Philharmonic Audio BMR Speaker Discussion

[Poseidons Voice]

One correction, Erin:

• Max SPL for 20Hz to 20kHz is approximately 99dB @ 1 meter. The compression threshold was exceeded above this SPL.
• Max SPL for 20Hz to 20kHz is approximately 102dB @ 1 meter. The compression threshold was exceeded above this SPL.

I think you mean 80Hz for the second line, not 20Hz.

Really Erin, awesome job. Take a bow. This is how reviews should be written. I particularly like that you had an extensive subjective evaluation along with PHOTOS of your room and a detailed room drawing. That helps with those who cry foul on objective measurements as being overly biased. You did both and you did both very extensively.

I suggest you include a panoramic video from your listening position. That would be cool to look at too.

Best,
Anand.

 

[hardisj]

One correction, Erin:

• Max SPL for 20Hz to 20kHz is approximately 99dB @ 1 meter. The compression threshold was exceeded above this SPL.
• Max SPL for 20Hz to 20kHz is approximately 102dB @ 1 meter. The compression threshold was exceeded above this SPL.

I think you mean 80Hz for the second line, not 20Hz.

Really Erin, awesome job. Take a bow. This is how reviews should be written. I particularly like that you had an extensive subjective evaluation along with PHOTOS of your room and a detailed room drawing. That helps with those who cry foul on objective measurements as being overly biased. You did both and you did both very extensively.

I suggest you include a panoramic video from your listening position. That would be cool to look at too.

Best,
Anand.

Ugh. I quit! LOL. Yes, I meant 80Hz. I'll fix that.


I am going to quit providing videos on my website and save that kind of thing for YouTube. Takes up a lot of server space. Cost me $20 last month when I added the 360° spin video of the speakers.

 

[B&WTube]

@Dennis Murphy
I built your BMR’s and love what you have done. I am in the process of tweaking my system, and was hoping to get your help on a passive high pass filter. I have dual binding posts, with the BMR/RAAL on one set of posts. I am wanting to bi-amp your speakers, and would love you advice on where to filter out the bass before it goes to the tube amp powering the mid/upper frequencies. Thanks in advance!

(Here are mine. I am building huge wall sconces for the speakers to get them apart, as we speak.)

 

[Dennis Murphy]

@Dennis Murphy
I built your BMR’s and love what you have done. I am in the process of tweaking my system, and was hoping to get your help on a passive high pass filter. I have dual binding posts, with the BMR/RAAL on one set of posts. I am wanting to bi-amp your speakers, and would love you advice on where to filter out the bass before it goes to the tube amp powering the mid/upper frequencies. Thanks in advance!

(Here are mine. I am building huge wall sconces for the speakers to get them apart, as we speak.)View attachment 77384

@Dennis Murphy
I built your BMR’s and love what you have done. I am in the process of tweaking my system, and was hoping to get your help on a passive high pass filter. I have dual binding posts, with the BMR/RAAL on one set of posts. I am wanting to bi-amp your speakers, and would love you advice on where to filter out the bass before it goes to the tube amp powering the mid/upper frequencies. Thanks in advance!

(Here are mine. I am building huge wall sconces for the speakers to get them apart, as we speak.)View attachment 77384

They're looking good. Did you build the cabinets from scratch, or do those use the flat pack? If you've wired the crossovers so there is no common ground between woofer circuit and the mid-tweet circuit, you should be good to go. Others might want to chime in on the necessary conditions for a bi-amp operation like this to work, but just from a crossover standpoint, if the woofer circuit is completely separate from the mid-tweet, you already have the proper low pass and high pass filters in place. Perhaps I'm not understanding your question?

 

[B&WTube]

They're looking good. Did you build the cabinets from scratch, or do those use the flat pack? If you've wired the crossovers so there is no common ground between woofer circuit and the mid-tweet circuit, you should be good to go. Others might want to chime in on the necessary conditions for a bi-amp operation like this to work, but just from a crossover standpoint, if the woofer circuit is completely separate from the mid-tweet, you already have the proper low pass and high pass filters in place. Perhaps I'm not understanding your question?

Thanks for your quick response. I bought the flatpack. I am a novice woodworker, which is why I used a hammered spray paint finish and wrapped the sides in leather...BTW, those nice looking leather trim pieces on each side of the middle leather are extra long purse straps from Amazon, with the hardware cut off, lol- but my friends all love it.

Yes, the crossovers are already wired separately as you have outlined-so they are good to go. I am wanting to build a passive, line level, high pass filter to remove those large bass/woofer frequencies before they are sent to the amp powering the mid/high. This way the amp is not trying to reproduce frequencies that the mid/tweeter aren’t going to play, anyways. The problem is that I’m not entirely sure what frequency to cut the signal at, and I don’t know how to do this with a line level signal. Here is an example, but I know the values aren’t right for your BMR’s

 

[R Swerdlow]

Yes, the crossovers are already wired separately as you have outlined-so they are good to go. I am wanting to build a passive, line level, high pass filter to remove those large bass/woofer frequencies before they are sent to the amp powering the mid/high. This way the amp is not trying to reproduce frequencies that the mid/tweeter aren’t going to play, anyways. The problem is that I’m not entirely sure what frequency to cut the signal at, and I don’t know how to do this with a line level signal. Here is an example, but I know the values aren’t right for your BMR’s View attachment 77413

Your question about a high-pass filter upstream from the amp that will drive the mid-range/tweeter is based on an incorrect idea. There will not be a sound quality benefit if voltages from frequencies below their pass band are absent.

Due to the passive crossover filters inside the BMR speaker, that amp will see high impedance below the woofer-to-mid crossover frequency. Similarly, the amp driving the woofer will see high impedance above that crossover frequency.

The signal voltages going into those amps may be unfiltered, but the amps will pass little or no current outside of the pass bands of the respective drivers. No current = no watts = no work.

 

[B&WTube]

Your question about a high-pass filter upstream from the amp that will drive the mid-range/tweeter is based on an incorrect idea. There will not be a sound quality benefit if voltages from frequencies below their pass band are absent.

Due to the passive crossover filters inside the BMR speaker, that amp will see high impedance below the woofer-to-mid crossover frequency. Similarly, the amp driving the woofer will see high impedance above that crossover frequency.

The signal voltages going into those amps may be unfiltered, but the amps will pass little or no current outside of the pass bands of the respective drivers. No current = no watts = no work.

That diagram I posted is from a different speaker designer (don't want to name names), but he is someone well known in the community and whose DIY speaker kits you can buy from Meniscus. When he explained it to layman me, it made perfect sense. Basically, an amp is not smart- if you give it a signal, it will amplify whatever is sent to it, full range or partial. If an amp does not have to create a portion of the signal (especially the larger low frequency waves) it can more precisely create the portion that it is responsible for. He also stated that it is not worth doing a low pass filter, due to the complexity and losses from the design, plus it is the smaller, high frequency, waves don't effect the bass as much as what happens in the inverse.

In a tube amp, with a tube rectifier (where you get bass sag), elimination of the lower frequency responsibilities is even more beneficial/important than it is on a SS amp.

 

[hex168]

Question on BMR drivers; folks in this thread may know:

If one were to use two conventional midrange drivers, one would get lobing above a frequency defined by the center-to-center distance. My understanding is that BMRs, above a certain frequency, do not radiate as a piston and that the wavefront is somewhat uncorrelated. Would lobing/comb filtering from two BMR drivers playing high frequencies be reduced compared to conventional drivers? Is it practical to double-up (or more) on BMR midranges to get the efficiency up, or is this a non-starter?

 

[R Swerdlow]

That diagram I posted is from a different speaker designer, and when he explained it to me, it made perfect sense. Basically, an amp is not smart

Yes, amps are not smart. You're right so far.

- if you give it a signal, it will amplify whatever is sent to it, full range or partial.

That depends on whether you are talking about voltage or current. Saying the word "signal" is not enough. If the amp tries to amplify voltage while the speaker/crossover combination creates high impedance, the current will be very low. Under those conditions – amplifying voltage while at the same time not amplifying the current – results with the amp not passing an amplified the signal to the speaker. When you are trying to drive loud speakers, high voltage and low current translates to no signal at all.

Think about a garden hose as an analogy to electricity. Water pressure inside the hose is analogous to voltage, and the amount of water flowing through the hose is analogous to current. Imagine there are two valves, one is between one end of the hose and the hose bib, and the other is a nozzle at the other end of the hose.

• If the hose bib valve is open, and the nozzle is also open, then water flows through the hose and out the nozzle.
• If the hose bib valve is closed, there is no pressure in the hose and no water flows through it.
• If the hose bib valve is open, but the nozzle is closed, there may be water pressure inside the hose, but no water flows through it, and no water comes out of the nozzle.

To control water flow you need one valve or the other, but not both.

Of course, analogies like this usually break down before long. Electricity flow can have both direct or alternating current, AC or DC. AC electric circuits have frequencies. Water flow through a hose doesn't depend on frequency. In an AC circuit, as we have with sound systems, there are capacitors and inductors which act as high-pass and low-pass filters. Nothing like those exist in garden hoses.

 

[B&WTube]

Yes, amps are not smart. You're right so far.
That depends on whether you are talking about voltage or current. Saying the word "signal" is not enough. If the amp tries to amplify voltage while the speaker/crossover combination creates high impedance, the current will be very low. Under those conditions – amplifying voltage while at the same time not amplifying the current – results with the amp not passing an amplified the signal to the speaker. When you are trying to drive loud speakers, high voltage and low current translates to no signal at all.

Think about a garden hose as an analogy to electricity. Water pressure inside the hose is analogous to voltage, and the amount of water flowing through the hose is analogous to current. Imagine there is both a valve between one end of the hose and the hose bib, and a nozzle at the other end of the hose.

• If the hose bib valve is open, and the nozzle is also open, then water flows through the hose and out the nozzle.
• If the hose bib valve is closed, there is no pressure in the hose and no water flows through it.
• If the hose bib valve is open, but the nozzle is closed, there may be water pressure inside the hose, but no water flows through it, and no water comes out of the nozzle.

To control water flow you need one valve or the other, but not both.

Of course, analogies like this usually break down before long. Electricity flow has both direct and alternating current, AC or DC. AC electric circuits have frequencies. Water flow through a hose doesn't depend on frequency. In an AC circuit, as we have with sound systems, there are capacitors and inductors which act as high-pass and low-pass filters. Nothing like those exist in garden hoses.

Thanks for taking the time to write up such a great description, for a novice, like myself. I would like to muddle through my limited knowledge base, if you will indulge... It is my understanding that the line level signal is both voltage and current, and basically voltage is what causes current, which is the force that moves the speaker (or is amplified). So, by limiting the voltage waveform with a high pass filter, the voltage and current from the source (a DAC in my case) is altered to a smaller, less dynamic, version/signal. This, in turn, makes the DC voltage in the amp easier to precisely control, and by default making the audio clearer...Now please don't hesitate to tell me how I am totally wrong, as long as you tell me why

 

[Dennis Murphy]

Question on BMR drivers; folks in this thread may know:

If one were to use two conventional midrange drivers, one would get lobing above a frequency defined by the center-to-center distance. My understanding is that BMRs, above a certain frequency, do not radiate as a piston and that the wavefront is somewhat uncorrelated. Would lobing/comb filtering from two BMR drivers playing high frequencies be reduced compared to conventional drivers? Is it practical to double-up (or more) on BMR midranges to get the efficiency up, or is this a non-starter?

I'm currently working on a design with two BMR mids in an MTM configuration with a RAAL ribbon, but I'll be crossing the mids at around 3800 Hz, and I'm assuming you would be using the BMR's full range into the high treble. For what it's worth, my full-range plot for the BMR's looks exactly like the plot for a single BMR when measured on the tweeter axis. I haven't taken off-axis plots yet. You can't run two BMR's in parallel to get sensitivity up because they're 4 ohm drivers. I'm running them in series to raise power handling but not sensitivity.

 

[hex168]

I'm currently working on a design with two BMR mids in an MTM configuration with a RAAL ribbon, but I'll be crossing the mids at around 3800 Hz, and I'm assuming you would be using the BMR's full range into the high treble. For what it's worth, my full-range plot for the BMR's looks exactly like the plot for a single BMR when measured on the tweeter axis. I haven't taken off-axis plots yet. You can't run two BMR's in parallel to get sensitivity up because they're 4 ohm drivers. I'm running them in series to raise power handling but not sensitivity.

 

[hex168]

Thank you! How about four, series/parallel? Or is that just asking for trouble? I have no idea to what extent a set of BMRs would act like one Distributed Mode loudspeaker, which seem to be able to radiate from a large panel without serious lobing. (Unless that is what causes DML's frequency response hash. I would love to see a DML on Amir's Klippel rig. However, in the few measurements I've seen of DMLs, they have higher distortion than the BMR drivers you use. OK, end of rabbit-hole-dive.)

 

[Dennis Murphy]

Thank you! How about four, series/parallel? Or is that just asking for trouble? I have no idea to what extent a set of BMRs would act like one Distributed Mode loudspeaker, which seem to be able to radiate from a large panel without serious lobing. (Unless that is what causes DML's frequency response hash. I would love to see a DML on Amir's Klippel rig. However, in the few measurements I've seen of DMLs, they have higher distortion than the BMR drivers you use. OK, end of rabbit-hole-dive.)

I've also built a monstrosity with 4 BMR's configured as you describe, and it sounds great, but it was designed with a specific vertical dispersion pattern. Again, I wasn't running them all that high for the mid-tweet cross.

 

[R Swerdlow]

Thanks for taking the time to write up such a great description, for a novice, like myself.

You're welcome.

If you're interested, you can read a better explanation of the water flowing in a pipe as an analogy to voltage and current in Wikipedia.

I would like to muddle through my limited knowledge base, if you will indulge... It is my understanding that the line level signal is both voltage and current, and basically voltage is what causes current, which is the force that moves the speaker (or is amplified).

Yes, all electricity has both voltage and current.

Ohm's Law says that electrical current flowing between two points of a circuit is directly proportional to the voltage across those same two points. I is the current in units of amperes, V is the voltage in units of volts, and R is the resistance of the conductor in units of ohms. Ohm's Law tells us how I and V can vary, but R is constant, and is independent of I and V. (All this is basic electronics, adapted from Wikipedia.)

Ohm's Law is a generalization from many experiments that showed that I is approximately proportional to V for most materials and conditions. Mathematically, Ohm's Law says I and V are directly proportional to each other. If you change V, I must also change. Similarly, if you change I, V must also change:

I = V/R​

Rearranging, you get:

V = I×R, or​

R = V/I​

To address your question, voltage does not cause current, nor does current cause voltage. They work together, in direct proportion to each other. If V goes up, so must I. And if I goes down, so must V.

To go back to your bi-amped BMR speaker, the mid-range driver's high-pass filter causes I to go down to nearly zero below the woofer-to-mid crossover frequency. If I goes down, V = IR tells us that V must also go down to nearly zero.

Neither V nor I, alone, are the force that moves a speaker. It takes both voltage and current. Power (P), expressed in Joule's Law, moves speakers. Power is V multiplied by I. Electrical power (in watts), is expressed mathematically as P = V×I. Substitute I×R for V (from Ohm's Law), and you get:

P = V×I = I×R×I = I²R​

So, by limiting the voltage waveform with a high pass filter, the voltage and current from the source (a DAC in my case) is altered to a smaller, less dynamic, version/signal. This, in turn, makes the DC voltage in the amp easier to precisely control, and by default making the audio clearer...Now please don't hesitate to tell me how I am totally wrong, as long as you tell me why

I'd comment, but I can't make any sense from what you said. Specifically, the part I marked in bold, is the part I don't understand.

All I can do is repeat what I said in an earlier post. The high-pass filter of the mid range driver in the BMR speaker is enough to block the voltages from lower frequencies than the woofer-to-mid range crossover frequency.

Blocking it twice, in a bi-amp set up, with two separate high-pass filters, one upstream from the amp and the other downstream, is unnecessary. At best, it might accomplish nothing, and do no harm. But it also might introduce unintended filtering or unwanted resonances that could alter or stomp on the excellent mid-range sound of these wonderful speakers. The best advice I can offer is, don't go there.

 

[B&WTube]

You're welcome.

If you're interested, you can read a better explanation of the water flowing in a pipe as an analogy to voltage and current in Wikipedia.
Yes, all electricity has both voltage and current.

Ohm's Law says that electrical current flowing between two points of a circuit is directly proportional to the voltage across those same two points. I is the current in units of amperes, V is the voltage in units of volts, and R is the resistance of the conductor in units of ohms. Ohm's Law tells us how I and V can vary, but R is constant, and is independent of I and V. (All this is basic electronics, adapted from Wikipedia.)

Ohm's Law is a generalization from many experiments that showed that I is approximately proportional to V for most materials and conditions. Mathematically, Ohm's Law says I and V are directly proportional to each other. If you change V, I must also change. Similarly, if you change I, V must also change:

I = V/R​

Rearranging, you get:

V = I×R, or​

R = V/I​

To address your question, voltage does not cause current, nor does current cause voltage. They work together, in direct proportion to each other. If V goes up, so must I. And if I goes down, so must V.

To go back to your bi-amped BMR speaker, the mid-range driver's high-pass filter causes I to go down to nearly zero below the woofer-to-mid crossover frequency. If I goes down, V = IR tells us that V must also go down to nearly zero.

Neither V nor I, alone, are the force that moves a speaker. It takes both voltage and current. Power (P), expressed in Joule's Law, moves speakers. Power is V multiplied by I. Electrical power (in watts), is expressed mathematically as P = V×I. Substitute I×R for V (from Ohm's Law), and you get:

P = V×I = I×R×I = I²R​

I'd comment, but I can't make any sense from what you said. Specifically, the part I marked in bold, is the part I don't understand.

All I can do is repeat what I said in an earlier post. The high-pass filter of the mid range driver in the BMR speaker is enough to block the voltages from lower frequencies than the woofer-to-mid range crossover frequency.

Blocking it twice, in a bi-amp set up, with two separate high-pass filters, one upstream from the amp and the other downstream, is unnecessary. At best, it might accomplish nothing, and do no harm. But it also might introduce unintended filtering or unwanted resonances that could alter or stomp on the excellent mid-range sound of these wonderful speakers. The best advice I can offer is, don't go there.

I greatly appreciate the detailed response. I agree with what you have said. Here is the part that I am still not understanding: If you give an amp less frequency to cover- how does that not make it much easier for the amp to do it's job? It seems like a waste to make an amp amplify and produce low frequencies that are then just going to be scrubbed by the speakers crossover. Cutting out the LF before the amp would make it more focused, by cutting out the large waves.

To your question: my understanding is you take AC wall power, becomes DC through the rectifier and then through the amplifying components. If you aren't amplifying those much more demanding LF waves (because that part of the signal is gone) the process is less demanding on the amplifying section and you get more precision in demanding songs (again, this is my layman understanding). Logically, it just seems like part of the work would be easier to do than all of the work.

Dennis I would still like to try this, just to see. If you can advise on where I might want to filter out the woofer in a line level cross- I would appreciate it.

 

[B&WTube]

The BMR's are great speakers, but I still like to experiment. I made some cabinet tweeks that made some marginal improvements, for anyone who cares (I'll leave out the failed ideas):
-Made 40 durometer sorbothane gaskets for the mid and tweeter to increase clarity on dynamic songs (hurt my bass response on the woofer)
-Used a sonic barrier from parts express (like dynomat) on the back of the baffle, and it really cut the resonances down (you can feel it with your hand)
-Extra layer of blue jeans in the mid range box helped slightly with tone and clarity (note major, but a little)

 

[Dennis Murphy]

I greatly appreciate the detailed response. I agree with what you have said. Here is the part that I am still not understanding: If you give an amp less frequency to cover- how does that not make it much easier for the amp to do it's job? It seems like a waste to make an amp amplify and produce low frequencies that are then just going to be scrubbed by the speakers crossover. Cutting out the LF before the amp would make it more focused, by cutting out the large waves.

To your question: my understanding is you take AC wall power, becomes DC through the rectifier and then through the amplifying components. If you aren't amplifying those much more demanding LF waves (because that part of the signal is gone) the process is less demanding on the amplifying section and you get more precision in demanding songs (again, this is my layman understanding). Logically, it just seems like part of the work would be easier to do than all of the work.

Dennis I would still like to try this, just to see. If you can advise on where I might want to filter out the woofer in a line level cross- I would appreciate it.

I'm afraid I have to background or experience in line level filtering. I can't model it in my design software, and I'm still not clear on why there should be any advantage to it, since the mid-tweet amp won't be called upon to amplify deep bass material. I don't have anything to add to what Richard has said on the subject. if there's a PhD electrical engineer out there, please chime in and comment on whether there could be any advantage to line level filtering. I misspent my youth and took political science (a true oxymoron) instead of EE.

 

[Dennis Murphy]

The BMR's are great speakers, but I still like to experiment. I made some cabinet tweeks that made some marginal improvements, for anyone who cares (I'll leave out the failed ideas):
-Made 40 durometer sorbothane gaskets for the mid and tweeter to increase clarity on dynamic songs (hurt my bass response on the woofer)
-Used a sonic barrier from parts express (like dynomat) on the back of the baffle, and it really cut the resonances down (you can feel it with your hand)
-Extra layer of blue jeans in the mid range box helped slightly with tone and clarity (note major, but a little)

Just a comment on the midrange chamber stuffing. The little square pads of EcoCore that Meniscus includes with the kit aren't sufficient to fully damp the BMR driver. The chamber needs to be stuffed fairly tightly and all the way to the top. And the speaker wire holes must be caulked to prevent any air leakage through the bottom, and thin gasket tape or rope caulk must be applied to the routing lip or driver frame to prevent any leakage out the top. Otherwise, the BMR will have a dip and a peak around 2 kHz. For non-kit types, the new BMR's and mini will be loaded on the container ship on Thursday, with arrival here in a little less than a month. I'll post details on my website.

 

[B&WTube]

Just a comment on the midrange chamber stuffing. The little square pads of EcoCore that Meniscus includes with the kit aren't sufficient to fully damp the BMR driver. The chamber needs to be stuffed fairly tightly and all the way to the top. And the speaker wire holes must be caulked to prevent any air leakage through the bottom, and thin gasket tape or rope caulk must be applied to the routing lip or driver frame to prevent any leakage out the top. Otherwise, the BMR will have a dip and a peak around 2 kHz. For non-kit types, the new BMR's and mini will be loaded on the container ship on Thursday, with arrival here in a little less than a month. I'll post details on my website.

Thanks, for the response! I did buy the full meniscus kit. I will repack the midrange cabinet. Would you recommend? I have on hand the meniscus stuff, and the following:
https://www.parts-express.com/sonic...ts-with-foil-155-sq-ft-315-x-181-4-p--268-040
https://www.parts-express.com/sonic-barrier-1-acoustic-sound-damping-foam-with-psa-18-x-24--260-525
I look forward to seeing your continued development. I may end up being interested in the new woofer you have coming out, which might be especially cool with the dual BMR's.

I already have the Mundorf Supreme caps in the tweeter. I am thinking of upgrading from the Audyn caps for the mid range and woofer crossovers. Do you have any recommendations? I was looking at the Audyn True Copper caps.