Goldpoint SA4M-47 passive 'preamp' winging its way to Marco Towers...

[Marco]

Incidentally, just a further note on the Shinkohs and AN Tant (non-magnetics). It was pretty clear to us that the latter were, musically, in a different league - and I can go into the required detail if necessary. Therefore, I would say that for those looking to use Tantalum types of resistor, the AN non-mags are the daddies!

Even though the Shinkohs weren't fully burnt in, they would've had to transform themselves at the level of going from almost black to white, to have competed with the ANs. The latent potential simply wasn't there, IMO, to have allowed them to usurp the ANs, even after a longer burn-in period. Furthermore, I had a judgement call to make, unless I wanted to spend weeks on end testing resistors!

There's a point during these experiments where you have to say: 'job done' and be confident that you've made the right choice.

In that respect, I'm totally satisfied that the Z-Foils are the best resistors to use in the Goldpoint. As ever, though, that may not be the case in other applications, where Shinkohs could be better. In that respect, it'll be interesting comparing Macca's NVA passive, with the Goldpoint, at the forthcoming NEBO event in November.

Marco.

[struth]

Guess it will depend on what else is in the system as to a possible winner unless its in your face better...interesting read it will make as obviously i wont be there

[awkwardbydesign]

There's a point where you have to say 'job done' and be confident that you've made the right choice. In that respect, I'm very satisfied that the Z-Foils are the best resistors to use in the Goldpoint. As ever, though, that may not be the case in other applications, where Shinkohs could be better. In that respect, it'll be interesting comparing Macca's NVA passive, with the Goldpoint, at the forthcoming NEBO event in November.

Marco.

And initially you liked the extra weight of the ANs. I wonder if the time or mood make a difference to our choices too. In my DAC I wanted a clean sound, but now some extra "authority" might be welcomed.
BTW, the Mills MRA5 wirewounds are available in 20k and 22k. Go on, you know you want to!

[lurcher]

One of the problems with quoting Wikipedia is that it often generalises and is not thoroughly checked. I was referring to manufacturers' specs.
Thermal noise probably isn't a problem at these levels, but then the idea that resistors sound different is - resisted!

I was quoting Wiki because it saved me typing. As it happens there is a error on the page I quoted from, but its not in that quote. You may have been quoting manufactures specs, I was repeating well known and understood physics. The thermal noise of a resistor is dependent on its resistance and nothing else. Makers can work at reducing other noise sources but they cant touch thermal noise, thats written into the fabric of the universe. I was not claiming that resistors don't sound. different, they do, I pointed this out several pages ago in fact.

I was just correcting your statement

"non inductive wire-wound resistors can exhibit some of the lowest Johnson noise"

which is flat wrong, every 20k resistor will produce exactly the same Johnson noise, what would have been correct is:

"non inductive wire-wound resistors can exhibit some of the lowest overall noise"

[awkwardbydesign]

OK, which might make them worth trying. Actually that strengthens my point, so thanks for that.

[Marco]

Guess it will depend on what else is in the system as to a possible winner unless its in your face better...interesting read it will make as obviously i wont be there

Oh, most certainly. Passive preamps are up there as one of the most system-dependent components there is, so much will depend on which one the test power amp 'likes' best, and vice versa. Most important of all is that the source used has a healthy (min) 2V output, in order to drive the various passives properly. That will make or break results further down the chain.

In that respect, it's probably advisable for the source used to be digital. The last thing you'd want is a 'weedy' cartridge and/or phono stage upsetting the apple cart.

Also crucial will be placing the partnering equipment in such a way that the shortest possible interconnects can be used with the passives being tested - and those interconnects must also be low-loss, low-capacitance types. However, returning to my situation, I'm satisfied that I've got the Goldpoint where I want it to be for now.

I now only have one 'audiophile-grade' resistor in series with the signal path (the Z-Foils), and with having maximised the available volume on the lower steps of the switches (by using the optimal value of resistor for my system for that job: 20K), and with how the Goldpoint attenuators work, with less shunt resistors used in the early steps, than further up, I've optimised things for use at my favourite listening levels, positions 3-10, on the available 47 steps.

There will be some experiments carried out in due course with the shunts, in terms of upgrading them if necessary. However, based on what I'm hearing now (the sound is so open, detailed, dynamic and goddamn musical), I'm confident that their sonic effect is not as significant as that of the series resistors, and also that the shunt resistors currently fitted to the Goldpoint attenuators (SMT Nichrome) are already very good.

Quite simply, if they were seriously crap and holding things back to a notable degree, Duncan and I would've been able to hear it. Judging now by the performance of the Goldpoint, there's just no way that's the case. However, that's not to say that they can't be improved, but they're definitely NOT acting as a significant bottleneck

Here's an interesting article to read on the subject, and also on Audio Signal Attenuators in general: http://www.blackdahlia.com/html/tip_67.html

In terms of this discussion, this bit is most relevant:

The advantage of a series attenuator is that you only have one switch contact in the path. The disadvantage is that all 24 resistors are in the signal path all the time. Normally one would observe that the disadvantage out weighs the advantage, but it’s not so simple, as this type has received a new lease of life with the likes of DACT (which is a highly regarded piece of kit), and the much better priced Goldpoint attenuator. DACT and Goldpoint both take advantage of surface mount technology (SMT) type resistors, which are very small in size, meaning the 24 resistors can be tightly packed in a manner that almost resembles a single resistor. What I mean to suggest is this. Imagine a single resistive tract cut into 24 small pieces, and then soldered together, with switch contact between. If you can implement this neatly – tightly spaced and clean – then you start to feel a whole lot better about the series attenuator approach. DACT and Goldpoint do exactly this, thanks to SMT.

The emboldened text reflects my own experience of using controls with SMT-type resistors, as the DACT series-type attenuator I use in the Croft outperformed any shunted 'boutique resistor'-based attenuators I'd used in the past, which certainly confirms the theory that not all SMT-type resistors are bad. Again, as with many things in audio, it's all about context and implementation!

Therefore, given the above, it's no surprise then that the Nichrome resistors, employed by Goldpoint on their attenuators, aren't letting the side down. Indeed, quite the opposite, as they appear to be pretty much sonically transparent, which is why I'm in no hurry to replace them.

In terms of this bit from the above article, I quote:

"The advantage of a series attenuator is that you only have one switch contact in the path. The disadvantage is that all 24 resistors are in the signal path all the time", whilst unquestionably a truism, what Duncan has done, by rewiring the switches on the Goldpoint (as he outlined earlier), is provided me with a 'shunt-type' device, where as mentioned before, only one series resistor is line-line with the signal path, at any point on the switches, and depending on that position, will result in how many shunt resistors are being used: at position 1, one, and at position 47, forty seven, which is why it pays for me to achieve maximum available volume on the early steps!

Now, with having maxed out the low-level volume available on the early steps of the dials, by fitting the optimal value of series resistor (for my system), I will rarely, if ever, need to use the controls beyond position 12, which means that a maximum of 12 shunt resistors would be in use at any one time. However, in reality only 3, to a maximum of 10 resistors will be used, as positions 3-10 on the dials is where 95% of my listening will be done.

Therefore, if we consider that because the shunt resistors currently being employed in the Goldpoint are high-quality Nichrome SMT-types (as favoured after extensive research by Goldpoint), and that the principle behind using SMTs is, as the article above says, so that: "resistors can be tightly packed in a manner that almost resembles a single resistor", then effectively with using a maximum of 10, tiny, tightly-packed SMT-type shunt resistors, at the loudest positions on the dials I'm liable to use, then it seems feasible that, as claimed, those are indeed behaving similar to a single resistor.

That, coupled with the way Duncan has rewired the Goldpoint switches, so that only one series resistor is in line with the signal path, is probably why the sound I'm getting now is so good!!

It's also why I'm now in no hurry to change anything else, and simply allow all the new resistors to burn-in. Later on, I will experiment with replacing the Nichrome shunts, with Z-foils, which would probably be the ultimate, but for now, the stunning sound I'm getting is, shall we say, 'more than adequate'!

Marco.

[Lodgesound]

All very interesting reading this.

My only observation is that I do not understand why equipment at this level is not all balanced signal line - we use it in the industry for very good reasons.

[Marco]

Hi Stewart,

Indeed, but the advantages of a balanced signal line are only fully realised when the whole system, from input to output, is designed that way, including the cables connecting it all together, otherwise for me, it's not worth doing. It's 'all or nothing', really.

And while that's the norm in studios, etc, and thus easily and cost-effectively implemented with professional equipment, that's not the the case with 'domestic' hi-fi, as rarely are the circuits employed in such equipment designed to take advantage of a fully balanced signal line - so we make the best of what we have

Marco.

[lurcher]

All very interesting reading this.

My only observation is that I do not understand why equipment at this level is not all balanced signal line - we use it in the industry for very good reasons.

Yep, but to be fair, those reasons are not particular applicable for home use.

[lurcher]

"resistors can be tightly packed in a manner that almost resembles a single resistor"

Sorry, but that makes no sense. just because its small and if you squint it looks like a single resistor doesnt make it a single resistor. unlike the owner, audio doesnt care what it looks like, the reality is that SMD resistors are used because they are small and can be assembled to PCB's by automatic pick and place equipment. Then trying to reverse argue that because they are small and take up no space makes them in any way better sonically is just marketing bullshit. If you have 20 surface mount resistors or 20 of any type resistors you have 20 different resistive elements, and 40 contacts between the resistive element and the end conductor, and 40 end cap to PCB solder joint. With a single resistor you have 1 resistive element, two end caps, and two solder joints.

When it suits them they use the same but reversed argument when they make a thing of the advantage of only having the one switch contact in the signal.

Quite simply, if they were seriously crap and holding things back to a notable degree, Duncan and I would've been able to hear it.

But you could and did, thats exactly what you heard when you replaced the 30 or so surface mount resistors with a single shunt, the improvement you have heard (even with a cheap SINGLE metal film resistor) show that the SMD ones are not transparent as they claim.

Now, again, the shunt resistor(s) may have less effect on the sound than the series one, but I think its certain that it will have some. as you say, you may try this later so it will be interesting what you find.

If you are only using position 3-10 on the switch, if you decide to try single resistors, you may as well just use a 24 way switch and spread the values from to your current 10 over those points. You could use a 12 way switch, but I suspect psychologically you prefer the control to be in the lower half of its travel under normal use.

BTW (1), I am almost certain no engineer from Goldpoint has read thair web site, they would have noticed this mistake "The surface mounted resistors used on the Goldpoint V47s are 0.5% Nichrome THIN film with a thermal noise specification (TCR) of 25 PPM, same as on our V24s,"

BTW (2), "let's not lose sight of the fact that the best resistor is NO resistor", hence the point of inductive designs.