crackle NOT okay

Started by Stratomaster, May 10, 2007, 02:00:23 PM

Previous topic - Next topic

markm


Isaiah

Thanks, Sebastian!

Does anybody know which console the SHO circuit was supposedly taken from?

O

Quote from: Isaiah on May 24, 2007, 01:34:09 PM
Thanks, Sebastian!

Does anybody know which console the SHO circuit was supposedly taken from?

Neve?  ???

markm

Quote from: Isaiah on May 24, 2007, 01:34:09 PM
Thanks, Sebastian!

Does anybody know which console the SHO circuit was supposedly taken from?

I read it somewhere but don't remember where.....ZVEX site maybe?  ???

Isaiah

Quote from: O on May 24, 2007, 01:48:18 PM
Quote from: Isaiah on May 24, 2007, 01:34:09 PM
Thanks, Sebastian!

Does anybody know which console the SHO circuit was supposedly taken from?

Neve?  ???

Possibly...
http://zvex.com/hardon.html

"This is the perfect preamp pedal. The "Crackle Okay" volume knob is a negative-feedback control styled after classic 60's recording console inputs. (They crackled when adjusted too.)"

markm

Quote from: Isaiah on May 24, 2007, 04:15:57 PM
Quote from: O on May 24, 2007, 01:48:18 PM
Quote from: Isaiah on May 24, 2007, 01:34:09 PM
Thanks, Sebastian!

Does anybody know which console the SHO circuit was supposedly taken from?

Neve?  ???

Possibly...
http://zvex.com/hardon.html

"This is the perfect preamp pedal. The "Crackle Okay" volume knob is a negative-feedback control styled after classic 60's recording console inputs. (They crackled when adjusted too.)"

Okay.
Well, I was close!  :-\

Isaiah

Hmmm...

I don't know whether that means he took the whole front end of a channel on a console
or merely the gain idea (it's a generic gain control though).

I'd be interested to know if anyone has tried plugging a dynamic mic into it
(just something standard like a '57 or '58 using an XLR to 1/4" cable). Anyone?
I'm looking to build myself a little mixer (about 4 channels, briefcase-sized) for portable recording, so this circuit
could make a great mic pre-amp.

Thanks :)

WGTP

Is it correct, that the lower the source resistance, the higher the gain and  the more asymmetrical the distortion? :icon_eek:
Stomping Out Sparks & Flames

marcoharder

Quote from: spudulike on May 10, 2007, 03:51:34 PM
If you got the schematic from a certain japanese website, then the 1M resistors around the mosfet should actually be 10M.

Is there going to be a big difference in tone if 1M resistors are used?

remmelt

If I understand it correctly, the difference is that the input impedance is (a lot) lower. The original boasts a 5M (2x10 in parallel) input impedance as defining for the sound. As with everything, some people like it, some people don't. I like it quite a bit, especially for clean sounds: better string separation, a more "round" sound from the guitar, very pleasing. Other people don't like it at all. Add to taste! Just try the different values and see which one you like best. I think the circuit should be OK if those two resistors are equal. If you don't have the 10M resistors, just try a couple of your largest value in series. See if you can hear the difference between the 1M and the 10M version. Should be noticeable, even with something from 6M up, I guess. If you don't have one yet, buy a breadboard. It's really quite handy for stuff like this.

R.G.

Gentlemen:
- the resistor crackles because there is DC running through it and you change the current by moving the wiper.
- you can change the gain only by using a fixed source resistor and a capacitor to a pot which does not have DC running through it. However, this does not change the bias point like the crackling resistor does.
- An LDR would be a perfect replacement for the crackling resistor. However, it's hard to get LDRs down that low in resistance. A custom setup with several low-resistance LDRs in parallel should do it. You could servo the resistance against a real pot by using something like the VT5C3/2 and using one section as the source resistor and one section as the servo resistor, then driving the LED from an opamp which detects the difference between the second LDR section and the control pot. No crackle, same feel as the pot.
- the input impedance is, strictly speaking, only the gate-source capacitance, reduced by the amount of feedback. As long as the source resistor is significant compared to the gate-source capacitive impedance, the gate looks like an open circuit. When the capacitance comes into play, there will be some treble bleed.
- As you turn down the source resistor, the gain goes up and the bias point shifts. There comes a place where treble starts to roll off too.
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

puretube

fix`d?

hey, I still ain`t seen that schemo...  :icon_eek:

R.G.

R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

puretube


stm

WATCH OUT!  Input impedance is not 5 Mohms, and it is gain-dependent !!!

I ran several simulations to investigate the input impedance claim.  Yes, at first I expected to get 5 Mohms as a pair of 10Megs in parallel give that...  Wrong!

Input impedance in fact changes as the gain of the circuit (and bias point) are changed with the Crackle OK knob...

The truth is as follows:

1) At min gain you have unity gain and about 3 Meg input impedance, not 5 Meg (!)
2) At noon you have like 20 dB of gain and about 1 Meg input impedance (!!)
3) At max gain you have about 40 dB of gain and around 200k input impedance (!!!)

The above may seem counterintuitive, so those of you who aren't buying this should try to run your own independent simulations to verify.

After looking for an explanation to the impedance reduction phenomenon I noticed that the biasing network is not taken from a fixed voltage or bypassed by a capacitor.  Instead, it comes out from the drain where a negative version of the input is present.  This has the effect of reducing the input impedance.

Have you heard about boostrapping?  One of its uses is to increase the input impedance of an emitter follower by taking a posistive version of the output (at the transitor emitter) and feeding it back to the base of the transistor.fed back to its base to increase the apparent input impedance), which is, reduces input impedance as gain increases.  In case of the SHO the opposite is true, since a negative version of the output is taken from the drain back to the gate.

analogguru

Quotethe resistor crackles because there is DC running through it and you change the current by moving the wiper.

Why do you think, that the original Dallas Rangemaster used a wire-wound potentiometer ?

analogguru

R.G.

QuoteAfter looking for an explanation to the impedance reduction phenomenon I noticed that the biasing network is not taken from a fixed voltage or bypassed by a capacitor.  Instead, it comes out from the drain where a negative version of the input is present.  This has the effect of reducing the input impedance.
There are two things happening. The first is what I mentioned - the feedback from an unbypassed resistance effectively shields the input signal from the Cgs capacitance of the MOSFET. The second is that the Cgd is multiplied by the Miller effect. The sum of the two is what loads down the input signal.

With high gains, you probably want the signal loaded down, for the same reason that most fuzzes roll off treble before clipping it. The distortion sound is less harsh. You could check that out by running a MOSFET or JFET source follower in front of this thing and seeing what that does by unloading the input signal from the gate capacitance.

QuoteWhy do you think, that the original Dallas Rangemaster used a wire-wound potentiometer ?

... um... lemme see... because the designer owned stock in a nickel-chromium wire company?    :icon_lol:
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

analogguru

#97
Quote
Quote
Why do you think, that the original Dallas Rangemaster used a wire-wound potentiometer ?

... um... lemme see... because the designer owned stock in a nickel-chromium wire company?

hmmm.... I assume that it doesn´t crackle so much because by moving the slider acts like a switch shortening one wound after the other, you wanna try it out ?

analogguru

Quote
QuoteDoes anybody know which console the SHO circuit was supposedly taken from?

Neve? 

Forget to think about it, it´s only another " make-business lie":

Quote

Zvex wrote:
Quote
The console which I dissected to discover the use of a negative-feedback resistor adjustment was designed by Rupert Neve for Philips Corporation of Holland.  In his design, the microphone input transistor (bipolar) immediately following the input transformer had a selector switch on the emitter resistor, allowing the user to select the gain and headroom simultaneously.  The four positions on this switch were best selected with the monitor speakers turned down, of course, because the console's output made a horrific "POP!!!" when the switches were adjusted, as well as a crackling noise if they were wiggled.  I traced one of the mic pre modules to discover this, and carried this idea to an adjustable negative feedback resistor in my MOSFET gainstage design.

analogguru wrote:

hmmm ... did the "mic-pre module you traced" look like this or different:



This is the oldest Neve-design I am aware of.  Assuming that you don´t want to engage also in the studio-market and want to clone old Neve-designs for profit,  my request shouldn´t violate forum-rules:
Maybe it is possible that you post the traced schematic ?

This is really interesting new stuff for me what I read above.

analogguru

Zvex wrote:

QuoteI didn't trace the entire schematic.  I only traced out how the gain section worked.  The board was a Philips broadcast console, made for radio stations.  There were two in Minneapolis at one time and I purchased one of them for about $400 and traded it later for a couple of RCA tube mic preamps.  I don't have the schematics or pictures of the modules.  It was a glorious-looking thing, but it had lots of mechanical problems.


Maybe he "traced the gain section" of a Philips mixer but not designed by Rupert Neve.

as you can read here, "The first client for the new transistor equipment was Phillips Records Ltd", located in London not Holland.

Yes, Philips also had a record department, like Mitsubishi is selling semiconductors and cars.

So Philips Records (London) bought a Neve console looking like this:

http://rupertneve.com/wp-content/i/doubledesk.jpg

And this console was based on the modules shown above.

The first transistor stage had a fixed 22k from emitter to ground bypassed by a 100µF electrolytic, so no chance of switching, popping or crackling at the "transistor (bipolar) immediately following the input transformer".

What we can learn from this?:
"If you need to lie, you should look first, that noone can proof that you are lying."

analogguru

Gus

#98
I have posted in the past about the bias change and input resistance change with gain (drain gate R feedback) of the SHO.  This makes it different than other mosfet boosts.

It is a nice sounding circuit I like it. 

  One could use a Make Before Break rotary with 12 steps or more like some preamps use to switch the gain/bias source resistors

I have not seen another booster like it(the changing source resistor value,mosfet) made before the SHO.  It is different and the ones I have heard driving good tube amps sound very good. 

  Maybe a wirewound pot as posted by analogguru. 

stm look at the npn boost.

ignsk

very interesting thread
thank you all

Max
== May The Charge Be With You ==