Lead Extension - Switch Types - Signal Path - Other

[TurbOSquiD77]

Hello everyone,

I have a few things I'm having a tough time finding an answer for. Could I ask of your knowledge? I appreciate it!

1.) What is the difference between twisting leads, and wrapping one lead around another in terms of how they affect the circuit? Power IEC inlet to power transformer, or the diodes in S.E.X. 3.0 for example. Or extending leads on "high end" capacitors with silver leads so I don't ruin them from rolling caps. Copper set screw lugs and the lead extensions have worked great so far. Is there a proper way to extend? What about the tightness of wraps, like from the RCA inputs to the selector switch - is there a difference in physics between super loose and super tight?

2.) I'd like to implement a switch with 3 resistance values in the Crackatwoa, as some have done with their Crack in order to easily roll 12au7, 12bh7a, e80cc....etc. Is there a thread with this specific information? What would be the best switch to use for the job of 2 values? 3 values? Are these R1 resistors on the low current board in the signal path? I can mount, but just need to be sure on how to configure, and what switch to use.
Would there be switches that are better choices than these (materials used)?
http://www.partsconnexion.com/controls_toggle_switch.html
https://www.hificollective.co.uk/switches/tapesource_switch.html

3.) How could I determine what wires, resistors, and other components are in the signal path of the Crackatwoa? Or maybe a better question would be.... what isn't?
Maybe a list of part positions and/or terminals to pay attention to. If one would like to add a short description of what this part does without going into too great of detail they surely can, but do not feel obligated to do so! I can research on my own what the position does in detail if I know what it's function is, and what other parts it affects directly. 

4.) Can other resistor types be used in place of metal oxide parts on high current boards without negative consequence? What do these resistors accomplish? What rating of theirs is important in shunt regulation? What specifications of shunt regulation are important to pay attention to in regards to audible and non-audible performance? 5,000 - 10,000uF - 50,000uF shunt reg cap values - difference in performance?

5.) What would the difference in performance be between an inductive power supply resistor, and a non-inductive part (270r 5W - 300r 10W in my application)? What is recommended?

6.)  What would the difference in performance be between resistor types for the 270k 1W resistor in the power supply? Carbon film comes with the kit correct? Does the C-7X choke people sometimes use in the Crack replace this resistor? I understand this may not be very audible due to the shunt regulator in the C2A - just wondering.

7.) Is anything below a 5%, or 1% tolerance audible for resistors in the signal path? I don't plan to do it, but just wondering about those Vishay/Dale PTF(56) .01% series parts if I ever rebuild the TwoQuiet. However, this would be crazy expensive so I'm just curious about performance differences. 

8.) What type of resistor are the 2.49k 1/4W resistors on the low current C4S board? They look a bit different than most of the others.

9.) My Crackatwoa brings USB/Computer noise through the circuit, and my S.E.X. does not. Is this due to the C2A being OTL? I have pinpointed the issue to be USB noise, and I'm ok with that. It gets worse with computer load and input..... lots of other t-shooting steps have been taken.



Most of these questions I'm pondering have to do with my curiosity, and quest for higher knowledge about electronics more than the actual need to replace parts. I understand I may not have a good basic understanding of electronics, but these will help me understand the concepts I encounter along the way.

Thank you very much, and happy listening!

-T

[Paul Birkeland]

1.) What is the difference between twisting leads, and wrapping one lead around another in terms of how they affect the circuit? Power IEC inlet to power transformer, or the diodes in S.E.X. 3.0 for example.

You have to be a bit more specific here.  When you say "leads", I assume bare pieces of wire attached to a component like a resistor or capacitor.  When you talk about twisting leads, are you talking about leads or jacketed wires?

Or extending leads on "high end" capacitors with silver leads so I don't ruin them from rolling caps. Copper set screw lugs and the lead extensions have worked great so far. Is there a proper way to extend? What about the tightness of wraps, like from the RCA inputs to the selector switch - is there a difference in physics between super loose and super tight?

Again, you're talking about two very different things here.  In my opinion, it is never OK to extend the leads of a capacitor, especially a capacitor that isn't physically mounted to a chassis plate.

The general rule of thumb for a twisted pair of wires (like those running from RCA jacks or feeding tube heaters) is about three twists per inch.

2.) I'd like to implement a switch with 3 resistance values in the Crackatwoa, as some have done with their Crack in order to easily roll 12au7, 12bh7a, e80cc....etc. Is there a thread with this specific information?

I wouldn't generally recommend that, since the R1 value on the "A" side of the big C4S boards should move a little as well as the R1 values on the small center board.  Pick the tube you like the most and use that.

3.) How could I determine what wires, resistors, and other components are in the signal path of the Crackatwoa? Or maybe a better question would be.... what isn't?

The high voltage DC power supply isn't in the signal path (not in a significant way), nor is the AC supply that heats all the tubes.  The R2 resistors on the C4S boards are also not in the signal path.

Maybe a list of part positions and/or terminals to pay attention to. If one would like to add a short description of what this part does without going into too great of detail they surely can, but do not feel obligated to do so! I can research on my own what the position does in detail if I know what it's function is, and what other parts it affects directly. 

You'll have to ask specific questions about those parts that you're interested in.  For me to sit down and write a paragraph about what each resistor, capacitor, diode, and transistor does in the circuit would take weeks.

4.) Can other resistor types be used in place of metal oxide parts on high current boards without negative consequence? What do these resistors accomplish? What rating of theirs is important in shunt regulation? What specifications of shunt regulation are important to pay attention to in regards to audible and non-audible performance? 5,000 - 10,000uF - 50,000uF shunt reg cap values - difference in performance?

The metal oxide R2 resistors are not the signal path.  These resistors bias the two LED's in each constant current source.  The capacitors similarly are also not in the signal path.  They provide a bit of a slow delay during amplifier warmup.

5.) What would the difference in performance be between an inductive power supply resistor, and a non-inductive part (270r 5W - 300r 10W in my application)? What is recommended?

The inductive power supply resistor will do a slightly better job of filtering high frequency noise that isn't likely to be present. In power supply filtering, there is little difference between the two. (See my answer about what's not in the signal path)

6.)  What would the difference in performance be between resistor types for the 270k 1W resistor in the power supply? Carbon film comes with the kit correct? Does the C-7X choke people sometimes use in the Crack replace this resistor? I understand this may not be very audible due to the shunt regulator in the C2A - just wondering.

The C-7X is a 270 Ohm choke, the 270K 1W resistor is 270,000 Ohms.  If you replace it with that choke, you will draw power through the power transformer until the fuse blows, a UF4007 diode blows, or the power transformer blows.  I don't recommend this.  The C-7X is a popular replacement for the 270 Ohm power supply filtering resistors in the stock Crack, since the high voltage power supply is handling signal current.   In the shunt regulated C2A, this is not the case, and the addition of power supply chokes is not considered helpful.

7.) Is anything below a 5%, or 1% tolerance audible for resistors in the signal path? I don't plan to do it, but just wondering about those Vishay/Dale PTF(56) .01% series parts if I ever rebuild the TwoQuiet. However, this would be crazy expensive so I'm just curious about performance differences. 

You can calculate the differences in Decibels between 1% and .01% resistors in the TwoQuiet.  I would suspect that you are talking about a few 1/100th's of a dB in most positions.

8.) What type of resistor are the 2.49k 1/4W resistors on the low current C4S board? They look a bit different than most of the others.

They are Vishay/Dale metal film 1% resistors.

9.) My Crackatwoa brings USB/Computer noise through the circuit, and my S.E.X. does not. Is this due to the C2A being OTL? I have pinpointed the issue to be USB noise, and I'm ok with that. It gets worse with computer load and input..... lots of other t-shooting steps have been taken.

A DAC connected by a Toslink cable will eliminate this issue.

[Doc B.]

The radiated magnetic field of a wire is perpendicular to the direction of the current through the wire. So you want the two wires to be as perpendicular as possible when they are twisted, so that the magnetic fields of the two wires cancel as much as possible. Wrapping one around the other is not as optimal. Someone theorized at some point that three turns per inch is the optimal size for the twist. Is it really? I supposed it depends upon the gauge of wire and the thickness of the insulation. But it's a good starting point.

[TurbOSquiD77]

Thanks for the fast reply Paul! Thanks for the addition Doc! I understand some of this a bit better now.

Wire Twisting:
I see jacketed wires twisted tightly in some places (5-10 twists per inch), and some more loosely (2-3 twists per inch). Is one method better than the other for any position? Anywhere not to do this? I understand now why twisting input wiring would be beneficial.


Resistance selector switch:
As long as Kreg on the low current board stays within tolerances would it be okay? I do enjoy rolling between the three tubes I've tried. Not too often do I change them, but I want to prevent having to desolder and solder new values too many times on the R1 low current position. Kreg drops ~1mV in the current configuration (High Current R1A @ 12.1 ohms) when going from 12bh7a to e80cc, ~5.5mV and 4.5mV respectively. Havent run the ecc82 for a while! I'll have to try again soon - super lush mids with the Philips Miniwatt!


Signal Path:
I've been looking at Lee's thread for signal path wiring identification. I assume he's placed silver wire where audible. I can see what I need to here, except for what would wires would be connected to the C4S boards, and which parts/wiring connected are in the signal path. Currently I understand what you've said about the DC or AC power supply, R2 metal oxide resistors, and lytic caps.
http://bottlehead.com/smf/index.php?topic=8937.0


Power Supply Choke/Resistors:
Thanks for the explanation about the C-7X. Makes sense to me! 


Resistor Tolerances:
Difference between 1% and .01% tolerances for TwoQuiet resistors wouldn't be worth it - got it lol. Especially for the price of the PTF's! Didn't realize that since this position is audible that we would be calculating the tolerance difference in dB. I thought it might be speed change or decay or something. How about resistor types such as Takman carbon film or Kiwame vs using a tight tolerance type like metal film Dales? I most likely will not go the Takman/Kiwame route but just curious if there would be a noticeable tone change if used here.


USB Noise:
Definitely looking at SPDIF solutions in place of USB. The S.E.X. is absolutely dead silent I love it!


Lastly, are there any non-audible performance gains you choose to implement in circuit designs? I'm sure this ties into how the circuit is created in the first place, and could get quite complex, so no need to go into this unless something more broad comes to mind.

Again, thank you.

-T

[Paul Birkeland]

Wire Twisting:
I see jacketed wires twisted tightly in some places (5-10 twists per inch), and some more loosely (2-3 twists per inch). Is one method better than the other for any position? Anywhere not to do this? I understand now why twisting input wiring would be beneficial.

Sometimes we (meaning mostly me) get a little ham-fisted with the drill and twisting wires, so there is some variation in the number of twists per inch.

As long as Kreg on the low current board stays within tolerances would it be okay? I do enjoy rolling between the three tubes I've tried. Not too often do I change them, but I want to prevent having to desolder and solder new values too many times on the R1 low current position. Kreg drops ~1mV in the current configuration (High Current R1A @ 12.1 ohms) when going from 12bh7a to e80cc, ~5.5mV and 4.5mV respectively.

You'd ideally want the lowest value R1 on the "A" side of the big C4S board that was required to operate the amp properly.  The risk is that using less driver current in this configuration will make the 6AQ5 draw more current to take up the excess, which increases its dissipation and will tend to drop Kreg voltage lower.  If Kreg gets to or drops below 2.5V, then the regulator doesn't work anymore.  I wouldn't worry as much about 6AQ5 dissipation, those little things are tough as nails.

I've been looking at Lee's thread for signal path wiring identification. I assume he's placed silver wire where audible. I can see what I need to here, except for what would wires would be connected to the C4S boards, and which parts/wiring connected are in the signal path. Currently I understand what you've said about the DC or AC power supply, R2 metal oxide resistors, and lytic caps.
http://bottlehead.com/smf/index.php?topic=8937.0

Mostly the wires leaving the "O" pads on the C4S boards indicate the presence of signal current.

Didn't realize that since this position is audible that we would be calculating the tolerance difference in dB. I thought it might be speed change or decay or something. How about resistor types such as Takman carbon film or Kiwame vs using a tight tolerance type like metal film Dales? I most likely will not go the Takman/Kiwame route but just curious if there would be a noticeable tone change if used here.

Tight tolerance metal film Dale resistors are what's in the kit.  In the case of the attenuator resistors, if you start to stray from the values specified, you'll get uneven steps in the attenuator.  Also, the Kiwame resistors are way bigger than I would recommend, and I doubt the leads would fit through the holes on the attenuator switch.  Some people have had good results by replacing the two 34.8K resistors from the terminal strip to the fine attenuator with a custom made pair of foil resistors from Texas Components Inc.

USB Noise:
Definitely looking at SPDIF solutions in place of USB. The S.E.X. is absolutely dead silent I love it!

Toslink will break the coupling of the noise, SPDIF might, then again it might not.

Lastly, are there any non-audible performance gains you choose to implement in circuit designs? I'm sure this ties into how the circuit is created in the first place, and could get quite complex, so no need to go into this unless something more broad comes to mind.

Yes, we de-rate the dissipation of most of our components significantly to increase their lifetime.  We also look pretty carefully at the expected lifetime of the capacitors we purchase to ensure that we aren't providing "computer grade" components only made to last a year or two. 

-PB

[TurbOSquiD77]

Sometimes we (meaning mostly me) get a little ham-fisted with the drill and twisting wires, so there is some variation in the number of twists per inch.

Hah, all good. I just wanted to know if there's a difference in performance between how tight the twists are etc...

You'd ideally want the lowest value R1 on the "A" side of the big C4S board that was required to operate the amp properly.  The risk is that using less driver current in this configuration will make the 6AQ5 draw more current to take up the excess, which increases its dissipation and will tend to drop Kreg voltage lower.  If Kreg gets to or drops below 2.5V, then the regulator doesn't work anymore.  I wouldn't worry as much about 6AQ5 dissipation, those little things are tough as nails.

Understood. I'll study my voltage test document before making any changes, and continue to monitor Kreg voltages

Mostly the wires leaving the "O" pads on the C4S boards indicate the presence of signal current.

Cool. O pads also trace to other areas of the board which intersect with non O pads. Is signal current separated from these terminals/parts somehow?

Tight tolerance metal film Dale resistors are what's in the kit.  In the case of the attenuator resistors, if you start to stray from the values specified, you'll get uneven steps in the attenuator.  Also, the Kiwame resistors are way bigger than I would recommend, and I doubt the leads would fit through the holes on the attenuator switch.  Some people have had good results by replacing the two 34.8K resistors from the terminal strip to the fine attenuator with a custom made pair of foil resistors from Texas Components Inc.

Thanks. Why only the 34.8k parts? I've had my eye on the Z and S-Foil resistors - spendy! Duelund CAST as well.... candle sticks.

Toslink will break the coupling of the noise, SPDIF might, then again it might not.

Will experiment. Although, I'll have to play with my drivers/sample rates being used because I get crazy distorted popping when using toslink. Using ASIO Bridge and a Amanero USB 384. I'll post my findings.


Cheers!

[Paul Birkeland]

O pads also trace to other areas of the board which intersect with non O pads. Is signal current separated from these terminals/parts somehow?

The O pads trace to the collector of either an NPN or PNP transistor.  The impedance of the C4S is very, very high from here, and the signal current path is through each transistor, then through R1.