Recommend low-resistance 20 AWG hookup wire

[SY]

If by "jumper," you mean the plug-and-pin, then the difference is long term reliability, not resistance. You don't need any kind of thick wire- for the fractions of an inch you're talking about, 5 minutes with a wire table and a calculator will tell you how many millionths of ohms of difference you're dealing with.

 

[cliffforrest]

Nope, no snake oil. I am requesting your help in giving me an objective, quantified analysis. Please estimate the differences in resistance (or whatever relevant measure) between those connection options that I mentioned in my last question...

I would like to get an idea based on numbers whether the differences are negligible or not. Thanks!

They are negligible. TOTALLY NEGLIGIBLE! Already stated by several.

What else to you want?

 

[bob91343]

I completely agree with Cliff. I too am an engineer and snake oil isn't one of the courses I took in college.

By now perhaps you are beginning to realize that you are chasing your tail in this. There is no significant difference in audio circuits between various gauges, between stranded and solid, insulated and bare, different colors of insulation, and so on. You are wasting your time on trivia when there are much more significant concerns.

But if you enjoy it, you are welcome to it. However, if you are sincere about making good gear, I suggest learning more about the fundamentals of electricity and then of electronics. It's a wide, complex field, and oversimplifying by asking questions in a forum isn't going to get you anywhere.

 

[mtl777]

They are negligible. TOTALLY NEGLIGIBLE! Already stated by several.

What else to you want?

Well, we were talking about wires for the switch, which you and the others said the resistances were negligible. But now I am asking about comparing the different connection options...

My last question is this: Is there a non-negligible difference between using a jumper to engage two header pins vs soldering a thick wire across those pins? And is there a non-negligible difference between using a jumper to engage the pins vs using a short wire connected from the pins to a switch? (For this question let us consider resistances only and not any hum or screening issues with the wire.)

I would really appreciate your help in giving me quantified comparisons of these, even just rough estimates. I just want to have an idea of how much the differences are.

Peace out.

 

[cliffforrest]

I suggest YOU do the work!

Most here are, like me, very happy to help - especially in nurturing a budding interest in electronics and audio.

However your questions have already been answered if you BOTHER to read the thread.

No more from me.

 

[mtl777]

Alright, I get the message. Suddenly it's not so friendly in here anymore.

 

[SY]

Alright, I get the message. Suddenly it's not so friendly in here anymore.

Problem is, several people have told you how to figure this out. Instead of complaining that someone else hasn't done your work for you, why not try a few of the suggestions? (like the one I gave you a few posts up)

 

[mtl777]

Problem is, several people have told you how to figure this out. Instead of complaining that someone else hasn't done your work for you, why not try a few of the suggestions? (like the one I gave you a few posts up)

Sorry, I was responding to Cliff before I saw your post. Thank you then!

 

[mtl777]

I've done the calculations. Due to lack of electrical data on the switch and jumpers, I had to make a lot of assumptions. Here are the calculations:

(1) Using Jumper Plug - Assuming it is equivalent to a 5mm 12AWG wire with 1.65R/1000ft,

Rjmp = 5mm x (1.65R/1000ft) x (1ft/304.8mm) = 0.00002707R

(2) Soldering 5mm 10AWG Solid Wire

Rsol = 5mm x (1.04R/1000ft) x (1ft/304.8mm) = 0.00001706R

(3) Using 20AWG Wire to Switch - There are two resistances to calculate here, the wire and the switch. The wire is 10in long (round trip) so its resistance is,

Rw20 = 10in x (10.4R/1000ft) x (1ft/12in) = 0.00867R

For the switch, assuming the contacts plus the pair of pins on the switch have a total length of 21.6mm (3.6mm contacts + 9mm x 2 pins) and that this is equivalent to 20AWG wire,

Rsw = 21.6mm x (10.4R/1000ft) x (1ft/304.8mm) = 0.000737R

Therefore, the total for this option is,

Rw20sw = 0.00867R + 0.000737R = 0.009407R

(4) Using 23AWG Wire to Switch - The resistance of the switch is the same as was already calculated, so we only need to calculate the resistance of the 10in long wire.

Rw23 = 10in x (22.6R/1000ft) x (1ft/12in) = 0.01883R

Then the total for this option is,

Rw23sw = 0.01883R + 0.000737R = 0.019567R

Finally, here are the differences between the options:

Rjmp (1) vs Rsol (2) : 0.00001001R

Rw20sw (3) vs Rjmp (1) : 0.00938R

Rw23sw (4) vs Rjmp (1) : 0.01954R

Rw23sw (4) vs Rw20sw (3) : 0.01016R

Next, how do I calculate the error when I don't know the resistance of the source that the jumpers are selecting? One source is the input pot (stereo input level knob on the front panel of the unit) but its resistance is variable depending on what position the input level knob is at. Another source is the trim pot (internal 10-turn trim pot) but its resistance also varies depending on what position it is set. However, I know (from what the tech support told me) that the signal selected by the jumper goes to a 100R resistor before going into the input pin of the PCM4202 ADC chip. Can I use 100R as the basis for error calculation? Assuming that I can, then the error for the largest difference above would be 0.01954R / 100R = 0.0001954. I guess that's not too much of an error to worry about, is it? Sorry to ask again but I'm not sure. Newbie that I am, I don't know what the acceptable limits are -- what's negligible and what's not -- for this error calculation.

Thanks!

 

[SY]

Do an order-of-magnitude calculation. Assume the load to be 10k (which probably is close to the truth), remember, you're feeding the top of the pot, not the wiper, and the current draw through it is pretty damn small. What's the error? Less than a part per million?

 

[mtl777]

Do an order-of-magnitude calculation. Assume the load to be 10k (which probably is close to the truth), remember, you're feeding the top of the pot, not the wiper, and the current draw through it is pretty damn small. What's the error? Less than a part per million?

So, assuming a 10K load, the largest error would be 0.01954R / 10000R = 0.000001954 = 1.954 ppm. Very small indeed, assuming that load is correct! BTW, up to how much error is considered negligible? Is an error of 0.0001 or even 0.001 still negligible? I'd like to know the error acceptability limit that you guys use. Thanks!

 

[cliffforrest]

So, assuming a 10K load, the largest error would be 0.01954R / 10000R = 0.000001954 = 1.954 ppm. Very small indeed, assuming that load is correct! BTW, up to how much error is considered negligible? Is an error of 0.0001 or even 0.001 still negligible? I'd like to know the error acceptability limit that you guys use. Thanks!

A big cigar for doing the calculations

Now the more difficult bit of value judgements of what is significant. (Many try this without the calculations so they go round in circles for ever!)

First question is WHAT error you are trying to measure / calculate?

In your original post it was the gain error (AFAIR) due to the added resistance.

1 or 2 ppm of anything is very very difficult to MEASURE reliably, even in a good standards lab.

It is generally accepted that a 1dB difference is the smallest change in level that can be routinely detected. Ballpark.

a 1dB (decrease in gain) is roughly a 1:1.12 or 12% decrease in voltage ratio. There are lots of internet based calculators to do this if you have mislaid your slide-rule!

So a roughly 12% error in gain is audible, a 1.2% error is certainly not. Your 2ppm or .0002% error is absolutely not.

Which is what we were saying in the initial responses, without properly explaining to you why.

Next?

 

[SY]

What Cliff said. If I'm designing a single-photon detector, a ppm is a big deal. If I'm designing an audio gadget, it's not even discernible over the noise, much less the temperature change in the room. And if it's a bit level, it's hilariously unimportant.

 

[mtl777]

A big cigar for doing the calculations

Now the more difficult bit of value judgements of what is significant. (Many try this without the calculations so they go round in circles for ever!)

First question is WHAT error you are trying to measure / calculate?

In your original post it was the gain error (AFAIR) due to the added resistance.

1 or 2 ppm of anything is very very difficult to MEASURE reliably, even in a good standards lab.

It is generally accepted that a 1dB difference is the smallest change in level that can be routinely detected. Ballpark.

a 1dB (decrease in gain) is roughly a 1:1.12 or 12% decrease in voltage ratio. There are lots of internet based calculators to do this if you have mislaid your slide-rule!

So a roughly 12% error in gain is audible, a 1.2% error is certainly not. Your 2ppm or .0002% error is absolutely not.

Which is what we were saying in the initial responses, without properly explaining to you why.

Next?

Good to know, thank you so much! So I'll remember that for audio signals a gain error of maybe 7% or less is nothing to worry about.

 

[mtl777]

Next?

What about distortion? Are there any distortion issues with higher resistance? If so, how do you calculate it?

Thanks!

 

[SY]

What about distortion? Are there any distortion issues with higher resistance?

At that scale? No.

 

[ashok]

I guess this may not be a problem in some parts of the world , but I'm VERY picky about wire insulation as I find that our local market has/had wires that oxidises the surface of the conductor after a short time (black). After that you cannot solder it unless you scrape the surface clean. This is bad and depends on the poor plastic that they use for insulation. So I prefer teflon coated wires which in any case generally comes with silver plated copper wires.

However the market IS getting better . We have some who specifically talk about the superior plastics they use but I need to check that out. I've just bought some of those. The plastic insulation surely looks different ( by look and feel !).

 

[mtl777]

At that scale? No.

Thanks, that's good to know.

I went to the nearby Electronics City store looking for CAT6 cable for this project. Instead, much to my delight, I chanced upon a nice Belden 8773 cable that is just perfect! It has 27 twisted pairs of 22AWG wire (not that I need 27 pairs - I need only 6 pairs - but the wires are different colors, which is good for distinguishing them when you have many jumper connections). And the best thing is that each pair is individually shielded with 100% foil shield and includes a drain wire. This will help to protect against stereo cross talk (the issue I was having with ordinary LAN cable) and against EMI/RFI as well. I bought one foot of the cable for $3.28. If you do the math, that's 27 pairs x 2 = 54 feet of wire. More wire and much cheaper than a $5 25-ft spool of hookup wire, plus it has the shield and drain wire and different colors, too.

My next question is, where's the best ground point to connect the drain wires (since there are six jumpers, there will be six drain wires which will be tied together to a common ground point)? I'm seeing a screw in the center of the board which has continuity with the ground pin of the voltage regulators, the ground pin of the IEC power connector, the chassis, etc. This screw is 5 inches away from the jumpers. I'm not sure if the board is using star grounding and if this screw is the star ground point. However, there is another nearer screw in the corner of the board that's only 1.5 inches away which also has continuity with the grounds mentioned. Which screw should I connect the drain wires to: the center screw or the nearer corner screw?

Thanks again!

 

[cliffforrest]

No one will know.

From a verbal description, it is impossible to tell.

This is where YOU have to experiment and try different things.

Report back on what you find.

HINT: for 12" of wire inside a box it almost certainly makes no difference at all.

 

[SY]

Not to sound like a broken record but... what Cliff said. I'd recommend a close reading of Dave Davenport's article on grounding. It's a subject that is seemingly simple, has lots of subtleties and gotchas, and no universal answers.