Hi-Fi speaker resistor/capacitor question

[snk]

Hello,
I recently got a pair of vintage DUAL speakers. They look great, sound okay, but I find the overall sound a bif muffled/damped.
I see that there are a couple capacitors in the integrated  passive filter, and from what I know some people tend to recap them (like any other old gear). I have read quite often that it can help "improve the sound", but to me it means nothing (by that, I mean that it doesn't tell what it actually does)
From my basic understanding, I thought that the resistors in hi-Fi speaker filters was there for the crossover between each speaker/voice, and not acting as a global hipass or lowpass.

My question is the following : do filters in Hi-Fi speakers can help shaping the frequency range (hipassing or lowpassing the signal), or is it there only for crossover?
If I change the capacitors, I understand that it might change the sound (because changing the crossover might create a dip or a bump around the crossover point, or increase the frequency overlap, etc), but may it help getting a brighter sound?
I thought it was the "inner/core" features of the tweeter which was responsible for the hi-frequency quality, and not that some capacitors might damp it...

So, this is a general question, but if the question is too vague (there are thousands of speakers, maybe some are different than others), here is the Dual CL240 schematics :


I would be also curious to know what the 10µF cap does (because on some pictures, I have seen it being 6.8µF), as well as the 20µF (because I am wondering if using a 22µF would make a big difference)?

[iainpunk]

a few points;
1 those are not the TYPE of cap you replace when recapping
2 the values of the crossover are chosen depending on the speaker properties and the cabinet in which they are mounted, don't mess
3 for a tweeter, there is a cap in SERIES to block bass, probably a resistor in that part too, my guess is the 47ohm parallel to the tweeter, you might want to leave the values alone. if you want a bit more top end, the tweeter's part of the crossover is usually not the place to mess with.
4 if you want a brighter sound, get yourself an equalizer, to tailor the tone to your room/ears

cheers, Iain

[snk]

Thank you, Iainpunk.

2 the values of the crossover are chosen depending on the speaker properties and the cabinet in which they are mounted, don't mess

Yes, I know that
I began wondering because 1- I saw pictures of the same speaker model featuring different cap values (6.8µF on one, and 10µF on the other one), 2- I thought that 20µF & 22µF might be within the cap tolerance range...

1 those are not the TYPE of cap you replace when recapping

I went on a couple vintage audio forums, and most members seemed to replace capacitors in their speakers filters. I never replace caps blindly as a "routine", but got surprised to see so many people changing the filter caps there.
(here, for instance, on the Dual CL240)

4 if you want a brighter sound, get yourself an equalizer, to tailor the tone to your room/ears

Well, those are not my main speakers, and while I like to chain pedals, I don't want to add too many stuff between my vinyl player and my amp

[snk]

... For reference, this is where I found about these servicing :

here, it's 10µF and 22µF:
https://www.audiovintage.fr/leforum/viewtopic.php?t=8613

there, it's 6.8µF and 20µF :
https://vintage-audio-heritage.fr/viewtopic.php?t=573

... On the schematic above, it's 10µF and 20µF.
(And from what I know, they didn't change either the tweeter or boomer, or the cabinet. This is why I supposed there might have some tolerance and asked the question)

[iainpunk]

well, 20 and 22 uf can overlap in the tolerances, so i wouldn't worry about the value change there.
i think its a bit weird to replace non-electrolitic capacitors, since they don't go bad at all! i have foil caps from the 1930's that still measure the correct capacitance (500pF) and 0.03ohm ESR!

the 6.8uF might have a little influence, but it would probably be on the mid range, not the treble.

if it sounds dull, have you checked if the tweeters aren't blown? probably piezo tweeters, so you cant use the multi meter to measure resistance, since its meant to be infinite! i suggest running the box and feeling with your fingertips if the tweeter does its job (to vibrate).

cheers

[Rob Strand]

if it sounds dull, have you checked if the tweeters aren't blown? probably piezo tweeters, so you cant use the multi meter to measure resistance, since its meant to be infinite! i suggest running the box and feeling with your fingertips if the tweeter does its job (to vibrate).

Good advice and probably the best place to start.     You should be able to hear sound from the tweeter.

The next thing to check would be to make sure the grey disk part, a protection device,  is measuring 0.7ohms.     You can use a multimeter set to ohms.   Measure the resistance across the device.   You might see upto 1.7 ohm because of the meter  leads.  If you are seeing more than say 2 ohms I'd suspect those protection devices are damaged.    They don't normally blow.

It's unlikely any other parts are damaged.

For treble it helps to not have the speakers on the ground and listen at standing height.  It's (usually) best to listen to the speaker on the axis of the tweeter.

With that speaker design you will not be able to increase the level of the tweeter.

One way to make speakers sound better is to put more stuffing in the box.    Lightly fill the entire enclosure.  It doesn't make the tweeters louder but it certainly improves the midrange and makes it sound less confused and muddy.

If we assume the crossover is properly designed you shouldn't mess with it.    Speaker crossovers are hard to design.   They are designed to be correct at a certain listening point.    The time delays to the ear from each driver have a large impact on the sound.  The type of crossover in that speaker  should theoretically be wired with the speaker polarities as shown.  However, because of the time delays it can often work out better by flipping the polarity of the tweeter.      Normally that would be done by the designer and the crossover part values would be chosen to give the best results.

Now here's where I'm reluctant to make a recommendation:   You will sometimes find flipping the polarity of the tweeter will sound better.    That occurs when the speaker crossover has not been design correctly.    Technically the crossover needs to be redesigned however trying to make the best of a bad situation flipping the tweeters is the only way forward.   When the polarity is *wrong* typically speaker sounds weird, it can have a hollow or confused sound.   More often than not the better tweeter connection is when there is more treble:   the reason for that is the correct tweeter phase should add and the wrong tweeter phase should give quite a bit of cancellation.    Be warned it's easy for an untrained ear to choose the wrong connection!
[Perhaps focusing on the 'sh' and 'ch' in speech would be a good start.]

Suck-out due to wrong speaker phase around crossover frequency,


FWIW, "Dual" was a brand of HiFi equipment.

[GGBB]

FWIW, "Dual" was a brand of HiFi equipment.

And still is - sort of. http://www.dualav.com/

[antonis]

FWIW, "Dual" was a brand of HiFi equipment.

Yeapp..

I still have CS606 turntable "somewhere" in my basement..

[Rob Strand]

I still have CS606 turntable "somewhere" in my basement..

That's one of the nicer 80's models with direct drive.   I had an earlier Dual turntable with the idler wheel drive and AC motor.
Dual were know for their turntables more than anything else.

[amptramp]

Tweeter lead reversal is going to be necessary in a properly designed crossover because at the crossover point, there will be a phase lag of 90 degrees in the woofer and a phase lead of 90 degrees in the tweeter.  If they are connected in the same phase, there will be cancellation at the crossover point.

I obtained a couple of JBL G-730 instrument speakers for $8 each because the crossover components are mounted on a plastic piece that holds the input connectors.  This was broken and whoever owned it tried to fix it but couldn't and damaged the inductors.  I obtained a crossover board with the necessary inductors and capacitors but the capacitors are electrolytic, which are not likely to be able to take the current for very long.  The original JBL crossover had no discrete resistors but the new crossovers I obtained did have them, so I took them out.  These are there to keep the frequency response relatively level but for an instrument, power output trumps a level frequency response.  The person who tried to fix these speakers used a nice item that I hadn't see before - a circuit element that looked like a lamp filament in a glass enclosure (it might actually have been a lamp).  This way, the resistance was minimal until an overload occurred and the filament resistance went up.  Neat trick.

[Rob Strand]

Tweeter lead reversal is going to be necessary in a properly designed crossover because at the crossover point, there will be a phase lag of 90 degrees in the woofer and a phase lead of 90 degrees in the tweeter.  If they are connected in the same phase, there will be cancellation at the crossover point.

That's the textbook answer.  Real speakers and geometry change that quite a bit.

The delays due to different distances to each driver changes the phase.  That arises from two sources: the distance from each driver to the ear, and different acoustic centers of the driver.   Typically the acoustic center of the a woofer is around the voice coil position which effectively moves the woofer back from tweeter.   A 30mm delay at 3kHz is 90 degrees.

On top of that are phase shifts associated with the speaker roll-off.  The low-end cut-off of the tweeter adds to the tweeter crossover roll-off.   The high-end cut off of the woofer adds to the woofer crossover roll-off.  The additional roll-off has associated phase shifts which moves the phase further from the ideal.

When you combine all the effects it's not uncommon to have to flip the phase.

You can get a situation where the phase-shift adds up to where neither in-phase or out-of-phase work well.  That's where the phase shift is +/ 90deg.    In this case it's not uncommon to make the tweeter crossover third order and the woofer cross-over second order.   The additional phase-shift of the asymmetrical order crossover puts things right again.

Here's an example of a real crossover design,
http://www.troelsgravesen.dk/DiscoveryW18.htm

Pay close attention to the series resistor values on the caps to determine the true order of the crossover.   See note #1 " Use an asymmetrical crossover usually consisting of 2nd order to the bass and 3rd order to the tweeter to make "proper" phase integration."    His baffle step tries to fix the delay to some degree.

[Phend]

These things have gotten to complicated.

[Rob Strand]

These things have gotten to complicated.

Unfortunately speaker crossover are complicated.    Like most things the people that do this stuff all the time know the game.   There's a number of speaker crossover design programs around which take into account all the details.   They are essentially optimization programs they don't use textbook formulas at all.

[Rob Strand]

At the top of this thread is a frequency response of the CL-144.   The CL-144 seems similar to the CL-240.   The response looks fairly flat but there does seem to be a tendency for the bass, below 1500Hz, to be a little higher than the treble.
http://forum.vegalab.ru/showthread.php?t=41769


A second observation if you look at the schematic for the CL-240 the PCB has terminal markings + and - for the woofer and tweeter.  The pcb + connection for the woofer is the red wire and that wired to the + terminal on the actual woofer.
The pcb + connection for the tweeter goes to the black wire and the black wire is wired to the - terminal on the actual tweeter.

You photos seems to match the flipped *wiring*.

The question is does what does the schematic show?  Are the + and - connections on the schematic the PCB terminal markings or the actual connections to the driver?

It wouldn't be hard to check the PCB but the point here is the polarity on the schematic is ambiguous since + and - could mean the actual driver or the PCB terminals.

It's not clear if those speakers are 4 or 8 ohms.  If they are 4 ohms then the 0.7 ohm protection device could knock out more than 1dB.

Here's the response plot for the CL-240,
http://www.hifi-archiv.info/Test/Dual%20CL-Serie%20Test%201977/1977-2-112.jpg

Small amount of dip around the crossover, not severe,  but a small level imbalance between the tweeter and woofer.

Looks like 4 ohm.  So maybe shorting the protection device will help.
-----------------------------------------------------------------------------------------------------
Well, well, well, what's this?

This guy had a soft tweeter in one speaker and it turned out to be a faulty PTC (the grey disk).

https://audiokarma.org/forums/index.php?threads/need-crossover-recap-advice-dual-cl-240.570901/

[snk]

Wow, I wasn't expecting so many replies, Thank you all  for the knowledge and tips ! 


First, to be clear, I wouldn't want to mess with the crossovers, I know it's a tricky thing. My initial question was less about « willing to mess with carefully tuned crossovers », but more about « trying to understand the schematic » (like « capacitor X acts as a HP along with resistor Y to remove high frequencies, then it goes to the boomer », etc)

Yes, Dual is quite popular here in Europe. Their turntables and amps have a good reputation. Their speakers are less famous, but a neighbor was selling these and I wanted to give them a try.

I don't thing that the tweeter is dead or faulty, but I will check. I think it's just their sound, the way it was designed. It's vintage gear, and it sounds good on vintage music (like soul, early funk, psychedelic rock, jazz, etc). But I do mostly listen to modern electronic music with deep basses, sharp transients and sizzling hi-hats, and while the speakers claim to have a very wide frequency range (up to 25khz), I think that the upper octave is damped (compared to my other speakers), thus making the sound a bit dull.

On this forum, the original poster has the same listening impression :
https://www.audiovintage.fr/leforum/viewtopic.php?t=8613

« Les aigus sont très fins et pas agressifs, ils mériteraient, pour moi en tout cas, un niveau sonore un poil supérieure histoire de rééquilibrer. Je les trouve un petit peu en retrait, et du coup le médium est mis un peu en avant. Je pinaille, mais c'est pas dérangeant du tout, l'enceinte est agréable à écouter, les timbres sont très jolis. Elles manquent cependant d'ouverture »

« The overall sound is pleasing. Treble is sharp but smooth, and could be increased a bit for a better balance. I find them slightly rolled off, making the medium a tad upfront. »

[snk]

Looks like 4 ohm.  So maybe shorting the protection device will help.
-----------------------------------------------------------------------------------------------------
Well, well, well, what's this?

This guy had a soft tweeter in one speaker and it turned out to be a faulty PTC (the grey disk).

https://audiokarma.org/forums/index.php?threads/need-crossover-recap-advice-dual-cl-240.570901/

Thank you, very interesting information!
Yes,from the informations i have, it's 4ohm. I obviously drive them with a 4ohm amp (DUAL brand, too).

[Rob Strand]

I think that the upper octave is damped (compared to my other speakers), thus making the sound a bit dull.

You might make a small improvement by soldering a bridge across that protection device.    I'd say that would give worthwhile tweak to the sound, perhaps even enough.

The grills on those old speakers use quite heavy cloth which can absorb the high frequencies.   Taking that off would probably help a lot.  You would need to do a test.  From what I can see you cannot remove the grills alone.  So yeah a big job and it would be a shame to stuff-up the original look of the speaker.

The simplest way to do specific EQ is to do it electronically.   You can EQ the highs with a very simple circuit.  You place it between the preamp out and the power amp in.

[snk]

if it sounds dull, have you checked if the tweeters aren't blown? probably piezo tweeters, so you cant use the multi meter to measure resistance, since its meant to be infinite! i suggest running the box and feeling with your fingertips if the tweeter does its job (to vibrate).

Good advice and probably the best place to start.     You should be able to hear sound from the tweeter.

I will check. I think they work (i mean : i do hear some high fequencies, it's not like I had put a Moog filter on the output

The next thing to check would be to make sure the grey disk part, a protection device,  is measuring 0.7ohms.     You can use a multimeter set to ohms.   Measure the resistance across the device.   You might see upto 1.7 ohm because of the meter  leads.  If you are seeing more than say 2 ohms I'd suspect those protection devices are damaged.    They don't normally blow.

I'll check, too.

For treble it helps to not have the speakers on the ground and listen at standing height.  It's (usually) best to listen to the speaker on the axis of the tweeter.

With that speaker design you will not be able to increase the level of the tweeter.

One way to make speakers sound better is to put more stuffing in the box.    Lightly fill the entire enclosure.  It doesn't make the tweeters louder but it certainly improves the midrange and makes it sound less confused and muddy.

Thank you for the tip. I know quite well how to place speakers, and these are 50 cm away from the wall, at 1m height, which should be fine for the listening tests.
I have read about stuffing the inside further, it's something I am considering (but I like generous low end, and at the moment I like what i hear in the lows, ah ah !).

[snk]

You might make a small improvement by soldering a bridge across that protection device.    I'd say that would give worthwhile tweak to the sound, perhaps even enough.

I opened the case, and checked the circuit components.

- On my board, I have the 20µF and 10µF caps, so it's on par with the service manual specifications.
- I checked the grey disk resistor : as you wrote, i got 1.6 ohm with my DMM, so I guess it's good. But it looks ugly, with stains on the coating. Maybe it's only cosmetic?
- I tried to put a metal wire across the resistor legs to see if it sounded different : it didn't.
- Would you care to elaborate a bit about that 0.7 resistor? Its purpose is to act as a voltage protection, right? but, depending on the input impedance, it will also act as a smooth lowpass, that's it?
- I noticed that the wiring on my system is "antiphase", fwiw.
- I have glued some cork sheets on the sides of the cabinet, and a rubber damper sheet on the cab base. I'm waiting for it to dry, and i'll proceed to some listening tests.

The simplest way to do specific EQ is to do it electronically.   You can EQ the highs with a very simple circuit.  You place it between the preamp out and the power amp in.

Thank you for the tip, but i don't want to go too deep with these speakers, and I want to keep things simple. I mean, I often like to stay faithful to the way things were designed, and I feel that if I don't like the way the speakers sound, maybe I should just use other speakers (just like there is no point in trying to make a Tubescreamer sound like a Fuzz Face. You can tweak a Tubescreamer, but trying to change it into something that much different is quite pointless).
After a deeper listening, I don't think the Dual are faulty at all. Maybe it's to be expected to have 1978 speakers (with a sweet mid bump) sounding fine on 70s music, but not as good on 2020s electronic music. Also, maybe that I am trying too much to compare my Dual against my bigger Magnat speakers, and that I should enjoy the Dual for what they are.

[snk]

pictures :

_ the wiring :



_ the caps & disk resistor :