Leak Stereo 20 and a Thank You

[Mike P]

Alex, your ST20 is more or less a copy of Tony's. For me this is the right balance of retaining originality, whilst at the same time taking advantage of some modern components.

Yours is built using Takman carbon film resistors, Sprague Atom small electrolytics, F&T large electrolytics and Russian K40y-9 PIO caps.

Deviations from standard spec are that the dual section capacitor at C10 and C11 is 50uF + 50uF rather than the std 33uF+33uF and there's also a extra 33uF in parallel at C12 (easily removed in moments if you so wish).

 

[Mike P]

Though by saying that if it sounds different, it is different. Passive preamps do sound different from one another, it may not be night and day, but it is there and a Stereo 20 is more than good enough to hear it.

I can't argue with that but the subject of resistor sound quality is a debate for another day.

The point I was trying to make is that for some reason inline attenuators seem to really get peoples audiophile paranoia going, and they really shouldn't, especially for fans of passive pre-amps.

 

[Tony L]

The point I was trying to make is that for some reason inline attenuators seem to really get peoples audiophile paranoia going, and they really shouldn't, especially for fans of passive pre-amps.

I don’t understand it either, and I really want to! I have a pair of Rothwell RCA attenuators that suck the life out of any interface I’ve tried them on. I’ve not tried measuring anything so I don’t know what they are doing beyond adding attenuation (maybe capacitance?), but I can hear it exactly as Alex describes. Not night and day, but enough to be obvious and something I obviously don’t want.

 

[Alex S]

@Mike I understand. I certainly have found different passives to sound different although that might be the switch. I don’t know what resistor is used in Goldenjacks but my Khozmo is stuffed full of Takman Rex and Vishays. Whether they sound better or even different to any others I can’t say.

Most of my audio nervosa is behind me thankfully but since I can find a volume position on the Khozmo for all sources without attenuation I might as well. I added the Goldenjacks (which I find less damaging than Rothwells) to see if I could equalise the volume between sources but after an hour or two they had to go.

 

[Mike P]

I've read this stream with interest as I'm thinking a nice ST20 should be something to grace my living room, soon. I have a bonus incoming and thinking I'd like to assign a portion of it to laying my hands on one.

FWIW Alex, although not specifically related to the Leak (as I've never owned one, yet), I've recently moved on a maxed out modded Jolida DAC, and I spent quite a bit of time tube rolling. The EI Yugoslavia tubes mentioned above are special indeed and I rate them v.highly.

If/when I pick up a Leak that's the first thing I'll try to pick back up for it.

The Leak rebuild, looks awesome.

Keep an eye on the classifieds. I have another one, which will be available as soon as some new valves arrive for it.

 

[GT]

I have one too which I will fully restore later this year and will be offered for sale...

 

[Mike Reed]

I don’t understand it either, and I really want to! I have a pair of Rothwell RCA attenuators that suck the life out of any interface I’ve tried them on.

My only experience with attenuators was a few years ago when trying to stop the massive 3+ V output from my CDP from swinging the meters onto the stops on my pre's meters. A friend brought his -10 and -20 ones round and it flattened the dynamics; the -20 audibly more than the -10. I later took a punt on a pair of Goldenjacks -6 dB and couldn't detect any sonic deterioration; just sufficient attenuation to do the job. Still have these in situ.

I rather think, and this is what I've read in the past, that the amount of attenuation is the key. Of course, it could also be the specific application involved. I remember being advised on here at the time to get -25 or-30 dB ones; glad I didn't. I can't think that Goldenjacks (now ceased, I believe) is any better than Rothwell or vice versa but if attenuation is taking place, surely impedance is being affected. That's a totally non scientific and probably spurious thought, b.t.w. I use mine at the receiving end (pre.) as I believe is the more sensible option.

 

[Alex S]

Mike, out of interest, what is that extra 33uF at C12 doing? I am ignorant of such things.

 

[GT]

My only experience with attenuators was a few years ago when trying to stop the massive 3+ V output from my CDP from swinging the meters onto the stops on my pre's meters. A friend brought his -10 and -20 ones round and it flattened the dynamics; the -20 audibly more than the -10. I later took a punt on a pair of Goldenjacks -6 dB and couldn't detect any sonic deterioration; just sufficient attenuation to do the job. Still have these in situ.

I rather think, and this is what I've read in the past, that the amount of attenuation is the key. Of course, it could also be the specific application involved. I remember being advised on here at the time to get -25 or-30 dB ones; glad I didn't. I can't think that Goldenjacks (now ceased, I believe) is any better than Rothwell or vice versa but if attenuation is taking place, surely impedance is being affected. That's a totally non scientific and probably spurious thought, b.t.w. I use mine at the receiving end (pre.) as I believe is the more sensible option.

The thing to note is the input impedance of the Rothwell attenuators (11dB version) post series resistor, is about 10K ohm. Obviously, if you connect that attenuator to the input of a valve amplifier, the input impedance is now the input impedance of the valve amp, which is now in parallel with the 10K ohm of the attenuator, so the amplifiers input impedance will now be less that 10K ohms.

For example: if the amps input impedance was 100K, with the attenuator fitted the amps input impedance will now be about 9K ohms. This is more than likely the problem as these attenuators look like they were designed for use in solid state products, where input impedances are much lower, as in 5K to 20K ohms compared to valve/tube amps, which are normally 50K ohms and above, typically 100K ohms. For valve/tube amplifiers you could use a higher parallel resister in the attenuator, but then you might well get noise issues (hum) due to the higher series resister that would be needed. This also to some extent depends on the output impedance of the driving stage from the preamplifier which is driving the attenuator connected to the amplifier. A very low output impedance from the driving preamp would help enormously here.

The Leak Stereo 20 has an input impedance of 1Meg ohm, so with the 11dB attenuators fitted, the input impedance is now approximately 9.9K ohm. A far cry from 1 megohm and I suspect this is the problem that people are experiencing...

Note: I have ignored the series resistor in the attenuator for these calculations above.

 

[Mike P]

Mike, out of interest, what is that extra 33uF at C12 doing? I am ignorant of such things.

The power supply in the ST20 is a C-R-C arrangement and C12 is the second capacitor in that. Increasing the value will lower the ripple and help 'stiffen' the power supply.

You can't increase the first capacitor much as it will overstress the GZ34 (datasheet states 60uF max).

If it offends anyone it can be removed a seconds leaving no trace.

 

[Mike Reed]

The thing to note is the input impedance of the Rothwell attenuators (11dB version) post series resistor, is about 10K ohm.

Impressive, Graham; can't fault your analysis because it's way beyond my expertise in understanding it. . What you're saying is that the results from attenuators would differ depending on whether they are connected to a valve or s/s amp, I believe. I also think what you're saying (but less sure of this) is that my valve pre would be much more sensitive to attenuator resistance (in terms of s.q. as well?). Although I'm a technical numpty, I can maybe just about understand why from your analysis, so my -6 dB attenuator would have much less effect in reduced s.q. but still reduce the attenuation sufficiently, which is what we found.

Presumably the same resistance effects would be seen if the attenuators were at the source end, though I gather this is not best practice.

 

[Shane2468]

I saw these 2 videos last night. The repair man has an interesting voice for sure, lol.

S.

 

[toprepairman]

Good grief! that guy should be shot.

 

[Tony L]

I remember cringing at that one years ago! One can argue it saved the amp from landfill, but still…

 

[paulfromcamden]

I saw these 2 videos last night. The repair man has an interesting voice for sure, lol.

S.

His rug really doesn't tie the room together.

 

[GT]

Impressive, Graham; can't fault your analysis because it's way beyond my expertise in understanding it. . What you're saying is that the results from attenuators would differ depending on whether they are connected to a valve or s/s amp, I believe. I also think what you're saying (but less sure of this) is that my valve pre would be much more sensitive to attenuator resistance (in terms of s.q. as well?). Although I'm a technical numpty, I can maybe just about understand why from your analysis, so my -6 dB attenuator would have much less effect in reduced s.q. but still reduce the attenuation sufficiently, which is what we found.

Presumably the same resistance effects would be seen if the attenuators were at the source end, though I gather this is not best practice.

Like most things in audio Mike, its swings and roundabouts. You need to attenuate the signal, which the attenuators will do, but the downside is they will effect the audio performance, some more than others depending on the equipment design and how much attenuation is needed. You can help to minimise the audio degradation a little by using better quality resisters in the attenuator, but there will still be some form of sonic degradation. Remember that any component in the signal path will degrade or effect the signal passing through it.

I would not recommend using the attenuators other than the direction they were intended to be used in. So in a preamp to power amp situation. The attenuators should be sighted at the input of the power amplifier.

 

[Alex S]

I know nothing of electronics but the idea of living with that mains transformer bolted on the back gives me aesthetic shivers.

 

[Alex S]

For the last few days I've been rolling a few valves, only on the ST20 input ECC83 since I didn't want the exercise to become tedious or tiring (or too expensive).

Paulfromcamden very kindly popped round on Sunday with a few to try, some I haven't yet. I tried a JJ first and it was okay, fine given it's price and availability. Next I tried a GEC - I liked this enough to get one (similar but different) from Watford immediately: https://www.watfordvalves.com/product_detail.asp?id=681 This is an expensive valve and is in the Leak now. Following a recommendation here I also bought an EI elite smooth plate. These aren't cheap either so I have stopped there and will tell you what I think, in brief: NOS 1970s Mullard - inviting but fully detailed; 1960s used Mullard - v slightly preferable to the NOS 70s; GEC - like the Mullard but perhaps cleaner (if it's good enough for Stevie RV it's good enough for me); EI Elite - probably my favourite. Like the GEC but more drive.

None of my thoughts relate to reliability/longevity but having tried JJ, EH, Golden Dragon, I conclude that the price premium for good NOS valves is worth it (in the ST20 at least), and I can very happily live with any of the ones mentioned.

 

[Mike P]

Thanks for sharing your findings Alex.

 

[The Bish]

For the last few days I've been rolling a few valves, only on the ST20 input ECC83 since I didn't want the exercise to become tedious or tiring (or too expensive).

Paulfromcamden very kindly popped round on Sunday with a few to try, some I haven't yet. I tried a JJ first and it was okay, fine given it's price and availability. Next I tried a GEC - I liked this enough to get one (similar but different) from Watford immediately: https://www.watfordvalves.com/product_detail.asp?id=681 This is an expensive valve and is in the Leak now. Following a recommendation here I also bought an EI elite smooth plate. These aren't cheap either so I have stopped there and will tell you what I think, in brief: NOS 1970s Mullard - inviting but fully detailed; 1960s used Mullard - v slightly preferable to the NOS 70s; GEC - like the Mullard but perhaps cleaner (if it's good enough for Stevie RV it's good enough for me); EI Elite - probably my favourite. Like the GEC but more drive.

None of my thoughts relate to reliability/longevity but having tried JJ, EH, Golden Dragon, I conclude that the price premium for good NOS valves is worth it (in the ST20 at least), and I can very happily live with any of the ones mentioned.

That GEC is a 5751 which has 70% of the gain of a regular ECC83 - I really like them in ECC83 preamps as normally quieter and in my experience a bit more '3-D'. Prices have really jumped, they were £25 including VAT a couple of years ago.