jez1234]

The A48 certainly easily outperformed the Stereo 70 yes. They were fairly popular and had a good reputation back in the mid to late 70's but you rarely hear anything about them these days. The early ones in the wooden sleeve and with a single pair of TO3 output devices sound considerably better than the later ones.

The A21 is the classic one I guess. I rebuilt a few of these about 5 years ago, including an immaculate looking Richard Allan one. Virtually non existent production engineering and very awkward to work on. They have a strange quirk where the steel inner cabinet has various parts held together with insulators (1/2" circles of hardboard!) and nylon nuts and bolts to prevent the chassis forming a "shorted turn" to the transformers external field.

Another unit from the same era and similar performance to the A48 was the Armstrong 600 series, which IMHO get unfairly ignored. The receiver versions are quite interesting as the FM tuner is very good and if you get the one with AM as well then the AM section is double conversion!