Leak Stereo 20 questions

[Valvepower]

Hi,

The 270 Ohm cathode resistor dissipates approximately 0.6W so a higher power resistor is needed in this position. From memory Leak used a 1 Watt resistor.

I use the ROX3 series in production units, so I know how it behaves thermally with the kind of dissipation plus it builds in a safety margin. Thinking non electrically the size fits with the old style of the original parts as does the MFR5 resistors.

If you have a look at this thread, you’ll see a ZIP file containing my humble efforts in post #9, including ROX3 and MFR5 resistors. The large can 60+100uF can electrolytics did reform and test perfectly OK on these amplifiers – maybe the easy life helps here? It’s the only Leak where I’ve kept the reservoir and filter capacitors though.

https://www.vintage-radio.net/forum/...d.php?t=187240

Finally, I'm currently working on a TL25+

Regards
Terry

[Geoff Sherratt]

I've just had a poke around in the the leak varislope preamp that came with the poweramp and the markings appear to indicate that both were made in 1964. The preamp being made on the 1st of January '64 and the the poweramp being built on the 30th of December '63. Both appear to have been in for repair overhaul or servicing or something on the 2nd of November 1977. It's nice to see see all the old markings so well preserved, and it looks like business must have been good, with a nice bit of lucrative overtime for someone working on New year's Day!!

[Radio Wrangler]

Are you intending restoring and using the varislope preamp?

They really aren't in the same league as the Leak power amplifiers, so they usually wind up as museum pieces rather than as operational equipment.

David

[Valvepower]

Hello,

Ah, there’s the three figure works order number, batch number and unit number etc., written on the round yellow sticker in the first photo…

I may be wrong, but in my lifetime I can remember folk working on New Year’s Day as it's only in recent times it became a general ‘bank’ holiday – I stand to be corrected though.

I rather doubt It’ll have been serviced by Leak in 1977 as Leak was sold to Rank in 1969 and from memory ceased trading under the leak name in the late 1970’s.

I echo RW’s comments regards the preamplifier, however there is nothing wrong with restoring preamplifiers, however – for me – they are of the time so to speak, and should be used as such. I’ve used my preamplifiers to nostalgically listen to 60’s pop like Motown, Joe Meek, Philles, Trojan and alike and they do add the ambience if you see what I mean.

Terry

[Herald1360]

Originally Posted by Geoff Sherratt

.......

and it looks like business must have been good, with a nice bit of lucrative overtime for someone working on New year's Day!!

New Year's Day wasn't a bank holiday in England until 1974 so no overtime in 1964, lucrative or otherwise

[Geoff Sherratt]

Yes I forgotten that New year's Day wasn't a bank holiday!
With regard to the servicing dates I don't think Leak would have been doing the servicing of these amps anyway. It's much more likely that it was a local repair shop. It was almost certainly a servicing job though, rather than being sent in for repair due to a fault, because both were done at the same time.
I don't intend to restore the preamp, however that brings me to another question. I've read various discussions about the sensitivity of the power amp vs the output levels from modern input sources. The question is where do I stand if I want to use the amp in conjunction with a record turntable? Will I need some form of pre amplification of the output from a modern cartridge? You can see I'm a bit out of touch with all these technicalities!!

[Herald1360]

You'll need a suitable RIAA preamp unit if you want to use a magnetic cartridge. And a "passive preamp" (volume control) between the Leak and the RIAA box. Passive tone controls too, if you like!

[Valvepower]

Hello,

Herald1360 is bang on with the RIAA preamplifier and passive ‘pot in a box’ volume control/passive preamplifier.

Although the input sensitivity is higher than required the ST20 can sound rather nice with a good solid state moving magnet RIAA preamplifier. Personally, I feel it deserves a good preamplifier, nothing ‘silly’ expensive but a good sensibly priced commercial product. I’ve had good results using IC based two stage preamplifiers using NE5532 grade IC’s.

An ‘off the shelf’ commercial MM preamplifier intend for use with 'modern' amplifiers (typical input sensitivity of 150-250mV) will work fine.

I use an Alps RK27 50K pot for the volume control as I’ve a couple of RK27’s in the toy box from other projects.

Going back to a posts by bikerhifinut and Radio Wrangler the ST20 is a good amplifier and I’d just restore it using modern components and don't touch the circuit. For me its synergistic design encompassing industrial, commercial, and electronic design. All right it may have shortcomings due to the high input sensitivity but considering the overall design it does what it says on the tin.

A late friend of mine worked at Leak between 1959 and 1966 and chatting to him and also speaking to Steve Spicer the chap who wrote the ‘Leak’ book and also meeting one of Harold Leak Sons I was able to understand what was the Leak ideology. Thankfully some of these ideologies are still (just) alive in the hi-fi industry.

Terry

[Lucien Nunes]

Agree with most of the above.

Use larger-rated resistors as they will be physically nearer the original components in size and that will achieve the same surface temperature. Modern resistors are only more highly rated for a given size because they can tolerate higher temperatures themselves. A modern 3W resistor the same size as a vintage 1W will run at the same surface temp as the 1W in the same circuit. The 91k phase splitter anode load can be a deciding factor in what series to use, if you want them all to look alike, as many stockists only keep E24 values in more popular ranges. Of course you can cheat with 1M in parallel with 100K but I prefer a resistor from a different series to a parallel pair.

While on the subject of resistors, no specifically applicable to the S20 but in a few circuit positions in higher-power valve gear, the limiting element voltage of smaller modern resistors might be too low. Using the higher-power types even in low-dissipation positions gets you 350 or 500V ratings.

In theory, a well-sized coupling cap experiences very little voltage at frequencies of interest, therefore very little of the characteristics of the dielectric are manifest. As such, polyester, which is reliable but not the most perfect for distortion, should be perfectly OK. I tend to use polypropylene throughout, which is nearer ideal electrically, more for the psychological principle of not spoiling the ship for a ha'peth of tar than for any sonic benefit. I have just switched my dining room system from an S20 (fitted with polypropylenes) to a pair of TL12+ (fitted with 1970s SRC polyesters). The difference between the triode and pentode 1st. stages I can hear (with my ear next to the L/S) but the coupling caps I cannot!

For the reservoir and smoother I fit 47μF long life, high ripple industrial Kemets under the chassis, e.g. https://uk.rs-online.com/web/p/alumi...citors/2267312
These should give another lifetime of service and I see any economy here as a false one. The electrolytics further down the chain are not so critical and any good quality, 105°C part will be fine. The ones I used in the resto I linked are Panasonic NHG. I would be equally inclined to use Vishay axial 030s (formerly BC, formerly Philips).

As I mentioned before, don't touch the mica feedback caps, you won't improve on them unless actually damaged.

Like other posters above, I do not use the original preamps and do not generally restore them. The EF86 is a good valve in its context but it cannot equal the performance of a reasonable modern S/S phono stage, and it is not needed for line level sources.

[knobtwiddler]

I haven't spoken to him in a while, but I used to regularly speak with a learned engineer who also worked for Leak. Once the deal with Rank was finalised he got out. His room was down the hallway from where Harold would demo the newer transistor amps to clients, and he claimed he could always tell whether it was the valve or transistor amp being demoed! He would shout 'valve!' or 'transistor' down the corridor, much to Harold's annoyance! I didn't discuss how he discerned the difference. I seem to recall it was something to do with the bass response of the valve amps.

I've posted this factory tour link once or twice before, but for anyone that missed it here it is again:

http://44bx.com/leak/1968.html

Video of said tour (inc. footage of Harold and his Rolls):

https://www.youtube.com/watch?v=49He8QwlAU0

(I'm not connected to the fellow who authored the above btw! I just found them on the 'web)

[knobtwiddler]

solid state moving magnet RIAA preamplifier. Personally, I feel it deserves a good preamplifier, nothing ‘silly’ expensive but a good sensibly priced commercial product. I’ve had good results using IC based two stage preamplifiers using NE5532 grade IC’s.

If you've read Self's book, it shows that the NE5534 is still the quietest opamp - 40 years since its release - for MM preamps. I have confirmed this myself with an analyser. Mind you, in truth, this is academic - as the typical cart + record combo will have far higher noise sources to be concerned with, so the convenience of the dual 5532 will not be of concern!

[Geoff Sherratt]

Originally Posted by snowman_al

Obviously the GZ34 rectifier is dead too unfortunately.

Reading your comment about the relative loadings on the GZ34 MA12 versus Stereo 20 got me thinking....I was hoping to use this in the Stereo 20. Is it man enough?

[GrimJosef]

It should be fine, as long as the cryo treatment hasn't micro-cracked the glass around the feedthrough wires.

Cheers,

GJ

[Lucien Nunes]

Its internal design is supposedly based on Mullard and it's not Chinese, so it stands a good chance. In any case it will take longer to cook from frozen*.The original Philips GZ34 was rated for 250mA at 450-0-450 input and 160mA at 550-0-550 and it is here at the high-voltage end of things that problems tend to show up with modern examples. At the lower voltage used in the Leak it should be fine.

*No rants about cryo-treated valves here.
Somewhat simulposted with GJ.

[Geoff Sherratt]

Hi. I've given the amp a birthday with all new caps, apart from the four 32uF electrolytics which re-formed very well. Almost all new resistors (I'm just awaiting the arrival of a couple that were out of stock but in fact the originals measure well within tolerance). The only question I have relates to the voltage measured on V1 left and right anode. Both are reading identically at 190V. According to the figures given on the circuit this is about 30V high. All other voltages read almost spot on to the figures on the circuit. Any advice appreciated. Thanks, Geoff

[Lucien Nunes]

30V is not catastrophic but that seems quite high even with the preamp disconnected. Presumably you have confirmed that the replacement anode load and cathode resistors are of the correct value.

What are the voltages a) at the V1 cathodes and b) at the preamp socket HT pin? If b) is correct then the drop across the anode loads is low which means the anode currents are low. This will then be reflected in the cathode voltages being low by the same factor. If so it suggests that the valve is low in emission on both halves. Is it new / old / tested / unknown?

[GrimJosef]

The original circuit has 100k resistors as the V1 anode loads and they give a quiescent anode voltage around 160V. The revised circuit, after all the trouble with the EL84s going into thermal runaway, changed (among other things) the V1 anode loads to 47k (sometimes AC bypassed). This lower resistance drops fewer volts for a given current and the anode voltage therefore rises to around 190V. The curious thing is that although Leak changed the resistor value in their later re-issued version of the circuit, they left the voltage unaltered - just a slip, I guess.

Early circuit here https://www.44bx.com/leak/Leak/Circuits/Stereo20.gif

Later circuit here https://www.44bx.com/leak/Leak/Circu...ereo20_3lr.gif

Cheers,

GJ

[Geoff Sherratt]

Excellent...so all is well! My circuit does indeed show the 47k with the 135v/160v figures.
Thanks everyone!..What a superb forum this is!