When I graduated from U of Nebraska with an Electrical Engineering degree in 1992, I interviewed at Jensen in Chicago. They had a job opening in aftermarket (“Jensen 6x9 Triaxes with blue surrounds”) and an opening in OEM (mass production speakers for Honda, Mazda, Chrysler, Ford etc). I took the OEM job and was super glad I did.
Taking the aftermarket job would have meant reporting to a product manager whose previous job was marketing washers and dryers at Whirlpool, answering questions like “How come our subwoofer doesn’t go up to 20KHz like Kenwood’s subwoofer does?”
The OEM job gave me a ridiculous range of latitude, so long as we kept within a few constraints. Mostly price, but also stipulations about bandwidth or distortion. “Design a 16cm door speaker. Cost has to be $4.15 or less. Distortion has to be under 2.5% at 1 watt. Has to handle 15 watts RMS for a week in 100% humidity. SPL at least 88dB. Holes and mounting require blah blah blah.”
And that would be about it and it was up to me to make it sound great.
And yes, a design engineer given a $17 budget for an economy car can take great pride in doing the absolute best job he can of making a silk purse out of a sow’s ear.
So for the 1996 Honda Civic they gave us a ridiculously low price requirement for 4x16cm speakers, something like $15.00 for all 4. $3.75 each. Their purchasing department was cutting the price to the absolute bone. Like, you couldn’t even afford a cone with a surround. You could only afford a one piece crappy cone like organ speakers from the 1960s.
I said “For $15.00 I can give you four crappy 16cm speakers… but what if we bend the rules a little bit?
“What if I give you two really nice 10cm speakers and a good 6x9 subwoofer? Three GOOD speakers will sound way better than four cheap speakers.”
So I built them a prototype system with 10cm / 4” speakers for the front with rising response, light voice coils, “edgy” sounding whizzer cones, and super tight suspensions to minimize excursion. Then I designed a dual voice coil 6x9, also with a huge whizzer cone, but with a slightly downward taper in the response because of its long high inductance voice coil and high Xmax, which added some midrange and treble fill for the rear package shelf but most of all had really good bass. WAY better bass than you’d get from 2x 16cm speakers.
(The irony is, my Honda Civic subwoofer actually DID go up to 20KHz. Or at least 15K anyway!!!)
Cherry on top: Total cost was $14.00 not $15.00. We saved them an entire dollar, in a business where prices are quoted to the hundredths of a cent.
The sounded FANTASTIC - considering it cost $14.00. They had lots of bass punch, great extension, and crisp high end and great imaging in front. No one noticed that there was only one speaker in back if they were sitting in the front seat, and they might not even notice when sitting in the back. With the fader dialed all the way to the rear, it was warm and creamy but still articulate and had sufficient high end.
[All "good" sound systems sound fantastic, CONSIDERING…. SOMETHING. “Your $100,000 Wilson speakers with Class A Mark Levinson Amps sound fantastic… considering we’re in your living room and not at the Chicago Symphony Orchestra... considering Miles Davis is dead and recordings are the only way you can listen to him now.” Compared to the live CSO the $300K reference system is a joke.]
We drove our 3 speaker prototype system with beefy dual voice coil 6x9s down to Marysville Ohio.
They said “We would be giving the customer only 3 speakers instead of 4.” I argued back, “But you can sell this as having a SUBWOOFER which you certainly can’t with the 16cm speakers.” Honda wasn’t willing to retool the metal to accommodate it and the people I was dealing with were bean counters. They didn’t buy it.
So that project never saw the light of day and they chose to use crappy 16cm speakers with one piece cones in the 1996 Civic, which I designed.
The thing I liked about that job, which my 1996 3-Speaker Prototype story illustrates, is:
My dear friend John McGee, every time I build a speaker, knows to ask me:
“What’s the constraint this time?”
Of course in DIY your constraints are a lot wider than if you're a manufacturer who has to put things in a box and sell them in a store and deal with ignorant consumers and warranty returns. I build lots of stuff and my constraints are self imposed. They set boundaries for the art project. Sort of like “Hey Homer, tell me a story of a Greek hero. You have to deliver the verses in Iambic Pentameter.”
In the Single-Woofer Dipole Reflex Hybrid, the constraint was:
“You only get ONE 10” woofer. You don’t get a 2nd 10” woofer. You don’t get any LF crossovers or passive radiators either. I want it to have that 3 dimensional warm ambient Open Baffle sound, but you also have to force it to deliver actual bass from a small box.”
This was the result.
It doesn’t have extra subwoofers or crossovers. Why? Because that would violate the constraint. It wouldn’t be elegant. Throwing drivers and technology and crossover points at problems is not disciplined thinking. It also gives you lousy aesthetics. It has low WAF. A speaker made out of gorgeous white oak is a lot less gorgeous after you install extra woofers.
Throwing technology at problems allows us to be lazy and not think very hard.
This forced me to think.
I was happy about that because it inspired me to invent (or re-invent) something that very few people have ever seen before.
EVERY speaker design is a compromise. You all know that.
In this exploration I discovered: One of the compromises of a single-woofer Dipole Reflex Hybrid is: You DO get the best of both worlds (Open Baffle + Reflex), but you also get cone rocking and high IM distortion at high SPL’s. The distortion curve goes out of control faster in this design than in a standard ported box.
OK. So it’s another version of trading distortion for efficiency and bandwidth. Which is another one of the many “Hoffman’s iron law” triads that we encounter in speaker design.
Taking the aftermarket job would have meant reporting to a product manager whose previous job was marketing washers and dryers at Whirlpool, answering questions like “How come our subwoofer doesn’t go up to 20KHz like Kenwood’s subwoofer does?”
The OEM job gave me a ridiculous range of latitude, so long as we kept within a few constraints. Mostly price, but also stipulations about bandwidth or distortion. “Design a 16cm door speaker. Cost has to be $4.15 or less. Distortion has to be under 2.5% at 1 watt. Has to handle 15 watts RMS for a week in 100% humidity. SPL at least 88dB. Holes and mounting require blah blah blah.”
And that would be about it and it was up to me to make it sound great.
And yes, a design engineer given a $17 budget for an economy car can take great pride in doing the absolute best job he can of making a silk purse out of a sow’s ear.
So for the 1996 Honda Civic they gave us a ridiculously low price requirement for 4x16cm speakers, something like $15.00 for all 4. $3.75 each. Their purchasing department was cutting the price to the absolute bone. Like, you couldn’t even afford a cone with a surround. You could only afford a one piece crappy cone like organ speakers from the 1960s.
I said “For $15.00 I can give you four crappy 16cm speakers… but what if we bend the rules a little bit?
“What if I give you two really nice 10cm speakers and a good 6x9 subwoofer? Three GOOD speakers will sound way better than four cheap speakers.”
So I built them a prototype system with 10cm / 4” speakers for the front with rising response, light voice coils, “edgy” sounding whizzer cones, and super tight suspensions to minimize excursion. Then I designed a dual voice coil 6x9, also with a huge whizzer cone, but with a slightly downward taper in the response because of its long high inductance voice coil and high Xmax, which added some midrange and treble fill for the rear package shelf but most of all had really good bass. WAY better bass than you’d get from 2x 16cm speakers.
(The irony is, my Honda Civic subwoofer actually DID go up to 20KHz. Or at least 15K anyway!!!)
Cherry on top: Total cost was $14.00 not $15.00. We saved them an entire dollar, in a business where prices are quoted to the hundredths of a cent.
The sounded FANTASTIC - considering it cost $14.00. They had lots of bass punch, great extension, and crisp high end and great imaging in front. No one noticed that there was only one speaker in back if they were sitting in the front seat, and they might not even notice when sitting in the back. With the fader dialed all the way to the rear, it was warm and creamy but still articulate and had sufficient high end.
[All "good" sound systems sound fantastic, CONSIDERING…. SOMETHING. “Your $100,000 Wilson speakers with Class A Mark Levinson Amps sound fantastic… considering we’re in your living room and not at the Chicago Symphony Orchestra... considering Miles Davis is dead and recordings are the only way you can listen to him now.” Compared to the live CSO the $300K reference system is a joke.]
We drove our 3 speaker prototype system with beefy dual voice coil 6x9s down to Marysville Ohio.
They said “We would be giving the customer only 3 speakers instead of 4.” I argued back, “But you can sell this as having a SUBWOOFER which you certainly can’t with the 16cm speakers.” Honda wasn’t willing to retool the metal to accommodate it and the people I was dealing with were bean counters. They didn’t buy it.
So that project never saw the light of day and they chose to use crappy 16cm speakers with one piece cones in the 1996 Civic, which I designed.
The thing I liked about that job, which my 1996 3-Speaker Prototype story illustrates, is:
“Here is a constraint. Do anything else you want to work within it, and make a great speaker given the resources you are given.”
That is the definition of great engineering.
That is the definition of great engineering.
My dear friend John McGee, every time I build a speaker, knows to ask me:
“What’s the constraint this time?”
Of course in DIY your constraints are a lot wider than if you're a manufacturer who has to put things in a box and sell them in a store and deal with ignorant consumers and warranty returns. I build lots of stuff and my constraints are self imposed. They set boundaries for the art project. Sort of like “Hey Homer, tell me a story of a Greek hero. You have to deliver the verses in Iambic Pentameter.”
In the Single-Woofer Dipole Reflex Hybrid, the constraint was:
“You only get ONE 10” woofer. You don’t get a 2nd 10” woofer. You don’t get any LF crossovers or passive radiators either. I want it to have that 3 dimensional warm ambient Open Baffle sound, but you also have to force it to deliver actual bass from a small box.”
This was the result.
It doesn’t have extra subwoofers or crossovers. Why? Because that would violate the constraint. It wouldn’t be elegant. Throwing drivers and technology and crossover points at problems is not disciplined thinking. It also gives you lousy aesthetics. It has low WAF. A speaker made out of gorgeous white oak is a lot less gorgeous after you install extra woofers.
Throwing technology at problems allows us to be lazy and not think very hard.
This forced me to think.
I was happy about that because it inspired me to invent (or re-invent) something that very few people have ever seen before.
EVERY speaker design is a compromise. You all know that.
In this exploration I discovered: One of the compromises of a single-woofer Dipole Reflex Hybrid is: You DO get the best of both worlds (Open Baffle + Reflex), but you also get cone rocking and high IM distortion at high SPL’s. The distortion curve goes out of control faster in this design than in a standard ported box.
OK. So it’s another version of trading distortion for efficiency and bandwidth. Which is another one of the many “Hoffman’s iron law” triads that we encounter in speaker design.
Last edited:
So, one design that I think would be elegant would be a 15” coax in a Dipole Reflex Hybrid. It would look a lot like these
https://psma-website-assets.s3.amazonaws.com/live_edge_beryllium_dipoles.pdf
..but minus the tweeter, just one driver on the front; the driver would be moved to the upper half of the baffle. It would have a back panel below the halfway point of the woofer. 40-50 liters.
I’m pretty sure I could get the tuning frequency down to around 40-45.
Since a 15” woofer has a surface area of 880 cm^3 and I’m using half the surface area to drive the ported box, that means I’ve got 440 cm^3 of Sd (equal to an 11 inch woofer in a bass reflex) so I can get 100dB SPL with only a few mm of cone excursion.
And it would be a very beautiful design. With high WAF. And the gorgeous ambience of Open Baffle, all at the same time.
And relatively simple and won’t require a 17dB bass boost at 40Hz to sound right.
Very tempting to build them, using B&C 15CXN88 or the BMS drivers that @airvoid mentioned, or the Faital 15HX500.
Or maybe someone on this forum would enjoy tackling it.
https://psma-website-assets.s3.amazonaws.com/live_edge_beryllium_dipoles.pdf
..but minus the tweeter, just one driver on the front; the driver would be moved to the upper half of the baffle. It would have a back panel below the halfway point of the woofer. 40-50 liters.
I’m pretty sure I could get the tuning frequency down to around 40-45.
Since a 15” woofer has a surface area of 880 cm^3 and I’m using half the surface area to drive the ported box, that means I’ve got 440 cm^3 of Sd (equal to an 11 inch woofer in a bass reflex) so I can get 100dB SPL with only a few mm of cone excursion.
And it would be a very beautiful design. With high WAF. And the gorgeous ambience of Open Baffle, all at the same time.
And relatively simple and won’t require a 17dB bass boost at 40Hz to sound right.
Very tempting to build them, using B&C 15CXN88 or the BMS drivers that @airvoid mentioned, or the Faital 15HX500.
Or maybe someone on this forum would enjoy tackling it.
Last edited:
Hi,
I'm looking into making an open baffle. I just order 2 15ob350. And I might take some 15CXN88 to make a "mini" Bitches Brew with a single bass woofer. I just don't know if it will ok to pressurize a 65m2 living room. BUT I have 2 8nmb20 and tw29txn drivers that I like a lot. So maybe I will try to make something with all thoses drivers ....
Any way, your solution remind me a design from a french company that make an hybrid Bass-reflex/ very short transimission line.
They made a patent for this. The "transimission line" is only about 30cm and about SD area. Removing the pipe thay put on the magnet would make a rather big area and would be comparable to a U frame.
The advantage would be to be able to control the "open" section between the OB and the BR
I'm looking into making an open baffle. I just order 2 15ob350. And I might take some 15CXN88 to make a "mini" Bitches Brew with a single bass woofer. I just don't know if it will ok to pressurize a 65m2 living room. BUT I have 2 8nmb20 and tw29txn drivers that I like a lot. So maybe I will try to make something with all thoses drivers ....
Any way, your solution remind me a design from a french company that make an hybrid Bass-reflex/ very short transimission line.
They made a patent for this. The "transimission line" is only about 30cm and about SD area. Removing the pipe thay put on the magnet would make a rather big area and would be comparable to a U frame.
The advantage would be to be able to control the "open" section between the OB and the BR
I love this quote! Its key to minimising the gross time domain errors inherent in all electro mechanical transducers.
The more accurate the time domain performance the more life like to sound.
Hum I see there is no link in my previous post.
The enclosure type is called Solen HEL, there is an english page that explains how it works on there website.