Jump to content

Klipsch crossovers


Bacek

Recommended Posts

@mustang_flhtand others: 

 

In my attempt to modify CW III Crossover to H III with LPAD attenuation to compensate 3db difference with CW III, I first tried to see what Klipsch did with Tweeter circuit by inserting a 3 ohm series resistor before Tweeter.

I inserted 3 Ohm Series Resistor and I am surprised that measurements of voltages before and after 3 Ohm Series show not much difference at all. There is not even 1 dB attenuation in the tweeter pass band, let alone 3 dB. Here are my measurements. 

 

Note -Please do not look at the Absolute voltage values, as my meter cannot measure full voltage beyond 1 KHz. Above 1 K its readings get tilted down as we go up in freq. But what we are interested is only the Delta not absolute value.

What is wrong here? Also notice the Voltage across (3 Ohm + Tweeter) increased as opposed to Voltage across just Tweeter. Do not understand why Voltage is not same.

 

What is wrong here? 

 

image.png.28d9c52e0920fd63a530b69208123eb1.png

Link to comment
Share on other sites

On 7/21/2022 at 10:00 AM, radiogram said:

 

Also notice the Voltage across (3 Ohm + Tweeter) increased as opposed to Voltage across just Tweeter. Do not understand why Voltage is not same.

 

What is wrong here? 

 

"Math is the only subject that counts"

 

Are there any other series resistors in this circuit?

 

Assuming you have an 8 ohm tweeter, the circuit is designed for that specific load.  The series 3 ohm resistor increased the load, and now the circuit sees an 11 ohm load.  Think of it as replacing an 8 ohm tweeter with an 11 ohm tweeter.  In order to get the same voltage across the 11 ohm tweeter, you need to divide all the capacitors in the circuit by 1.375 and multiply all the inductors in the circuit by 1.375.   

 

The Lpad did not change the load on the circuit.

 

Mike

Link to comment
Share on other sites

@mboxler- There are no other series resistors other than the 3 ohm I added. if we assume an average impredance of 8 ohm for the tweeter then the total load impedance has increased from 8 to 11 ohms. Then the volatge across the tweeter should be 72% (8/11) translates to power drop of about -2.7 dB (20 * LOG(8/11)). But my measured date shows at no frequency it is  not even -2dB. In fact at 8K and 10K it pretty much -0.3dB only.

 

from My experiments I figure the mid driver K53 approximation of 16 Ohms hold good, but for the tweeter 8 ohm does not hold good. In 6K to 10K 10 ohm holds good. Beyond 12K Tweeter impedance increases as we up the frequency.

 

Anyway I just got Hand held OWON oscilloscope and will re do my measurements as my Fluke multimeter readings above 1K are suspect and post my findings again.

 

Also note that I just realised I replaced my Titatinum tweeter diaphragms with Bob Crite's diaphragms and they are no longer original Klipsch diaphragms.

Link to comment
Share on other sites

28 minutes ago, radiogram said:

@mboxler- If we assume an average impedance of 8 ohm for the tweeter then the total load impedance has increased from 8 to 11 ohms. Then the voltage across the load should 72% (8/11) translates to power drop of about -2.7 dB (20 * LOG(8/11)). 

 

Just so I'm clear.  At 4000hz the voltage across the resistor and tweeter is .122 volts.  The voltage across the tweeter alone is .0940 voltages.  That's a -2.27db drop, so the tweeter isn't a resistive 8 ohm load at that frequency.  

 

The added 3 ohm resistor dropped the current through the crossover circuit, which also lowered the voltage drop across the crossover circuit.  Since the voltage decreased across the crossover circuit, the voltage increased across the load (the resistor and tweeter).  

 

You should be able to verify this by measuring the voltage from common to before the tweeter alone, then measure the voltage from common to before the resistor/tweeter combo.

 

Sorry if I'm not following you on this.  If you post the schematic I could simulate my explanation.  

 

Mike

Link to comment
Share on other sites

@mboxlerThe attenuation we are looking at is after vs before inserting the resistor because that is the end result we want as Without changing the volume of my amplifier I want Tweeter attenuated. Yes if you just compare voltage across tweeter vs 3 ohm sure we get -2.27dB drop and the math is right. But this not the real attenuation we are looking for. .

 

So, the real attenuation is 20 * LOG((Voltage Across Tweeter w/o 3 ohm series) / (Voltage across tweeter with 3 ohm series) ). So for say 4K Hz it is 20 * LOG(.094/.1020) = -0.709 dB Only.

 

 

Link to comment
Share on other sites

Sorry I'm no help, perhaps someone else can chime in.

 

The point I'm poorly making is that the transfer functions of an LPad vs series resistor are different.  This is due to the load change on the network.

 

The plot below is a simple high pass to an 8 ohm resistor (driver). I'll use an 8 ohm series resistor to get -6db to make it simple.
 
Brown  is 20uf capacitor.  Voltage across the driver.
Red      is 20uf capacitor and 8 ohm attenuating resistor (-6db).  Voltage across the driver.
Blue     is 10uf capacitor and 8 ohm attenuating resistor (-6db).  Voltage across the driver.
Purple is 20uf capacitor and 8 ohm attenuating resistor (-6db).  Voltage across both the attenuating resistor and the driver.

 

Even though the voltage across the driver is always -6db less than the voltage across the series resistor/driver, it can't give you -6db I believe you are looking for across the entire frequency without changing the capacitor.

 

If you want to keep the 20uf capacitor AND get the blue plot, you need a -6db Lpad.

 

Mike

Screenshot (149).png

  • Like 1
Link to comment
Share on other sites

@mboxlerI think I understand what you are driving at. Perhaps I did not communicate my expectations clearly. A properly implemented LPAD will give desired attenuation and yet make the network see the same load as before and thus without altering the frequency response that was there before the LAPD was inserted. A simple series resistor will not maintain the same freq response and so attenuation at different frequencies will be different. So I am not expecting same attenuation at all frequencies.

 

What I am saying is the the insertion of 3 ohm resistor makes no real difference in output across tweeter (as compared to without 3 ohm)  in the tweeter's passband which is about 5.5K and up. Except for -1.9dB drop at 20K from 6K to 16K the final attenuation is not even -1 dB let alone -3 dB.  And the intent here was to lower the Tweeter output by 3 dB compared to the woofer.

 

So with the amplifier giving the same voltage to the network, the insertion of 3 ohm has not shown any meaningful measured difference  in the tweeter's most of the passband.

  • Like 1
Link to comment
Share on other sites

2 hours ago, Patrick D HiFi said:

I have a pair of 1975 Heresy  with Type D crossover. The capacitors are brown plastic. Does anyone know if these are original? Seems most other Heresy crossovers of this vintage have oil can capacitors.

IMG_0817.jpg

IMG_0820.jpg

IMG_0822.jpg

I have seen these  rare  burgundy  capacitors before ,  the metal holder brackets and screws look to be made very specifically for these capacitors  for a clean look , are there any other  holes in the wood board 

 

Link to comment
Share on other sites

On 7/30/2022 at 9:43 AM, Scott Analog said:

So I have a pair of 1963 H series speakers.

Type HH12

The crossovers are identified as type 4 on the warranty card.

I have not been able to find any info about this crossover, and I would like to replace the caps.

Help

your speakers are either 1962 or 1963  based on the serial numbers , 

 

open the back covers and look at the wood board of the crossover , you should see markings that designate the crossover type  , a picture can help

Link to comment
Share on other sites

1 hour ago, Patrick D HiFi said:

Thank you, great to hear you have seen these before. I will check for other holes in the board and update

 look to see if there are any screws holes that were used ,  that would tell you whether screws and  PIO  capacitors were removed  , also look at the solder joints , are they clean enough to look like a factory job .

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...