[SnoopyRocks]

You caught a typo. I neglected to remove R4 from the denominator when I cut and pasted the equations. The correct equation is on the first page of the pdf. You seem to have it all working regardless.

I updated my files with the correction.

Quote:

Originally Posted by morsel SnoopyRocks, you are a gentleman and a scholar. I massaged your 3db equations and noted that your Adc is what I call Abb, which is relative to unity gain, not Av. Since I was planning to use Abb .vs. unity rather than Ao .vs. Av, I merely substituted 1 for Av, and it worked. The new fc equation plots identically to Solve[Abb=3dB]. The new fs equation works, too. Thanks again!

Is Mountain View near Sunnyvale? I drove though there on Monday while I was visiting the bay area this week.

 

[morsel]

Apparently I got confused with the 2 different versions of Adc and thought after looking at your 3dB frequencies page that your Adc was my Abb. In any case, it seemed simpler to reference Abb to 1.

Mountain View is next to Sunnyvale. Please send me an email via the link on my profile page. I'd like to talk with you offline about the merits of Matlab .vs. Mathematica, filters, math, and whatnot.

 

[SnoopyRocks]

email sent

Quote:

Originally Posted by morsel Apparently I got confused with the 2 different versions of Adc and thought after looking at your 3dB frequencies page that your Adc was my Abb. In any case, it seemed simpler to reference Abb to 1. Mountain View is next to Sunnyvale. Please send me an email via the link on my profile page. I'd like to talk with you offline about the merits of Matlab .vs. Mathematica, filters, math, and whatnot.

 

[Frenchman]

Someone really should sticky this.

 

[morsel]

I am not a fan of stickies, as they eat up space on the first forum page that should be available to current threads. More than 1 or 2 stickies is a clutter. I would rather see a single sticky with important info and links to threads of note. If there is interest, I could be talked into summarizing this thread with a single document of the valid equations we came up with.

 

[Frenchman]

Why not throw it into the big list/faq section then? If somethings refered to often enough to be stickied, or is simply important enough that we don't want to lose it to the depths of the forum, then why not stick it someplace where it's still easily accessable but also out of the way. Of course, I'm refering to all the current stickies as well.

 

[Syzygies]

Quote:

Originally Posted by morsel A bit more massaging to put all these equations in similar form, and added fo to represent the 1/2 gain point where Abb < 6dB.

Would it be possible for someone who has this working to upload the actual Mathematica file? They use a nifty text format which ports well between machines. I could type this reading the gifs, and if I do, I'll post my file, but it would save time if someone already has the file handy.

(The calculators don't tell enough of the story...)

Thanks

 

[SnoopyRocks]

An old thread finds new life...

Here is a version of the bass boost mathematica notebook.

BassBoost.nb

I take it that you don't have/like matlab Syzygies. Everything is it the code I posted. Knock yourself out.

 

[morsel]

bassboost.nb is the Mathematica file I created during the process of deriving the equations we use for bass boost calculations from the equations SnoopyRocks posted. It generated the graphs I posted in this thread. Thanks for your scholarly assistance, SnoopyRocks, without which we would still be using the old numerical analysis model.

Maybe some day I'll learn out how to use Mathematica properly and clean up the notebook a bit. With that in mind, the link above points to my copy, but don't hold your breath about a new version showing up anytime soon.

 

[SnoopyRocks]

Morsel you don't need to clean anything up. It works just fine after all. I prefer the fo term you added to characterize the filter for low boost levels. It makes more sense than the 3dB frequencies, which are after all just a mathematically convenient way to characterize things.

You have all the equations posted on your BB calculator. AMB's plots the frequency response. People seem to like the default values that you're recommending for the filter. What more is there to ask? The design is a success.

I suggest also listing the DC gain with the bass boost on since it, not Av, amplifies the DC offset.

I could understand if people might be interested in seeing the phase, group delay, step response, etc also. I presume Syzygies is interested in the code to take a look at something along these lines. At this point though, I imagine many DIYers would probably find the extra information more perplexing than enlightening.

Just to clarify: the mathematica file that I posted was originally morsel's work. After playing with it for a bit, I decided to go back to matlab because the typeset GUI scared me.

 

[Syzygies]

Quote:

Originally Posted by SnoopyRocks I take it that you don't have/like matlab Syzygies. Everything is it the code I posted. Knock yourself out.

Quote:

Originally Posted by SnoopyRocks I could understand if people might be interested in seeing the phase, group delay, step response, etc also. I presume Syzygies is interested in the code to take a look at something along these lines.

Thanks, Snoopy and Morsel. No, I just wanted to stare at locally-generated curves to match the default recs using my choice for R4, to see how changing the cap works, the usual. None of my actual amps have needed debugging, but for some reason I couldn't sort out why my recently reinstalled, fully loaded Mac was running the rest of the planet's Javascript in every browser, but balking at DIY audio Javascript. I'd give up this hobby before I sat in front of Windows as my main machine, it's so corporate and such a poor interpretation of this GUI OS thing that the programmer in me finds it painful.

I get Mathematica for free at my University; I was actually the host once when Steve Wolfram came to give a talk announcing a new version. I used to prefer Maple, but I don't have any experience with Matlab, which never made any inroads here.

The Mathematica package I crave is Analog Insydes 2.1, but it's $900 academic. Spice is an amazing accomplishment from, what, 30 years ago? I'm learning it, but I find the interface antiquated.

If Matlab or Maple had a decent analog electronics add-in I could afford, I'm there.

 

[tangent]

Quote:

Spice is an amazing accomplishment from, what, 30 years ago? I'm learning it, but I find the interface antiquated.

Several of the commercial GUI SPICE packages have free demo versions available. These are usually only limited in terms of the number of components it includes and how many parts you can add to the design; neither of these limits will be a problem for simulating such a simple circuit. With SPICE, you get all those other curves SnoopyRocks mentioned in his previous post.

Unfortunately, I don't know if any of these SPICE demos run on OS X. The most popular are MicroCap and PSpice on Windows, for what that's worth.

 

[SnoopyRocks]

Quote:

Originally Posted by Syzygies The Mathematica package I crave is Analog Insydes 2.1, but it's $900 academic. Spice is an amazing accomplishment from, what, 30 years ago? I'm learning it, but I find the interface antiquated. If Matlab or Maple had a decent analog electronics add-in I could afford, I'm there.

Which version of spice are you using? Just curious. I use spectre. There is much more to spice than the interface. A nice GUI on top of spice does help though.

Analog Insydes looks like a neat toy which doesn't really cost that much (about 1/10th) compared to a Cadence (professional 'spice' for IC design) licencse set. In the end, it's still a symbolic solver though. I have always found the sybolic solvers only useful for very simple tedious stuff. For anything else, you have to trick them into giving you what you actually want and you'd have been better off just doing it yourself. Good luck with your designs if you trust the software to think (what Analog Insydes calls simplifications) for you. I'd be shocked if it didn't crash on anything more than a couple of transistors.

 

[morsel]

I found out why Mac users have problems with the bass boost calculator. The Apple Safari and IE browsers do not support the Number.toFixed() Javascript method. My recommendation for Mac users is to use a real browser like Firefox, Camino, or Mozilla that fully supports Javascript. I might get around to changing the Javascript at some point to format numbers with reduced precision manually.

 

[Syzygies]

Thanks! I was going nuts trying to figure out how this was something I'd done, though I truly do select every installation option for both basic OS and for development.

I want Safari to be a credible browser, so I will use the Apple feedback mechanisms to document this issue with them.

IE I can live without. By coincidence these were the two I tried, before moving to a Windows box. I was going to try the rest of my browsers, but these two failing made me believe it was something in my Java installation.