How hard to build a 4-ohm 3-way with ALL 4 ohm drivers

[a4eaudio]

I know that different impedances can work (4 ohm tweeter with 8 ohm woofer, 8 ohm tweeter with 4 ohm woofer, etc.) but I also have seen many posts when people run into trouble with impedance dropping to 2 or 3 ohms.

I am okay with a nominal 4 ohm speaker. While I don't know the "tricks" that might be necessary, if I am willing to post a build on the forum and get crossover advice as I go, should I be able to build a 3-way speaker with a 4 ohm tweeter, 4 ohm midrange and two 8-ohm woofers wired in parallel that will be okay for an amp that can handle a 4 ohm speaker?

If it helps, my initial thoughts are: (feel free to comment/critique)

• GRS PT2522-4 3-1/2" Planar Tweeter 4 Ohm (272-122) [I will use this tweeter for sure];
• Dayton Audio PC105-4 4" Full-Range Poly Cone Driver (295-158) [I can go with 8-ohm version or a different driver altogether if necessary, but I think I need the high sensitivity to go with the 2 woofers below];
• 2 x Dayton Audio DC200-8 8" Classic Woofer (295-310) [I'm willing to go with something different, but total drivers must stay below $200, so possibilities are somewhat limited.]

[And yes, I realize that "all" the drivers aren't technically 4 ohms, but I'm considering the two 8-ohm woofers in parallel as a single driver for simplicity.]

[Paul O]

Nothing wrong with that design idea and it's no harder to build than any other driver combo.

[tvrgeek]

Can your amp handle a 4 Ohm load? Many can't.
Why do you want to do this? Why not an easier to handle load?

Tweeters are often 4 Ohms, but more efficient so they gat padded to a higher impedance.
I would use 2 4 Ohm woofers in series. Probably need to pad the mid as well. For the typical AVR or receiver, you want to stay above 6 Ohms when done. There are lots of reasons from stability to distortion to do this.

Before you get too carried away, remember that efficiency as advertised is often a bit optimistic. Best to get one of each and test to verify. 2-8, 4 and tweet, nice combination. Crossovers like 300 and 3K.

[a4eaudio]

Can your amp handle a 4 Ohm load? Many can't.

Yes, I will use a class D amp that can handle 4 ohms.

Why do you want to do this? Why not an easier to handle load?
I would use 2 4 Ohm woofers in series.

I don't really "want" to do this, but if it is possible it seems to work with my design constraint. The main constraints are that I will use the GRS tweeter and I must stay below $200. To get good bass I would like to use two of the Dayton Classic 8" woofers and they only come in the 8-ohm variety.

IF this turns out to be a bad idea, then I can consider alternatives. I don't think I can get two 4-ohm woofers for $50 that I like, but I can look at single 8-ohm woofers up to $50 which offers some possibilities but I have to give up the slender TMWW.

Probably need to pad the mid as well. Before you get too carried away, remember that efficiency as advertised is often a bit optimistic. Best to get one of each and test to verify.

Yes, I may do that unless I use drivers that other's have already used and can rely on their measured efficiency. I know Wolf has used the PC105-4.

[isaeagle4031]

A true 4 ohm 3way can be a difficult task. As drivers overlap, so do their impedances. With a pr of 8" woofers, the mid and tweet won't require much padding. As such, impedance can drop to pretty low levels.

[billfitzmaurice]

Nothing wrong with that design idea and it's no harder to build than any other driver combo.

+1. A 4 ohm three way is no more difficult to realize than an 8 ohm three way. Frankly I'm surprised to see the question raised.

[a4eaudio]

I know it is "possible", but didn't know if it was particularly "difficult". I have seen various threads where someone ran into problems similar to what isaeagle4031 brings up, where they are below 3 ohms (and maybe closer to 2 ohms) but usually people were able to offer xo suggestions to lead to an acceptable solution. Keep in mind many of us are not very experienced; a common question I found was can a 4 ohm tweeter be used with an 8 ohm woofer...that one might have sent you into shock!

[devnull]

As long as the DCr is above 3 ohms you shouldn't have a problem. Any amp that says it can drive a 4 ohm speaker can drive down to at least a 3.2 ohm load. Old tube amps had issues driving low impedance speakers which is why you used to see a lot more 16 ohm speakers and compression drivers

[billfitzmaurice]

With respect to tube amps they're the opposite of SS. They have no problems with low impedance loads, even dead short circuits. They don't work well with impedance loads higher than the tap rating. The reason for using 16 ohm compression drivers with 8 ohm woofers, the usual configuration, was to lower the compression driver sensitivity to better match up with the woofer. I still employ that scheme.

[Chris Roemer]

There are a few "tricks" you can do to help hold up the Zmin.
One I usually use is (on the mid's bandpass) to place the series components ahead of the shunts. Not sure WHY this helps hold the impedance up, but I've observed (many times) that it does. Apparently SOME XO sims aren't flexible enough to allow this topology. (You'd have to run the HP portion - cap/shunt coil - ahead of the LP section - coil/shunt cap, or the other way around.)
The narrower the bandpass, the lower the load drops (due to "greater paralleling" of the 3 sections). Also, the lower the filter "orders", the worse the situation. You should aim for at least a 3-octave wide midsection, or target the mid/tweet's Fc to be 8x to 10x the mid/woofer's Fc.
Having the woofer(s) (on a 3-way) down by the floor will minimize the padding of the mid(s) and tweeter, due to less of a BSC requirement (so it's better not to do that).

[tvrgeek]

Depending on if you are crossing to a sub may suggest if you want a single woofer or a pair. Do your simulations. Watch your excursion. Watch your impedance.

Speaker building in not a cheap hobby. You may be causing yourself more problems by limiting a budget build too much. Do your models and price everything. Sometimes a better driver means a cheaper crossover. You can go old school and use a couple bolts and nuts for the terminals instead of fancy cups and banana jacks. One size thicker particle board may be cheaper than MDF. A slot vent can be cheaper than buying a fancy tube and so on. ( or if crossing to as sub, sealed easier and cheaper. ) A budget built can be quite hard to do well.

Also look at the Silver Flute woofers. They seem to do quite well for their price.

[LOUT]

Just a quick visualization of what ChrisRoemer is talking about:

Not that there's anything inherently wrong with the lower version....But it tends to create a slightly lower impedance curve while the upper example often helps to keep impedance in check.

[billfitzmaurice]

What affects the impedance in the region of the crossover is the filter Q. The higher the Q the lower the minimum impedance. When you model the filters if the impedance goes too low make the Q lower. Q of 2nd order Linkwitz-Reilly is .49, Bessel .58, Butterworth .707, Chebyshev 1. You don't have to use any of these specific alignments, just like box alignments they're just reference starting points. If you wonder why we don't just use low Q filters they give reduced sensitivity in the crossover region. As with all aspects of speaker design filter Q is a compromise, in this case between response and impedance.

[Chris Roemer]

Your XO is ready.
Fcs near 400 and 3000 Hz.
Only 2-3dBSC, so you can keep the woofers nearer to the floor.
Runs approx. 89dB/2.83v
4 element HP, 6-el BP, 2-el LP.

Look to Dave Tenney's "Dayton 8" for box inspiration.

[a4eaudio]

Your XO is ready.
....
4 element HP, 6-el BP, 2-el LP.

Hi Chris, did you simulate a full cross-over? I'm not implying that you "should" do that for me, but your post makes me think that you may have done that but left out the xo details from the post.

[Chris Roemer]

2nd order HP w/L-pad: 15uF series cap, 0.30mH shunt coil (to gnd), "L": SR=2n(ohms) / PR=6n
2nd order (notched) Band-Pass (in about 0.10cf - sealed & stuffed): 0.50mH series coil and 47uF series cap,
then 2 shunt elements - a 2.0mH coil and a 7uF cap.
For the "notch" filter (near 8kHz), use a 4n resistor in series w/a tiny 0.68uF cap, BOTH across (in parallel with) the initial 0.50mH coil
2nd order LP using a low DCR (iron core) 4.0mH series coil, then a 120uF shunt cap feeding a paralleled pair of Classic 8"ers.
(reverse polarity is "indicated" on the mid - but I'd listen to just the woofers and mid and determine which way sounds best, THEN do the same w/just the mid and planar)