Hi,
Basically in several ways:
Twisting a lead that carries audio signal and feeding it to sonething doesn't feel happy with too much capacitance on the input.
Be aware that this capacitance may vary from channel to channel which in turn may alter frequency response between channels.
There's more to say but let's say that twisted leads may be used irresponsibly.
Cheers,
Basically in several ways:
Twisting a lead that carries audio signal and feeding it to sonething doesn't feel happy with too much capacitance on the input.
Be aware that this capacitance may vary from channel to channel which in turn may alter frequency response between channels.
There's more to say but let's say that twisted leads may be used irresponsibly.
Cheers,
Frank,
I understand your concern and where you're heading.
However, this thread is heading in a slightly different direction
We were mainly talking about the power supply conductors, and with these impedance levels, the naughty boy "Mr. Miller" can go and play elsewhere. (To put things as politically correct as possible here *grin*)
You're totally correct that twisting can give you a headache if one is dealing with small signal cables. (Mr. Miller then goes:
)
However, when it somes to signal cables, I would choose to use a coaxial signal cable (not meening the stiff antenna cable), but the flexible stuff. Just make sure it's with a weaving (word?) structure shielding (the stuff that goess criss-cross). If you use the shield that is just spinning around in one direction along the inner conductor axis (rolled on) , you may get into problems.
Jens
I understand your concern and where you're heading.
However, this thread is heading in a slightly different direction
We were mainly talking about the power supply conductors, and with these impedance levels, the naughty boy "Mr. Miller" can go and play elsewhere. (To put things as politically correct as possible here *grin*)
You're totally correct that twisting can give you a headache if one is dealing with small signal cables. (Mr. Miller then goes:
)However, when it somes to signal cables, I would choose to use a coaxial signal cable (not meening the stiff antenna cable), but the flexible stuff. Just make sure it's with a weaving (word?) structure shielding (the stuff that goess criss-cross). If you use the shield that is just spinning around in one direction along the inner conductor axis (rolled on) , you may get into problems.
Jens
Re: Twisting the field away...
It's worth nothing that in a class-B amp, current flows in a loop between PSU +ve and ground one one half-cycle, and PSU -ve and ground on another. So it's especially critical that +ve and -ve are close to each other, otherwise the resulting magnetic field will be a half-wave rectified version of the signal. This will contain lots of distortion products, which are much more likely to be objectionable (than, say, the field produced by the speaker output lead and its ground return).
Cheers
IH
Jennice said:
When a positive signal is applied to a non-inverting class B amp, the current runs from the positive supply rail through part of the output stage, then to the speaker, and back to the speakers negative terminal. This terminal is a ground line. NOW the ground is "actively" participating as is conducts current
It's worth nothing that in a class-B amp, current flows in a loop between PSU +ve and ground one one half-cycle, and PSU -ve and ground on another. So it's especially critical that +ve and -ve are close to each other, otherwise the resulting magnetic field will be a half-wave rectified version of the signal. This will contain lots of distortion products, which are much more likely to be objectionable (than, say, the field produced by the speaker output lead and its ground return).
Cheers
IH
Re: Re: Twisting the field away...
Exactly
With all due respect, Ian, you just proved yourself wrong in your posted statement.
Let's look at the "positive" half cycle:
The current flows from PSU+ through the amp and load, and then back to GND.
In this situation, the high current flows "out" (as seen from the supply wires) at PSU+ and returns through GND. Therefore the "couple" to keep together would be PSU+ and GND.
The wire at PSU- is not "active" (apart from the relatively small circuit supply current) and can therefore be neglected.
In the negative half, the current runs "out" from GND and returns at PSU-, which has a relatively lower potential than GND. In this half cycle, PSU+ is not "active", and the twisted couple should therefore be GND and PSU-.
The only way to obtain both is to twist all three together.
If you only twist PSU+ with PSU-, you obtain exactly what you describe in your text (quoted above). As you said yourself, only PSU+ OR PSU- is active at any given time. The resulting field is then a result of only one active conductor, wile the current's return path to the PSU is not included in the twist you describe. At any given time (except when crossing through zero or when no signal is applied), the GND wire is an active conductor at "full time", while the two PSU + and - take turns being active.
Please read my original post again, and think it over again. The sum of currents from one point (electrons arriving + electrons leaving) is always zero. Otherwise you'd have an eternal power source, which I'm sure you could sell to the world for a good price.
Jens
IanHarvey said:
Exactly
With all due respect, Ian, you just proved yourself wrong in your posted statement.
Let's look at the "positive" half cycle:
The current flows from PSU+ through the amp and load, and then back to GND.
In this situation, the high current flows "out" (as seen from the supply wires) at PSU+ and returns through GND. Therefore the "couple" to keep together would be PSU+ and GND.
The wire at PSU- is not "active" (apart from the relatively small circuit supply current) and can therefore be neglected.
In the negative half, the current runs "out" from GND and returns at PSU-, which has a relatively lower potential than GND. In this half cycle, PSU+ is not "active", and the twisted couple should therefore be GND and PSU-.
The only way to obtain both is to twist all three together.
IanHarvey said:
If you only twist PSU+ with PSU-, you obtain exactly what you describe in your text (quoted above). As you said yourself, only PSU+ OR PSU- is active at any given time. The resulting field is then a result of only one active conductor, wile the current's return path to the PSU is not included in the twist you describe. At any given time (except when crossing through zero or when no signal is applied), the GND wire is an active conductor at "full time", while the two PSU + and - take turns being active.
Please read my original post again, and think it over again. The sum of currents from one point (electrons arriving + electrons leaving) is always zero. Otherwise you'd have an eternal power source, which I'm sure you could sell to the world for a good price.
Jens
Re: Re: Re: Twisting the field away...
To clarify what I meant: if you twist PSU +ve and PSU -ve, you end up with the equivalent of one conductor which carries the output current going to the speaker. If the return ground wire is not near this, there will be a resulting field, but it will be linearly related to the signal (i.e. sine wave in = sine wave out).
If, however, the +ve wire is not near the -ve wire, there will be a resulting field and it will be non-linearly related to the signal. My general point is that you can tolerate much more of the former than the latter: a 1% deviation in frequency response (corresponding to a purely linear error signal) is vastly less audible than 1% distortion.
Cheers
IH
Jennice said:
To clarify what I meant: if you twist PSU +ve and PSU -ve, you end up with the equivalent of one conductor which carries the output current going to the speaker. If the return ground wire is not near this, there will be a resulting field, but it will be linearly related to the signal (i.e. sine wave in = sine wave out).
If, however, the +ve wire is not near the -ve wire, there will be a resulting field and it will be non-linearly related to the signal. My general point is that you can tolerate much more of the former than the latter: a 1% deviation in frequency response (corresponding to a purely linear error signal) is vastly less audible than 1% distortion.
Cheers
IH
I know this is a necro but I'm just posting a couple of links for my own benefit (easier for me to find), so some further supporting info is added to this thread:
What is the basic idea behind the twisted pair? Why are the two wires twisted? How does this arrangement compensate undesirable disturbances?
Twisted pair - Wikipedia
What is the basic idea behind the twisted pair? Why are the two wires twisted? How does this arrangement compensate undesirable disturbances?
Twisted pair - Wikipedia
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.