Passive preamp buzzing?

[Arkless Electronics]

The so-called "series" resistor and "shunt" resistor in a potential divider are both doing the same job and one cannot be more more important than the other. Imagine as an example a poor quality resistor which has a lot of parasitic inductance: if you put it in the shunt position where, it is claimed, it doesn't matter it will still form a high pass filter. So it must matter! Of course it does or else why have it there at all? There's no rational explanation as to why one of the two resistors should be more important than the other - other than expectation bias based on a misunderstanding of electronics and a belief in "the signal path".


I think you're being a bit unfair there, Jez. It's perfectly possible to have too much gain at the front end of a circuit - so it distorts - but also have some attenuation further down the line so the actual signal level doesn't appear to be overly "hot". Less gain at the front and more gain (or less attenuation) later on would cure it.
BTW, I'm not familiar with the exact circuit details of the Jolida but I have read that it does suffer from piling on the gain and then simply cutting the level with attenuation.

I was waiting to see if anyone who knows what they're talking about would come along and say that! Yes technically there is such a thing as gain structure and it is important. What I'm ridiculing is the concept of there being a subjectively important gain structure. So long as overload margins and noise have been taken care of by proper design it makes no difference if the vol control has to be at 5 or 8 for the same volume.

[Marco]

The so-called "series" resistor and "shunt" resistor in a potential divider are both doing the same job and one cannot be more more important than the other. Imagine as an example a poor quality resistor which has a lot of parasitic inductance: if you put it in the shunt position where, it is claimed, it doesn't matter it will still form a high pass filter. So it must matter! Of course it does or else why have it there at all?

Of course it matters. Who said it didn't? *Everything* matters!

Unfortunately, there was no practical way of removing and upgrading the surface-mounted shunt resistors on the existing switches (primarily because of their minuscule size, and fitting something else better, but unavoidably much larger, such as Z-Foils). Especially on a pair of attenuators with 47 different steps!

Therefore, the only way to have done that would've been to start again with two new Elma switches, and populate them with Z-Foils, and even for me, that was a bridge too far, which is why the Goldpoint was eventually sold.

There's no rational explanation as to why one of the two resistors should be more important than the other - other than expectation bias based on a misunderstanding of electronics and a belief in "the signal path".

Ha - now you're beginning to sound like Jez!

Fortunately, I don't sign up to such absolutist thinking, and simply judge what my ears tell me [which was that replacing the existing series resistors with Z-Foils massively improved the sound. Indeed, such was the level of sonic improvement, you'd have needed to have been deaf not to have heard it].

And I wasn't alone either, as three other people were present at the time when the resistors were changed, with the Goldpoint first being listened to in stock form, and then after the Z-foils had been fitted, using the same music and test system.

Expectation bias is always a possibility, but the same upgrade was carried out afterwards to two other passives, with the same result, so I'm confident that the improvements were genuine.

If I can find it, I'll dig out the thread where I reported on this, complete with pictures of each type of resistor that was tested. From memory, there were at least five of the same value from different manufacturers - and all sounded markedly different!

It's perfectly possible to have too much gain at the front end of a circuit - so it distorts - but also have some attenuation further down the line so the actual signal level doesn't appear to be overly "hot". Less gain at the front and more gain (or less attenuation) later on would cure it.

Precisely, and that's essentially what I refer to as the 'gain structure' of a system. It's entirely valid too, as are the results when alterations are made to how it operates, which Oliver and I (not to mention plenty of others) can clearly hear.

Marco.

[Bigman80]

It's quite important to configure the attenuator in a way that never puts an undue load on the source or has an unduly high output impedance. However, that's quite simple - just wire the attenuator as a conventional volume control rather than trying to have "just a single resistor in the signal path". I set out my reasons for this in post #24 of this thread: http://theartofsound.net/forum/showt...tenuator/page3

At its simplest, you could make a passive preamp with just a pot. I'd suggest a pot between 10k and 25k. 10k will have the advantage of a lower output impedance. 25k will have the advantage of a higher input impedance.
There are plenty of pots to choose from here:
https://www.hificollective.co.uk/com...tiometers.html

Next level up is to use a stepped attenuator instead of a pot. There are plenty to choose from here:
https://www.hificollective.co.uk/com...tenuators.html

25 years ago when I started to make passive preamps there were hardly any stepped attenuators available - now there seem to be hundreds! I used to make my own with 132 steps If there's so much interest in stepped attenuators, how come I never got rich?

Thanks Andrew. My current pre has a 50k Alps blue. Ideally, I'd like to Taylor it to the Jolida

Sent from my EVA-L09 using Tapatalk

[RothwellAudio]

[/B]
I was waiting to see if anyone who knows what they're talking about would come along and say that!

I've never claimed to know what I'm talking about

[Marco]

From earlier:

For the record, there is no such thing as gain structure as in something to be optimised.... other than technically.

And just now:

Yes technically there is such a thing as gain structure and it is important.

So, if I change the electrical value of the series resistors in a passive preamp, thus making a technical change, in order to optimise the gain structure [perhaps it's better referred to here as 'loudness' rather than gain?] and achieve a similar effect to what Andrew has described (subsequently hearing the obvious difference), then according to you there is no such thing as that?

The fact that I succeeded in optimising results, as described (by moving the position on the volume controls where maximum gain is achieved), disproves your statement above in bold.

It is undeniable that Oliver would've been able to use the Goldpoint with his Jolida and Firebottle monos, optimally, without hardly being able to turn up the volume on the Goldpoint before it was just too loud (as full gain was arriving too early on the attenuators), if it had been fitted with a different value of series resistors - and it would be an easy one to demonstrate.

Quite simply, going from the existing 10K, to something such as 50K, would've meant he'd have had much more usable range on those controls. Maybe, in this instance then, we should call it 'loudness structure', rather than gain?

I agree that the gain level itself, present in the circuit or system remains unchanged, but how it arrives, in terms of where on the volume controls maximum gain is achieved, does

Marco.

[RothwellAudio]

Of course it matters. Who said it didn't? *Everything* matters!

Unfortunately, there was no practical way of removing and upgrading the surface-mounted shunt resistors on the existing switches...

Sorry, my mistake - I was putting words into your mouth. I should have said that it is often assumed that one resistor is in the signal path and the other one isn't, therefore the one that isn't is of lesser importance.
BTW, I have some Elna switches with through hole PCBs (unpopulated) instead of surface mount resistors. They're available cheap if anyone wants one.

[Marco]

Sorry, my mistake - I was putting words into your mouth. I should have said that it is often assumed that one resistor is in the signal path and the other one isn't, therefore the one that isn't is of lesser importance.

No worries. Yes, apparently the shunt resistors aren't supposed to matter, sonically, as technically they're not in the signal path. If that's incorrect, then I'm simply quoting what I've often seen mentioned.

However, tests proved that this wasn't the case, and that they were clearly influencing the music signal, which is another good reason to ultimately trust your ears, and if necessary, throw theories out of the window!

Marco.

[RothwellAudio]

No worries. Yes, apparently the shunt resistors aren't supposed to matter, sonically, as technically they're not in the signal path. If that's incorrect, then I'm simply quoting what I've often seen mentioned.

However, tests proved that this wasn't the case, and that they were clearly influencing the music signal, which is another good reason to ultimately trust your ears, and if necessary, throw theories out of the window!

Marco.

Sorry, I'm confused now. Did the tests show that the shunt resistors "were clearly influencing the music signal"? If so, that's what I was saying.
I'm not aware of any theory which would explain why one resistor should be more important than the other, so no theory needs to be thrown out of the window - only the internet rumour that the "signal path" resistor is the important one needs to be thrown out of the window.

[Ali Tait]

Oli, have a look here at the 200 dollar autoformers. One of the best bargains in audio IMO.

http://www.intactaudio.com/atten.html

These make a superb passive pre.

[farflungstar]

Oli, have a look here at the 200 dollar autoformers. One of the best bargains in audio IMO.

http://www.intactaudio.com/atten.html

These make a superb passive pre.

Yes they do - I had the ready built modules for a while before selling to a member. Easy to hook up and imo better than mfa.