You are here

Focusrite pots & linear law

Customising, building or repairing your own gear? Need help with acoustic treatment or soundproofing? Ask away…

Focusrite pots & linear law

Postby DAGGILARR » Thu Jan 17, 2013 6:51 pm

I have a Saffire Pro24 dsp which suffers with the linear law gain pot problem meaning all the gains piled up at the end.

I am wondering about the potential for a mod for this.

Is it possible to put non linear pots in it ? audio taper I think they are called

Can some electrickery be achieved by the judicious placement of resistors?

Has anyone tried such a thing ?

Is there an AI that has (what I call) normal gain pots? that is in the same price range ?
User avatar
DAGGILARR
Frequent Poster
Posts: 711
Joined: Tue Sep 21, 2010 11:00 pm
Location: Exeter, Devon.

iMac 27" i7 2.9gh 2010 12GB ram, Focusrite Pro24DSP.


Re: Focusrite pots & linear law

Postby Hugh Robjohns » Thu Jan 17, 2013 6:59 pm

DAGGILARR wrote:I am wondering about the potential for a mod for this. Is it possible to put non linear pots in it ?

No. If it was that simple the manufacturers would have done it already! The gain versus feedback resistance of simple gain stages has an inherently non-linear scaling, but it's not a simple log, antilog, linear or audio taper law either! So there is no simple replacement pot mod that will cure the bunching effect.

Audient managed to improve matters with their Mico preamp by completely changing the entire gain structure of the various stages of that preamp, but even so the gain is still bunched at the high end -- just not quite so badly.

This is such a common problem precisely because there is no easy solution.

hugh
User avatar
Hugh Robjohns
Moderator
Posts: 16767
Joined: Thu Jul 24, 2003 11:00 pm
Location: Worcestershire, UK

Technical Editor, Sound On Sound


Re: Focusrite pots & linear law

Postby DAGGILARR » Thu Jan 17, 2013 7:11 pm

But there are audio interfaces that do not have this issue or is this just wishful thinking on my part
User avatar
DAGGILARR
Frequent Poster
Posts: 711
Joined: Tue Sep 21, 2010 11:00 pm
Location: Exeter, Devon.

iMac 27" i7 2.9gh 2010 12GB ram, Focusrite Pro24DSP.


Re: Focusrite pots & linear law

Postby Hugh Robjohns » Thu Jan 17, 2013 7:59 pm

Different amplifier circuit designs work differently. The simplest cheapest designs tend to suffer most from gain bunching. There are better designs with more linear gain controls, but they are often much more expensive and elaborate. Some designs require reverse log or reverse taper pots, but these are relatively rare and expensive -- if not impossible to get in the right value -- so manufacturers often use slugged linear pots instead, with some compromise to the gain law.

Whatever the case, it would always involve more than just swapping a pot -- other circuit changes would probably be needed to arrive at the desired linear gain law.

These factors are one reason why so many high end preamps use gain switches rather than pots -- it's much easier to configure exactly the right resistance values for each gain step!

H
User avatar
Hugh Robjohns
Moderator
Posts: 16767
Joined: Thu Jul 24, 2003 11:00 pm
Location: Worcestershire, UK

Technical Editor, Sound On Sound


Re: Focusrite pots & linear law

Postby ef37a » Thu Jan 17, 2013 11:34 pm

Even the vastly experienced and talented Duggy Self admits in his book Small Signal Amplifiers that this is a virtually unsolvable problem and as Hugh has mentioned, the best one can do is carefully split the gain over two stages.

The problem is somewhat less for mixers since you have a gain pot plus a channel level pot, so you could buy a mixer! Not a totally daft idea but a more practical one might be the Cloudlifter? This gives you another 25dB or so of gain and thus you can crank back the AI channel gain to a more useable point. If the the problem is too much gain (with say a capacitor mic) some XLR inline attenuators will perform the same task but the "other way".

The Saffire Pro24 does not unfortunately have inserts on the mic amps but my Fast track pro does and I built a potbox to put the gain pot into a more "linear" point (I then modified it for 20dB gain with an NE5532)

AIs will always be a compromise in this area I fear. The SoS tests might have shown very little sonic advantage in expensive pre amps but you cannot get operational finesse on the cheap!

Dave.
ef37a
Jedi Poster
Posts: 7079
Joined: Sun May 28, 2006 11:00 pm
Location: northampton uk

#They did not listen, they are not listening still...Perhaps they never will?#


Re: Focusrite pots & linear law

Postby Folderol » Thu Jan 17, 2013 11:57 pm

Fiddly, and expensive, but with something like a 20 way rotary switch you can get just about any law you want. Chose resistors for (say) 2dB steps and you've got a precision 40dB 'pot'.
User avatar
Folderol
Jedi Poster
Posts: 4542
Joined: Sat Nov 15, 2008 12:00 am
Location: Rochester, UK

Save paradise, Pull up a parking lot!


Re: Focusrite pots & linear law

Postby DAGGILARR » Sun Jan 20, 2013 5:51 pm

So if a gain control was described as having a 'continuous linear curve' would this be different ? this is how Motu describe theirs
User avatar
DAGGILARR
Frequent Poster
Posts: 711
Joined: Tue Sep 21, 2010 11:00 pm
Location: Exeter, Devon.

iMac 27" i7 2.9gh 2010 12GB ram, Focusrite Pro24DSP.


Re: Focusrite pots & linear law

Postby James Perrett » Sun Jan 20, 2013 11:26 pm

One of the biggest issues seems to be that designers want to save the cost of a pad switch so they use these padless preamp designs which try to accomodate such a wide gain range. If preamp designers would go back to using a pad switch the problem wouldn't be half as bad.

James.
James Perrett
Moderator
Posts: 5762
Joined: Sun Sep 09, 2001 11:00 pm
Location: The wilds of Hampshire

JRP Music - Audio Mastering and Restoration.
http://www.jrpmusic.co.uk


Re: Focusrite pots & linear law

Postby ef37a » Mon Jan 21, 2013 6:41 am

DAGGILARR wrote:So if a gain control was described as having a 'continuous linear curve' would this be different ? this is how Motu describe theirs
.
I once butchered a well known phrase to "Lies, damn lies and specifications". For the above it could be modified to "Specifications. BAD specifications and BOLLOX!"

James.Pads degrade the CMRR and whilst this probably would not bother 99% of the people 99% of the time it looks bad on the specs! Peeps gotta sell'um you know!

This is an intractable problem, you either have to have a multistage switch or relay resistor ladder and software control?

Dave.
ef37a
Jedi Poster
Posts: 7079
Joined: Sun May 28, 2006 11:00 pm
Location: northampton uk

#They did not listen, they are not listening still...Perhaps they never will?#


Re: Focusrite pots & linear law

Postby Hugh Robjohns » Mon Jan 21, 2013 10:24 am

The original Audient Mico had a serious problem with gain-bunching. I raised this problem with the designer and persuaded him of a need to address it. Clearly, with a product already in production the ability to change the design was heavily restricted, but by altering a few component values to alter the gain structure a very worthwhile improvement was obtained. it's not a perfect solution, but it does illustrate James' points well. here's what I said in the review:

However, I found the gain control law on the original model suffered a ‘dead zone’ from about 11 o’clock through about 3 o’clock, in which the gain barely changed at all, followed by a mad rush of gain at the end — over 15dB in less than five degrees.

The problem was essentially due to the characteristics of the potentiometer being used (which was different to those used in the consoles and rackmount preamps). After I discussed the issue with Dave Dearden, various alternative configurations were tried and tested before settling on the version described above, which has a slightly different gain structure but far better control linearity, and is more easily adjusted at high gain settings as a result. The only obvious differences between the original and revised models are the pad attenuation and the gain control scaling. The original had a 10dB pad, with a gain range starting at 6dB and reaching 30dB in the middle. The revised version has a 20dB pad, and the gain range starts at 18dB, with 33dB in the middle.


In both the original and revised versions the maximum gain was/is 66dB.

While a poorly specified pad could affect the CMRR figure, it isn't usually a problem in practice -- it just comes down to low-tolerance resistors and some small adjustment on test.

H
User avatar
Hugh Robjohns
Moderator
Posts: 16767
Joined: Thu Jul 24, 2003 11:00 pm
Location: Worcestershire, UK

Technical Editor, Sound On Sound


Re: Focusrite pots & linear law

Postby Folderol » Mon Jan 21, 2013 6:53 pm

When I was a (poor) callow yoof, I used to raid the scrap merchant in Ash Vale for all sorts of electronic goodies - they had a contract with the R.A.E I believe. Mostly I just stripped stuff for parts, and lots of valves came to me that way.

An interesting module I picked up was some kind of amplifier, which I tried to puzzle out. There was a control marked 'gain' which was a WW pot with its wiper connected to ground. As best as I could make out, one end shunted the feedback circuit, the other end shunted the output.

Has anyone seen anything like this? I haven't a clue what sort of law it would have. I guess the term would be 'exotic'!
User avatar
Folderol
Jedi Poster
Posts: 4542
Joined: Sat Nov 15, 2008 12:00 am
Location: Rochester, UK

Save paradise, Pull up a parking lot!



Who is online

Users browsing this forum: No registered users and 1 guest