AR-11 Crossover Frequency Response Simulation

[Pete B]

I measured the input impedance of the AR-11 drivers many years ago.  Recently remeasured the midrange

and decided to take a look at the crossover frequency response in simulation for each of the switch positions.

I used the AR inductor values including DC resistance and estimated the NPE midrange cap ESR at .1 ohms,

the simulator allows these values to be included in the component model but I decided to show it for the caps

and not for the inductor:

S1, red, is the 0 dB or max output switch position, it is not very flat and there is peaking in the network response 

around 3KHz.  Curve S2, yellow, is the -3dB position and shows a reasonably flat response.  S3, black, is the -6dB

position with a very flat response and quite a bit of droop below 1 KHz.  I thought that the -6dB position sounded 

best but would have preferred something between -6 and -3dB.

S4, brown,  shows the response of the AR-3a network set to the dot position, the .8 ohm series resistance is the

.5 ohm resistor with .3 ohms added for ESR of the large wax box capacitor.  This was my estimate and I found a

value of .27 ohms in the AR-3a restoration manual - close enough.

The impedance curve shows the input impedance of the midrange driver alone which is important since it

loads the crossover network.

 

[Pete B]

Here is a larger zoom of the frequency response curve:

Note that the rolloff, as expected, is about 12dB/oct around 200 Hz, but it approaches 6 dB/oct

well below 100 Hz, this is due to the DC resistance in the .88 mH shunt inductor.

 

[Pete B]

Here is a simulation of the AR-11 mid network S1 is the normal .88 mH inductor with .35 ohm DCR,

S4 is with a .88 mH with only .01 ohms DCR, both are 0dB level.  The LF rolloff can be seen to

maintain the slope below 100 Hz with lower DCR for the inductor:

 

[fedeleluigi]

1 hour ago, Pete B said:

Recently remeasured the midrange

Hello Pete, 

Thank you for sharing all these simulations.

Did you measure the anecoic frequency response of the midrange?

If you didn't, what midrange frequency response did you use to make these simulations?

Thank you

Luigi

[Pete B]

No, I did not measure the acoustical response and it is assumed to be completely flat

and this is why the responses are the electrical crossover only.  The total response is the cascade

(complex multiply) of the electrical and acoustical responses.  The acoustical response is fixed

so this is the change to be expected between settings.

 

[fedeleluigi]

Thank you for the clarification, Pete

 

[roundhome]

Pete  half of my LST speakers have been salvage from the dreaded strippers of drivers to sell on ebay I now have enough AR 11 tweeters and mids also AR eleven crossovers  do you see a problem with splitting the crossovers and bi amping the speakers

[Pete B]

So if I'm reading you correctly, you are building LSTs with AR-11 tweeters and crossovers?

I don't see any problem at all setting them up for bi-amp input.

Might be better to start a new thread in the mods section if you want to have 

an ongoing discussion about them.

Some info from Mr. Allison himself:  http://www.classicspeakerpages.net/IP.Board/index.php?/topic/4198-roy-allison-ar-3a-ar-lst-mark-levinson-cello-amati-speakers/

 

[fedeleluigi]

On 7/16/2018 at 6:20 PM, Pete B said:

S4, brown,  shows the response of the AR-3a network set to the dot position, the .8 ohm series resistance is the

.5 ohm resistor with .3 ohms added for ESR of the large wax box capacitor.  This was my estimate and I found a

value of .27 ohms in the AR-3a restoration manual - close enough.

Pete,

I can't understand why you put the ESR values before and not after the capacitors (40 uF for the AR 11 and 50 uF for the 3a). This changes the load (zero Ohm vs ESR value) seen by the capacitors . Doesn't it?

Anyway, as regard the 3a,  shouldn't the 0.5 Ohm resistor be situated between the 50 uF cap and the inductors?

thank you for any clarification.

Luigi

 

[Pete B]

Hi Luigi,  In a series leg resistors can be added and moved to any location, the transfer

function remains the same.  I remembered the 40 uF in the AR-11 as a large can type 

that are usually very low ESR and that is why I used .1 ohm, if it was the more typical

type I would have used .3-.5 ohms.

R7 in the AR-3a section is the .5 ohm resistor plus .3 ohms for the ESR of the large paper

box type cap.  .3 ohms was a guess but I found .27 R in the rebuild guide which suggests

that .3 is close enough.

[Pete B]

Just opened up my AR-11's and noticed that the 40 uF is not a large can type rather

it is a cheap Callins so I should redo this with .3 ohms for the ESR.  I should probably 

measure the Callins but who knows how far the ESR has drifted since they were new.