[Rick Craig]

The stators are about 7mm apart, the diaphragm thickness is about a third of a millimeter.  Tensioning the diaphragm is most critical to proper performance.  Over the years we have seen buzzing from three sources:

1. Loss of tension in the diaphragm, either from stretching or glue coming loose.  Relatively infrequent.

2. Magnets dislodged during transport.  A loose magnet can touch the diaphragm causing problems. Most common failure mode.

3. Thermal failure of trace or diaphragm, caused by overpowe ...

Actually I wasn't talking about reliability but distortion products. Planar drivers like that can take a certain amount of abuse but pushing them that low (166hz) with a shallow slope really isn't a good idea whether you have one or ten of them.

[Rick Craig]

Brian, how much amplitude error and lobing anomolies due to combfilter induced cancellation at the crossover transition point is allowable before you consider it objectionable? Which is worse the phase error due the amplitude error or the phase inversion resulting from a reversed driver? Also do you have any information on distortion products as a function of drive level in the form of graphs showing harmonic and intermodulation distortion with regards to frequency?  Thanks, Scotty

You hit the nail on the head  The only first-order filters that really follow the theory are quite complex (ever see a Thiel crossover?). Most first-order advocates (Brian isn't the only guilty one) don't execute the filters so that the "advantages" are realized. A simple 1st order electrical combined with a driver will result in acoustic slopes that usually have poor phase behavior, lobing problems, off-axis response peaks / nulls, and inadequate driver protection. Studies have shown these to be much more audible than the "benefits" of first order slopes.

[Danny Richie]

Actually I have to agree with Rick (and this is a rarity coming from me).      

The only first-order filters that really follow the theory are quite complex (ever see a Thiel crossover?). Most first-order advocates (Brian isn't the only guilty one) don't execute the filters so that the "advantages" are realized. A simple 1st order electrical combined with a driver will result in acoustic slopes that usually have poor phase behavior, lobing problems, off-axis response peaks / nulls, and inadequate driver protection. Studies have shown these to be much more audible than the "benefits" of first order slopes.

Hey I even saw on axis response peaks and nulls...

Scotty, good questions again. Somehow though I don't think you'll be getting any answers.

[TheChairGuy]

This discussion is teeter-tottering into nasty...please keep it civil and control the cat swipes - whatever your reason for them.

I recognize that for several of the participants of the Topic, you are engaged in speaker-building as a livelihood...and competition is fierce.

Most of us reading, however,  are enthusiasts interested in learning a bit from industry experts....opinions and facts, both.  I know it's difficult to stay neutral when you have differing opinions with another competitor/speaker-builder....but try to state facts and opinions without swiping.

Purdy please  

John / Moderator

[morricab]

I have been working with the smallest RD driver (the 28.1 which now sadly seems to be discontinued).  I have found that this driver equalizes up nicely with some digital eq action.  Also i have found that a 2nd order filter at 300Hz gives a very nice blend with the woofer and doesn't seem to suffer greatly from compression issues due to lack of driver travel.  Also someone mentioned that the big one measure only about 83db but I am finding more like 86-87 db from my ribbons.  The highs can be compensated for nicely with eqing and of course so can the resnonace.  When these items are addressed the sound is quite fantastic.  I am even finding it equivalent or better in many ways to my Apogees.

[Jon L]

I have been working with the smallest RD driver (the 28.1 which now sadly seems to be discontinued).  I have found that this driver equalizes up nicely with some digital eq action.  Also i have found that a 2nd order filter at 300Hz gives a very nice blend with the woofer and doesn't seem to suffer greatly from compression issues due to lack of driver travel.  Also someone mentioned that the big one measure only about 83db but I am finding more like 86-87 db from my ribbons.  The highs can be compensated fo ...

Hmm.  Which digital EQ are you using and what kind of boost are we talking about?  3-6dB at 16-20kH or..  Also are you using digital EQ to act as notch filter instead of passive x-over notch filter?

[ekovalsky]

The current VMPS high pass filter to the planars is a second order design -- signal passes through a bank of caps then through an inductor shunted to the ground.  Not sure what the frequency is, probably around 166 hz.

The low pass filter to the planars is first order -- a hand wound iron core one at that    Based on the measured combined tweeter (also on a second order filter) and midrange response there do appear to be some response anomalies.  

I have already bypassed the planar high pass filter and replaced with DSP 6th order at 250hz in the TacT RCS.  Very nice improvement!  I will probably bypass the low pass coil also and implement it via DSP in the TacT S2150.  Hopefully this will flatten the measured combined response.

By the way the TacT is really great for drivers needing a notch filter or that have other response quirks like some of the BGCorp products.

[tkp]

The current VMPS high pass filter to the planars is a second order design -- signal passes through a bank of caps then through an inductor shunted to the ground.  Not sure what the frequency is, probably around 166 hz.

The low pass filter to the planars is first order -- a hand wound iron core one at that    Based on the measured combined tweeter (also on a second order filter) and midrange response there do appear to be some response anomalies.  

I have already bypassed the planar high pass fil ...

Eric,

You lost me on this one.  The planars have both high pass and low pass filters to them?  Did you mean highpass filter to the planars/tweeter first and then lowpass filter to the plannars only to create a bandpass filter for the plannars?

[John Casler]

To add an update, I think B, mentioned that he no longer X's at 166Hz. I don't think he has for some time.  Most all x-overs were raised higher around the time when the RM30 was introduced.

I think all are over 200Hz.

[ekovalsky]

To add an update, I think B, mentioned that he no longer X's at 166Hz. I don't think he has for some time.  Most all x-overs were raised higher around the time when the RM30 was introduced.

I think all are over 200Hz.

My pair pre-dates the RM30.  The change I remember reading about involved the low pass filter on the mids, changing from 7khz to 6.9khz or something like that.  It could be upgraded in the field with a coil swap.  I didn't realize the high pass filter had been changed on non-RM30 models, but if it has that is a good thing!

I think the RM30 was designed with the higher bass-mid frequency since it has smaller woofers that are better above 200hz.  The 10" non-Megawoofers pretty much give up above 350hz.  The Megawoofers probably only work below 100hz or so.

[ekovalsky]

Eric,

You lost me on this one.  The planars have both high pass and low pass filters to them?  Did you mean highpass filter to the planars/tweeter first and then lowpass filter to the plannars only to create a bandpass filter for the plannars?

Coming from the binding post, there is first a high pass filter (eliminating signal below 166hz, or whatever) which then goes to the L-pad, the output of the L-pad then goes through a coil which is a low pass filter (eliminating the signal above 6.9khz or so).  The two filters in series act like a bandpass but with different slopes (2nd order high pass then 1st order low pass).

Anway, I now have some expensive caps and a two cheap coils going unused in my speakers  

[audioengr]

In short, the spacing between your drivers (center to center) is relative to how high they are allowed to play. The spacing must be shorter than the wavelengths that they are playing to avoid comb filtering. Say your center to center spacing is 5.25", then comb filtering will be present down to as low as 2,560Hz. If the crossover point is below that then you are home free..

]

So, Why wouldn't I just butt the drivers against each other(midrange drivers)?  Would this not give me the highest crossover point?

Or do they need to be spaced further for another reason?

Steve N.

[Danny Richie]

So, Why wouldn't I just butt the drivers against each other(midrange drivers)? Would this not give me the highest crossover point?

Or do they need to be spaced further for another reason?

Yes, mounting them as close as possible keeps comb filtering effects higher up in the frequency range. There is really no reason to space them further apart.