Sound vs. Components

[Hulk]

What component or components of an analog circuit has the biggest influence on the overall sound - lets say of a SSL clone or a G9 preamp (trafo, sum of resistors....)???

 

[gyraf]

IMHO, Component type influence on audio quality - in order of significance:

1 - Audio transformer quality
2 - Correct grounding
3 - Overall component count/number of active stages in audio path
4 - User interface - front panel layout
5 - VCA quality (for stuff based on such)
6 - Electrolytic capacitor quality in audio path
7 - Layout
8 - Poly capacitor quality
9 - Powersupply
10 - Resistor types
11 - switch/pot/wire quality

All this is general categories - specific applications will always call for component optimizing.. Like the resistors quite low on the list - for high-gain, low noise circuits, resistor quality may sometimes be more important.

 

[zebra50]

Nice summary Jakob - I think most would broadly agree.

Whilst quality of post and switches may not have a big influence on sound, I would argue that they contribute a great deal to the robustness & durability of the project and ensuring it stays sounding good for a long time.

 

[gyraf]

You're off course right, Z50 - the low placement of pots and switches is weighting "immediate" bonus.

Jakob E.

 

[Hulk]

Great list Gyraf!!

This might be a stupid question but isn't the trafo (nr. 1 on the list) just part of the power supply(nr. 9)?? How do you differentiate between the two??

 

[zebra50]

Many audio circuits use transfomers in the signal chain as well as for power supplies. Check out a few schematics - try the G9 mic pre for a good example.

 

[gyraf]

Yes - Audio transformers are not the same as power transformers.

 

[Hulk]

aahhh I really learned something today :idea:

I'm still working on my first project (SSL clone) so I have only read the schematics for the SSL and there's only a trafo in the PSU :roll:

 

[Gus]

Hulk

The best way is to do the tests yourself don't believe anything on the web.

After some time tesing and reading others posts you will have an idea of what some peoples ear/brain like and then you can start to find out who likes the things you do.

FWIW the other day when testing microphones on a nice guitar I learned something that I would have never found out from reading other's posts or books.

You realy need to do you own tests and trust your own ear/brain.

If you look at my posts here I will sometimes post what I like in parts but I never say it is the best. Its what I LIKE.

 

[Greg]

This post is exactly relevant to a GSSL I just finished building for a local mastering guy. Let me briefly explain:

On my first GSSL, I used DBX2151 VCAs in the sidechain and channels. I used generic (can't even remember the type) electrolytics and polyesters (MKT). On the GSSL I build for him, I used a DBX2151 in the sidechain and THAT 2180LBs in the channels. I also used Panasonic FC and Nichicon UPW electrolytics, and I used Panasonic B Series polyesters. All other components were the same.

Yesterday we spent some time A/Bing the two compressors, and I must say the later of the two sounds much more "hi-fi," just meaning that high end is much more extended. In a way, I'm glad that they have a slightly different flavor. The first one, though less "hi-fi", brings clarity to the midrange that the second one does not. But on the other hand, if I'm going for less midrange definition and extended top end, I know which one to grab.

Anyway, just thought I'd mention this and perhaps start some discussion. I'd contribute these sonic differences to the electrolytics and the polyester differences. I wouldn't suspect the the VCA, but possibly? Any comments?

Also, I should be building another GSSL (possibly 2) in the near future that have DBX202Cs in the channel VCAs...

 

[Gus]

Jakob

Looking at your ssl type schematic I wonder why you use one electro instead of two electros setup as nonpolar for coupling caps.

I think this makes the different electros sound different in the gssl . There is no forming voltage. The real SSL looks like it uses two electros setup as a nopolar

Greg

I find the differences in electro chemistrys shows more without forming voltage.

 

[gyraf]

The real SSL looks like it uses two electros setup as a nopolar

I don't think they do so - other than in special areas like the input of the mix-amp and such.

I've tried with the double-electrolytic scheme for this, but didn't find much difference (or rather, I was unable to detect any difference), so I decided that fewer components in signal path is better than more..

A possible real-upgrade would be to use 4u7/16V Wima MKS polyesters at the input in stead of the electrolytics. THAT is noticeable - but these caps are not easy to find DIY-wize

Jakob E.

 

[Gus]

In the USA Mouser carrys up to 10uf 50V PETs by CDE. Nice PET caps IMO.

 

[gyraf]

Aah yes - but they'll be too large. The WIMA's can actually fit the PCB, if you're a bit violent..!

:shock:

Jakob E.

 

[Gus]

I forgot to add I heard Emperor-TK build of the gssl sounds very nice.

 

[Greg]

Regarding my previous post... just a little update.

I spoke with the guy I built the GSSL for, and he decided to replace the Panasonic FC caps (which I had on the inputs) with some Nichicon caps (not UPW series), and now he likes the sound of the GSSL much better now. He has so much gear that already has a huge open clear top end, that he was looking for something that did more to the midrange.

I remember reading that soundguy wasn't a fan of the Panasonic FC.

They seemed to have less midrange, but the top was very bright and open.

 

[Hulk]

Does the input trafo influence the sound just as mush as the output trafo - or is the input trafo mostly for balacing the input???

 

[digichild]

[quote author="gyraf"]

The real SSL looks like it uses two electros setup as a nopolar

I don't think they do so - other than in special areas like the input of the mix-amp and such.

I've tried with the double-electrolytic scheme for this, but didn't find much difference (or rather, I was unable to detect any difference), so I decided that fewer components in signal path is better than more..

A possible real-upgrade would be to use 4u7/16V Wima MKS polyesters at the input in stead of the electrolytics. THAT is noticeable - but these caps are not easy to find DIY-wize

Jakob E.[/quote]i guess you mean the 22 u/35 v?

 

[Hulk]

[quote author="Hulk"]Does the input trafo influence the sound just as mush as the output trafo - or is the input trafo mostly for balacing the input???[/quote]

Maybe I should add that I'm building a G9 and I want one OEP channel and one Lundahl channel but if the main function af the input trafo is to balance the input and if the input trafo does not affect the sound as much/if any then I could use the less expensive OEP input trafo on the "Lundahl channel" (so using an OEP A262-A3E input trafo and a LL5402 output trafo)??????????

 

[radiance]

Hulk: as I understand it, transformers in general color the sound, not only the IN or the OUTPUT transformer. So make one channel al the way Lundahl (IN & OUTPUT) and one al the way OEP. That way both channels wil sound more "different" to each other...