My Lockdown Project

[dcathro 272]

My system is down, my speakers are gone, and my room is a mess!

 

 

I am using the lockdown to finish a project that I started years ago: my partial open baffle speaker system.

 

The idea being to have a high efficiency speaker based around the Supravox 215 widebander in an open baffle, supported by a sealed 15" bass and a helper tweeter.

 

The speakers were started many years ago, but never completed.

 

The drivers are listed below:

 

 

 

The Wolf Von Langa A150 15" bass field coil

 

 

The Supravox 215 EXC field coil wide bander

 

 

The Seas Exotic T35 tweeter:

 

 

 

The bass is in a sealed and brace 130 litre cabinet, while the Supravox and tweeter are mounted on a baffle - more pics to follow.

 

I have been using omnimic to do measurements, and Xsim to model the crossover.

 

I have then been listening to the one speaker, adjusting the crossover then remodeling.

 

here is the crossover schematic snapshot from Xsim (S1 tweeter, S2 Widebander, S3 Bass)

 

Here is the current modelled frequency response 1/3rd octave:

 

 

Here are a couple of closeup pics

 

 

 

More to follow.

[dcathro 272]

Here are the individual driver responses (modelled in Xsim)

 

[mwhouston 4,891]

Great to see you have dug your teeth into a lockdown DIY project. Now you have to finish it before the pandemic is over. I’m watching. 

[davewantsmoore 6,502]

10 hours ago, dcathro said:

Here are the individual driver responses (modelled in Xsim)

What are your target responses? (and their sum?)

[dcathro 272]

10 hours ago, mwhouston said:

Great to see you have dug your teeth into a lockdown DIY project. Now you have to finish it before the pandemic is over. I’m watching. 

Thanks Mark,

 

Hopefully this will keep me sane.

[dcathro 272]

35 minutes ago, davewantsmoore said:

What are your target responses? (and their sum?)

Hi Dave,

 

The bass drivers have an Fs of 36hz and sensitivity of 97db, but in the sealed box I loose some sensitivity to have an F3 of 40hz and usable bass in room down to 30hz.

 

The 35mm tweeter's response drops off at 15Khz and is 5db down by 20Khz.

 

The Supravox widebander has reasonably flat output from about 50hz to about 5Khz, where it then drops off about 15db. The open baffle rolls off the low end somewhat, although I am getting significant overlap with the bass driver in the mid bass / lower midrange. I am pushing the top end down a bit to get a better integration with the tweeter.

 

I am doing this iteratively - measuring/modelling/listening to get a flat  response, whilst also making subjective judgment calls. 

 

The summed frequency response modelled from the raw driver measurements is in the first post.

 

Cheers

 

David

[davewantsmoore 6,502]

2 hours ago, dcathro said:

The summed frequency response modelled from the raw driver measurements is in the first post.

Yes.... sorry if the question wasn't clear.      That's what you've got now.  I was asking about where you are trying to get to.

 

I have assumed that the phase response above 1khz is not intentional.

 

[dcathro 272]

1 hour ago, davewantsmoore said:

Yes.... sorry if the question wasn't clear.      That's what you've got now.  I was asking about where you are trying to get to.

 

I have assumed that the phase response above 1khz is not intentional.

 

I am relatively happy with the frequency response, although we will see if it changes after some more experimentation. The phase above 1k is not something that I quite understand yet, given that this is the Supravox widebander with nothing that should cause any large phase swings. 

[davewantsmoore 6,502]

1 hour ago, dcathro said:

The phase above 1k is not something that I quite understand yet

All three drivers are playing in that range.

The phase warping indiatcates that they are not moving in unison with each other.

[dcathro 272]

1 hour ago, davewantsmoore said:

All three drivers are playing in that range.

The phase warping indiatcates that they are not moving in unison with each other.

I'm not so sure. When I take other 2 drivers out of the simulation leaving only the widebander, the phase issues don't change at all

[davewantsmoore 6,502]

57 minutes ago, dcathro said:

I'm not so sure. When I take other 2 drivers out of the simulation leaving only the widebander, the phase issues don't change at all

I see.

 

In that cas.... You have captured "corrupted" data for your driver.    This could be due to data which includes reflections from the room, etc.

 

If you're confident in the frequency response ... then you could extra the "minimum phase" from it.   (ie. calculate the phase). 

 

This would take away the phase rotation between 1 and 3khz.... and show you the real phase.     The you can design a crossover where the (actual) phases of the 3 drivers line up in time (and so all 3 are moving together in unison).

 

Similarly at ~200Hz.

Edited August 9, 2020 by davewantsmoore

[dcathro 272]

I measured the drivers at 1 meter, with the speaker pretty close to room boundaries. I am going to measure it all again, as I have modified the baffle and changed the field coil strength since my measurements.

 

Here is the manufacturer provided impedance and phase response for the driver:

 

Edited August 9, 2020 by dcathro

[dcathro 272]

Got both speakers up and running.

 

Some more pics:

 

[dcathro 272]

Have made some good progress this past week, mainly working on the crossover. The overlap of the wide bander with the bass did not work, so have pulled the top of the bass down more and also the bottom of the Supravox.

 

 

[dcathro 272]

One issue that I have is that the resistor infront of the Supravox is having to dissipate a lot of power. It is the black curve in the following graph. An alternative option is to use an autoformer, but this reduces the overall impedance down to between 2 and 3 ohms in a few spots. If I proceed with the resistor approach, I will need to parallel up a few power resistors.

[dcathro 272]

I am getting close to finishing the crossovers. I have put in about 50 hours of listening, and remodeling in the past few weeks. Most of the component values have changed, and I have added a little RC filter across the bass inductor to tame the bump at 450 hz and allow a shallower bass roll off. I can probably spend another 20 hours on fine tuning before I move on to other work.

 

Here is the current circuit:

 

 

The driver contours are now like this:

Here is the current overall response (3rd octave) :

 

The impedance and phase are relatively benign,  with an average impedance between 4 and 7 ohms and minimums of 3.7 ohms at 65hz and 3.4 ohms in the high treble. The phase angle has a maximum of 16 deg at 160hz and a minimum of 23 deg at 5.5Khz

 

I have to say that I am pleased with the result, although I would love a bit more efficiency ( I estimate I am getting around 91 db) and deeper bass

 

When I have finished fine tuning the crossover, my work list includes:

• some work on the current power supplies for the electromagnets;
• improving the diffraction around the open baffle area

Down the line, I would like to build some nicer looking open baffles, and also build some tungar power supplies for the Supravox's.

[dcathro 272]

35 minutes ago, dcathro said:

I have to say that I am pleased with the result, although I would love a bit more efficiency ( I estimate I am getting around 91 db) and deeper bass

I should add that I could have had an extra 3-4 db efficiency plus a bit more bass with a ported bass. I just prefer sealed.

[mwhouston 4,891]

2 hours ago, dcathro said:

I should add that I could have had an extra 3-4 db efficiency plus a bit more bass with a ported bass. I just prefer sealed.

Sealed for me too. 

[Red MacKay 1,198]

Jeez - reading all this I think it best I not even try to build speakers!

 

[mwhouston 4,891]

36 minutes ago, Red MacKay said:

Jeez - reading all this I think it best I not even try to build speakers!

 

Qwazy, do it. 

[dcathro 272]

On 20/09/2020 at 7:43 PM, Red MacKay said:

Jeez - reading all this I think it best I not even try to build speakers!

 

But Red, the more I work on these the more you are going to hate them

[Melon_Head 15]

Wow, you must be kicking yourself for not finishing these speakers sooner.

 

I built some open baffles about 7 years ago, I could never go back to boxed speakers.

I even went open baffle in the bass but I have dual 15" drivers taking care of things there.

[dcathro 272]

1 minute ago, Melon_Head said:

Wow, you must be kicking yourself for not finishing these speakers sooner.

 

I built some open baffles about 7 years ago, I could never go back to boxed speakers.

I even went open baffle in the bass but I have dual 15" drivers taking care of things there.

I didn't have a room that would allow them to work properly, or allow me to get the crossover correct. My old room had every problem you could ask for:

• one of the side walls all thin glass windows;
• other side having only a half height wall - with kitchen on the other side
• missing corners with stairs going down to garage and up to bedrooms
• a sloping ceiling.
• an opening in the middle of the room leading to the kitchen/dining room, where I measured 30db suckouts at certain frequencies!

We moved house 18 months ago which allowed me to get my system working OK.

[Melon_Head 15]

Oh yeah.  You're gonna love it.

 

If the crossover gives you headaches, just leave off the tweeter and enjoy it with bass driver and FR for a while.  Then come back to the tweeter later.

 

On mine I think I went 2nd Order Low pass on the Bass and first order High pass on the FR but separated the poles one octave apart, eg 120Hz 2nd Order LP Woofer, 240Hz 1st Order HP FR.  It works beautifully like that.

I have adjustable active crossovers so it was fairly easy dialing it in.

Edited September 21, 2020 by Melon_Head

[dcathro 272]

Have you got a thread with some pictures?

[Melon_Head 15]

No thread.

I am new here.

At some point I guess I should do a thread.

[dcathro 272]

Yes please!

 

What drivers did you use?

[Melon_Head 15]

I used Full Range drivers from Audio Nirvana USA.

I am extremely happy with them.

I have some Seas FR drivers too, but I much prefer the Audio Nirvana Drivers

 

When this bloody lock down is over, I guess I could bring some drivers around for you to compare etc.

Edited September 21, 2020 by Melon_Head

[Melon_Head 15]

One thing you could try, to keep it nice and simple, I noticed Zu Audio run their FR without any upper Low pass filter but utilizing the natural high frequency roll off of the FR Driver and just using a first order high pass around 10kHz on the tweeter, seems to work well for them.

 

Just ideas to try out and compare with what you already have.

 

[dcathro 272]

6 hours ago, Melon_Head said:

I used Full Range drivers from Audio Nirvana USA.

I am extremely happy with them.

 

When this bloody lock down is over, I guess I could bring some drivers around for you to compare etc.

Where are you located?

 

I will definitely come round for a look and listen.

[dcathro 272]

6 hours ago, Melon_Head said:

One thing you could try, to keep it nice and simple, I noticed Zu Audio run their FR without any upper Low pass filter but utilizing the natural high frequency roll off of the FR Driver and just using a first order high pass around 10kHz on the tweeter, seems to work well for them.

 

Just ideas to try out and compare with what you already have.

 

That is how I started off -running the full range with only a baffle peak correction and allowing the natural roll of at the top and the baffle termination at the bottom. However, it sounds better with the top and bottom pulled down a bit. Note that my filters are very mild, less than first order.

[Melon_Head 15]

Make sure you do plenty of AB listening tests.

Sometimes going for what might seem like a near perfect response curve is sometimes not the most enjoyable result.  

I found trying to EQ the Seas driver with extra crossover components led to less transparency.  So in the end took the EQ out.  I guess it kind of depends where the issues are in the actual frequency range.  Some areas are not so troublesome, other areas will sound terrible if not corrected.

[dcathro 272]

Nearly all my adjustments are done by listening. I measure and model to verify the result.

[Melon_Head 15]

Excellent.

[dcathro 272]

Have pretty much finished these, and am now just polishing.

 

Here are the current graphs:

 

[dcathro 272]

I am pretty pleased with the measured and simulated results as well as the subjective impression.

 

The crossover points are 200hz and 7khz, which are not in the sensitive hearing zones. The drop in the high frequencies above 12khz is partially due to the measurements being taken off the tweeter axis - they were done on the Supravox axis.

 

The phase angles are benign, and the impedance only drops a little below 4 ohms in the treble, meaning that the speakers are easy to drive.

 

Some pictures of the crossover: