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

[Marco]

This level thing as perceived on a volume control has changed from the 50's and 60's. Back then, volume pots and the overall gain was designed to have optimum listening at TWO THIRDS of the way up, so you have loads of spare travel for lower level listening - look at older Naim stuff using vinyl...

Yup, and I hate that, as I think it sounds shit!

Marco.

[Ali Tait]

Ok, just a little update on the above... As the unit has effectively been changed from a series to a shunt type (as it sounds markedly better that way), so that the signal, at any position on the dials, now only goes through one resistor in each channel, the value of those two resistors, as well as their inherent quality, is therefore critical.

In that respect, it's been rather interesting playing around with those resistor values (initially using inexpensive but decent quality metal-film types), in order to tailor the gain structure of my system to the volume levels available, at various positions on the controls of the Goldpoint, where it became obvious that the 50K value of the resistors originally fitted on the stepped attenuators were the reason for why I wasn't getting enough volume available on the early steps [e.g, 0-8 was virtually silent]...

Therefore, changing the value of the two resistors, initially to 10k, completely cured that problem, but made it too loud, so we changed them to 22K, which was just about right, but not quite, so I've now settled on 20K, which is perfect. It now means that normal listening levels, for me, occur at the 8 to 9 o'clock positions on the dials, and not the 11 to noon positions previously, which I found unnatural and somewhat counter-intuitive.

What this means now is that there is an opportunity to experiment with some 'audiophile quality' 20K resistors, and to that end I'm going to order from Hi-Fi Collective two each of the following:

Charcroft Z-Foils, AudioNote Tantalum (non-magnetics), Caddocks, PRPs, and for something a bit different, Amtrans AMRG carbon films, just to see what carbon sounds like compared with metal... All will be either 1 or 2W versions. You can read all about the various resistor types here: http://www.hificollective.co.uk/catalog/-c-61_62.html

The resistor 'bake-off' will be taking place on Tuesday this week, and the results of that will be posted here later on that evening, so it'll be very interesting hearing what happens and what turns out to be the winner!!

Marco.

I've done this in a couple of previous preamps Marco. I ended up with the AN resistors as my preference, though I never tried the Z foils and I suspect they may be better.

It's interesting how much of a difference it makes.

[awkwardbydesign]

I've done this in a couple of previous preamps Marco. I ended up with the AN resistors as my preference, though I never tried the Z foils and I suspect they may be better.

It's interesting how much of a difference it makes.

As I/V resistors in my DAC they are, but not used them elsewhere.

[Marco]

I've done this in a couple of previous preamps Marco. I ended up with the AN resistors as my preference, though I never tried the Z foils and I suspect they may be better.

It's interesting how much of a difference it makes.

Indeed, and I've never played around with resistors before, even in the Croft - only caps. Therefore, I'm really looking forward to hearing the results obtained, with the various types of resistors selected, in such a critical position on a passive preamp.

If it works well, then I may consider fitting the resistors deemed as sounding best in the phono stage of the Croft and/or in the Copper amp. That would be fun!

Btw, do you understand what I'm getting at here, and if you do, could you try and explain it to Macca:

Essentially, it's about the 'input level'/loudness [I can't think of a better description] a power amp 'sees', from a preamp (passive or active), to be amplified, and that's a bit like setting the recording levels on a tape deck.

If the 'input level' of the music signal, in that respect is insufficient, then no matter what you do with the volume control afterwards, you'll still end up with a sound lacking in 'balls'...

And that 'input level' of the music signal, which the power amp 'sees', is controlled (at least that's what my ears tell me), by the value of resistors used on the output of the preamp (passive or active), which in my system is how I ended up with a value of 20K, thus allowing me the normal listening levels I desire at around the 8 or 9 o'clock position on the volume controls).

The lower you go in the 'K' value of those resistors, the louder things get, and the higher that value, the less loud it gets (just like lowering or increasing the recording level on a tape deck), thus influencing the playback level available (lower down on the volume controls) and overall sound accordingly.

Therefore, employing the use of, say, 47K resistors would result in there being much less 'loudness level' available, in the early stages of a stepped attenuator, than there would be had, say, 10K resistors been used instead.

Cheers!

Marco.

[Ali Tait]

Yep I get what you're driving at, I did the same to get a travel on the volume pot I liked. It may be that you weren't getting the sound you wanted because there were more resistors in the signal path(before the shunt mod)?

For me it was about getting the gain structure in my own system right as you say.

[Marco]

Good - I'm glad that I'm making sense to someone!

Yes, there were two things originally that were sub-optimal with the Goldpoint (even though as standard it was far from shabby, and I could hear its potential), and that was the amount of resistors that were in-line with the music signal at low to medium volumes, thus degrading the sound, and that the 'loudness input level', controlled by the 'K' value of the resistors used, was too high, and so the resulting sound was somewhat 'limp and weedy'.

Therefore, both shunting the switch and lowering the resistor values has allowed the full potential of the device to be released - and it's not a subtle improvement either!

Like you say, it's about getting a "travel on the volume pot you like", but more importantly for me, supplying a 'juicy' enough signal to the power amp, to enable the overall sound produced by the system to have enough 'drive and balls'

Now that that those issues have been resolved (and perhaps it's one of the reasons why some folk find passives lacking in 'cojones'), it's all about maxing out SQ, by fitting the best sounding resistors. That done, I *suspect* the Goldpoint will oust the Croft for line-level duties, and that the latter will be employed for use as a phono stage only.

Marco.

[awkwardbydesign]

it's all about maxing out SQ, by fitting the best sounding resistors. That done, I *suspect* the Goldpoint will oust the Croft for line-level duties, and that the latter will be used as a phono stage only.

Marco.

I buy my Z-foils direct from Texas http://www.texascomponents.com/store...?idProduct=461 Up to 100k in this version, or 250k in t'other. Just watch the max voltage! And I get 0.1% tolerance; that should be close enough. But you will have to decide for yourself if they sound "best".

[lurcher]

A couple of things come to mind. Though in theory the passive should do nothing to the sound, so at one third the sound should be the same as at two thirds but just quieter. Practice often doesn’t match theory, mainly because the theory is wrong.

Firstly, the input impedance seen by the source and the output impedance into the load will be different in the one and two thirds position, that will alter how everything works together, and also will alter how the cables interact with the components.

Secondly, at two thirds, even though the control has been converted into a shunt, I assume the attenuator is still a bunch of resistors in series, so the number of resistors that the signal sees will be different in both cases. Resistors have a sound. If there are 47 steps on the attenuator, at one third there will be 15 in series to ground, at two thirds that will be 30 resistors. I am willing to bet that 30 resistors in series will have more of a sound than 15 of them.

A pure shunt control avoids the last problem (though not the first) by only having the single resistor to ground at each position.

[Marco]

A pure shunt control avoids the last problem (though not the first) by only having the single resistor to ground at each position.

That's what I have in place now, Nick, as Duncan has wired the dual-mono switches that way

Agreed about cables, which is why I use nothing longer than 0.5m interconnects, with the Goldpoint - and high quality, low-capacitance (minimum loss, low attenuation) ones at that.

Marco.

[lurcher]

That's what I have in place now, Nick, as Duncan has wired the switches that way

Agreed about cables, which is why I use nothing longer than 0.5m interconnects, with the Goldpoint - and high quality, low-capacitance (minimum loss, low attenuation) ones at that.

Marco.

Yes, resistive passives will be very sensitive to capacitance on the output side.

I thought the goldpoint was made using surface mount resistors on a PCB, how has it been converted to a shunt with a single resistor to ground?