Thread for the passive XO gurus (RTi A7)

[gp4jesus]

This goes out to those who know more than I do. Another thread for passive XO gurus and RTi A7 owner/modifiers.

Bottom line: I don't see a series cap much less a shunt inductor on the Mid to provide the HP filtering suggested on Polk's site. 12dB/octave HP&LP for all drivers.

Read what I saw upon inspection of XO; I hope it makes sense:

Tweeter: 2 caps* paralleled**, shunt inductor, 2 caps* paralleled** - 18 dB/oct, right?!!
* in series w/tweeter
** w/each other; under 20 micro farads total

Mid: series inductor, shunt cap & resistor - 12 dB/octave LP, right?
NO SERIES CAP or SHUNT INDUCTOR on the mid

Sub: BIG series inductor, shunt cap 12 dB/octave LP, right?

This also suggests 4 ohms below 125hz. If I did the math right, 125hz HP series cap should be about 2500 microfarads.

Polk Tech spt stands behind the web site specs. How can Polk claim any Mid HP filtering w/out a series cap?
Unless they know something I don't, the site's not accurate.

anyone's comment please?!!

thnx for bearing w/me; tony

[ShinAce]

I'm gonna copout and say 'not enough info'.

I could read through your description and layout the crossover, but everything you got there adds up. Tweeter, as described, is 18dB HP, mid is 12dB LP and woofer is 12dB LP.

There are other factors to consider. If the inductance of the woofer is enough, we can use its natural high eng roll to crossover to the mid. So it may not need a HP(or in this case, bandpass). Other factors may apply to the mids. If they lose sensitivity in the very low end, playing them may not affect the oversall response of the speaker by more than, say, 0.5dB.

Keep up the good fight!

[FTGV]

OK gp I'm not a guru but I'll through my $.02 in.

.
Tweeter: 2 caps* paralleled**, shunt inductor, 2 caps* paralleled** - 18 dB/oct, right?!!

Correct it is 18 db electrically,but the actual acoustic slope could be closer to 24 db(or greater) when the natural low frequency roll off of the tweeter is factored in.

Mid: series inductor, shunt cap & resistor - 12 dB/octave LP, right?

Correct ,but again the acoustical slope maybe be steeper.

Sub: BIG series inductor, shunt cap 12 dB/octave LP, right?

Yes.

This also suggests 4 ohms below 125hz. If I did the math right, 125hz HP series cap should be about 2500 microfarads.

You have one too many zeros.For your 4 ohm assumption text book value the shunt cap would be closer to 250uf not 2500uf.

How can Polk claim any Mid HP filtering w/out a series cap?

If? there is no HP filter present on the mid then its low frequency roll off will be determined by the woofer/box tuning just like a normal woofer and will operate in parallel with the bass drivers.Does the mid have it's own separate enclosure?I,m not familiar with this speaker but from the pic I seen online it appears the mid may have it's own port thus it's own enclosure?

[Face]

ShinAce and Fred already did an excellent job answering your questions. Let me just add a little more.

Acoustic and electrical filters aren't the same thing. You can have a 2nd order electrical filter(1 cap, 1 coil), with 6 order acoustic rolloff. Not as common, but it can also go the other way, 2nd order acoustic roll off with a third order electrical filter, etc...

As for Polk's 2nd order HP on their midrange with no parts in series with the driver, here's an example of another midrange that has a similar response: http://www.madisound.com/catalog/pro...oducts_id=1315

And as Fred mentioned, the cabinet tuning can also have a big influence on a mid's high pass.

FYI, textbook/website crossover calculators are wrong 90% of the time.

Polk Tech spt stands behind the web site specs. How can Polk claim any Mid HP filtering w/out a series cap?
Unless they know something I don't, the site's not accurate.

:D

[FTGV]

FYI, textbook/website crossover calculators are wrong 90% of the time.

:D

Yeah its a rare instance when a "text book" design gives the desired response.

[gp4jesus]

thanks everyone. I neglected to mention I'm looking only @ the electrical (filters). Your responses became reminders: the mid's inductance, cabinet tuning specifics, the different driver's natural rolloffs.

I ran some tests Wed pm. Heard some things things that told me the subs & (6.5") mid may share the same enclosure space. Let me make one point clear: I do not see any HP filtering on the mid*. It wouldn't suprize me if Polk took the RTi A3 drivers & XO, put it in a bigger cabinet, added a 2nd order LP & 2 musical sounding (4 ohm in series) 7" drivers, and called it the A7. I don't consider that a bad thing.
* I don't like that

to FTGV & Shinace: Thanks professors! Looks like I aced the exam! I needed to be sure I recognized the different filter designs, interaction w/the driver aside. Also I remembered incorrectly the formula to calculate the value for a HP filter (@ 125hz)

Early on I planned to use an EXO to bi-amp these even b4 I started this thread. In a msg I sent to Face I mentioned using tri-amping to determine the source of a brightness/edginess issue. Since then I realized I may run into a phase issue with the mid/tweeter filtering (12dB LP/18dB HP passive) and 24dB HP/LP active @ 125hz. I plan to continue the research to include cap upgrades* following the "cap shootout" results.
* if I can get them into the cabinet!

When I'm fininshed messing around I will have tri-amped rewired* A7s w/out passive filtering. Check out my system. I didn't list everything...
* the stock wire is lame, particularly the mid, & especially the subs

thnx again, stay tuned

[Face]

What's wrong with no high pass filtering on the mid? The less components in the chain, the better.

As for fitting upgraded components in the cabinet, it can be a chore. Make your own board to mount to the bottom of the cabinet, and run wires to it directly from the binding post cup, bypassing the OEM crossover board.

[FTGV]

...the subs & (6.5") mid may share the same enclosure space.

That would be an unusual approach as most designers of 3 ways prefer the midrange be completely isolated from the high pressure backwave of the woofers.

It wouldn't surprise me if Polk took the RTi A3 drivers & XO, put it in a bigger cabinet, added a 2nd order LP & 2 musical sounding (4 ohm in series) 7" drivers, and called it the A7.

From your description of the 7 that may not be far off the mark.

Early on I planned to use an EXO to bi-amp these even b4 I started this thread. In a msg I sent to Face I mentioned using tri-amping to determine the source of a brightness/edginess issue. Since then I realized I may run into a phase issue with the mid/tweeter filtering (12dB LP/18dB HP passive) and 24dB HP/LP active @ 125hz.

I'm a big advocate of active crossovers and think atleast for experimental purposes you certainly could try an active unit between the mid and woofers.However for optimum results you would need to remove the passive LP section on woofers and since there is no HP section on the mid you won't need to remove it .

A 24 db LR type should work well and I would try to set the xover point in the 2-300hz range.The addition of the steep 4th order roll off at the bottom of the midranges response should reduce distortion in the important midband as the excursion demands on the mid driver will be substantially reduced.
There should be no phase issues if 1)a 24 db active network is used 2)the passive LP section on the woofers is removed and 3)crossing over in the 2-300hz range.

Keep in mind if amplifiers with different gains are used for the woofer and mid/tweeter sections then the systems frequency balance will be altered.Most active crossovers will have a means to adjust the high pass and or low pass sections so you can restore the proper balance.

[gp4jesus]

What's wrong with no high pass filtering on the mid? The less components in the chain, the better.

I agree w/ the "less components" discussion. In this case at very high levels the mid* gets pushed* too hard while the subs loaf. HP @ 100hz+ would help a lot.
* popped a couple of times

As for fitting upgraded components in the cabinet, it can be a chore. Make your own board to mount to the bottom of the cabinet, and run wires to it directly from the binding post cup, bypassing the OEM crossover board.

If I read this right, run from the amp to "new" XO, then to BPs? Great tip! Easy "fix" for sure. That's what makes this forum great. Now I have to make up my mind exactly where to place the speakers & use the spikes for more clearence. Part of my interest in cap upgrade is curiosity for myself and other A7 owners. Read on for the "end product."

It wouldn't surprise me if Polk took the RTi A3 drivers & XO, put it in a bigger cabinet, added a 2nd order LP & 2 musical sounding (4 ohm in series) 7" drivers, and called it the A7.

From your description of the 7 that may not be far off the mark.

Since my last post I pulled the mid & upper sub to learn the above (disappointingly) is true. Otherwise the cabinet seems well braced/built. Now if I could install a divider, creating 2 chambers... It so happens Polk braced the cabinet just below the mid port. Of course the trick is getting in there and attaching it. W/6 kids now, Natale Grace born Apr 22d, MY spare time is a very sparse, precious comodity. I'll resign myself to the easy, quick stuff* early on, like rewiring the cabinet, bypassing the XO, bi/tri-amp experiment, getting to the more complex stuff later.
* have the wire, 10 ga, 12ga OFC, 18 ga OFC etc

I'm a big advocate of active crossovers and think at least for experimental purposes you certainly could try an active unit between the mid and woofers.

Most active crossovers will have a means to adjust the high pass and or low pass sections so you can restore the proper balance.

Been huge on active XOs since the early '80s. The Audiocontrol Richterscale 3 has a control to turn up or down the sub. The other has filter input gain plus separate adjustments for both HP & LP. Got more equipment than listed.

Some went down for a nap; I'll do what I quietly can...

thnx; stay tuned, tony

[FTGV]

I agree w/ the "less components" discussion. In this case at very high levels the mid* gets pushed* too hard while the subs loaf. HP @ 100hz+ would help a lot.

IMO assuming the use of a good active crossover ,the performance gains made by the reduction in midband distortion and superior damping in the bass should/may? overshadow any downside of the added complexity in the signal path.

Now if I could install a divider, creating 2 chambers... It so happens Polk braced the cabinet just below the mid port. Of course the trick is getting in there and attaching it.

Alternatively installing some acoustic foam or fibreglass between the woofers and mid should provide a measure of isolation between them.

W/6 kids now

Keep up the good work and you will soon catch the Duggars:D

, Natale Grace born Apr 22d,

Congrats on the new arrival.

Been huge on active XOs since the early '80s.

Likewise,I have been running a fully active setup for about 4 yrs.

[gp4jesus]

FTGV: spent half the evening last night editing a reply to the previous one.

You're VERY right about active XOs

the A7s need the divider. Polk put a bunch stuffing in the speaker. Unfortunatly it didn't work. When I listen to just the subs, I can hear their bass thru the mid. My curiousity over that prompted me to pull the upper sub & mid.

Wednesday pm I may get to some rewiring, bypassing the XO, and bi-amping the mid/tweet. I, along w/some other A7 owners, want to know if the edgy brightness is the XO or the drivers. My $ is on the XO.

anyway I agree w/your points. would love to see the details of your system.

stay tuned, cheers tony

[gp4jesus]

Upgraded to AR 12ga OFC over the weekend; have 100+ hours on them. Listened to the same Donald Fagen/Steely Dan tracks the first time I cranked them up the day after they arrived about 5 weeks ago. They are noticably softer w/reduced brightness at the same (loud!) volume levels.

Couldn't make time to do full blown mid/tweet bi-amp tonight. Did make time to do the following:

Ran thru a lively Steely Dan track 4 times. (Kid Charlemenge)
1. Right side w/the mid alone, full range; left side as Polk built it with a pillow over the tweeter: little difference; right side fractionally softer

2. right side-4 order active LP @ 2.2K*; left-same. Big difference; way cleaner

3. Right side @ 2.8K*; left: as above. As clean, w/ more top end (of course)

4. Right side @ 3.5K*; left: as above. As clean, still more top (of course)
* Polk's site has them @ 2.7K 12dB, HP & LP...


Will try full blown bi-amp w/both channel to confirm above results.
Otherwise my conclusion: Either upgrade caps or bi-amp the mid/tweeter.

Stay tuned. Cheers tony

[FTGV]

1. Right side w/the mid alone, full range; left side as Polk built it with a pillow over the tweeter: little difference; right side fractionally softer

2. right side-4 order active LP @ 2.2K*; left-same. Big difference; way cleaner

3. Right side @ 2.8K*; left: as above. As clean, w/ more top end (of course)

4. Right side @ 3.5K*; left: as above. As clean, still more top (of course)
* Polk's site has them @ 2.7K 12dB, HP & LP...

By the description above it appears you were cascading the active xover with the passive or did you remove the passive?The results of using them both on top of each other will be highly unpredictable and chances are good that serious response errors will occur.

At the very least you will need the ability to make accurate measurements to see what is really happening. Crossing over in the 2-3k range is a lot more difficult than crossing over in the 2-300hz because of the drivers natural roll off's and break up modes etc are usually within an octave of the intended xover point.These will all have an affect the final response.

However as I mentioned in a previous post I would stick to using your active in the 2-300hz range( along with the removal of the passive crossover from the woofers) as this should work well.In this lower frequency range the unfiltered response of the woofers and mid should have atleast an octave of overlap aswell there shouldn't be any cone breakup modes or large peaks or dips in the drivers response.Thus the response with the active should be close to predidiction.

[gp4jesus]

Thanks for your interest and help! Sorry for coming across vague

To clarify: I connected the amp DIRECTLY to the mid driver alone for all tests. It was quick and simple. Pulled the driver. Disconnected the XO. Ran speaker wire through the port below the mid. reinstalled the driver. First test was full range w/zero filtering.

My goal: to get a feel for the reason for the edgy brightness. Is it the XO or drivers? Seems like the XO; posibly the XO overlaps the drivers.

This weekend I hope to try the tweeter by itself with active HP* starting about 5K, going down to about 2K.
* 4th order

Again, thnx for your help. stay tuned tony

[gp4jesus]

They're softer, less fatiguing these days.

RTA* testing several weeks ago showed a broad 3-5dB* peak around 10K. Life has kept me from posting results.
* low res display

Still commited to bi/triamping as final, no-cost* solution. May pursue passive mid/tweet XO cap upgrade**-$ & which brand to buy, stumbling block
* for me anyway
** purely out of curiousity

[zingo]

I also have been considering an active crossover, but the problem is tube XOs aren't cheap...

[gp4jesus]

A quick RTA test update from several weeks ago:

found RTA screen resolution setting & took careful readings at different volume levels. all dB readings relative from first measurement

1. Peak centered @ 3550-4K hz confirming what my ears have told me all along.
2. -5 dB @ 10 K then plateaing to above 20K
3. greater than 10dB* dip centered below 2K* gently rising to -5dB plateau @ 500hz
* posible XO interference

Feel free to direct questions, comments, or flamouts to me
tony

[Face]

How close was the mic when you measured?

[FTGV]

On what axis did you take the measurements?Response can change markedly with small changes in mic position.

[gp4jesus]

distance: about 1 meter, on-axis

[Face]

Was the mic tweeter level?

[gp4jesus]

Was the mic tweeter level?

Yes. Got similar measurements between T & M too.

[gp4jesus]

upgraded the wire between the mid/tweet BPs & their filtering to AR 12ga OFC

upgraded the wire between the mid & its filter to 14 ga OFC

removed the sub LP filtering
upgraded the sub wire to 8 ga, this channel's subs connected to its own XL280 w/10 ga MC

Currently evaluating with music. Modded channel's bass noticebly more percussive

will use RTA this weekend for comparative readings & tune sub/mid balance

[gp4jesus]

latest configuraton:
Adjustable Mid HP output & frequency on both channels. XO'd about 150hz
Adjustable sub output & frequency for the channel w/upgraded wire.

My ear tells me the range just above the mid HP needs a little help.

hope to get time to take RTA measurements later today.

cheers tony

[mdaudioguy]

Tony - Great stuff, but how about choosing one thread, instead of dumping the same info into three threads every time? I realize they are all related, but...

[Toolfan66]

Can someone explan to me what an active crossover is?? what it does as opossed to a what I would call just a crossover??

Thanks!!

[gp4jesus]

Active XOs are between a pre-amp & 2 or more power amps; they handle filtering the different frequency ranges to designated amps.

For example some might use a tube amp for mids & highs; to a tube amp; transistor amp for lows.

[Joe08867]

Can someone explan to me what an active crossover is?? what it does as opossed to a what I would call just a crossover??

Thanks!!

gp4jesus has it right. But they can also be tailored better or tuned better to the speakers, preamp and amplifier that is being used.

I have used electronic crossovers in PA work and honestly they are amazing in how much you can tweak the sound to fit the venue or room the system is in.

[gp4jesus]

...they can also be tailored better or tuned better to the speakers, preamp and amplifier that is being used.

great point Joe! You reminded me to mention: the better ones allow you to adjust XO frequencies &/or outputs to the different frequency ranges - very helpful if you have a "hot" tweeter or need to fatten the low end for a dance mix.

cheers tony