Understanding this crossover system

0 Members and 1 Guest are viewing this topic. Read 19513 times.

mboxler

  • Full Member
  • Posts: 300
Re: Understanding this crossover system
« Reply #20 on: 3 Dec 2018, 04:00 pm »
That's perfect thanks.

Now to understand the low pass crossover.

I will make a guess. The inductor itself is low pass and the capacitor is there to bypass the driver of any high frequencies.

That's somewhat correct.  As frequency rises, the impedance of the capacitor decreases.  Without taking into account phase shift, think of the capacitor as a variable resistor in parallel with a 8 ohm resistor (the driver).  As frequency rises, the resistance of the variable resistor decreases. 

At 292 hz, a 68 uf capacitor has an impedance of 8 ohms.  So, if we treat it as an 8 ohm resistor in parallel with another 8 ohm resistor, you get the equivalent of a 4 ohm resistor.  At 877 hz, the impedance of the capacitor is 2.66 ohms.  This in parallel with the 8 ohm driver is 2 ohms.

Given a fixed inductor value, the voltage drop across the inductor will increase and the voltage drop across the parallel capacitor/driver will decrease as the signal frequency increases.  The voltage changes occur twice as fast as they would without the capacitor.

The same voltage will always pass thru the capacitor and the driver at ALL frequencies.

In reality, the impedance of a 8 ohm resistor in parallel with a 68 uf capacitor at 292 hz is 5.66 ohms with a -44.9 degree phase shift. 

Here's a calculator you may find fun to play with

https://keisan.casio.com/exec/system/1258032649

Here's some "light" reading on complex AC circuits. 

http://www.bcae1.com/compleximpedance.htm

 

Bumpy

  • Full Member
  • Posts: 444
Re: Understanding this crossover system
« Reply #21 on: 4 Dec 2018, 07:42 am »
Thanks guys for taking such trouble to help me out. I think so far I have much better understanding of these particular crossovers. Hopefully no more questions for a while :)  :thumb:

Jeff

  • Facilitator
  • Posts: 278
  • Test...
    • Sonic Craft, Inc.
Re: Understanding this crossover system
« Reply #22 on: 8 Dec 2018, 08:37 am »
Bumpy,

My apologies for the late reply.  Initially, I was going to recommend another circle as my focus is less on the design side, and more on the flavor of the ingredients.  However, it appears that you more or less received what you were after.  Even though, there is a steeper curve...

While most of the information here is vague, some slightly incorrect, and a tad misleading, it does represent what you are up against.  One thing that was touched on in part, but not completely fleshed out is the phase rotation.  None of those reactive components completely block the signal.  The tweeter does not stop playing below the upper XO point, nor does the mid stop above it.  So we have two drivers playing the same information at a fixed distance from each other.  That means they will create both peaks (coupling), and dips (cancellation).  These can be minimized by rotating phase.  Inductors, and capacitors rotate phase in opposite directions.  They rotate almost 90 degrees as they move in and out of full reactance.  So, the cap on the tweeter has to work with the inductor on the mid.  The cap on the mid has to work with the inductor on the woofer.  Electrically, the cap on the woofer even finds itself in series with the mid an octave or two above and below the lower XO point.

The best advice I can give at this point is this.  During your listening tests, stick with the same brand/series of cap and coil.  It is OK to use a different cap on the tweeter from what you are using on the woofer, but keep that tweeter cap the same brand and type as you change values.  Do not bypass caps, but stacking similar values to achieve a custom value is OK.  Burn new parts in before you do your listening test.  When you have arrived at your final values (and/or topology), you will be ready to select the specific parts that possess strengths for your subjective and unique requirements.

Bumpy

  • Full Member
  • Posts: 444
Re: Understanding this crossover system
« Reply #23 on: 8 Dec 2018, 11:38 pm »
Thanks Jeff valuable info there.  :thumb:

Bumpy

  • Full Member
  • Posts: 444
Re: Understanding this crossover system
« Reply #24 on: 3 Jul 2020, 06:44 am »
Bumpy,
Do not bypass caps,

I come back to this thread time and again to refresh my understanding, but this one sentence stands out. Can you explain why one should not use bypass caps. Its often recommended elsewhere?

David Ellis

  • Full Member
  • Posts: 1044
    • http://www.ellisaudio.com
Re: Understanding this crossover system
« Reply #25 on: 3 Jul 2020, 12:30 pm »
I'll make a completely non-scientific guess at Jeff's assertion.

Adding a bypass capacitor of .1uf can be good, bad, or just wierd.  The sound will change depending upon the "marraige" between the primary and bypass capacitor AND, the placement in the crossover circuit.  Something that works behind the tweeter may not work in a notch filter behind the woofer and the subjective impact might be VERY different.  Adding bypass capacitors during TESTING would really change the listening experience resulting from the amplitude and phase of the curve.  Hopefully this make sense.

I am a huge fan of bypass capacitors, but adding them during initial testing introduces another variable that would make the testing more difficult.

My 2c.


 

Bumpy

  • Full Member
  • Posts: 444
Re: Understanding this crossover system
« Reply #26 on: 5 Jul 2020, 08:05 am »
That makes a lot of sense :)

Bumpy

  • Full Member
  • Posts: 444
Re: Understanding this crossover system
« Reply #27 on: 14 Apr 2021, 08:53 am »
I have returned to this topic as it has proven extremely valuable and my speakers
now sound just 10% short of perfect. Thanks so far.





My last focus is on the midrange driver which is a Supravox alnico T215 RTF 64. I find the driver performs great at its lower frequencies and doesn't need a filter there, so I removed the series capacitor.

I am therefore left with single quality inductor in series. Over extended listening there is a very slight raggedness to the upper frequencies of the Supravox so I would like to lower this low pass filter frequency a little.

Here are my questions

1. With just a single series inductor what type of filter is this (first or second order, Butterworth? etc)

2. If I add a parallel capacitor to it,  what type of filter do I get and presumably the low pass frequency is lowered.

Thanks

« Last Edit: 14 Apr 2021, 10:48 am by Bumpy »

S Clark

  • Full Member
  • Posts: 7356
  • a riot is the language of the unheard- Dr. King
Re: Understanding this crossover system
« Reply #28 on: 14 Apr 2021, 01:28 pm »
"I find the driver performs great at its lower frequencies and doesn't need a filter there"   
How did you determine this?  By ear? By measurement?  Removing the cap means the mid will reach well down into the woofer range and result in superposition addition (and cancellation when out of phase).

1. Remove the cap and you have a first order on the top range of mid and nothing on the bottom.  The overlap with the woofer will be unpredictable.
2. Put the cap parallel and you have next to nothing on the mid.  The inductor passes lows, the cap will pass highs and the mid will see the sum of both- or play sort of a screwed up full signal.

Bumpy

  • Full Member
  • Posts: 444
Re: Understanding this crossover system
« Reply #29 on: 14 Apr 2021, 02:45 pm »
"I find the driver performs great at its lower frequencies and doesn't need a filter there"   
How did you determine this?  By ear? By measurement?  Removing the cap means the mid will reach well down into the woofer range and result in superposition addition (and cancellation when out of phase).

1. Remove the cap and you have a first order on the top range of mid and nothing on the bottom.  The overlap with the woofer will be unpredictable.
2. Put the cap parallel and you have next to nothing on the mid.  The inductor passes lows, the cap will pass highs and the mid will see the sum of both- or play sort of a screwed up full signal.

Some complex stuff there so I'll try and unravel it in stages

"I find the driver performs great at its lower frequencies and doesn't need a filter there"   
How did you determine this?  By ear? By measurement?  Removing the cap means the mid will reach well down into the woofer range and result in superposition addition (and cancellation when out of phase)."


I have tried using microphone measurement techniques but have found, with open baffles, that they don't offer much value in defining what's happening. I'm not sure why this is, but put it down to the increased dependence on the rear wave compared to conventional boxes.

So I do prolonged listening tests often over a number of weeks. The way I have wired my speakers allows me to listen to each driver one at a time in perfect isolation. The midrange extends down nicely without drama and certainly no apparent excessive movement of the cones at low frequency - in fact no movement at all!  When played together with the bass driver there in no evidence of cancellation or addition. Phase has been set at the best sonic overlap (generally better base performance). It should be noted that the base drivers are of good quality (vintage Altec 416)  and contribute beautifully to the base regions, unlike the Eminence Alphas that are often recommended and which did not match the quality of the Supravox.

I tried also with the capacitor in place and SQ suffered.



Bumpy

  • Full Member
  • Posts: 444
Re: Understanding this crossover system
« Reply #30 on: 14 Apr 2021, 02:51 pm »
"I find the driver performs great at its lower frequencies and doesn't need a filter there"   
How did you determine this?  By ear? By measurement?  Removing the cap means the mid will reach well down into the woofer range and result in superposition addition (and cancellation when out of phase).

1. Remove the cap and you have a first order on the top range of mid and nothing on the bottom.  The overlap with the woofer will be unpredictable.
2. Put the cap parallel and you have next to nothing on the mid.  The inductor passes lows, the cap will pass highs and the mid will see the sum of both- or play sort of a screwed up full signal.

1. Remove the cap and you have a first order on the top range of mid and nothing on the bottom. 

OK that confirms what I thought, that the top frequencies will have a be first order (6dB) roll off with just the inductor.  Given that the driver will have an impedance is this what the call an LR filter?

Bumpy

  • Full Member
  • Posts: 444
Re: Understanding this crossover system
« Reply #31 on: 14 Apr 2021, 02:56 pm »
"I find the driver performs great at its lower frequencies and doesn't need a filter there"   
How did you determine this?  By ear? By measurement?  Removing the cap means the mid will reach well down into the woofer range and result in superposition addition (and cancellation when out of phase).

1. Remove the cap and you have a first order on the top range of mid and nothing on the bottom.  The overlap with the woofer will be unpredictable.
2. Put the cap parallel and you have next to nothing on the mid.  The inductor passes lows, the cap will pass highs and the mid will see the sum of both- or play sort of a screwed up full signal.

2. Put the cap parallel and you have next to nothing on the mid.  The inductor passes lows, the cap will pass highs and the mid will see the sum of both- or play sort of a screwed up full signal.

I'm sorry, but I'm struggling with this. The filter is the same orientation as the low pass on the bass driver albeit with different values ie inductor in series and capacitor in parallel with the driver. Its probable that I did not make it clear where the capacitor was going.


BTW thanks so much for your help.

S Clark

  • Full Member
  • Posts: 7356
  • a riot is the language of the unheard- Dr. King
Re: Understanding this crossover system
« Reply #32 on: 14 Apr 2021, 03:13 pm »
On the woofer, the second order crossover has an inductor in series and a cap that bypasses.  The inductor blocks high frequencies going to the woofer.  The bypass cap passes high frequencies and blocks low frequencies, letting the highs pass the woofer.  The textbook effect is a 6 dB reduction in highs by filtering by the inductor and another 6 dB reduction by letting highs bypass at a chosen crossover point.

If you put both in parallel to each other and in series to the mid, one will block lows and pass highs and the other will block highs and pass lows... kinda cancelling the effect of each other--- obviously depending on values chosen as to which frequencies are affected.  As designed, with a cap and coil in series, they work together to allow the passage of a mid range frequency (high treble blocked by coil, low bass blocked by cap).    But I may be misreading what you are proposing....

S Clark

  • Full Member
  • Posts: 7356
  • a riot is the language of the unheard- Dr. King
Re: Understanding this crossover system
« Reply #33 on: 14 Apr 2021, 03:21 pm »
given that the driver will have an impedance is this what the call an LR filter?
Yes, L for inductance, R for resistance.

Bumpy

  • Full Member
  • Posts: 444
Re: Understanding this crossover system
« Reply #34 on: 14 Apr 2021, 03:32 pm »
On the woofer, the second order crossover has an inductor in series and a cap that bypasses.  The inductor blocks high frequencies going to the woofer.  The bypass cap passes high frequencies and blocks low frequencies, letting the highs pass the woofer.  The textbook effect is a 6 dB reduction in highs by filtering by the inductor and another 6 dB reduction by letting highs bypass at a chosen crossover point.

If you put both in parallel to each other and in series to the mid, one will block lows and pass highs and the other will block highs and pass lows... kinda cancelling the effect of each other--- obviously depending on values chosen as to which frequencies are affected.  As designed, with a cap and coil in series, they work together to allow the passage of a mid range frequency (high treble blocked by coil, low bass blocked by cap).    But I may be misreading what you are proposing....

I think you may be.  :)

The inductor is in series as now (6dB reduction of highs) and the capacitor would be parallel to the driver and therefore bypass it (further 6dB reduction of highs). Is this a Butterworth second order (12db) low pass?

S Clark

  • Full Member
  • Posts: 7356
  • a riot is the language of the unheard- Dr. King
Re: Understanding this crossover system
« Reply #35 on: 14 Apr 2021, 04:28 pm »
Yes, not sure how Butterworth is defined, but it is a traditional second order, 12dB filter. 

Bumpy

  • Full Member
  • Posts: 444
Re: Understanding this crossover system
« Reply #36 on: 14 Apr 2021, 05:26 pm »
Thanks again

I'll try some calculations and see what I come up with

The single inductor of 0.39mH and the 8 ohm driver gives a low pass of about 3200Hz.

With the added cap I would like to bring that to about 2000Hz, (and from 6db to 12dB slope) but the calculators I have found suggest that the inductor is not in the required range, so I may have to start over with both components - it gets a bit expensive experimenting with quality components :(

I have tinnitus and I think it is triggered by 'distortions' at about 4000 Hz so there is an added reason to pursue this.

S Clark

  • Full Member
  • Posts: 7356
  • a riot is the language of the unheard- Dr. King
Re: Understanding this crossover system
« Reply #37 on: 14 Apr 2021, 08:26 pm »
If available, check the impedence of the driver at the crossover frequency.  It may not be 8 ohms.

See in the graph below, the impedance changes as frequency changes.  This is a mid made by Peerless/Tympani, a respected driver.


Bumpy

  • Full Member
  • Posts: 444
Re: Understanding this crossover system
« Reply #38 on: 15 Apr 2021, 12:03 pm »
Here is the impedance curve for the Supravox driver

http://rutcho.com/speaker_drivers/supravox_t215_rtf_64/images/supravox_t215_rtf_64_z.png

There will be a small tweak to my calculation for the slightly raised impedance at about 3KHz

What are the consequences of allowing the driver to be open to very low frequencies. Obviously this will present a higher impedance, but how will that manifest itself?
« Last Edit: 15 Apr 2021, 01:28 pm by Bumpy »

Jeff

  • Facilitator
  • Posts: 278
  • Test...
    • Sonic Craft, Inc.
Re: Understanding this crossover system
« Reply #39 on: 19 Apr 2021, 08:07 pm »
Quote from: Bumpy
I tried also with the capacitor in place and SQ suffered.

This almost always occurs in the mid.  It is a critical position.  While it can, and will, degrade mid performance, many times the alternative is worse as it is necessary to control the response of the mid.  If you are able to omit it, consider yourself lucky.

Quote
Given that the driver will have an impedance is this what the call an LR filter?

It would not be called a LR here.  By definition, impedance is frequency dependent reactance, not DCR. 

Quote
Is this a Butterworth second order (12db) low pass?

Butterworth describes the characteristics of the rolloff (Q=.707).  Does your rolloff approximate this Q?

Quote
I'll try some calculations and see what I come up with. The single inductor of 0.39mH and the 8 ohm driver gives a low pass of about 3200Hz. With the added cap I would like to bring that to about 2000Hz, (and from 6db to 12dB slope) but the calculators I have found suggest that the inductor is not in the required range, so I may have to start over with both components - it gets a bit expensive experimenting with quality components :( I have tinnitus and I think it is triggered by 'distortions' at about 4000 Hz so there is an added reason to pursue this.

When switching from a single pole to a two pole mid LP, you should also consider the tweeter cap value.  The issue that you site, is it better, worse, or the same when the tweeter is connected?

With this fly by the seat of the pants design approach, you should not be using expensive parts.  However, as mentioned before, they do need to stay the same for each position as you adjust values.  While it may not sound wonderful as a whole, the best sounding topology with cheap parts will also sound the best with better parts.  I would recommend using film everywhere, and air core inductors.

Quote
What are the consequences of allowing the driver to be open to very low frequencies. Obviously this will present a higher impedance, but how will that manifest itself?

The higher imp results in rolling the driver off (woofer). 

If the boost or cancellation is not detrimental to the response, you only concern is excessive cone excursion (mid).