Peak to peak

JaredC79

Banned
2015-09-02 4:32 am
NJ
Can someone please let me know the math behind getting to a a specific line out peak to peak on a preamp - supposedly 3.4 to 2 is ideal and will put you at 20hz to 20khz. I am using an elf filter on my subs but if I wanted to remove it and get naturally lower f3 then the amp and pre was capable what is the down side of the peak to peak change?
 

twest820

Member
2009-06-24 10:49 pm
I think it's

Q: How to forecast required swing as a function of filter selection?
A: Sweep the integrations across a music power spectral density spread. Numerical is usually easier than analytical.

Q: What is all that stuff about THD+N, clipping, and SnR?
A: Start with texts on amplifiers, such as Walt Jung's op amp handbook and Bob Corell's power amp book.

Q: Little help with gain structure?
A: See the DIY Audio article on the topic.
 

JaredC79

Banned
2015-09-02 4:32 am
NJ
I think it's

Q: How to forecast required swing as a function of filter selection?
A: Sweep the integrations across a music power spectral density spread. Numerical is usually easier than analytical.

Q: What is all that stuff about THD+N, clipping, and SnR?
A: Start with texts on amplifiers, such as Walt Jung's op amp handbook and Bob Corell's power amp book.

Q: Little help with gain structure?
A: See the DIY Audio article on the topic.

Yes, I would like to know the math/formula behind calculating rail to rail swing - I understand in practice things may need to be adjusted after testing so that my circuit drawing are more accurate in regards to what I want. I am not referring to opamps - most of the data for one be found in the white pages, in referring to discrete components and being able to work backwards and say I want 2 or 3v swing peak to peak then base my voltage, transistors and currents based off of what I would like the end result to be.

I don't know all the ways and tricks to minimize THD Max SNR but I am a fast learner and pick things up quickly. I did order Cordell's books. I am a finance geek by day and always loved physics, built almost anything by hand at one point or another - I know how things should sound and what is pleasent but coming from an analytical side numbers and values are how I understand things best. that being said I am sure there's a formula to calculate peak to peak swing and if you can please provide it, it would be greatly appreciated.

I've seen a lot of circuits that are very similar, even using the same or almost the same components. If I knew the math behind each stage i feel that I would be able to possibly come up with something I can call my own, picking the proper component to give me the end result I am looking for. Working backwards I feel would be ideal because then i have some goal to aim for rather then, I'm going to build a preamp or a crossover and the end results would be based off how I started rather than what I originally intended to create. I can then improve on the THD and other weak areas with compensation and techniques listed in these forums.
 

JaredC79

Banned
2015-09-02 4:32 am
NJ
Hi,

Sorry, I've tried but cannot extract anything meaningful from it.

Other than saying ELFs are a very good idea for subs, should
have nothing to do with the -3dB point, removal is a bad idea.

rgds, sreten.

Hopefully my previous reply cleared things up. Basically I was looking for formulas. I was also asking if changing the filtering where line level was lower then 20hz - possibly single digits while keeping the highs at 20khz would have any adverse effects. ELF and other devices have been helpful getting to subsonic F3s but if the typical line level output was widened, what would the ramifications be? Any down side? Also since most people can't hear over 20k to 22k why do some of the Raal ribbons go as high as 100khz? Supposedly anything over something like 23khz is fatiguing and annoying as opposed to on the low end where sub bass is felt and not heard.
 
JaredC79 said:
I am a finance geek by day
As a finance geek I guess you would want to know which currency was used if someone said that something cost 100. As a physicist I want to know what units are used when someone says
supposedly 3.4 to 2 is ideal
Now you have explained a little more I suspect you meant to say 'supposedly 3.4V to 2V is ideal' - although it is conventional to put the lower figure first.

and will put you at 20hz to 20khz.
still doesn't mean anything to me. I assume you meant 'Hz' not 'hz', but it still doesn't mean anything because I don't know what you mean by "put you" at a frequency range.

You are asking for formulas. You can't use formulas until you gain some circuit understanding. Some clarity in your questions would help us give meaningful answers.
 

juma

Member
2005-11-08 11:27 am
Zemun
.. which currency was used if someone said that something cost 100....
If you soberly consider the contemporary state of financial affairs in the world you'll see that your question doesn't really need an exact answer - financial experts are used to getting away with anything. Physicists mostly can't do that - or we can try to follow the logic behind the question asked in post #1:
f (delta Vpp / f3) = a
where a = something, anything or almost nothing
:clown:
 

JaredC79

Banned
2015-09-02 4:32 am
NJ
As a finance geek I guess you would want to know which currency was used if someone said that something cost 100. As a physicist I want to know what units are used when someone says

Now you have explained a little more I suspect you meant to say 'supposedly 3.4V to 2V is ideal' - although it is conventional to put the lower figure first.


still doesn't mean anything to me. I assume you meant 'Hz' not 'hz', but it still doesn't mean anything because I don't know what you mean by "put you" at a frequency range.

You are asking for formulas. You can't use formulas until you gain some circuit understanding. Some clarity in your questions would help us give meaningful answers.

Appreciate the grammar lesson. So during the planning for a new circuit for a preamp, what's the standard work flow to plan a build with certain parameters? Surely they aren't simply saying run it at 32v use a cfa or LTP to get to line level, allow the voltage to swing as much as it should and filter all the nonsense out. I feel their has to be a more academic approach.
 
Designing a circuit is a creative activity, so there is not a "standard work flow". You start from the requirements, and rough out a circuit which might meet them. Then you calculate/simulate to see how close it gets. Modify as necessary, or scrap and start again. Maybe modify the requirements, if they cannot be met in a reasonable fashion. Build it, measure it, listen to it.

Going from requirements to an initial circuit involves knowledge and experience. You are a finance geek so I guess you understand mathematics. Given a function to integrate, there is not a standard way to do it so you can't just crank a handle and get the answer. Instead, you use your experience to try what seems to be a useful method (substitution, parts, standard integral?), then try a different method. Maybe try a combination of methods. Or cheat and google it. Circuits are similar.