Modelling compression drivers

[philpope]

DE250

The above is quoted from Audio Psychosis • View topic - Simulating a B&C DE25

There are only two things I don't agree with in Patrick Bateman's (John) DE25 parameters. I've actually measured the Sd of a DE25 diapgragm and know it is 18.09557cm^2. The second thing is there is no way BL can be 18 T*m. It has to be somewhere between 7 T*m and 9 T*m. I do agree that Patrick's numbers for Fs, Qes, and Mmd are probably better than mine - Hey this was early on and we've learned so much since then.

I think Mms is about 1g. seen specs for a JBL 4" diaphragm that is 3g. area is a quarter but the voice coil weight is probably only half so 1g seems reasonable. surely Fs is between the impedance peaks at 720Hz and 1900Hz so 1170Hz seems a good guess for Fs.

The impedance and response of the model is close to published for the ME25 horn but have you tired modelling conical horn. HF goes through the floor - I think it has something to do with correctly specifying the radiation cone but not too sure. Any ideas?

cheers
Phil

 

[JLH]

HF goes through the floor - I think it has something to do with correctly specifying the radiation cone but not too sure. Any ideas?

cheers
Phil

That's the nature of the conical horn. Better directivity control, but falling high frequency response. Use of a low pass filter set to around 15KHz to 18KHz brings the high end back up.

 

[whgeiger]

Note on Sd

Beware, [Sd] is not the surface area of the diaphragm of a compression driver, it is [Vd]/[Xmax] or the projected area of an equivalent piston. I have seen this mistake made in some of the papers published in the AES on the subject of phase plugs as well as here.
Regards,
WHG

 

[gedlee]

surely Fs is between the impedance peaks at 720Hz and 1900Hz
cheers
Phil

Phil

This isn't necessarily the case if you are talking about the free air resonance of the diaphragm. Its a coupled acoustic system and none of the peaks or dips correspond directly to the diaphragms mechanical parameters. Its identical to a woofer in a ported box - neither peak is the drivers free air resonance and neither is it always right in the middle either.

I have found that blocking the throat with the back off can give a good idea of the diaphragms resonance since the front air is now simply a closed box compliance. If the rear box cannot be taken off then its just the compliance of the two springs added to the drivers compliance.

 

[philpope]

conical horn response

That's the nature of the conical horn. Better directivity control, but falling high frequency response. Use of a low pass filter set to around 15KHz to 18KHz brings the high end back up.

well that was what I thought originally but seen all these plots of unity horns flat up to 16kHz with no mention of corrective eq or LPF. I can see how the LPF would work but think I have worked out what others have been doing

My Unity Horn Speakers

if you look at the crossover pictured on the above link it appears to be at 3rd order HPF filter with a zobel and a notch filter. but the ratio of the capacitors is not 3:1 as is usual and the values of the capacitors and coil are way too small for a 1.2khz crossover point into a 16R load. so my guess is that it is a non-standard alignment crossed at about 15kHz to give an acoustic crossover of 1.2kHz. will give this a go in akabak and see how it works out

 

[JLH]

well that was what I thought originally but seen all these plots of unity horns flat up to 16kHz with no mention of corrective eq or LPF. I can see how the LPF would work but think I have worked out what others have been doing

My Unity Horn Speakers

if you look at the crossover pictured on the above link it appears to be at 3rd order HPF filter with a zobel and a notch filter. but the ratio of the capacitors is not 3:1 as is usual and the values of the capacitors and coil are way too small for a 1.2khz crossover point into a 16R load. so my guess is that it is a non-standard alignment crossed at about 15kHz to give an acoustic crossover of 1.2kHz. will give this a go in akabak and see how it works out

In that crossover the 3.3uF cap and 100 ohm resistor provide most of the high frequency "gain". It’s actually not providing gain, but providing a path for the high frequencies to get around the 100 ohm attenuation resistor. I like using the low pass filter because the coil size is really small and easy to make yourself, and the cap is a shunt element that doesn’t impart its signature on the sound.

 

[Kindhornman]

GedLee, Not so sure that what you are proposing with a blocked throat would at all give you the free air resonance of the diaphragm.

As you stated:"This isn't necessarily the case if you are talking about the free air resonance of the diaphragm. Its a coupled acoustic system."

You wouldn't measure the free air resonance of a driver in an enclosure and why would you do that for a compression driver? I would think that perhaps if you knocked the phasing plug out of the device in question you could come close, but you would still have the loading of the throat section and if that was removed you would still have the blockage of the magnetic motor changing the air loading on the diaphragm. As you stated the response and loading on the compression driver is going to change with every different waveguide that is mounted in front of the driver. The air column loading is going to shift the loading dependent on the flare rate and geometry.

For those who want to understand the response of these drivers I would suggest a plain wave tube to understand and analyze the response function of these devices. Not easily made but the way it is done.

 

[Patrick Bateman]

GedLee, Not so sure that what you are proposing with a blocked throat would at all give you the free air resonance of the diaphragm.

As you stated:"This isn't necessarily the case if you are talking about the free air resonance of the diaphragm. Its a coupled acoustic system."

You wouldn't measure the free air resonance of a driver in an enclosure and why would you do that for a compression driver?

<snip>

The reason why you would do this is simple. You can't remove the phase plug from the equation without trashing the compression driver.

Removing the back from the compression driver and measuring the impedance isn't ideal, but it gets you 'closer to the ballpark' then you would if you simply measured the impedance on a plane wave tube.

On a plane wave tube you have a lot more variables. You have the sealed enclosure on the back side of the compression driver. You have the mass of air in front of the diaphragm and the effect of the phase plug. You have the mass of air in the plane wave tube. You have resonances due to the quarter wave resonance of the plane wave tube.

That's a ton of variables!

Doing what Geddes suggest basically eliminates half of those variables. More importantly, you can calculate the air mass with a tape measure.

 

[Kindhornman]

Patrick Bateman,
With the 10/1 ratio common with a compression driver and the rarefaction of the air directly adjacent to the diaphragm along with the length of throat in the driver, what number are you really getting and how do you propose to use that Fs if you could accurately measure it. I was not suggesting that anyone measure the fs in a plain wave tube. I was suggesting that if you wanted to study the behavior of the driver that would be a better place to start than on any waveguide without the multiple changes that can create in response curves without doing polar response curves.

 

[philpope]

getting better

In that crossover the 3.3uF cap and 100 ohm resistor provide most of the high frequency "gain". It’s actually not providing gain, but providing a path for the high frequencies to get around the 100 ohm attenuation resistor. I like using the low pass filter because the coil size is really small and easy to make yourself, and the cap is a shunt element that doesn’t impart its signature on the sound.

doubled the mouth size and narrowed the dispersion angle a bit which has helped the top end sensitivity. still not getting what I want fro the DE250 so prob going to use the BMS 4550 instead which is speced as 113dB on a 90/75 horn and still does 105dB at 15kHz. can't find anything more sensitive without going dual voice coil.