John Curl's Blowtorch preamplifier part III

Status
Not open for further replies.
As specifically about reed relay distortion figure about 1 uV of distortion if the coil is feed from perfect DC. Tested at a signal level of 10mV.
0.01% ? More than what I was expecting, but not surprising I never noticed a big degradation ;-)
i'm just curious about your 10mv choice for signal level. I would have chosen something like 200mV.
About sensitivity of the mix desks inputs, I have seen all kind of things. Variable resistances, pots (in the feefback signal), step rotary switchs (the best solution) and even photo-resistances, like in some old limitor compressors.
 
Even here on diyAudio if you can imagine it!

Have a look at one or more of the threads here, entitled diyAudio <<project XYZ>> Build Guide. These are step by step sets of photos, showing how you go from a bare PCB that you purchased from the diyAudio Store, to a finished working <<project XYZ>>. Sort of a Monkey See, Monkey Do tutorial. But have a look at the questions that are asked. Cowabunga! Some question askers are so unencumbered by common sense, that they demonstrate "their elevator doesn't stop at MOST of the floors."

On a thread for an amp that I helped get into the Store, one guy asked a super weird question about the PCBs he received from the store. After some back and forth weirdness-decoding, it became clear that what he ACTUALLY wanted, was someone to tell him how to distinguish between the "Component Side" and the "Solder Side" of a two layer, thru hole, printed circuit board. This is a guy who successfully built other amps previously, he says. Cowabunga.
:rofl:
 
T, I don't want to challenge you about reed relays. We did a lot of work about 40 years ago on relays, and found that only gold on gold contacts to be adequate for signal level work. We found Shallco silver switches even better! Not just me, but Dave Wilson, and others, when they made careful listening comparisons. The most surprising relay that failed my listening test was mercury contact relays. They actually seemed to remove a whole layer of low level information, much like digital often does. We tested this before digital was popular, so we only used analog sources at the time. They made every piece of electronic equipment sound the same! No wonder that the guy who designed the mercury switches into a double blind listening box went into another profession.
 
The Schumann resonances are real and part of the a whole world of audio frequency radio, science is way more fun than audio.

Yes the resonances are real, but on a percentage basis DIY audio I suspect has more reasonable adherents than early Schumann fans.

My first exposure to them came from the folks who claimed they were evidence of aliens etc. They had other natural phenomena also impressively misinterpreted. (Roswell weather balloons, polished rocks that when rocked begin to spin etc.) (Worse yet MIT folks among them!)
 
s.

My first exposure to them came from the folks who claimed they were evidence of aliens etc. They had other natural phenomena also impressively misinterpreted. (Roswell weather balloons, polished rocks that when rocked begin to spin etc.) (Worse yet MIT folks among them!)

I can't speak to your experiences, I never saw anything but what made sense, I got my thank you right here from Rai Weiss and Kip Thorn of LIGO. Frankly I don't know what you are talking about, of course there are kooks everywhere without exception.

BTW the DIY adherents are as out there as one could imagine Schumann and scalar wave generators, garbage physics on parade.
 
Last edited:
Disabled Account
Joined 2012
Regarding reed relays.... the old HP 339A distortion analyser reads down to -100dB -- THD+N and it uses relay contacts both reeds and a big old clunky DPDT frame relay at the output. The ShibaSoku 725's use hermet sealed relays all over the place.

But I dont know how any of them sound


-THx-RNMarsh
 
Last edited:
T, I don't want to challenge you about reed relays. We did a lot of work about 40 years ago on relays, and found that only gold on gold contacts to be adequate for signal level work. We found Shallco silver switches even better!
You do not challenge-me about reed relay: I never made extensive listening comparisons on this subject like you. So I'm interested by your report.

No wonder gold is good, while it has less conductivity than cooper, it do not oxidize. No wonder silver is better when new, it is the metal with the best conductivity. But the problem with silver is they are sensible to ambiant pollution and deteriorate over time. So, the best relay on paper should be with silver contacts, silver or cooper conductors in a sealed enclosure with vacuum or neutral gaz inside. Mercure has 56 time less conductivity than silver.

This said, a lot of factors can be at work, by example, what was the other metal used in the mercure relay you dislike ? The one you tested was-it a wet contact (very thin layer of mercure) or plain mercure ?
An other thing is to not confuse relais used for HP lines, like in your double blind listening box (lot of currents and relatively high voltages), with line levels. The impedance are not comparable.
May-be it should be interesting to make a serious study on this matter, Without forgetting solid state relays, with distortions and noise measurements and listening comparisons. Both on line levels and HP ones. And new VS used ones.
If, like-you, I "trust my ears", I'm never satisfied when I have no technical explanation of something i can hear. Your "They made every piece of electronic equipment sound the same" concerns me. Why ?
 
Last edited:
My life ;-)
I never was an audiophile. When I was working as a sound engineer, I was in concern with the most accurate system i could find in my own listening room, to can judge the quality of my own work in studios as a known reference. Even if this system was not so agreeable. (high definition, attentive listening and high levels comes often with listening fatigue).
No pleasure in the listening moments, playing again and again the same little parts of music, or the same tune at various levels hunting my mistakes or trying to figure out what did my colleagues.
Same thing when I was on the design of some electronic or acoustic gear, as an audio designer, or when I was in need to listen to a mike, by example, as a studio technical manager, in order to complete a set of equipment.
And it is not the same activity, listening to music and listening to technical details.

Now, as I said previously, because I'm retired, I don't have the same needs.
I can enjoy listening music from youtube on my PC with poor (but balanced) speakers. Or to listen to music in my car despite my last one do not have a fantastic audio system (Old Volvo with his stock radio modified to add bluetooth). As well as walking in the country, with my Koss porta pro plugged in my smartphone playing MP3s copies of my favorite records.
I would add that I never liked to make listen friends to *my system* in order to impress them. When they see my horns, they often ask to listen to them, and I often answer my system is, for the moment, out of work.

The problem with my big system is it is so "defined" that, may-be 10% of the records sound fantastic on them and 10% OK. Means that 80% of others are less enjoyable because I can listen to recording or mix errors, artificial reverbs or an incoherent sound stage, bad balances, etc. As an example, because the localisation of the instruments are quite sharp, I'm exasperated when suddenly, a drums solo takes the whole width of the stage.

The last thing is that, because each and every piece of gear that I use in my system are home made or modified by me, I'm always, like an audiophile, looking unconsciously for faults in the system or something that could be improved.
Reason why i had ordered those Kef little self powered enclosures, in the hope I can forget about all this, and listen to the music in a more relaxed and indulgent way, concentrating on the playing of musicians without being "disturbed" by other preoccupations.

A philosophical questioning about the limits of the "audiophile" quest, or, if you prefer, when the quality of a "high end" system ends up harming our pleasure. If you prefer how a too transparent system can bring a wall between the music and you.
 
Last edited:
Bruno P had an article on his Mola Mola site that discussed the distortion in reed relays. The ferromagnetic contacts are the issue if I recall correctly, and good electromechanical relays measure better as far as distortion is concerned.
Bruno and Samuel published measurement results on reed relay distortion in Linear Audio Vol. 13:
volumes | Linear Audio
 
You do not challenge-me about reed relay: I never made extensive listening comparisons on this subject like you. So I'm interested by your report.
Yes, interesting findings.

Mercure has 56 time less conductivity than silver. This said, a lot of factors can be at work, by example, what was the other metal used in the mercure relay you dislike ?
Good question, good point. Mecury has (relatively) low conductance, which implies (relatively) high degree of scattering. This scattering has spectral and dynamic characteristics and will cause a masking response/reaction in the system which will cause overriding 'signature' sound. Additional materials in the relay may or may not play nice with the mercury and cause additional discord.

If, like-you, I "trust my ears", I'm never satisfied when I have no technical explanation of something i can hear. Your "They made every piece of electronic equipment sound the same" concerns me. Why ?
Agreed, 'house sounds' and other manufacturer characteristic/signature sounds are intriguing/interesting. IME yes this 'sounds the same' is possible (and normal/to be expected) when a dominant sounding material is in the signal path, and mercury is such a material, and in this case it is a dominant 'bad' sound and will be noted/noticed (and rejected). Such mercury whetted switches are in effect another BQP of sorts just the same as every other element/compound in the system, some elements/compounds have stronger effect than others. Nothing mysterious, it's just how things are !. The fun part is working out what does what to what and what plays nice with what, why this should be so and why we have preferences, after that it's easy lol.

Dan.
 
Last edited:
0.01% ? More than what I was expecting, but not surprising I never noticed a big degradation ;-)
i'm just curious about your 10mv choice for signal level. I would have chosen something like 200mV.
About sensitivity of the mix desks inputs, I have seen all kind of things. Variable resistances, pots (in the feefback signal), step rotary switchs (the best solution) and even photo-resistances, like in some old limitor compressors.

Good ears, bad math? 100 x 1uV / 10,000uV =?
 
Last edited:
Status
Not open for further replies.