UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Audio (record players, hi-fi etc)

Notices

Vintage Audio (record players, hi-fi etc) Amplifiers, speakers, gramophones and other audio equipment.

Closed Thread
 
Thread Tools
Old 9th Jan 2021, 4:11 am   #1
gkargreen
Triode
 
Join Date: Nov 2006
Location: Fairfax, Virginia, USA.
Posts: 25
Default Leak Stereo 20 rebuild questions

Hi all, got one of these I am in the process of restoring/repairing a monkey-tech's job. One thing noted is that the monkey installed what appears to be a voltage divider on V1's pin 2 and 7, guessing it is to drop the signal input because of the high gain of V1. Is it better to live with the high gain or is there a reasonable circuit change for dropping the gain of V1 without sacrificing the sound quality? Another question is the monkey replaced the two can caps with some bullshit cans that were not even close to the original 32-32 values, but those original cans came with the amp. A quick check show good ESR but they are bulged on the bottom (no breakthrough the black rubber coating) and using my Sencore cap checker at 500 volts to test for leakage I can't seem to get a low leakage value but that may be because it needs a bit of time to reform the can. Any ideas on that? Use the old cans which match up correctly, or just replace with appropriate F&T cans? Thanks!
gkargreen is offline  
Old 9th Jan 2021, 8:41 am   #2
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,637
Default Re: Leak Stereo 20 rebuild questions

One thing noted is that the monkey installed what appears to be a voltage divider on V1's pin 2 and 7 what values? You could try triode strapping the EF86 & dropping the anode resistor as well as/or lowering the FB resistor,the latter only if the amp remains stable.

A cap leaks more before reforming,what values are fitted? Fit decent F&T's for worry free music listening.

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 9th Jan 2021, 10:08 am   #3
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: Leak Stereo 20 rebuild questions

The circuit diagram I turned up for the Stereo 20 has 1/2 12AX7 (=ECC83) as the input stage of each channel. Pins 2 and 7 are the grids for the two channels.

The overall feedback from the output is applied to the cathode of this valve, so the gain it gives is within the loop.

If the gain of the stage is reduced, the feedback arrangement will see it as a reduction in open-loop gain and will partially compensate for the reduction. So, the sensitivity of the amplifier will not be reduced much, though the reduction in feedback will reduce what people generally call 'the amount of feedback', the surplus of open loop gain over the intended closed loop gain. So as you reduce the gain of the first stage, the sensitivity won't change much, but distortion will increase and the frequency response will get less flat, and the output impedance will rise.

Attenuating the signal before it enters the grid of the first valve is the best way of reducing the sensitivity.

The sensitivity of the closed loop amplifier could be reduced by a change to the feedback network, but doing so will yield an amplifier with more loop gain than originally designed and this will reduce the stability margin causing ringing if not outright oscillation. So to handle this the stage gains will need to be reduced.. and this will change the turnover points and phase shifts of the stages, also affecting stability.

So, yes you can reduce the sensitivity of the amplifier, but it amounts to pretty much a re-design. What you have left afterwards isn't really a Leak Stereo 20 any longer. You might like the result, but the resale value plunges - if that is of any importance to you.

Leak's power amps were pushed quite hard to get them down to the advertised distortion figures. This meant using as much feedback as could be done and still retain stability. Artful transformer design was crucial.

The reason they were designed with so much sensitivity in the first place was to allow the preamp designs to be simplified and just have two valve stages per channel.

For use with modern signal sources, the high sensitivity is a nuisance in so much as volume controls can't be used over much of their span. If attenuation has to be done, then doing it right at the power amp input is a good place, because it will also attenuate any unwanted pick-up from the cables from the preamp, ground loops etc.

If a capacitor is bulging, there has been pressure inside from heating. Even if the ESR is still OK, it's an indication that it needs changing before it makes a mess.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 9th Jan 2021, 10:09 am   #4
snowman_al
Octode
 
snowman_al's Avatar
 
Join Date: Dec 2018
Location: Llandeilo, West Wales, UK.
Posts: 1,092
Default Re: Leak Stereo 20 rebuild questions

There should be a 22kΩ resistor on pins 2 and 7 of the ECC83, before that there should be a 1MΩ input to ground. If someone has installed a divider I would put it back to normal, a passive pre (pot in a box) will be fine for the Leak.

Definitely replace the can caps. (I've not had a ST20 for years where the caps were good.) You can spray them to match the chassis pretty well. Strip the plastic coat, aluminium primer and a bronze (or charcoal) touch up can, job done...
Attached Thumbnails
Click image for larger version

Name:	stereo20.jpg
Views:	318
Size:	96.8 KB
ID:	224180   Click image for larger version

Name:	IMG_0584.jpg
Views:	292
Size:	73.2 KB
ID:	224181   Click image for larger version

Name:	IMG_0629.jpg
Views:	256
Size:	50.5 KB
ID:	224182  
__________________
Never Leave Well Enough Alone...
snowman_al is offline  
Old 9th Jan 2021, 10:56 am   #5
GrimJosef
Dekatron
 
Join Date: Sep 2007
Location: Oxfordshire, UK.
Posts: 4,310
Default Re: Leak Stereo 20 rebuild questions

Quote:
Originally Posted by gkargreen View Post
... and using my Sencore cap checker at 500 volts to test for leakage I can't seem to get a low leakage value ...
I'm afraid it may be too late for these caps. Reforming has to be a gentle process of relatively slowly raising the voltage on the caps, and certainly not going above their rated voltage. In fact many of the ones I've reformed seem to have a 'memory' of their actual operating voltage, and trying even gently to drive them above that causes the leakage steadily to get worse, not better. It's possible that these caps could have been saved. But if they've been driven up to 500V then I fear they're beyond saving now.

You can increase the values of the HT smoothing capacitors a bit and the main effect of that will be to lower any 100Hz hum arising from imbalance between the quiescent currents through the EL84s. 50uF for C13 and 100uF, or even more, for C10, C11 and C12 are popular choices.

As far as changing the sensitivity goes, David's described the situation perfectly. The Stereo 20 is a high-sensitivity amp. You can lower the sensitivity by making it into a different amp but then it will be, well, a different amp. Or you can attentuate the input signal. The price you pay then is in signal-to-noise. Reducing the 100Hz hum helps with that, as does selecting the quietest ECC83/12AX7 you can find for the first valve.

Cheers,

GJ
__________________
http://www.ampregen.com
GrimJosef is online now  
Old 9th Jan 2021, 4:09 pm   #6
bikerhifinut
Octode
 
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
Default Re: Leak Stereo 20 rebuild questions

Quote:
Originally Posted by GrimJosef View Post
Quote:
Originally Posted by gkargreen View Post
... and using my Sencore cap checker at 500 volts to test for leakage I can't seem to get a low leakage value ...
I'm afraid it may be too late for these caps. Reforming has to be a gentle process of relatively slowly raising the voltage on the caps, and certainly not going above their rated voltage. In fact many of the ones I've reformed seem to have a 'memory' of their actual operating voltage, and trying even gently to drive them above that causes the leakage steadily to get worse, not better. It's possible that these caps could have been saved. But if they've been driven up to 500V then I fear they're beyond saving now.

You can increase the values of the HT smoothing capacitors a bit and the main effect of that will be to lower any 100Hz hum arising from imbalance between the quiescent currents through the EL84s. 50uF for C13 and 100uF, or even more, for C10, C11 and C12 are popular choices.

As far as changing the sensitivity goes, David's described the situation perfectly. The Stereo 20 is a high-sensitivity amp. You can lower the sensitivity by making it into a different amp but then it will be, well, a different amp. Or you can attentuate the input signal. The price you pay then is in signal-to-noise. Reducing the 100Hz hum helps with that, as does selecting the quietest ECC83/12AX7 you can find for the first valve.

Cheers,

GJ
Yet another Stereo20 query...............

Agreed with GJ Snowman Al etc.

As I said in a previous post, many have tried to "improve" The stereo20, few have succeeded. It's a brilliant amp and if properly refurbished there should no practical noise issues assuming sensible loudspeaker choices etc.
The most effective way to handle any input sensitivity issues is a potential divider at the input, and indeed many enthusiasts simply use the "pot in a box" passive attenuator approach.
I run mine quite happily off either my commercial silicon preamp (rega Cursa3) or either the homebrew CF unity gain buffer or the homebrew transformer output pre which has a 6dB gain. None of these give me any issues, and of course I do use the original leak preamps when in nostalgic mood.

The PSU capacitors can be safely increased to 50uF off the GZ34 cathode and downstream of the 100 ohm filter resistor you can push it to as much as 200uF with no ill effects. In fact the TL12 plus did use a larger value here.

Again, I am wondering if a sticky should be started as a general information resource on the leak Valve amps.
There's a few of us here with ACTUAL practical experience and as a group I would imagine a fairly encyclopaedic knowledge of these fine products.
I'm in possession of copies of most of the leak circuit diagrams, I imagine GJ and Al amongst quite a few others also have access to them, along with a working knowledge of the factory mods along the products production life.

Andy, (hope I am not treading on any toes)
bikerhifinut is offline  
Old 9th Jan 2021, 6:25 pm   #7
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: Leak Stereo 20 rebuild questions

The Leak power amplifiers are all rather sensitive, in order to go with their preamplifiers. These power amplifiers could be re-designed to go with modern preamp and CD player sorts of levels, but it would need to be an extensive redesign taking a lot of effort. The excess gain that needs to be lost can, in a redesign, be traded for some small improvements in a number of areas, but I stress 'small'. These tradeoffs aren't available when the surplus gain is thrown away by a simple attenuator on the front. There are several of us on the forum who I think could make a good job of re-designing the Stereo 20 (or one of its siblings), but none of us have. Partially because it would just be yet another Stereo 20 mod and would wipe out a lot of its monetary worth, partially because not that many people with those amplifiers would understand just what had been done, and why, and partially because that amount of effort might as well be spent on something new.

The unmodified stereo 20, like the other Leaks, is a nice amplifier which works well. If you want something of higher performance in some respect, then you might as well not use one as a starting point. This is why I suggested the input attenuator route.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 9th Jan 2021, 7:29 pm   #8
Richardgr
Heptode
 
Join Date: Oct 2019
Location: Täby, Sweden
Posts: 692
Default Re: Leak Stereo 20 rebuild questions

I thought the 'monkey' referred to at the start of this thread got the thin end of the wedge. Did not seem like he/she had done anything untoward.
Richardgr is offline  
Old 9th Jan 2021, 10:45 pm   #9
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: Leak Stereo 20 rebuild questions

The attenuator is sensible, and I think others have suggested it on here. So monkey status hinges on the quality of the rest of the work. It sounds from the OP that there was some dodgy work around the reservoir/smoothing capacitors.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 9th Jan 2021, 11:40 pm   #10
gkargreen
Triode
 
Join Date: Nov 2006
Location: Fairfax, Virginia, USA.
Posts: 25
Default Re: Leak Stereo 20 rebuild questions

Thanks to all, I thought I had read somewhere that changing the front end of the amp was not the way to go, that the monkey did other things, like coupling cap value changes, PS cap value changes (100 ufd/section!!!) and changed several of the resistors makes it look like they were trying to do something, but to my mind, could not have been good. Not unusual to find monkey techs here in the US, most of the good ones have left the business to actually make some livable money, myself included, just doing what I can for a few extra shekels on the side and for fun.
gkargreen is offline  
Old 9th Jan 2021, 11:51 pm   #11
bikerhifinut
Octode
 
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
Default Re: Leak Stereo 20 rebuild questions

Welcome back.
can you just clarify, the resistors off pin 2 and 7 of V1 are they 22k, to the input and a 1meg from there to earth?

If so they are part of the design, 22k grid stopper to reduce/prevent oscillation, and the 1 meg gridleak, both important.

As to PSU capacitors, You are fine with up to 50uF off the GZ34(5AR4) cathode, the valve is well up to this and personally that's how I would do any rebuild.
After the 100R filter resistor You can usefully increase the value. The benefit is arguable however and I would just use a good quality e.g. the F & T 50 + 50 capacitor rated at 375V or above.

What value coupling capacitors and where did you find the altered values?

Can I suggest some photos of the underside as theres plenty in here can make constructive advice from them.

I would make myself familiar with the circuit, and take note of the later alterations which were made to improve reliability.

Regards.

Andy.
bikerhifinut is offline  
Old 10th Jan 2021, 1:01 pm   #12
Michael Maurice
Moderator
 
Michael Maurice's Avatar
 
Join Date: Feb 2006
Location: Wembley, Middlesex
Posts: 7,219
Default Re: Leak Stereo 20 rebuild questions

I would suggest that the use of the word 'monkey' could have racial overtones and could be very offensive to some people.

You could use the words 'third party' or 'other tech' even 'cowboy' Monkey may be acceptable in parts of the US, but not here in the UK.
__________________
Forum Moderator

http://www.michaelmauricerepairs.co.uk/
Michael Maurice is online now  
Old 10th Jan 2021, 10:28 pm   #13
Valvepower
Octode
 
Valvepower's Avatar
 
Join Date: Aug 2008
Location: Rayleigh near Southend-On-Sea, Essex, UK.
Posts: 1,852
Default Re: Leak Stereo 20 rebuild questions

Hi,

For me, the ST20 is a synergistic design and best left stock. On the ST20’s I’ve done I’ve replaced the smoothing and reservoir capacitor with a 50+50uF F&T part.

If you want to understand the ST20 and any variants you could do no worse than read the Steve Spicer Leak book.

Regards
Terry
Valvepower is offline  
Old 11th Jan 2021, 12:37 am   #14
gkargreen
Triode
 
Join Date: Nov 2006
Location: Fairfax, Virginia, USA.
Posts: 25
Default Re: Leak Stereo 20 rebuild questions

Well taken, Maurice, and you are correct, I can see how that could be construed as racist, my apologies. I think "cowboy" should be fine, didn't want to really use a more expletive deleted term for these crummy rebuilds, been seeing them on vintage radios for decades and now in tube/sand amps as well, sheesh, you think these people could learn some basic electronics before ruining these amps! At any rate, the coupling caps to the driver tubes were changed to 0.01 ufd, I changed them back to 0.1 ufd, not sure how anyone could make a mistake like that! BTW, I assume since this cathode-bias I should have all 4 output tubes closely matched, correct?
gkargreen is offline  
Old 11th Jan 2021, 1:05 am   #15
GrimJosef
Dekatron
 
Join Date: Sep 2007
Location: Oxfordshire, UK.
Posts: 4,310
Default Re: Leak Stereo 20 rebuild questions

Close-matching of the output valves for Ia will minimise the 100Hz hum.

In a perfect world you'd have them close-matched for gm as well, and the 12AX7 driving them would have close-matched triodes inside it. This will minimise even-order distortion. You can achieve much the same effect though with gm-mismatched valves as long as any mismatch in the EL84s is compensated by an opposite mismatch inside the 12AX7 (try swapping the EL84s between the sockets while looking at the second harmonic).

Cheers,

GJ
__________________
http://www.ampregen.com
GrimJosef is online now  
Old 11th Jan 2021, 6:13 am   #16
gkargreen
Triode
 
Join Date: Nov 2006
Location: Fairfax, Virginia, USA.
Posts: 25
Default Re: Leak Stereo 20 rebuild questions

Thanks GJ, I am matching up a quad on my maxi-matcher to get them within 1.5 ma each, so that should be good. I still have some things to replace that the cowboy wonked up, the 270 ohm resistors as well as the 3 watt 100 ohm are on order. There is also a 15 pfd cap in the feedback loop that needs to be changed to 200 according to the schematic. As noted in the photo, there are a couple of things I am not sure about, those black cylinder things, not sure what they are supposed to be, can't match them with anything I have seen on other original pics of the amp and it's a bit tough to match via the schamtic since there are wires running under the board that can't be followed easily, too bad there isn't a pictorial of the tagboard with the part numbers, would make things easier. It's running right now and the voltages are close on everything, so that is good
Attached Thumbnails
Click image for larger version

Name:	IMG_6410.jpg
Views:	314
Size:	120.2 KB
ID:	224316  
gkargreen is offline  
Old 11th Jan 2021, 6:45 am   #17
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: Leak Stereo 20 rebuild questions

I can't see what you mean by 'black cylindrical things' in that photo. I might just be being a bit dim at the moment.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 11th Jan 2021, 10:03 am   #18
snowman_al
Octode
 
snowman_al's Avatar
 
Join Date: Dec 2018
Location: Llandeilo, West Wales, UK.
Posts: 1,092
Default Re: Leak Stereo 20 rebuild questions

Quote:
Originally Posted by gkargreen View Post
...As noted in the photo, there are a couple of things I am not sure about, those black cylinder things, not sure what they are supposed to be, can't match them with anything I have seen on other original pics of the amp and it's a bit tough to match via the schamtic since there are wires running under the board that can't be followed easily, too bad there isn't a pictorial of the tagboard with the part numbers, would make things easier. It's running right now and the voltages are close on everything, so that is good
The 'black things' are C1 on the schematic (0.001uF cap) across R4.
There is speculation as to why they are needed, but stopping RF from long speaker leads is often quoted.

Plenty of layout pictures on the web too.

Alan
Attached Thumbnails
Click image for larger version

Name:	tlayout.jpg
Views:	189
Size:	50.6 KB
ID:	224324   Click image for larger version

Name:	C1.jpg
Views:	206
Size:	176.1 KB
ID:	224327  
__________________
Never Leave Well Enough Alone...

Last edited by snowman_al; 11th Jan 2021 at 10:11 am.
snowman_al is offline  
Old 11th Jan 2021, 1:30 pm   #19
snowman_al
Octode
 
snowman_al's Avatar
 
Join Date: Dec 2018
Location: Llandeilo, West Wales, UK.
Posts: 1,092
Default Re: Leak Stereo 20 rebuild questions

Had time to study your picture in more detail now.
Looks as if the output transformers are '3921' types, if so it is an earlier version of the ST20.
Important to check that R12 and R13 (the 4 x EL84 grid leak resistors) have been changed to 470KΩ and not still 1MΩ. I cannot see them clearly in your picture.

Re the 15pF caps, it looks as if someone has put caps in parallel with R18 (the 12kΩ feedback resistor) They are not 'standard'. Can you read a value? Perhaps they add up to 200pF with the 15pF.

And a picture of the wiring side for info.

Alan
Attached Thumbnails
Click image for larger version

Name:	IMG_6410-002.jpg
Views:	235
Size:	93.0 KB
ID:	224348   Click image for larger version

Name:	IMG_7012-001.jpg
Views:	238
Size:	90.8 KB
ID:	224349  
__________________
Never Leave Well Enough Alone...
snowman_al is offline  
Old 11th Jan 2021, 3:01 pm   #20
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: Leak Stereo 20 rebuild questions

Ah, yes, we are firmly into black magic territory.

This amplifier has a feedback loop with R18 shunted by R4 trying to control the overall gain of the amplifier.

With any feedback loop, controlling the gain around the loop, along with its roll-off at higher frequencies and overall phase shift is critically important to keeping the thing stable. The devil is in avoiding the build-up of too much phase shift before the gain is low enough that oscillation cannot occur. It's a race. The things you do to bring the gain down at higher frequencies also increase the phase lag. Things you could do to reduce the phase lag, slow the rolloff of gain. It amounts to walking a knife edge, and on top of it all are the effects of strays in the circuitry and especially the output transformers.

Leak as a company advertised low distortion, and that meant having stronger than usual feedback control in their amplifiers. They were pushing the limits of what could be done.

C9 bridges the series feedback resistor and adds a phase lead to the feedback loop in order to cancel out some of the growing lag (it also slows down the rate of roll-off) and it rolls off the closed-loop gain of the amplifier. Aficionados whose hearing has no upper frequency limitation may rail at even the presence of such a component, but it is needed.

200pF across 12k means a zero at about 66kHz trying to make the overall amplifier 3dB down at that frequency. This should not be offensive to non-golden ears. It also means that the feedback loop gets a 45 degree lead at 66kHz, compensating for some of the stray lags coming in up there.

So now someone adds C1, 1000pf across R4 (100 Ohms)

The resistor ratio 12k/100 is trying to flatten the overall voltage gain at 120 (this is definitely a very sensitive amplifier input!)

The capacitor ratio 200p/1000p is trying to flatten the ultrasonic gain at about a factor of 5. So the addition of C1 doesn't completely take out the effect of C9, but it reduces it.

1000p//100 Ohms comes to 1.59MHz so it's probably out of the way as far as loop stability concerns.

If the speaker leads act as antennae for radio signals, C9 will route anything fairly efficiently back to the cathode of the input valve where it could get rectified. C1 will act to filter this a bit. It won't do much for medium wave, though it'll help with shortwave signals.

So I think C1 is more likely a bit of fine tuning to keep the feedback maths good with changes in transformer strays.

To do these things properly involves the 'Control theory' and Laplace transform side of maths, and having good modelling data for the output transformer as well as reasonable estimates for other strays. Leak should have done this. Not that many people in the downstream audio world are up to it.

So I'm always very nervous of 'mods' that affect feedback loops. Most students doing electronics degrees run in fear from control theory and plan answering other options in their exams. A few of us saw it as a saleable skill and something that could not be avoided once you started designing things.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 12:16 am.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.