Calculating RC time constants...and why you should.

Info for maintaining and tweaking your amp to perfection.

Moderator: VelvetGeorge

User avatar
novosibir
Senior Member
Posts: 4650
Joined: Tue Nov 22, 2005 2:32 pm
Just the numbers in order: 7
Location: Nuernberg, Germany
Contact:

Re: Calculating RC time constants...and why you should.

Post by novosibir » Tue Dec 23, 2008 10:23 pm

f (-3dB knee) = 1 / 2*pi*R*C

Larry
The fault almost always is sitting in front of the amp :wink:

Larry's Website now with included Pix's Gallery

mogo
New Member
Posts: 6
Joined: Tue Mar 10, 2009 9:56 pm
Just the numbers in order: 7
Location: Disneyland

Re: Calculating RC time constants...and why you should.

Post by mogo » Wed Mar 11, 2009 11:03 am

roberto wrote:
VelvetGeorge wrote: For your typical .68/2.7k combination the formula is this:

F= 1
------------------------------- = 87Hz
6.28 * .00000068 * 2,700

Meaning that every frequency above 88Hz is at full gain for that stage.
This is not completely correct. As you can see here: http://aikenamps.com/RC-FHz.htm" onclick="window.open(this.href);return false;
the frequency rolloff is higher than calculated as above. This is due to the fact that the real resistance to ground is dependant on the anode resistor.
Keep a typical 12ax7 for exemple: Ra=62500 and mu=100, and Rp= anode resistor = 100k
The real resistance to ground will be the parallel of the physical resistor (ie 2k7) with the internal resistance, calculated as:

Rk'= (Rp+Ra) / (mu+1)= 1k61

So 2k7//1k6 is 1k, and then 2k7//680n with 100k Rp has a cutoff frequency of 232Hz.
As you can see different tubes have different mu and Ra (that depends also on voltage), and this can change the cutoff frequency of the amp (ie more mu means more gain and more midrangey)
MattPete wrote:There are no harmonics below the fundamental. No matter how hard I bash or whisper sweet nothings to the low E string, it will never produce a 40Hz frequency when in standard tuning.
So how can you explain that 330u on plexis sounds different than 100u?
4u7 should be enought to obtain full guitar bypass, but the amp sounds WAY different. Fourier was wrong? NO, but an amp is an interaction of purely electronic with some physical elements as speakers, strings, and so on, and this produce some harmonics below the fundamental.
Correct and the RC give you the 3db point in most of these cases, in a first order filter as most of this stuff is the 235khz would be the knee at 3db where it starts boosting signal and 6db per octave after
So 235hz =+ 3db, 470hz is 9db, 940hz is 15db, 1.8khz, 24db
Here is the cheater method 159200 / R*C(in uf)

Roe
Senior Member
Posts: 5050
Joined: Thu Apr 13, 2006 1:36 pm
Just the numbers in order: 7
Location: Drontheim. Norwegen
Contact:

Re: Calculating RC time constants...and why you should.

Post by Roe » Thu Mar 12, 2009 5:40 am

I love that cheater method
http://www.myspace.com/20bonesband" onclick="window.open(this.href);return false;
http://www.myspace.com/prostitutes" onclick="window.open(this.href);return false;
Super 100 amps: 1202-119 & 1202-84
JTM45 RS OT JTM50 JMP50 1959/2203/34/39

User avatar
Grosh_Guitars
Senior Member
Posts: 1001
Joined: Tue Oct 24, 2006 9:11 am
Just the numbers in order: 13492
Location: Longmont CO. by way of Southern CA.
Contact:

Re: Calculating RC time constants...and why you should.

Post by Grosh_Guitars » Sun May 03, 2009 11:47 am

This is a great book for understanding a circuit,

http://www.pentodepress.com/contents.html
Sound Clips: http://www.soundclick.com/bands/default ... tent=music" onclick="window.open(this.href);return false;

User avatar
roberto
Senior Member
Posts: 59
Joined: Mon May 05, 2008 9:44 am

Re: Calculating RC time constants...and why you should.

Post by roberto » Sat Oct 17, 2009 1:53 pm

just seen the reply to my post.
mogo, the Ck bypass makes a shelving boost. So you have a totally different boost that goes between an unbypassed cathode gain and a bypassed cathode gain. :wink:

parkhead
Senior Member
Posts: 140
Joined: Thu Jun 21, 2007 8:35 pm
Just the numbers in order: 7

Re:

Post by parkhead » Fri Feb 19, 2010 9:13 am

MattPete wrote:
flemingmras wrote:And that's due to the harmonics below the string's fundamental frequency. But NO...those aren't important remember? :lol: :lol: :lol: :lol: :lol: :lol: :lol: :lol:
There are no harmonics below the fundamental. No matter how hard I bash or whisper sweet nothings to the low E string, it will never produce a 40Hz frequency when in standard tuning.

unless you are driving the string with a Feather, and Ebow or swelling the volume after picking, and have carved stone guitar clamped to 2 ton workbench ... there is tons of stuff going on below the fundemental tone of the string pitch

simply consider the act of picking, unless you can stimulate the string without touching it the attack of the pick and its twang of release are happening well below 40hz with all of the coresponding harmonic overtones

eddie van halen
Jimi Hendrix
angus young
and Gordie Johnson
Jeff Beck

are just a few of the Many players who make excellent expressive use of this
in fact any time you think the flick of a selector switch is kinda cool
or heard Stevie Ray POP his low e string (the note eventally settles to e or e flat, but the pop is probably 10hz)
... you've heard why its important

just look at a tuner display and see how long it takes to settle on the note you're testing, that whole delay is mostly the string being unsettled by sonic info below the pitched of tuned note of your string, the faster tuners are just better at "filtering" out that extra info faster

You've been BUSTED

p
replica ?? I don't need no stinking replica ...

Roe
Senior Member
Posts: 5050
Joined: Thu Apr 13, 2006 1:36 pm
Just the numbers in order: 7
Location: Drontheim. Norwegen
Contact:

Re: Calculating RC time constants...and why you should.

Post by Roe » Sun Feb 21, 2010 3:16 pm

what's the frequency the .022 pi input coupling cap pass. does it pass everything down to a few hz? could the cap size be reduced without much difference in bass response?

EDIT: aiken show that 1500pf gives 53hz. so a 22nf does indeed pass down to a few hz
http://www.myspace.com/20bonesband" onclick="window.open(this.href);return false;
http://www.myspace.com/prostitutes" onclick="window.open(this.href);return false;
Super 100 amps: 1202-119 & 1202-84
JTM45 RS OT JTM50 JMP50 1959/2203/34/39

User avatar
novosibir
Senior Member
Posts: 4650
Joined: Tue Nov 22, 2005 2:32 pm
Just the numbers in order: 7
Location: Nuernberg, Germany
Contact:

Re: Calculating RC time constants...and why you should.

Post by novosibir » Mon Jun 07, 2010 7:05 pm

Roe wrote:EDIT: aiken show that 1500pf gives 53hz.
But that's already the -3dB point, where the signal voltage is diminished to 71% and the point, where 45 degrees of phase shift occurs :wink:

Larry
The fault almost always is sitting in front of the amp :wink:

Larry's Website now with included Pix's Gallery

Roe
Senior Member
Posts: 5050
Joined: Thu Apr 13, 2006 1:36 pm
Just the numbers in order: 7
Location: Drontheim. Norwegen
Contact:

Re: Calculating RC time constants...and why you should.

Post by Roe » Tue Jun 08, 2010 3:29 am

I must admit I don't understand the phase shift part.
http://www.myspace.com/20bonesband" onclick="window.open(this.href);return false;
http://www.myspace.com/prostitutes" onclick="window.open(this.href);return false;
Super 100 amps: 1202-119 & 1202-84
JTM45 RS OT JTM50 JMP50 1959/2203/34/39

User avatar
StuntDouble
Senior Member
Posts: 1589
Joined: Tue Dec 18, 2007 6:26 pm
Just the numbers in order: 7
Location: the left coast

Re: Calculating RC time constants...and why you should.

Post by StuntDouble » Sun Feb 27, 2011 12:42 pm

I'm sorry if this is a dumb question but...

If I understand this formula correctly, my .68uf/820 ohm cap resistor combo at V1 should yield around a 285hz cutoff, and if I understand ohm's law correctly, I should get much less voltage drop across a smaller resistor value.

Here's my question:

If I use a 20uf/28 ohm C/R setup at V1, I calculate that I'll have roughly the same freq cutoff point as .68uf/820 ohms. Will I get more gain at V1 due to less voltage drop across the resistor witha 20uf/28 ohm? Has anyone tried this; will it sound like ass?
V=CxR...so if I measure the current in ma at the input for the two values and then subtract the difference I should have the difference in votlage drop of the two respective values. :?

I've used a 620 ohm resistor with a .82uf mustard at the cathode follower, but I couldn't hear any real appreciable difference.

User avatar
6burgh
Senior Member
Posts: 521
Joined: Wed Oct 21, 2009 10:14 am
Just the numbers in order: 7
Location: Sniffing Solder Fumes

Re: Calculating RC time constants...and why you should.

Post by 6burgh » Sun Feb 27, 2011 2:19 pm

I'm not an expert, but I think I've read in several places and technical explanations that bad things start happening like grid current limiting (clipping) at somewhere near or lower than 600 ohm cathode resistance given the other characteristics of a standard 12ax7 with a 100k anode/plate resistor. The attachment is overkill, but very informative in general. I hope this helps, and sorry for my lack of details.
Attachments
Common_Gain_Stage.pdf
(701.43 KiB) Downloaded 415 times

User avatar
StuntDouble
Senior Member
Posts: 1589
Joined: Tue Dec 18, 2007 6:26 pm
Just the numbers in order: 7
Location: the left coast

Re: Calculating RC time constants...and why you should.

Post by StuntDouble » Sun Feb 27, 2011 4:29 pm

6burgh wrote:I'm not an expert, but I think I've read in several places and technical explanations that bad things start happening like grid current limiting (clipping) at somewhere near or lower than 600 ohm cathode resistance given the other characteristics of a standard 12ax7 with a 100k anode/plate resistor. The attachment is overkill, but very informative in general. I hope this helps, and sorry for my lack of details.
ahhhh, so that's why; thanks. :)

yladrd61
Senior Member
Posts: 248
Joined: Thu Jun 17, 2010 5:03 am
Just the numbers in order: 7

Re:

Post by yladrd61 » Tue Aug 09, 2011 11:53 pm

MattPete wrote:
flemingmras wrote:And that's due to the harmonics below the string's fundamental frequency. But NO...those aren't important remember? :lol: :lol: :lol: :lol: :lol: :lol: :lol: :lol:
There are no harmonics below the fundamental. No matter how hard I bash or whisper sweet nothings to the low E string, it will never produce a 40Hz frequency when in standard tuning.
It will if you drop some Jimi bombs with the whammy bar :jimi:

RobBozic
New Member
Posts: 26
Joined: Fri Sep 15, 2006 1:47 am

Re: Calculating RC time constants...and why you should.

Post by RobBozic » Mon Feb 27, 2017 10:54 pm

For coupling caps you have to take into account the mixer resistors and the other channels Volume pot setting.
Assuming it's on zero, it would approximately be (470+470)*1000/(470+470)+1000=484k,plus 40k output imp.
F=1/6.28*.000000022*524000 = 13.8Hz for a 22nf, and 138Hz for a 2.2nf capacitor.
This is for the Volume on '10', and the other Volume on '0'.
Last edited by RobBozic on Wed Mar 01, 2017 5:44 am, edited 1 time in total.

Roe
Senior Member
Posts: 5050
Joined: Thu Apr 13, 2006 1:36 pm
Just the numbers in order: 7
Location: Drontheim. Norwegen
Contact:

Re: Calculating RC time constants...and why you should.

Post by Roe » Tue Feb 28, 2017 4:43 am

http://www.myspace.com/20bonesband" onclick="window.open(this.href);return false;
http://www.myspace.com/prostitutes" onclick="window.open(this.href);return false;
Super 100 amps: 1202-119 & 1202-84
JTM45 RS OT JTM50 JMP50 1959/2203/34/39

Post Reply