Hello all
Is there a general concensus on wire thicknesses for a 3-way speaker.
I will make a few assumptions.
The tweeters will benefit from a smaller guage, i.e. 26 guage?
The mids a slightly higher guage, i.e. 20-22 guage?
The woofers the biggest guage, 16?.
Is there a general concensus on wire thicknesses for a 3-way speaker.
I will make a few assumptions.
The tweeters will benefit from a smaller guage, i.e. 26 guage?
The mids a slightly higher guage, i.e. 20-22 guage?
The woofers the biggest guage, 16?.
I use 2.5mm>2 almost exclusively because I get it for the right price, that will handle 15 amps at 240Volt and very few speaker drivers will handle more than that.
If for any reason I want to lower DCR I just double up the wires, although running thinner wire to tweeters would be OK I think it is simply easier to use all the same wire.
I get mine from industrial power leads that don't pass the annual tagging test and get chucked in the bin, hundreds of metres of cable discarded every year and usually it is only for small nicks in the outer sheath, I use the same discarded cables for speaker leads for parties too.
You can find lots of useful stuff in dumpsters.
regards
Ted
If for any reason I want to lower DCR I just double up the wires, although running thinner wire to tweeters would be OK I think it is simply easier to use all the same wire.
I get mine from industrial power leads that don't pass the annual tagging test and get chucked in the bin, hundreds of metres of cable discarded every year and usually it is only for small nicks in the outer sheath, I use the same discarded cables for speaker leads for parties too.
You can find lots of useful stuff in dumpsters.
regards
Ted
3-4 years ago I viewed the Magico website. At that time, there was a picture of the Magico Mini's crossover. In that picture it was obvious Mr. Wolf used much thinner transformer wire leading to the tweeters then he used for the woofers. He also used multiple runs of varying guages to both these drivers
At AudioGon I have read 'smearing ' occurs with thicker wire, especially going to the tweeter's. Thus the thinner the better.
So my next question is; for a 140 watt amplifier, going into a 3-way, with a x-over at 2800hZ, 'What is the amount of power going to the tweeters? (is this a volt or amp question?).
I have read at the Goertz website that they recommend different guages of speaker wire depending on the power of the amps. The lower the power, the smaller the cable.
Second question; what is the power going to the mids, x-over at 250hZ?
JadeM at Audiogon did an exhaustive study on cable thicknesses, composition (copper, silver, gold), and seperation (this included braided, lightly twisted, and seperated with 3/4" air gap). He found the thicker the wire, the less detail, too thin, loss of body).
I have read hear, and at AudioGon, about copper, silver plated copper, silver, and gold wire. Some people use one kind for their tweeters, and yet another for the lower drivers.
Considering I am using Accuton drivers, attention to this detail should be worth the effort.
At AudioGon I have read 'smearing ' occurs with thicker wire, especially going to the tweeter's. Thus the thinner the better.
So my next question is; for a 140 watt amplifier, going into a 3-way, with a x-over at 2800hZ, 'What is the amount of power going to the tweeters? (is this a volt or amp question?).
I have read at the Goertz website that they recommend different guages of speaker wire depending on the power of the amps. The lower the power, the smaller the cable.
Second question; what is the power going to the mids, x-over at 250hZ?
JadeM at Audiogon did an exhaustive study on cable thicknesses, composition (copper, silver, gold), and seperation (this included braided, lightly twisted, and seperated with 3/4" air gap). He found the thicker the wire, the less detail, too thin, loss of body).
I have read hear, and at AudioGon, about copper, silver plated copper, silver, and gold wire. Some people use one kind for their tweeters, and yet another for the lower drivers.
Considering I am using Accuton drivers, attention to this detail should be worth the effort.
I'll take the liberty to re-phrase your questions.Cousin Billy said:
If all three drivers in the 3way speaker are of the same sensitivity and similar impedance then the questions become:-
What is the peak transient voltage fed to each driver?
I'll contend that the peak voltage to each driver is the same and further since the drivers are of the same impedance then the peak currents must also be the same.
Finally, since the sensitivity of all the drivers is the same and each can be fed by the same voltage and the same current then the peak SPL of all the drivers should be the same.
But, we rarely build speakers with drivers of equal sensitivity and equal impedance.
So let's start from the same peak SPL from each driver and then work back to the peak transient voltage and peak transient current required by each driver. Then you can estimate the cable requirements for each driver.
Having done this, I suspect the over-riding requirement becomes the source impedance seen by the bass driver to maintain Q of the box within the range assumed in the initial design and that all other drivers can be fed with relatively small gauge wiring.
NB.
this argument revolves around my contention that all the drivers should be capable of the design maximum SPL of the speaker. One or two drivers can exceed the design target but that will not allow the speaker to play louder on general wideband and very varied audio signals.
Exactly the type of answer I was looking for. Clear, concise and to the point. No guessing.
I will experiment. I will start with larger thicknesses for the mids and tweeters and work my way towards thinner wires.
Any input as to whether multiple strands are better then one strand?.
The output impedance of the amps is 2.3ohms. The crossover and box volumes are being calculated with this impedance in mind.
More input would be appreciated.
Thank you AndrewT
I will experiment. I will start with larger thicknesses for the mids and tweeters and work my way towards thinner wires.
Any input as to whether multiple strands are better then one strand?.
The output impedance of the amps is 2.3ohms. The crossover and box volumes are being calculated with this impedance in mind.
More input would be appreciated.
Thank you AndrewT
I used to use # 14 wire But due to the thickness come the difficulty in wiring them and they also seem to stress the terminals more compared to thinner wires, So from now I decided I'd just use #18 wire to wire the inside of the speakers.
Yes AndrewT, they are Atma-Sphere MA1, 140 watt OTL (Output transformerless) Tubed mono block amps.
They deliver lots of current, and wide, wide, wide bandwidth.
They are brutally revealing of any flaw in the signal path. This is why I am going to extremes in my speaker construction.
I am building a similar speaker to the Humble HomeMade HiFi 'Humble Statement' speaker. With this type of time and money investment, I will leave no stone unturned.
They deliver lots of current, and wide, wide, wide bandwidth.
They are brutally revealing of any flaw in the signal path. This is why I am going to extremes in my speaker construction.
I am building a similar speaker to the Humble HomeMade HiFi 'Humble Statement' speaker. With this type of time and money investment, I will leave no stone unturned.
Depends on who you talk to.
Stranded is more flexible and easier to work with.
Some contend that "skin effect" should be taken into consideration, and that stranded is what should be used, especially HF is involved.
Syd
you may find that silver plated copper wire of smallish gauge works well for the treble driver. Not too much more expensive than copper wire.
I would never go to the expense of silver wire, not even as an experiment.
Tinned (Pb/Sn) or tin plated (Sn) copper wire is an alternative that prevents the copper from corroding.
Most drivers with a very low Q will not work as expected with a 2r3 source impedance. High Q drivers will be less affected but must still account for Rs as you seem to be aware.
You can convert a low Q driver to medium Q and a medium Q driver to high Q using the Rs technique.
The tube amp and the crossover will both benefit greatly from near constant impedance across each individual driver. Spend some time trying to equalise out the impedance variations using Zobel etc.
This will waste output power but that is a consequence of using high Rs.
I don't know any way to avoid it with a single amplifier and passive crossover. Comments please?
I would never go to the expense of silver wire, not even as an experiment.
Tinned (Pb/Sn) or tin plated (Sn) copper wire is an alternative that prevents the copper from corroding.
Most drivers with a very low Q will not work as expected with a 2r3 source impedance. High Q drivers will be less affected but must still account for Rs as you seem to be aware.
You can convert a low Q driver to medium Q and a medium Q driver to high Q using the Rs technique.
The tube amp and the crossover will both benefit greatly from near constant impedance across each individual driver. Spend some time trying to equalise out the impedance variations using Zobel etc.
This will waste output power but that is a consequence of using high Rs.
I don't know any way to avoid it with a single amplifier and passive crossover. Comments please?
I would suggest something like 14awg for bass, 16awg for mid and 18awg for the tweeter.
I've heard some plated wire that sounded harsch to me, so I would suggest that you test them first if you want to go that route.
I've heard some plated wire that sounded harsch to me, so I would suggest that you test them first if you want to go that route.
AndrewT
The drivers have a Qms of 4.49, Qes of 0.29, and Qts of 0.27.
They will be driven in series, not parallel.
How would you classify this Q?.
The drivers have a Qms of 4.49, Qes of 0.29, and Qts of 0.27.
They will be driven in series, not parallel.
How would you classify this Q?.
Lemme guess:
http://www.madisound.com/catalog/product_info.php?products_id=1032
Driven by a 2.3 ohm source impedance you are looking at a Qes of
2*pi*20.4*.032*(5.9+2.3)/9.1^2 = Qes = 0.41, So then Qts= 0.38
and your 1W/1m sensitivity is about 86.5, and that doesn't even include inductor DCR if you are using a passive xo. So your driver goes from lowish to medium Q, just because of your amp.
Wire Gauge Resistance per foot from
http://www.radiolocman.com/shem/shem-cache.html?di=18899
AWG Ohm/ft
4 .000292
6 .000465
8 .000739
10 .00118
12 .00187
14 .00297
16 .00473
18 .00751
20 .0119
22 .0190
24 .0302
26 .0480
28 .0764
seems like wire resistance in the 12-18gauge range doesn't amount to much unless you are wrapping the inside of your speaker when building a crossover.
http://www.madisound.com/catalog/product_info.php?products_id=1032
Driven by a 2.3 ohm source impedance you are looking at a Qes of
2*pi*20.4*.032*(5.9+2.3)/9.1^2 = Qes = 0.41, So then Qts= 0.38
and your 1W/1m sensitivity is about 86.5, and that doesn't even include inductor DCR if you are using a passive xo. So your driver goes from lowish to medium Q, just because of your amp.
Wire Gauge Resistance per foot from
http://www.radiolocman.com/shem/shem-cache.html?di=18899
AWG Ohm/ft
4 .000292
6 .000465
8 .000739
10 .00118
12 .00187
14 .00297
16 .00473
18 .00751
20 .0119
22 .0190
24 .0302
26 .0480
28 .0764
seems like wire resistance in the 12-18gauge range doesn't amount to much unless you are wrapping the inside of your speaker when building a crossover.
Thank you Ron E
That sounds like good news.
Any more information would help.
What is your views on running the drivers is series?.
Bill
That sounds like good news.
Any more information would help.
What is your views on running the drivers is series?.
Bill
Cousin Billy said:
That's your decision. The OTL amps would probably prefer the 16 ohm load to a 4 ohm load. You could always put them in parallel and use autoformers.
To me it is a contradiction, when constructing a crossover, to use heavy guage wire when the leads of the crossover components (caps and resistors in particular) are so thin...
Heavy guage wire for long runs between the amp and speakers seems to make sense.
Ray
Heavy guage wire for long runs between the amp and speakers seems to make sense.
Ray
Ray
I will take your comment one step further.
The leads of the driver itself are very small. It looks to me like 26 or even 28 guage leading from the driver terminal into the voice coil.
Like you said, the leads of the Duelund Graphite resistors look to be 26 - 28 guage. The leads of the Duelund VSF caps 24 - 26 guage, etc.
The inductors obviously are much bigger.
I will take your comment one step further.
The leads of the driver itself are very small. It looks to me like 26 or even 28 guage leading from the driver terminal into the voice coil.
Like you said, the leads of the Duelund Graphite resistors look to be 26 - 28 guage. The leads of the Duelund VSF caps 24 - 26 guage, etc.
The inductors obviously are much bigger.
- Status
- Not open for further replies.