Compression ratio/new build

[Dag Johansen]

Hi, first post in this section.
Thought i would share my project with you.
First off i would like to say that it is not my intension to disrespect Tom Danley and his work, he is my gratest inspiration, but i am not doing this for money - just for fun.
Even if this project turn out very well i will not sell these speakers.

This is a 2-way project.

I have searched these forums to find out some more about compression ratio in this Synergy/Unity type horns.
Studying the photos of SH 50 one can see the holes for the twelves and they are not big.
The holes in my horn is 12,5cm2, two holes per woofer, so 25cm2.
That correspond to 1:9 compression with eigth inch woofers.
Would that be a problem?
I am aiming for 90-100Hz downwards.

So.... any input?
As i said, its just for fun!

My english is not that good, but i think you understand!

Dag

 

[pk]

Hi Dag,

That's very impressive woodwork! I can easely understand what you are after here. I really wish I had the woodworking skills and technical knowledge to clone the Danley SH-50's for a serious home rig.

What crossover frequency are you after? How do you intend to make the transition from square to circular for the compression driver? And which drivers will you be using?

Please keep us informed about the progress on this project (measurements, etc.!)

Thanks and good luck!

Best regards
Peter

 

[Dag Johansen]

Thanks for nice words PK.
The transition between the driver and the horn will be grinded to a round smoth surface.
To begin with i will try Eminence Alpha-8 for bass/midrange and Eminence PSD2002A - 1" Compression Driver for tweets.
I will have to experiment a lot with crossover frequency.
For this i have a plate-amplifier with buildt in DSP from Digmoda.
</title> <meta http-equiv="Content-Type" content="text/html; charset=utf-8" /> <meta http-equiv="X-UA-Compatible" content="IE=EmulateIE7" /> <title>Digital amplifier
I will post some meassurements when i get to that stage!

Dag

 

[pk]

Hi Dag,

Thanks for responding. It definately sounds like an interesting project! I may be wrong, but if I understand it correctly, then the crossover point between the mids and the tweeter is more or less determined by the positioning of the holes for the 8" mids (the holes should be placed at distance specified by 1/4 of the wavelength of the crossover point).

I am currently using the Beyms TPL-150H ad tweeters and like them a lot! Given the rectangular profile of the TPLs they should be great candidates as tweeters in a SH-clone (no need for making transition from rectangular to round).

BTW: the IcePower amps in your Digmoda's sound great!

Best regards
Peter

 

[panjilaras]

i believe the midrange port is frustoconical not just simple hole. JLH is the man for explaining synergy theory in layman term.

 

[jeno]

Hi!

Cool project!

Have you read the patents on the Unity horn and the Synergy horn? If not I recomend reading them before doing more work. They are easely found by the google patent search. If you are going for a trial and error aproach I think it may be very difficult to make this work.

If you read the Unity patent you will see that the area of the insert holes should equal the area of the horn at the insert point to match the impdance at the insert point. But in the Synergy patent you will see that the holes/ports may be reduced in size if you use a frustroconical shape or a stepped port. I don't think Tom Danley has given away the exact answer to this, so here trial and error may be the only way to do it anyway.

Anyway, good luck with your project! Would love to hear the result if you make it work.

 

[doug20]

I definitely have to follow this, its on the list of projects I would love to complete some day!!

 

[Dag Johansen]

Thanks for all the inputs guys!
Ok, i have now read the patent part about the holes and the frustoconical/stepped port.
If i understand it correctly it has to do with the bandpass effect!?
Would it not be better to make some kind of phaseplug to minimize the "chamber" in front of the driver!?
Or maybe thats the whole meaning of the construction, to make a small bandpass system, firing in to the horn?

I have been fiddeling around with loudspeakers for 35 years, and i know a lot, but i also realize that there is so much i dont know.
(and that many of you guys know a lot more than i do )

Anyway, this is a work in progres and i try to understand as i go along.

If there was a dealer here in Norway i would most likely bought a pair of SH-50 as they appeal to me.... they are downright sexy!

Dag

 

[pk]

Thanks for all the inputs guys!
Ok, i have now read the patent part about the holes and the frustoconical/stepped port.
If i understand it correctly it has to do with the bandpass effect!?
Would it not be better to make some kind of phaseplug to minimize the "chamber" in front of the driver!?
Or maybe thats the whole meaning of the construction, to make a small bandpass system, firing in to the horn?

I have been fiddeling around with loudspeakers for 35 years, and i know a lot, but i also realize that there is so much i dont know.
(and that many of you guys know a lot more than i do )

Anyway, this is a work in progres and i try to understand as i go along.

If there was a dealer here in Norway i would most likely bought a pair of SH-50 as they appeal to me.... they are downright sexy!

Dag

Hi Dag,

If I were to start all over again, I think I would go for a pair of SH-50s as well! You may already be aware of this, but DSL has recently got a dealer in Europe (Poland).

soundmission.eu - Home

At least we now have a fair chance to listen to these beasts!

Best regards
Peter

 

[panjilaras]

JLH/GM/Bateman/Synergy experts, please slap me,

the frusto thingy is designed to:
1. low pass filter uper harmonics by air chamber inside port plus inside cone
2. of course as midrange sound exit, but it is small enough to make it invisible to HF, yet large enough to avoid port noise / high port velocity.

did you do simulation of this box? why not using delta pro 8a?

 

[GM]

did you do simulation of this box? why not using delta pro 8a?

I've simmed the basic HF horn and the sealed back (4th order) and band-pass (BP) alignments with theoretically ideal' reflex ports (didn't know they were called frustoconical though) for Unity/Synergy concepts, but not the complete horn system or the SH-50's TH (6th order) mid-bass system. AFAIK only AkAbak can do either of these and I haven't taken the time to learn it yet.

Frankly, while I find the SH-50 technically interesting, I'd rather have a larger three way horn with all BP alignments and limit the TH to (sub)woofer duty.

GM

 

[JLH]

That’s some nice wood working, but just looking at things will not get you anything close to a Unity or Synergy horn. I really don’t see how you plan to pull off a two-way with a 1” comp driver and some 12” woofers. You are going to have a bad notch in the response. You are better off using a smaller woofer to bridge the gap between the 1” comp driver and some 12 woofers.

There are some very specific requirements that have to be met with the design in order to make it work. You are unlikely to approach anything worth listening to unless you model it with a higher end simulation software like Akabak. The biggest issue with any Unity like horn is the phase relationship between the drivers. You have to balance the bandpass effects of the entry ports with the transfer function of the crossover in relation to the physical placement of the drivers all at the same time. This is practically impossible without software modeling to at least get you close. Once I get the entry ports tailored to what they need to be, then I use Akabak to look at the phase. (See pic for phase simulation) Using the components in the crossover, I phase align the drivers and contour the power response. Even with good software it takes several hours or days to get a good combination that is stable within the tolerance ( +/- 5%) of the crossover components. Then you have to actually build what you modeled and measure it. Once measured, you can make the final changes to get the response nice and flat.

It’s a LOT of work to iron out a good Unity/Synergy horn. With the introduction of the SM-60F Synergy horn, it is unlikely I will bother building anymore. After looking at my time and material costs, it’s actually cheaper to just purchase the SM-60F.

http://www.danleysoundlabs.com/pdf/SM 60F spec sheet.pdf

 

[doug20]

What will be the price on the SM-60F?

 

[John Sheerin]

I agree with John. When I built my first few unities based on looking at Tom's and the Lambda ones and doing some very basic modeling, the response was horrible. It wasn't until I simulated the horn in a Matlab script I wrote that I started to get decent response out of the mids. On a 3 way unity I built, it was still surprising how good it sounded after I eq'd it to have roughly flat response (with active filters). I will say that using a dsp crossover makes life a lot easier - you can get the response smooth during the modeling and then use time delay to tweak any positioning errors instead of having to potentially build another iteration of the design.

 

[pk]

That’s some nice wood working, but just looking at things will not get you anything close to a Unity or Synergy horn. I really don’t see how you plan to pull off a two-way with a 1” comp driver and some 12” woofers. You are going to have a bad notch in the response. You are better off using a smaller woofer to bridge the gap between the 1” comp driver and some 12 woofers.

There are some very specific requirements that have to be met with the design in order to make it work. You are unlikely to approach anything worth listening to unless you model it with a higher end simulation software like Akabak. The biggest issue with any Unity like horn is the phase relationship between the drivers. You have to balance the bandpass effects of the entry ports with the transfer function of the crossover in relation to the physical placement of the drivers all at the same time. This is practically impossible without software modeling to at least get you close. Once I get the entry ports tailored to what they need to be, then I use Akabak to look at the phase. (See pic for phase simulation) Using the components in the crossover, I phase align the drivers and contour the power response. Even with good software it takes several hours or days to get a good combination that is stable within the tolerance ( +/- 5%) of the crossover components. Then you have to actually build what you modeled and measure it. Once measured, you can make the final changes to get the response nice and flat.

It’s a LOT of work to iron out a good Unity/Synergy horn. With the introduction of the SM-60F Synergy horn, it is unlikely I will bother building anymore. After looking at my time and material costs, it’s actually cheaper to just purchase the SM-60F.

http://www.danleysoundlabs.com/pdf/SM 60F spec sheet.pdf

Have you, John (or anybody else), had a chance to listen to the new SM-60F and maybe compared them to some of their bigger 'siblings' (e.g., the SH-50s)?

Thanks!

Best regards
Peter

 

[Dag Johansen]

OK guys, you have effectively killed my hope and joy!
I have looked at AKBAK but i have not gotten in to learn how it work.
What i do not understand is how one can simulate something like this...!?
I mean.... you are talking about frustoconical ports, how do you simulate that?
And if you can, how to measure Vb in front of the membrane?
Surely 20% off in volume will do more harm than if you not cones the port-openings?
Well, i will proceed anyway, having the ability to do some time alignment in the dsp i hope to get some results.
(and the fact that i am a carpenter also, i have 3 hours of work on these horns)
This is not an attempt to duplicate any of Danleys work - a 3 way would be even harder to make - as i said earlier, i am just doing it for fun.
Even if it sound terrible, i will end up as a wiser man!

@ JLH: i am using eigth's, not twelves.
@ panjilaras: I have the Alpha-8's laying on the shelf.

Thanks again for input, i really appreciate it!

Dag

 

[GM]

Have you, John (or anybody else), had a chance to listen to the new SM-60F and maybe compared them to some of their bigger 'siblings' (e.g., the SH-50s)?

Don't remember reading about it, but methinks I need to make the 45 mi road trip up to Gainesville for an audition before gas prices ramp up too much, if possible. The last time I was there they only had the SH-69's IIRC set up as mains for comparison.

GM

 

[GM]

OK guys, you have effectively killed my hope and joy!

I mean.... you are talking about frustoconical ports, how do you simulate that?

And if you can, how to measure Vb in front of the membrane?

This is not an attempt to duplicate any of Danleys work - a 3 way would be even harder to make - as i said earlier, i am just doing it for fun.

Thanks again for input, i really appreciate it!

Sorry about that! For sure, its design is no trivial pursuit, but get a good grasp of compression/TH horn design to get its basics right and as noted, with DSP you can dial one in to get good performance at HIFI/HT SPLs.

It should become transparent once you've studied up on theoretically ideal Helmholtz resonator ports.

Well, you can calculate it by carefully measuring it as a truncated cone or mounting it to a thin plexi-glass plate with a hole in it to drip in a pre-measured amount of alcohol tinted with food coloring same as is done with engine cylinder head combustion chambers.

Understood, just a concept as opposed to a clone...... Right, if you're not having fun, find a different project or hobby. Life's too short and tends to be too full of unpleasantness to intentionally add to it.

You're welcome!

GM

 

[John Sheerin]

The typical method to simulate in a program like matlab or akabak (never used akabak personally...) or even Excel is to make an equivalent circuit model. You model all the physical parts of the horn and drivers as electrical quantities. Then you calculate what the impedance would be of all the elements. Then you calculate what the current would be for a given input voltage to that impedance. Then you calculate back the other way to see what the current would be in the horn which is actually a flow. From that you can get the pressure. You do that for a bunch of frequencies (hence the use of a computer program) and then you can plot the response curve. There are books written that describe this sort of thing - see Beranek's Acoustics or Olson's Acoustical Engineering for example. There are examples of horn speakers in there. The unity style horns need some additional elements like the volume of the horn throat in parallel with the horn from the midrange ports to the mouth and the port from the midrange chamber to the horn.

I think JLH posted his akabak script somewhere though, so maybe you could try looking for that. Here it is:
http://www.diyaudio.com/forums/multi-way/109568-unity-horn-script-akabak.html

 

[JLH]

My Akabak scripts have evolved quite a bit since then. The better I got to understand Akabak, I began to use more of its tools to make better designs. Being able to compare the phase of each driver at any point along the horn is a incredible tool.