Leak Stereo 20 questions

[Geoff Sherratt]

Hi. I am just about to start restoring a Leak Stereo 20 that I have had for a couple of years. It is in pretty much pristine condition physically and I want to try to keep it as original as possible. There are three dates and initials presumable indicating regular servicing? The large canned smoothing caps appear to be in good external condition with zero bulging of at the base. I'm pretty sure that the amp hasn't been powered up since it was put away years ago, so I was thinking of seeing if they will reform. I'd be interested in the expert view.
I've been reading most of the stuff on here about the Stereo 20 and there's been a fair amount of discussion about a couple of factory modifications to the circuitry. It looks like there may be on on this specimen, I'm referring to the three resistors stacked in parallel shown in the attached closeup (last) photo...am I right and what is the idea of this mod?
Thanks

[Radio Wrangler]

Those aren't three resistors.

The 47k resistors are probably anode loads, and across them is bridged a resistor-capacitor series pair.

This is to shape the gain and phase of the amplifier at high frequencies with a view to optimising stability and keeping good closed-loop bandwidth.

These are exactly the things the threads on stereo 70 mods are discussing and trying to link to the different variants the factory produced. They were fiddled with in synchronism with mods to the output transformers.

Welcome to the fray!

David

[snowman_al]

That is a Very Nice example cosmetically. Most have a few scuffs and scratches by now.

Yours has the 'latest build' spec: 8162 mains transformer, 8778 output transformers, 470KΩ EL84 grid leaks and 47kΩ ECC83 anode resistors.
The 4.7kΩ + 330pF step network (as David describes) were added later than 1963 from memory, so it has been 'updated' at some stage. (I've added a late build picture for reference, December 65.)

You might be lucky and reform the can caps, worth a try. Though from experience I seldom find one that works these days...
Do not even try the other 50uF electrolytics they will be shot. That goes for the 0.25uF and 0.1uF ''Metalpack'' and ''Visconol' coupling caps too.

A word of warning - be careful unsoldering the old components, the tags can become brittle and break off unless you are quite gentle.

Obviously the GZ34 rectifier is dead too unfortunately.

First step is to measure the output transformer primaries resistance just in case. A few original resistors might be in spec. though most drift off over the years.

They are a pleasure to work on and easy to fix up, enjoy it.

[Valvepower]

Hi,

I've attached two pictures of my later 8778 ST20’s.

I normally replace both can capacitors with F&T or similar. I fit a 50+50uf in the reservoir and filter capacitor positions and a 32+32uF in the first stage and phase splitter decoupling positions. Strangely I’ve been able to reform the 60+100uf reservoir and filter capacitor in the TL12+'s though.

I have used the can capacitors as they come, however if I feel like embellishing the amplifier and making it look like an original, I spray them a similar colour to the chassis.

One 8778 ST20 has a step network however the other hasn’t.

I usually replace all the resistors and capacitors. These two have be fitted with parts from the workshop stash of parts, however I normally I use Welwyn MFR5 resistors, BC electrolytics and polypropylene ICEL and Arco film capacitors.

As regards the dates and number… there is the manufacturing date... 30/12/63 on your ST20. I was told by ex Leak engineer the other number relates to a works order number, batch number and the number of the unit... 2/11/77 on your ST20.

Terry

[Lucien Nunes]

IMO there is a direct conflict between maintaining originality and achieving optimum results and reliability. I have an example that is in original condition with original passive components. It must not be powered-up and I do not ever intend to power it up or overhaul it; it is a static museum piece. The Leak amps that I use have all been overhauled and few original passive components remain, because I want good sound and minimum concern about damaging the transformers.

You can (and should) keep the original green wirewound smoothing resistor; the mica feedback caps should be OK and are nice to retain, and if you want to try salvaging the main electrolytics you might get lucky. If the HT chain resistors are in-spec they aren't going to hurt the sound, and it looks like carbon film resistors have already been fitted in a few places including the output valve cathodes. But pretty much everything else is either dead or suspect if you want a high-end audio device. Who wants a noisy carbon comp in the feedback path or as a 1st stage anode load?

If the electrolytics don't reform, or you decide to decommission them for safety, you can see my solution in this thread: https://www.vintage-radio.net/forum/...ad.php?t=52492
I recall the original capacitors actually tested reasonably sound on that chassis but it was against the ethos of the overhaul to retain anything that was not 100%. I fitted top quality wire-ended parts under the deck as shown in the pics, because I value the original appearance above the chassis even on an overhauled unit. I later added 100KΩ resistors to apply HT to the original caps, to keep them formed while the amp is in use without jeopardising the rectifier or transformer should a short occur.

I rarely re-stuff components these days as I believe it better to retain the original components intact with the unit, and fit honest replacements. I could turn that amp back to how it left the factory if I wanted to, which cannot be done once the original parts have been mutilated. In any case my preferred industrial long-life high-ripple caps do not fit inside the cans.

Have fun, beware broken contacts in the valve holders and take every precaution to protect those transformers.

[Geoff Sherratt]

Thank you all for the help and advice. I have now reformed the the large electrolytics. They weren't bad at all, as far as I can tell. There was no evidence of either of them getting remotely warm and the meter readings dropped very quickly to below 0.08 mA. Given that outcome I wonder whether I should leave them in as they are or replace them with modern units, under the chassis??
I've also no checked the resistors and almost all of them are well within tolerance. However I will be replacing all the carbon ones with metal film as, despite being within tolerance, some are at one side of the spec and others are at the opposite side so there is a range of differences between left and right channels. Having read Lucien's advice I think it's obvious that they ought to be replaced with modern metal film resistors.
I'm interested in Lucien's comment about adding in 100k resistors to the old capacitors in order to keep them 'formed'.... How would I do this? Also could I ask for more details of the the type and manufacturer of your replacement main electrolytics and if possible where I might obtain some?
Again, many thanks!

[Lucien Nunes]

adding in 100k resistors to the old capacitors in order to keep them 'formed

All I did was feed the decommissioned original caps with HT through the resistors. As long as their insulation holds up, they will be subjected to near working voltage. If they short or leak excessively, the resistors simply get warm without anything going bang. My reasoning was that they appeared to be usable even though I had elected not to use them in-circuit at the present time. A future owner might choose to test them, find them good and reconnect them in place of my wire-ended ones. In that case I have done my best to ensure their survival up to that point by keeping them fed and watered.

Apart from excess noise, another argument in favour of replacing certain resistors even while apparently in-spec, especially the anode loads that experience significant signal voltage, is the possibility of voltage coefficient of resistance causing non-linearity of amplifier transfer function. I have measured remarkable VCoR in some old carbon comps, although I cannot say whether in actual use the effect could ever be bad enough to become an issue (noting that the global NFB will correct for it). But I am aware of my own subjectivity, so until I test and prove conclusively that the effect is insignificant, the thought that I have allowed the performance to be compromised for the sake of a handful of resistors, could affect my listening pleasure.

[1954Fender]

The original large electrolytics in my Stereo 20 reformed ok and were carefully monitored but over a few years of use the hum levels gradually rose as the capacitors finally gave up the uneven struggle. It is possible to "cut and shut" the cases, inserting new capacitors. This requires careful work but no respray or underside mods. All the other previous advice is solid gold. Be careful about "audiofool" modifications. Originally my ST20 had a few of these and worked well as a 10 W oscillator....

[Radio Wrangler]

Leak were competent designers and were quite good at optimising what they got out of their amplifiers. There isn't really any room for improvement. Not without sacrificing some other aspect.

Consequently almost all the modifications going around should be viewed suspiciously. A few things like more reliable capacitors, parts with improved ratings and lower-noise resistors are sensible.

THe amplifiers are a bit too sensitive for modern inputs, so some people have tried to reduce their gain. Even this isn't as simple as it might seem. Adding an attenuator is safe, but attempts to cut the gain of the base amplifier itself has interactions with the maths used to ensure the thing is stable.

Given that these modifications are originated by, approved by and propagated by people who pooh-pooh science, engineering and maths, uncontrolled oscillation is no surprise. The perpetrators were probably delighted "I can hear things on my records I never knew were there!"

David

[bikerhifinut]

Sage words David.
I make no excuse for being a fan of the products of HJ Leak, and as David posits, I have for decades now (yes really!) advised against anything other than a sensible and sensitive refurbishment of these amps. I think my friend Terry (Valvepower) would agree here, we're both veterans of the old Yahoo email group run by Steve Spicer back in the day and now and then you'd get some evangelist trying to convince the rest of us about the benefits of turning a Leak power amp into something else. Over to you Terry!
As stated previously they are an excellent example of good production design based on solid fact. If Harold Leak had been able to source good quality metal film resistors and plastic film capacitors at the time and at the right price, he certainly would have used them.
I'm less enamoured of the matching preamps however, worth having as objets d'interest, and passable as line level control units, but the power amps are far more capable and I still use mine regularly.
Seconded about the sensitivity at 125mV for full output is a bit high for modern sources, but easily sorted with an attenuator and saves a lot of brass on preamps with gain cos you dont need any.

Andy.

[Geoff Sherratt]

Hi I've just been looking online to source resistors and capacitors I've looked at CPC farnell and found metal film resistors manufactured by 'Multicomp pro'. It looks like I have to purchase about 400 components to obtain 30 but still costing less than £10.00 BUT are they OK quality wise?

I'm really struggling to find capacitors online of the right physical size and axial lead outs. CPC farnell don't seem to have anything suitable and it's a similar story elsewhere.
Can anyone point me in the right direction?

[GrimJosef]

Originally Posted by Radio Wrangler

Leak were competent designers and were quite good at optimising what they got out of their amplifiers. There isn't really any room for improvement. Not without sacrificing some other aspect ...

There are, perhaps, a couple of exceptions.

High-value, high-voltage electrolytics were a good deal more expensive and space-consuming in 1960 than they are now, and back then 32uF for C11 (the HT reservoir) and C12 (the first HT smoother) in the Stereo 20 was a good compromise between economy and keeping the 100Hz HT ripple down. Nowadays there is a case for increasing the value of C12. Putting another 47uF, or maybe even 100uF, in parallel with it will reduce the 100Hz HT ripple significantly. If your EL84s aren't closely matched for Ia then the push-pull action in the output transformer won't cancel the 100Hz hum resulting from that ripple. In which case reducing it with a larger C12 can produce an audible improvement.

As has been mentioned, the Stereo 20 is very sensitive. This means that even tiny amounts of hum picked up in the input circuit of the first valve can be amplified to audible levels at the output. The mains transformer is working quite hard, so can be prone to leaking 50Hz and 150Hz magnetic field components. The chassis is steel and so is capable of coupling that field quite efficiently around the amp. The fact is that the layout of the V1 grid and cathode circuits and also of the earth paths (very unusually signal ground is connected to the chassis at two separate points) doesn't look as though it's been optimised in terms of hum pickup. The advertised figure of 80dB below 11W can be very hard to achieve in practice and I suspect there is room for improvement if someone would be prepared to roll their sleeves up and really get to work on the component layout.

Cheers,

GJ

[joebog1]

Originally Posted by Geoff Sherratt

Hi I've just been looking online to source resistors and capacitors I've looked at CPC farnell and found metal film resistors manufactured by 'Multicomp pro'. It looks like I have to purchase about 400 components to obtain 30 but still costing less than £10.00 BUT are they OK quality wise?

I'm really struggling to find capacitors online of the right physical size and axial lead outs. CPC farnell don't seem to have anything suitable and it's a similar story elsewhere.
Can anyone point me in the right direction?

Hello Geoff,
I buy very high quality metal film 1%'ers from China, They come in various sizes and I find the 2 watters are a near perfect match for old one watters and one watters are perfect for replacing half watters.

They come in various sized bags starting with 10 per value at a cost of about $2 for me,
Australian, so about 12 bob for you. They usually take 10 days to reach me from China, which is half the time it takes to send a letter to the next town to me which is 35 kliks away. Freight or postage is mostly free.
There are some here that dont like shopping that way, but after 20 years I have yet to have a problem.

Joe

[Radio Wrangler]

Originally Posted by Geoff Sherratt

I've looked at CPC farnell and found metal film resistors manufactured by 'Multicomp pro'. It looks like I have to purchase about 400 components to obtain 30 but still costing less than £10.00 BUT are they OK quality wise?

'Multicomp' is Farnell/Newark's in-house brand. Actual parts bought froma variety of sources, to their declared specification.

I've had numbers of these components in equipment go through serious environmental testing (+80C, -55C, 80,000ft altitude, high humidity, salt spray, -1000ft altitude, explosive decompression, stresses from low battery voltage start, and massive supply voltage surges from alternator field dumps, fast rates of temp and altitude change etc etc)

I've experienced no problems with them.

I hope your lounge is a more human-compatible place

I agree with G-J about the possibility of increasing some decoupler capacitances..... within the ratings of the rectifier for the reservoir.

David

[Geoff Sherratt]

Originally Posted by Radio Wrangler

'Multicomp' is Farnell/Newark's in-house brand. Actual parts bought froma variety of sources, to their declared specification.

I've had numbers of these components in equipment go through serious environmental testing (+80C, -55C, 80,000ft altitude, high humidity, salt spray, -1000ft altitude, explosive decompression, stresses from low battery voltage start, and massive supply voltage surges from alternator field dumps, fast rates of temp and altitude change etc etc)

I've experienced no problems with them.
David

That's all reassuring then... I'll get them ordered!

Any suggestions on a similar vein with regard to capacitors?

[GrimJosef]

As far as capacitors go, for the ones feeding the valves' grids 630V rated plastic film ones will be a 'fit and forget' option. Despite their higher voltage rating they will probably be physically smaller than the originals. These days radially-leaded ones predominate because they're better suited to stuffing into pcbs, so they're cheaper, but axials are still available.

If you're looking at Farnell or CPC then here's a polyester 0.22uF one as an example https://uk.farnell.com/vishay/mkt181...pet/dp/1166879

Here's a polypropylene https://cpc.farnell.com/lcr-componen...1kv/dp/CA05657 (actually this one is quite large)

There are others out there which might be a little cheaper. In the end we just have to trawl through the suppliers' websites.

If you want to try ones which look a little more original without breaking the bank then there are often Russian ex-military paper-in-oils for sale on eBay. They're often designated K42Y or K40Y.

Cheers,

GJ

[Valvepower]

Hello,

I’ve used the LCR polypropylene capacitors, and they are very good.

As I said in a previous post, I use the F&T can electrolytic capacitors. They can be bought from the BVWS stall, however there are online suppliers like Watford valves etc. 50uF+50uF for reservoir and filter duties and 32uF+32uF for decoupling further downstream.

F&T make a 47uF 63V axial capacitor, which is nice and matches the original style.

I like the MFR5 Meatal film resistors for all positions except for the EL84 cathode bias resistor where I’d use a Nehom ROX3.

https://uk.farnell.com/tt-electronic...?ost=mfr5-100k

https://uk.farnell.com/neohm-te-conn...0+ohm&cfm=true

I've used the BC 47uF 63V for EL84 cathode bypass duties.

https://uk.farnell.com/vishay/mal202...c%2047uf%2063v

Hope this helps a tad.

Terry.

[Richardgr]

Something I discovered on the tag board of a Quad, which looks similar the ones in that Leak, is be careful with that solder sucker! When it recoils it can knock a tag off. They seem to be very brittle, not sure if they get case hardened with age?

[Valvepower]

Hello,

I have repaired a damaged tag by fitting a replacement tag carefully removed from a donor RS components tag strip.

I carefully remove the damaged tag and fit the replacement by opening up the two ‘prongs’ of the RS tag in the existing hole.

You have to careful when connecting to it as it’s not as robustly held in the board as the original, however, it does an OK job of replacing a damaged tag.

I’ve attached a picture of the RS tag and strip. The two prongs on the RS tag can also be clearly seen in the picture.

Terry

[Geoff Sherratt]

Originally Posted by Valvepower

Hello,

I’ve used the LCR polypropylene capacitors, and they are very good.

As I said in a previous post, I use the F&T can electrolytic capacitors. They can be bought from the BVWS stall, however there are online suppliers like Watford valves etc. 50uF+50uF for reservoir and filter duties and 32uF+32uF for decoupling further downstream.

F&T make a 47uF 63V axial capacitor, which is nice and matches the original style.

I like the MFR5 Meatal film resistors for all positions except for the EL84 cathode bias resistor where I’d use a Nehom ROX3.

https://uk.farnell.com/tt-electronic...?ost=mfr5-100k

https://uk.farnell.com/neohm-te-conn...0+ohm&cfm=true

I've used the BC 47uF 63V for EL84 cathode bypass duties.

https://uk.farnell.com/vishay/mal202...c%2047uf%2063v

Hope this helps a tad.

Terry.

Very helpful Terry, many thanks. Just one question, being pretty much a novice about the finer detail of component choice, could you explain why you choose a different resistor type, (Nehom ROX3), for the EL84 cathode bias?

Geoff