Crossover resistors - does watt rating matter?

[wank]

How important is the wattage rating in resistors for crossovers?
I'm fixing the crossovers on my AR-9LS - upgrading to the LSi(improved) version. This entails adding two resistors, one on the lower midrange woofer, and one on the upper midrange.
The schematic calls for 1.5 ohm / 22 watt on the lower, and a 2.5 ohm / 22 watt on the upper. I am unable to find better than a 12 watt in the 2.5 ohm. Does this make a difference?
Since I have to order the 2.5 (I can only source a 2.7 locally, and that attenuates the upper mids too much), I thought I'd go with Mills resistors from PartsExpress.
Advice?

[Paul C]

Yes, it does make a difference. Most likely, the 12 w resistors would do fine for most casual listening. And for midranges, where there is low power going to them anyway, 10-12 w is really plenty.

But if you are worried about the mid resistors, order 2 resistors at double the value, and connect them in parallel.

For example, if 2.5 ohms is required, get two 5 ohm 12 w, and connect parallel. Mills (from Parts Express) has 5 ohm 12 w. Two of those would give you 2.5 ohms 24 w.

Or 4.7 ohm 10 w Dayton + 5.1 ohm in parallel = 2.44 ohms.

Two 3 ohm 10 w Daytons (004-3) or Mills (005-3) in parallel would give you 1.5 ohms 20 w, or 24 w.

[fede_vr]

Edited.. I answered to the wrong thread... sorry ,too many windows..
:sigh: :sigh:

[dgwojo]

Wank, I use the ones listed on the bottom right of this page, they're dirt cheap and avaialable in 25 watt ratings, as mentioned before, you can pair up 3ohms to get 1.5ohms and pair up 5ohms to get 2.5ohms, Dave. Xicon @ Mouser (http://www.mouser.com/catalog/629/511.pdf)

[wank]

Excellent. Thanks for the replies.
What's the danger of having a too low rated resistor? Failure at higher power input? Does that "blow" or open the resistor?
Also, I assume whether you can hear the difference between Mills wire-wound vs. Xicom or generic cement resistors is similar to cable debates?

[jaymanaa]

Excellent. Thanks for the replies.
What's the danger of having a too low rated resistor? Failure at higher power input? Does that "blow" or open the resistor?
Also, I assume whether you can hear the difference between Mills wire-wound vs. Xicom or generic cement resistors is similar to cable debates?


I use Mills 12w quite a bit, and they can get pretty hot when you crank up the volume. Never burned one out, but all my stuff is 70w or less.

[tom1356]

The 12 w Mills will be fine. If you are putting more than 12 watts into a midrange you have other problems to worry about.

The mills sound much better than the sand cast resistors. Better still are Graphite based resistors like the ones available from Duelund Coherent Audio.
Or you can make them yourself.

[jaymanaa]

The 12 w Mills will be fine. If you are putting more than 12 watts into a midrange you have other problems to worry about.

The mills sound much better than the sand cast resistors. Better still are Graphite based resistors like the ones available from Duelund Coherent Audio.
Or you can make them yourself.


Well, now you have to tell us how.:yes:

[tom1356]

Well, now you have to tell us how.:yes:
You can buy graphite sticks at Art/craft supply stores. They are usually ~4" long 1/4" sq sticks.
You need to anneal them. I use a propane torch and a vise. Be prepared, they drip a small amount of liquid during annealing and may catch fire for a short time. After cooling I wrap a fine silver wire a few times around one end. Twist the wire to secure it to the graphite and solder it all the way around the resistor so the wire will hold fast.
Use a DVM meter to determine where to attach the second wire in the same manner as the first. Make the fine adjustments before soldering the second wire.
Coat them with your favorite nail polish or enamel paint. Or better still, in a low power tube set up, dip them in bees wax.
This method works perfect for resistors up to ~4 ohms. Tie two together for up to 8 ohms. I'm still looking for a reasonable way to make 30-180 ohm ones.
Good Luck!

[Paul C]

Also, you have to understand the distribution of energy in most music.

The dividing point is about 350 hz. About half of the energy is below 350 hz, and half above. If you have a crossover of, say, 800 hz, about 80% is below 800 hz in the woofer, and 20% above.

And the higher you go, the less energy in the upper driver.

So, say you have a 3-way with crossovers at 800 hz and 5000 hz. 80%-90% of the energy will go to the woofer. 10%-15% will go to the midrange. And only about 5% or less will go to the tweeter... more like 2-3%.

Not only that, but music is not a continuous flow of energy, it is peaks and valleys. A bass note will draw a lot of energy, but will taper off. Even if you were running your 100 w/ch amp very near clipping, it would only produce about 80 w on the peaks, and in between it would only be producing 10 w - 20 w.

The energy going to the tweeter is very intermittent. These sounds are the shimmery highs in cymbals, the little tinkly sounds of percussion, the high overtones of strings and woodwinds.

You just aren't going to strain 10 w resistors in your living room system. They may get warm, but they aren't going to burn out.

Even for very high power PA gear the place where you really need higher power resistors is in the zobel of the woofer, or in a low midrange. I use 20 w - 40 w (ganged 10 w resistors) to make those. 20 w resistors do fine in the padding for high power horns.

If you use 20 w (or combination adding up to 20 w) for your home stereo speakers you will have no problems.

[wank]

Awesome Paul. Thanks for the background on this stuff. Makes future problem solving easier in the future (you know, give a fish, teach to fish...). Since this is a 4-way crossed over at ~1200, I can work from there.

I'll probably order some Mills next week (12w for the upper mids, but ganged lower mids), if I get a problem sorted out.
I might wait til I get the lower mids looked at - one of them has a serious dropoff at about 260 Hz, or the other has a real hot spot there. I haven't measured SPL yet, and I've only gotten up to 300Hz, but I'm guessing 3dB difference between the two. They're 8" ARs, but were reconed a few years back.
Is this likely a difference in re-coning (both reconed by the same guy at the same time), or could a bad XO cap cause this? I'll have to see if their specs are very close to other AR 8" drivers, since I have some AR-17 drivers in the garage.

[jaymanaa]

You can buy graphite sticks at Art/craft supply stores. They are usually ~4" long 1/4" sq sticks.
You need to anneal them. I use a propane torch and a vise. Be prepared, they drip a small amount of liquid during annealing and may catch fire for a short time. After cooling I wrap a fine silver wire a few times around one end. Twist the wire to secure it to the graphite and solder it all the way around the resistor so the wire will hold fast.
Use a DVM meter to determine where to attach the second wire in the same manner as the first. Make the fine adjustments before soldering the second wire.
Coat them with your favorite nail polish or enamel paint. Or better still, in a low power tube set up, dip them in bees wax.
This method works perfect for resistors up to ~4 ohms. Tie two together for up to 8 ohms. I'm still looking for a reasonable way to make 30-180 ohm ones.
Good Luck!

Thanks Tom, I'm going to give that a try. Jay