• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Leak Stereo 20 - solid state rectifier?

With reference to: http://www.hilberink.nl/codehans/leak1.htm

in which Tim Bailey modifies his ST20 for, among other things, solid state HT rectifier and huge capacitance.

Anyone here tried this? I'd like to, but I'm scared that if I simply substitute silicon diodes for the GZ34 then the HT rail will be way too high.

I tried mailing Mr Bailey, but he's no longer at the address on this page.
 

EC8010

diyAudio Moderator Emeritus
2003-01-18 7:57 am
Near London. UK
I'd be scared too. With all that HT it's no wonder he can obtain more power with lower distortion. I wouldn't expect the output valves to last long. And the strain on the mains transformer must be considerable. Mind you, it won't help the fact that the output transformer begins to saturate at 10W at 50Hz.
 
EC8010 said:
I'd be scared too. With all that HT it's no wonder he can obtain more power with lower distortion. I wouldn't expect the output valves to last long. And the strain on the mains transformer must be considerable.

Tim's HT rail is at 320V, about the same as the original. Which implies that he did something other than simply replace the GZ34 with silicon diodes.

Mind you, it won't help the fact that the output transformer begins to saturate at 10W at 50Hz.

Really? Maybe I should stick it on Ebay and spend the cash on a clone based on Sowter transformers...
 

EC8010

diyAudio Moderator Emeritus
2003-01-18 7:57 am
Near London. UK
The going rate for a Stereo 20 (in good condition) is about £250 to £300. A pair of output transformers and a mains transformer from Sowter will probably cost about that. I know people will call it heresy, but it's best to view a Stereo 20 as a transformer kit with a chassis thrown in for free. It all depends on how handy a cloth-eared old git of a biker is at metalbashing.

To get more power, either the output stage has a higher HT, or the bias has been changed towards Class B - probably with grid bias. Oh, or the output stage could have been changed to pure pentode.
 
EC8010 said:
The going rate for a Stereo 20 (in good condition) is about £250 to £300. A pair of output transformers and a mains transformer from Sowter will probably cost about that. I know people will call it heresy, but it's best to view a Stereo 20 as a transformer kit with a chassis thrown in for free.

Except that the Sowter transformers (UA21) look a lot better...

cloth-eared old git of a biker

Have we met?;)
 
EC8010 said:
No, I just looked up your profile.

Ah, the old RTFM ploy. Cunning.

Are you sure you mean UA21 and not U064? U082 looks very nice.

Their site lists UA21 as being suitable for the ST20 but right enough U064 looks more likely.

I think I might be out of my depth here - tell you what, I'll go and read 'toobs for noobs' and take it from there. Right after I put these 1N4004s into my ST20...
 
Some realities / confession

(':smash:')(':smash:')

If you all re-read the article you'll note the word WE a lot.

For a long time now I've stuck to a promise to my wife ......
(':rolleyes:')(':rolleyes:')(':xeye:')(':xeye:')
to avoid risking my life working on tube amps, so they tend to be collaborative efforts. I'm ADD or kind of.

We modified two of the little darling St20's, JBTW. The other St20 is in triode mode, and does R & L channel tweeter duty.

I do the thinking and the simpler/safe stuff. I don't do a thing with them ON, except check voltages, and the PSU stuff happened in round 2 of the restore/mod process, a long time ago. So, I forget exactly how we did it.

Note that when we put in the soft recovery diodes, types recommended by TPC in Canada, we had about 340-50 volts HT coming off the array. There are resistors in the PSU, which keep the HT to near or just above the original level in the circuit.

One is now in pentode mode and does bass duty. 20 watts?! 20 to 20 k .....

So, I think the problems posted above at 10 watts - 50hz in the output trannies are individual to the trannie measured, or possibly elsewhere with the PSU.

WHY?

Well we measured 20 watts @ 20 hz 1.5 % TD, and 20watts @ 20khz, when in pure pentode mode into 8 ohms and into 4 and into 16.

Another support for this view is that the trannies for the St50 and St60 are hardly any bigger - metal core mass wise than the 20's. And, saturation is closely related to core size.

The amplifiers have been on duty now for well over 5 years and I have NOT had any trouble. Mind you this is in inland Australia, and in summer things do get very hot, over 40 C ambient outside on occasion, and this affects the trannies, especially the PT. But it is never over the safe rating at the casings.

I suppose I should check some voltages, ..... soon?!
(';)')
(';)')

I'm not really asking that much more from the amps, and yes the PT is pulsed a bit harder, but the duration of each pulse IS much shorter. I also use EL84M's now, and they last and last.

When I was first using them I really wanted to go original, or better but still PIO's etc and CCR's, and I think the piece covers the problems I had. Watching Chinese 5AR4's fail was enough to put me off valve rectification, pretty though the light show was.

It also MAY mention the blow up yers before of a loaner Radford STA25 - done over by Trevor Lees in Melbourne. They came with SS diodes OEM, and the ones TL used were well snubbed, to boot, as are mine in the 20's. The smell from a 'house-full' of smoke, took weeks to go away, ...... PCB's from the burning OPT? I don't know.
An NoS PIO cap failed., took out a resistor on standees,a nd then the OPT. The noise of the blow-up was VERY loud. So, my focus on the 20's included reliability.

I know that it is currently possible to approach my very low level of ripple, and get similar levels of storage, but only with TWO 'good' octal rectifiers, plus a choke/cap/ etc/ ladder PSU. I know that some new GZ34's are almost as good as $$$$$$$ NoS Mullards are.

Reliability is also why we didn't use any especially exotic circuit caps or resistors, NO PIO's - just 400 volt and 630 volt Wima MKP's plus MKS(polystyrene). the NFB circuits used the best resitors and caps I could afford. The last service and check 'round 4/5', we put non inductive R's in for cathode bias duties, added a 4th cap to the bass~mid amp, and put in a matched octet of EL84M's. Packed / marked as 2 sets of two pairs!

I would also like to point out that the OEM small signal tubes are high gain tubes and that lower gain/gruntier tubes allow you to drop NFB in their part of the circuit, a LOT. And sensitivity needn't fall all that much, ie still ~ 200 mv.

Of my then available options, the 12BH7, the 6DJ8 family, the 6CG/FQ7, and the 5687, I chose the 6CG7's for the splitters, (and 6DJ8 gold pin mullards for shared input tube duty), a friend had 6 NoS Aussie made, and I got them for virtually nothing! These used the same internals the local 6SN7's had. The Mil-spec version of these, is now gaining a good rep. This was round 2, too!

You might all like to fall about laughing that some idiot posted what was I doing using 'just' a B&W TV line-flyback driver?! He'd obviously not read any CJ or ARC circuit layouts!

Later when we added a CCS to the 6CG7's, we found again that we we needed even less NFB - again! Round 3.

We DO have a fair bit around the OPT and the 84M's.

If I ever build a separate PSU chassis it will STILL use SS diodes, soft recovery types, well snubbed, but there will be chokes, and I might even go for a fixed bias rail. Probably a dual winding PT!

24 watts from 7189's/EL84M's, in UL! maybe even 30!

Roger Modjeski better watch out!

While I'm not a self identifying objectivist, most of whom wouldn't know a beta from an alpha, I just don't get 'the latest super cap / super resistor thing'

BTA I did use Vishay R's and polystyrene caps in the passive valve RIAA stage, IIRC this was when we added AW's 50K turnover point.
It had MKP's before that, just like Bottlehead uses.

Has anyone here tried CCS's for their phono stage? If it uses two twin triodes you'd need 4? yes? and you'd need to match them too, no?

Timbo in Oz















:xeye: :xeye:
 
Has anyone here tried CCS's for their phono stage? If it uses two twin triodes you'd need 4? yes? and you'd need to match them too, no?

It depends on the circuit. For example, you couldn't use one on a cascode input. Bottlehead does use CCS's as an option for their Seduction phono stage. It requires a tweak of the RIAA...probably the series resistor...but it's supposed to make a marked improvement. I think that Gary Pimm also uses them in his phono stage, so you might want to take a look at his site.
 

ray_moth

diyAudio Moderator Emeritus
2004-01-27 8:55 am
Jakarta
I know people will call it heresy, but it's best to view a Stereo 20 as a transformer kit with a chassis thrown in for free.

Love it! However, I doubt that the Stereo 20 "transformer kit" is worth buying. The OPTs are nothing special and the power tranny gets hot. Leak was definitely out to save a buck with this one. I agree with the comment that replacing the thermionic rectifier with SS could put too much strain on the power tranny.

Reading the 'LEAK " POINT ONE " AMPLIFIERS' booklet, which was published to commemorate the TL/12 (of which the Stereo 20 is a descendant), is like reading a religious tract! It's hard to reconcile Leak's "triple loop" claim with the schematics of the Leak amps. There is obviously a global FB loop but where are the other two loops? Leak claims they're in the first and second stages but I can't see them.
 

SY

diyAudio Moderator Emeritus
2002-10-24 10:19 pm
Chicagoland
www.SYclotron.com
ray_moth said:

It's hard to reconcile Leak's "triple loop" claim with the schematics of the Leak amps. There is obviously a global FB loop but where are the other two loops? Leak claims they're in the first and second stages but I can't see them.

There is bootstrapping of the inputs of the long-tailed pair which will lower the effective cut-off frequency of the RC coupling from the first stage. That's probably what they meant.
 
Agreed, the output transformers are nothing special, but (sadly) I've seen far worse. The main worry is the mains transformer, as you've pointed out.>>

The OPTs do sound better with the secondaries rewired. I agree that there are better, but this mod certainly helps, and I'm glad I rewired mine. I believe I got this info from EC8010 in a previous post, in fact, if my memory is correct (should be in the archives). I've rebuilt my Stereo 20 to use 8 of 6S4A triode outputs in total, with ss rectification (HT is about 360v). I'm very pleased with the results. I was worried about the notorious transformer overheating, but with this circuit it actually runs cooler! Nice surprise. I used the 5v supply of the GZ34 to run a pair of 5BK7s as inputs. After years of different mods to my stereo 20 (I've had it since 1962!!) I think this is the final version. I personally was glad to see the backs of the EL84s, though I know this valve has a big fan club!! I love the look of Leak equipment, but I'm sentimental and a sucker for retro-chic. Andy
 


I've rebuilt my Stereo 20 to use 8 of 6S4A triode outputs in total, with ss rectification (HT is about 360v). I'm very pleased with the results. I was worried about the notorious transformer overheating, but with this circuit it actually runs cooler! Nice surprise. I used the 5v supply of the GZ34 to run a pair of 5BK7s as inputs. After years of different mods to my stereo 20 (I've had it since 1962!!) I think this is the final version.

I am full of questions: Is that just a pair of silicon diodes in place of the GZ? Is R21 (the 100R between the reservoir and smoothing cap) still in place? Did you increase the caps? Can we see a schematic? Should I buy a pair of Monitor Gold 12s? Oh, sorry, wrong forum...
 
The circuit with 6S4As is totally different. Input is 5BK7 as a diff pair with a CCS under it, then Russian teflon .1 coupling caps to parallel PP output triodes in fixed bias but with a small cathode resistor unbypassed of about 47 ohms on each cathode. the power supply is full wave ss rectifier, 100uf cap, then splits into one choke per channel of 10H 100mA, which feeds the outputs. then another couple of 100uf caps and one choke per channel of 10H 100mA and finally another couple of 100uf caps per channel for the input. Total four chokes and five caps, CLCLC. Getting all that under the chassis was an exercise in ingenuity. On the top of the chassis there are 10 valves in total which take up all ten holes, having removed the two PSU caps and the rectifier. I made a plate on each side to take the new valve bases so it looks pretty nice. I also raised the amp up on nice chunky gold feet so the chokes would clear. I have a draft schematic, but I'm not sure how to post it here. Ideally I should revise it so it's up to date. You also really need a picture of the underside to see how it all fits. Andy
 

ray_moth

diyAudio Moderator Emeritus
2004-01-27 8:55 am
Jakarta
There is bootstrapping of the inputs of the long-tailed pair which will lower the effective cut-off frequency of the RC coupling from the first stage. That's probably what they meant.

That's only one "loop", not two, assuming that it even deserves to be called feedback. It's a daft idea anyway, IMHO, because it incurs an extra coupling cap in the global loop with no attempt to ensure LF stability. Seeing that Leak went to the extent of actually naming their range of amplifiers TL/12, TL/25 etc., one would expect to see evidence of genuine triple feedback loops.

I used to like Leak gear and, back in 1962, I even built myself half of a Stereo 20, using a Parmeko transformer, together with half of a slightly simplified Varislope preamp. They seemed to work OK. However, when I got to learn a little more about these things, I came reluctantly to the conclusion that much of Leak's self-congratulatory publicity was just advertising puff.

[sorry for the rant]
 

SY

diyAudio Moderator Emeritus
2002-10-24 10:19 pm
Chicagoland
www.SYclotron.com
Well, it's really two loops since both sides of the diff amp are similarly bootstrapped. But philosophically, I'm in agreement with you- "triple loop" seems a bit puffy. I think the "TL" might have meant "trough line," though that might just be the tuners. Anyway, we see where the Japanese craze for throwing fancy names at prosaic circuits may have originated...

The bootstrapping does lower the f3 of the coupling significantly, which does help stability.