Jump to content

Deep End DIY - My first speaker project


acg

Recommended Posts



Anthony.    You went with the SS woofers ?!   Have you put any more consideration into your SPL expectations from them?

 

If you're wanting low excursion you might be disappointed   ;)

We modelled these drivers a number of pages back. With six a side I think output at 15W was 105dB/1m at 20Hz. I have expanded that to eight a side so will have an extra dB or so. Output at xMax was 115dB or something at 20Hz. I would have to go back and check the sims and our discussion from that time, but I am sure things will be ok.

Link to comment
Share on other sites



Excuse the blunt question(s).   I'm about to make this same decision myself.    ie.  choose a subwoofer solution  (just finished eighth-space 50hz horns).

 

 

So many competing priorities.

 

 

We modelled these drivers a number of pages back. With six a side I think output at 15W was 105dB/1m at 20Hz

 

Yeah, but the enormous assumption then was around your 'room gain'.   How are you going to ensure that the 12dB/octave rolloff of the woofers, matches your in room bass target?   (You said you won't have EQ)

 

Assuming the gain you get from the room boundaries does translate it back to flat.    Then for 8 drivers it's 15w for 109dB/1m ... and +/- 3.5mm cone movement  (about half xamx).

 

You mentioned real ULF before, like < 20Hz ....   will your xfmr go that low?

Link to comment
Share on other sites

Dave, it's not that I won't have Eq, but I'd like to get away without it if I can.

Regarding the trafo I was thinking about a 6Hz -3dB point if feasible. Hopefully an upper knee around 1kHz if possible.

Link to comment
Share on other sites



More goodies turned up today:  I finally received my film capacitor order from the US.  All up 114 Electrocube 950 Series Polypropylene and Foil caps ranging in size from 0.0010uF to 3uF.  The smaller values will be used in the crossovers whilst the larger ones will be coupling caps and power supply filter caps.  I have purchased more sizes than I will need and more of those sizes as well so that I can play with my line level passive crossovers to my hearts content.  Hopefully I will not need another order because these caps are difficult to find and the manufacturer will not sell directly.  I was lucky enough to find them all in one place and although I am missing a couple of values that I would like in the large caps I am sure I cam make do with the ones that I have got.

 

post-139669-0-39739300-1438916195_thumb.

 

 

I even have some exotics to trial once everything is set up and on song so to speak.  Both the Duelund Cast and Jupiter Copper will be trialled as coupling caps in different channels.

 

post-139669-0-27493400-1438916191_thumb.

 

 

I will find it very interesting to do these exotic capacitor trials down the track because I have definitely heard the contribution of these caps in other electronics and wonder what they could bring to this new system.  Of course I was listening to them in important positions in full-range equipment but the 6 Channel SET amplifier that I am building is most definitely NOT full-range so I do wonder if they will have any impact when being used across a narrow frequency range.  My gut says perhaps a small impact, maybe, if I am unlucky.  The retail on just one of those Duelunds is more than I have paid for the rest of the caps you see in the top photo...and normally I would not have purchased the Duelunds because they are just so expensive...but I just happened to call the right guy at the right time and picked them up for a lot less than Parts Connexion are asking...still though, those two Duelunds cost more than the rest of the caps on the table combined.  The amplifier circuit and speaker system is voiced using these Electrocube capacitors throughout, so maybe it is wishful thinking that the exotic caps will bring any significant improvements.

  • Like 3
Link to comment
Share on other sites

I wonder if one could roll caps the way one rolls inductors?

 

Your table looks impressive by the way, oh and 

 

"I have purchased more sizes than I will need and more of those sizes as well"

 

Ha ha ha ha! Been there too. I have a stash of SMD resistors I might never use....

  • Like 1
Link to comment
Share on other sites

Feeling a little bit proud of myself this morning.  Wife is away playing with ponies and kids so I have spent a relaxing morning trying to figure how to calculate out a third or fourth order LC line level Butterworth passive filter to sit in the amplifiers.  This is the crossover for the ribbon tweeter that operates above 10kHz.  I am not sure of the final crossover point for this transducer nor even how steep I want the slope so I was thinking of buying some air variable capacitors so I could keep my options open.  The aim of the mornings work is to run several crossover and circuit scenarios so that I know how large these variable caps should be.

 

Anyway, Google was getting me nowhere and all of the online discussion and filter calculators I could find for this used an equal source and load impedance whereas I "think" that I will have an infinite load compared to source so the capacitor values will change from one position to another in the crossover.  I remember looking at this about 6 months ago and giving up disheartened.  Anyway, I found a book called "Electronic Filter Design Handbook" by Arthur Wiliams and Fred Taylor which is fricken' awesome.  It steps through the process of calculating all kinds of variations of these filters and as a result in short order I have produced a spreadsheet so that I can vary load impedance and crossover point and number of poles and can size up not only the air variable caps but I also have some good info regarding likely inductor sizes when I next talk to Lucas at Black Art.  He will be able to tell me whether inductors of these sizes are doable or if I should be aiming for something larger.  The filter is likely to be third or fourth order Butterworth with capacitors in series and inductors as shunts.

 

Anyway, a productive morning.  I feel like the cat that got the mouse (and quite the nerd)!

 

 

  • Like 1
Link to comment
Share on other sites

At least someone is getting progress!  ;)  :P    (I'm stuck with work and home renos)

 

The easy thing about a passive filter for the ribbon is that it's impedance is constant with frequency .... making a filter for moreregular drivers is much more difficult because as the drivers impedance changes, so does the filter response ...  and so if you (for example) put in a 3rd order butterworth electrical filter ... the slope of the resulting response (from the speaker driver) will change with frequency  (and won't sum together with the adjacent driver properly).

 

 

Have you decided what overall crossover response your're going for?  (ie.  the slope of the output of the drivers, rather than the actual electrical filters used to facilitate those slopes)

  • Like 1
Link to comment
Share on other sites



Feeling the exact same pain here ... incl 12 hours of antenatal classes last weekend.

 

Geez Chris..12 hours????  I sat though a total of about four hours for my first kid and that was like drawing teeth...but twelve.  I feel for you man.  But congrats...looks like the procreation task is bearing fruit, so to speak.

  • Like 1
Link to comment
Share on other sites

At least someone is getting progress!   ;)   :P    (I'm stuck with work and home renos)

 

The easy thing about a passive filter for the ribbon is that it's impedance is constant with frequency .... making a filter for moreregular drivers is much more difficult because as the drivers impedance changes, so does the filter response ...  and so if you (for example) put in a 3rd order butterworth electrical filter ... the slope of the resulting response (from the speaker driver) will change with frequency  (and won't sum together with the adjacent driver properly).

 

 

Have you decided what overall crossover response your're going for?  (ie.  the slope of the output of the drivers, rather than the actual electrical filters used to facilitate those slopes)

 

 

 

No I have not decided as yet.  I figured that I would start at Romys values and see what I end up with.  It is difficult to exactly predict the acoustic roll-off due to the horns so I figure that I will get them finished and installed and measured and then see where I end up.  I have not modelled the entire system in any software, but have given individual channels a run in Hornresp.  I am not so hung up about getting crossovers to sum exactly properly but I am hung up about getting things to sound right.  Hopefully those two things will align to make tuning the system a little easier.

 

 

The scale of this project is quite daunting at the moment with both the speakers and the amps being built from scratch, and me knee deep in getting it all put together, but they say things are darkest before the dawn and as soon as I can nail down the iron for power supplies, filters and speaker output I will be be seeing the first glimpses of the dawn.  I am making progress with Lucas at Black Art and I am very thankful that he is so thorough and not just giving me what I think I want but rather trying to figure out what I actually need.  I also spent some time in Eagle and today have ordered some pcb's to be made for the bridge rectifiers and power supply bleeder circuits in the amplifiers.  Thanks Chris for your help especially given your ante-natal classes this weekend! 

  • Like 1
Link to comment
Share on other sites

It is difficult to exactly predict the acoustic roll-off due to the horns so I figure that I will get them finished and installed and measured and then see where I end up.

 

Ok.     To get you thinking, for a proper horn, you will have 24dB rolloff below cutoff, and 6dB above.

 

 

I am not so hung up about getting crossovers to sum exactly properly but I am hung up about getting things to sound right.

 

One of the most important things for "sound right"     ;)

Edited by davewantsmoore
Link to comment
Share on other sites

 

One of the most important things for "sound right"     ;)

 

It completely agree that it is one of the most important things, but if I have to add a lot of extra components or opamps or DSP into the signal path then just where the sound ends up is anyone's guess, but it won't necessarily be great just because I make the crossovers to sum perfectly, there are other things to think about.  Some of the major decisions that I have made in this project so far are not necessarily based on the best technical or scientific merit.  So many drivers for a start, single ended valve amplification, a channel or two that is there to do nothing more than inject a little tonal richness into the sound if necessary and so on.  If I went for just the technically "correct" options I would have fewer channels and a true point source speaker fed by ultra low distortion chip-amps...would it sound any better?  I guess I'll never know but so long as I can get the sound that I am trying to build I will be very happy.

 

Saying that Dave, I am hoping that you will stick around for the rest of the project because I learn so much when I am confronted with a voice of reason.  The biggest reason for starting this thread is to be able to get input and ideas from people that know more about this stuff than I do. 

Link to comment
Share on other sites



 

The easy thing about a passive filter for the ribbon is that it's impedance is constant with frequency .... making a filter for moreregular drivers is much more difficult because as the drivers impedance changes, so does the filter response ...  and so if you (for example) put in a 3rd order butterworth electrical filter ... the slope of the resulting response (from the speaker driver) will change with frequency  (and won't sum together with the adjacent driver properly).

 

 

 

Lets talk about the electrical filters for a while.  The filters as I intend to use them are line level passive and they will physically reside inside the amplifiers so that there are no components between the amplifier and the transducers.  These filters are applied a little differently depending on whether the particular SET amp is single or dual stage, but in all cases the filtering is finished before the final tube in each amplifier circuit.  With a valve and an output transformer in between the filter and the transducer these filters should "see" the transducer a lot less than if they were applied in the standard location between the amplifier and transducer, especially as we get away from bass frequencies.  So I think that the impedance curve of the transducer is not really in the equation, at least that is my opinion based on my current electrical knowledge.  The nature of the final valves impedance...well that is another story.

  • Like 1
Link to comment
Share on other sites

What will you use to shape the frequency rolloff? And eq if there's any peak?

 

Both the horns and the electrical filters will shape the frequency rolloff.  I have not measured these compression drivers in the horns of my construction yet, but I do have measurements of these particular drivers in other horns (part of the purchase process) and there were no peaks in those responses to be concerned about, so I do not expect any issues above say 500Hz that will require any form of "eq" other than a simple electrical crossover.

 

Is this what you were asking Henry?

 

 

EDIT:  I should add that in addition to the straight electrical filters applied to the input signal of the amplifiers, that I can also affect things to a certain degree with interstage coupling capacitors (high pass filter) and output transformer (eg. midrange transformer rolling off below 1kHz).

Edited by acg
Link to comment
Share on other sites

"The scale of this project is quite daunting at the moment with both the speakers and the amps being built from scratch, and me knee deep in getting it all put together, but they say things are darkest before the dawn and as soon as I can nail down the iron for power supplies, filters and speaker output I will be be seeing the first glimpses of the dawn.

 

The full project is in my opinion a ten year plan for a productive DIY maestro.

 

To make the project fun again I suggest splitting the stages up into doable chunks.

 

First I would suggest getting just enough horns set up to make music from say 120Hz -8kHz. Get a quality old cheap Marantz low power second hand multichannel AVR as the DAC, peamp and amp. Use computer audio to run active crossovers eg Foobar running VST's daisy chained.

 

That way you get to listen to your horns and drivers and have some quality time. You will also be able  to start mucking around with software crossovers to find what you enjoy. Fun huh?

 

Then when you add eg the ribbon or later the woofer stack you can enjoy how the sound evolves

 

Later you can add your DHT amps just one channel pair at a time using the pre-outs of the AVR

 

Finally you will end up being able to design the crossovers based on what you have found is best with the CA digital filters.

 

What do you think?

  • Like 2
Link to comment
Share on other sites



  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...
To Top