Would like some help with Infinity Kappa 400 crossover

[Dejau]

Hi,


I know very little about crossover design so please bear with me.


I bought used Kappa 400 after owning Kappa 200 and liking them quite a bit but kappa 400 had issues – one was that their crossover was most definetely "tweeked" (no idea what was "tweeked") by the previous owner as tweeter was far too hot (and owning kappa 200 I know how tweeter should sound in relation to midrange driver).


Upon opening these up (sorry, no stock photo) I saw that crossovers looked fine (everything looked "stock") apart from burnt resistors (now changed to new 1 and 12 Ohm cement resitors). 1 ohm resistor was at least recognisable (both measured 1.2-1.4 Ohm, color coding was recognizable to some extent) but other resistor (now 12 Ohm) in both crossovers was badly damaged and was hard to measure correctly (it was burnt, color codes made no sense, resistors measured 13 Ohm for the better looking one and around 27 Ohm for badly burnt one; I will try to find them because I know I put them somewhere).


I really like pronounced high frequencies (I also use monitor audio GS10 and Elac jet 310 speakers and AKG K812 headphones to give you an idea) but this kappa treble is far far too bright even compared to them. My last straw was new Kef LS50 (which should be somewhat bright) which sound like sennnheiser HD650 (veiled) when compared to Kappa 400 treble monsters.


1) So basically I now have a trouble of making treble far less hot but dont know how to do it correctly since I only know hot to add in-series resistor and here resistors are in parralel as far as I can see. Please bear with my lack of knowledge but the only ther thing I can think if is to replace that 1 Ohm resistor (higher value) to reign in tweeter and replace 12 Ohm resistor (lower value) to keep midrange driver at same-ish level (bass is way too shy at the moment but that may be because of tweeter murdering everything else)
(what would be a good start?)...


2) Also I think that polarity of midrange and tweeter was reversed (as far as my skills of reading circuit boards go, that is definetely the case here, dont know of it was designed to be so...) as I get phase error every single time in audyssey XT32 when trying to calibrate kappa's (do not get this error with my other speakers)?


Would really appreciate your help as I want to my these kappa's great again. Thanks!


Edit: wasn't able to add images to a thread so here are google drive links:


Top view – Infinity kappa 400 top.jpg - Google Drive
Back view (C – capacitor; R – resistor) – Infinity kappa 400 back.jpg - Google Drive

 

[picowallspeaker]

Mmmmhhh, just figured out ( after googling) what those Infinity Kappa 400 look like: two- way or three way, Two-way with side firing woofer.
So the 150 uF cap is the HP ( together with the coil and the 5 W 10 Ω resistor> this R determines a less...determined intervention of the coil so "softer" slope, thus a different Q)
The 1 Ω resistor is in series to the tweeter> it determines ...well, NO!
I'm lost
It's your exercise to derive a good looking circuit schematic from that PCB

Hope that helps...

 

[picowallspeaker]

Ok, now I got it...still I cannot figure out how a side-fire woofer...never mind!
The mid-woofer is high passed ( it goes together with the suspension/closed box approach) by the 150 uF cap, the LP ( one big coil )- the other big coil serves the side woofer. The tweeter is CLC ( third order) and has no attenuation; the polarity is right, what else? Uh, the color bar code...R U kidding?!? There's only J or K for tolerance, for those cemented resistors.

 

[Dejau]

Thanks for your help and I'm really sorry that I dont know how crossover is designed (I understand how simple crossover works but this is far more complex than what I'm familiar with) – side-woofer is only passed through iron core inductor (B in scheme). What did you mean by "color bar code"? I measured resistors at shop to be at 1 Ohm and 10 Ohm exactly. That 10 Ohm resistor goes to both tweeter and mid-woofer and 1 Ohm resistor goes just to mid-woofe (here is my terrible paint job – Infinity kappa 400 back(1).jpg - Google Drive )



By the way there is pretty much nothing in series in this crossover (apart from side-woofer). My main goal is just to tame that tweeter because it most definetely should not be so bright (and I like bright sound signature).

 

[picowallspeaker]

Well, it was you that was complaining about color codes:

color coding was recognizable to some extent) but other resistor (now 12 Ohm) in both crossovers was badly damaged and was hard to measure correctly (it was burnt, color codes made no sense, resistors measured 13 Ohm

But...well, nevermind (... the bar ?!).

The 1 Ω resistor makes me think...

 

[AllenB]

After looking at your drawings and photos, I don't see resistors going to the tweeter.

The 1 ohm is part of the midrange low pass and the 10 ohm is part of the midrange high pass. They do not change the level.

When you design an L-pad, you should de-solder T- and put it in there.

 

[Dejau]

So did I understand that correctly – as far as crossover is concerned – these is no resistors in circuit to lower level of any of these speaker drivers (meaning they should be about the same sensitivity)? Can I use QED WM14 to test out what tweeter level I like?

On another note I found my "original" resistors (and I personally think they are original crossover components). Their remeasured values are – 1.4 and 1.6 Ohm; 9 Ohm and 23 Ohm. Any idea of what value resistor should be of second resistor pair (they are same but one measures 9 and another 23 Ohm). Link to photo – Resistors.jpg - Google Drive

 

[AllenB]

Brown red gold should be 1.2 ohms. Brown grey black should be 18 ohms. I am not sure of the colours as brown is different in each case. Sometimes it gets like this with heat damage, age fading etc. What colours do you see?

 

[Dejau]

Brown red gold should be 1.2 ohms. Brown grey black should be 18 ohms. I am not sure of the colours as brown is different in each case. Sometimes it gets like this with heat damage, age fading etc. What colours do you see?

In "dry" parts of resistor it looks like it is – brown, light brown/brown, black, gold. In "wet" (covered in dark goo) parts – only dark goo/greyish, greyish, blackish, goldish.

 

[AllenB]

Is it time for some guesses? These resistors are to modify the filters, you might say it changes them a little away from 2nd order toward first order. In the case of the 1.2 ohm this is not much, it probably just tweaks a small peak or dip and adjusts phase a little. The larger resistor probably has more effect on the filter by increasing the upper bass.

Your idea about the QED is good. I don't know what circuit the QED uses, so I don't know if it is the best choice.

 

[picowallspeaker]

For 120 £ it is easily replaced with a 12 £ L-Pad
( I still keep the variable rheostats that equipped my original Infinity RS 4001...the ones you used to find in JBL and Altec speakers )
Or...with a fixed L-pad made with two resistors
Resistance Divider For Level Adjustment

 

[Dejau]

I just meant that it was really hard to find out what actual value of resistor should be (its obvious that both resistors should be the same but one (9ohm) is within range of 11 ohm and another (23 Ohm) is somewhat within a range of 18 Ohm). In turn I gambled on 11 Ohm to be a correct value and falsely assumed that they may play some part in dampening sensitivity of either a tweeter or midwoofer.



I will now play with audyssey XT32 (I have no calibrated mic) and L-pad (as I see it is really easy) to find how much attenuation I need on tweeter.


There are now only two things I dont get:
1) How come polarity is reversed (amplifier + (positive) goes to tweeter and midwoofer - (negative) but side woofer is in correct polarity (please keep in mind I get phase errors with both speakers in audyssey).
2) As I understand in this stock crossover there was no attenuation done on either tweeter or midwoofer?


Thank you for your patience.

 

[AllenB]

The phase of the response changes with frequency. Even looking at each driver on its own, the phase varies. Changing polarity is just a way to keep this matched up, so this polarity change is not wrong.

If you get these resistors wrong, it will affect phase. Can you show the audyssey results?

 

[Dejau]

Measurements show that I'm looking for problems that aren't there (apart from that mid-range dip from ~2 to 3 kHz (BBC dip?)) – Audyssey measurements.jpg - Google Drive

 

[AllenB]

Some thoughts. There appears to be a small peak above 300Hz and a falling below 300Hz. Could be room modes, but also looks like a mistuned crossover.

Not sure if the 2k dip is intentional, wonder where the upper cross is (without knowing the inductor values).

And knowing where the upper cross is might show whether the woofer is running too high in frequency. Perhaps a flat response might give the impression of too much treble.

 

[Dejau]

I'm still experimenting but at least now I know that the problem (at least it looks like) lied in side-woofer doing very little side-woofing (lets just say both elac 310i, kef LS50 are outperforming my side-woofers in 40-80 Hz range and it's not even comparable to Monitor audio GS10 and silly when compared to wherfedale evo2-50 (LS50 and 310i are less efficient than my kappa's ). Both L and R side-woofers are ok physically, its just that they play full range signal very well (12 000 Hz and I still hear at what appears like half or 1/3 volume or what it was like at 100 Hz). They only have beefy inductor in series (no capacitor in parallel) and do little in that bass range of 40-80 Hz (in adyssey they are classified as small speakers).


At least I have a power amplifier so I will be able to test side-woofers powered seperately (bypassing inductor) and connected with active crossover at 100hz. Will report my findings for any other curious soul.

 

[picowallspeaker]

Hi, you can eliminate the 150 uF high-pass capacitor for the midwoofer, since you can play with the active crossover!

 

[Dejau]

Thanks, will do just that as there is big dip in 120-150 Hz range (and audyssey sets crossover point at 150 Hz) with side-woofer disconnected.

It sounds far better with active side-woofer crossover (bass sounds authoritative even at low volume).

And that causes another question – shouldn't there be a gradual roll-off after crossover point for side-woofer? In passive crossover there is only inductor (quite big, dont have any equipment to measure inductance) connected in series but sound level is pretty much the same at 100 Hz and 1000 Hz and there is little difference even at 7000 Hz (after that roll-off is sharper but that I think has more to do with a driver itself and not a crossover).

 

[AllenB]

The inductor by itself is not enough because the impedance rises against it. Maybe they were expecting that on the side you wouldn't hear the highs. Obviously not the case.