Harmony - My attempt to clone the Synergy Horn

[goldyrathore]

I did a lot of trials with not so good results initially. I had almost given up with these drivers. After incremental tweaking, finally managed to get some good results.
Here is the 2" plot. As you can see there is pretty good response from 500hz to over 1k.



And here is the Sony 2.5" plot. This one goes from 200 hz to about 1k.


As mentioned earlier these results are after sealing the rear chambers. Considering that I ve not used any simulation so far (for the mids) and only relied on my bandpass sub experience from past, I think these are very nice results. The setup is temporary so not posting the images.

This gets me closer to tap the horn:clapping:

Goldy

 

[goldyrathore]

Before we tap the horn

The synergy horn is a conical horn. A conical horn has a variable flare rate that is dependent on the axial distance. At the throat of the horn the flare rate is very high and so is a good place to load the high frequency driver. But its not a good point for the mids. However as one goes further down the horn (towards the mouth), the flare rate lowers and the mids can be loaded at such a point.
Basically, the vertical/horizontal angles of the horn and the axial distance dictate the ranges where the midranges need to play.
Since there are multiple variables, so to ease the math, I am preparing an excel sheet that allows to play with all these variables.

Stay tuned.

 

[goldyrathore]

Below is a snapshot of the horn calculation from the sheet I prepared. This horn is a 72 degrees (h) by 60 degrees (v) horn. I started with a 75 degree x 60 degree horn. After woodworking, it came to 72x60




The columns H/J represent the frequency ranges where the mids must play. C represents the axial distance where the horn must be tapped. K is the frequency of the notch frequency (cancellation from the refletion of the throat of the horn). The notch freq is well above the highest freq produced by the midrange, so its not an issue.

So the computed range of mid is 258hz to 1315hz. The tap is at 5 cm from the throat. Next, need to check whether the driver can fit at the distance.

 

[soundnovice]

hi goldy,
good going...
since mids and tweeter will not be on same axis, will any delay settings needs to be applied for mids (time alignment)!?

 

[Hari655]

Below is a snapshot of the horn calculation from the sheet I prepared. This horn is a 72 degrees (h) by 60 degrees (v) horn. I started with a 75 degree x 60 degree horn. After woodworking, it came to 72x60

Sir may be my answer looks like silly!!

Can you add some card board or think ply or wood which can reduce the 3 degree area

Sorry if this is really silly.

Thanks,
Hari

 

[ketansethi1987]

I did a lot of trials with not so good results initially. I had almost given up with these drivers. After incremental tweaking, finally managed to get some good results.
Here is the 2" plot. As you can see there is pretty good response from 500hz to over 1k.



And here is the Sony 2.5" plot. This one goes from 200 hz to about 1k.


As mentioned earlier these results are after sealing the rear chambers. Considering that I ve not used any simulation so far (for the mids) and only relied on my bandpass sub experience from past, I think these are very nice results. The setup is temporary so not posting the images.

This gets me closer to tap the horn:clapping:

Goldy

How to get the FS Graph?

Which software/devices to use?

I want to do it for my DIY Sub

Thanks

 

[goldyrathore]

A bit about the offset of midranges

hi goldy,
good going...
since mids and tweeter will not be on same axis, will any delay settings needs to be applied for mids (time alignment)!?

The mids in the Syngery Horn are offset in all x,y and z planes.
The x and y planes dont matter since the mids are fairly 'close' even if we see them from exterme off -axis, they are hardly offset by any margin when one compares the relative distances of each mid from the listening position. Whatever that margin is, it is always small as compared to the 1/4 wavelength of the highest frequency produced by the mids. So the summation of the mids is coherent and no comb filtering kind of effects are observed at any angle.

The z plane has an offset (mids are placed somewhere down the horn) which is actually an advantage. The mids use a bandpass alignment which has an inherent delay. The offset delays the compression driver's response too. Thus these two are delayed (acoustically) resulting in a more coherent mid/high blend.

Compare this situation with a normal TM where both the drivers are on the same baffle. The tweeter's acoustic center happens to be forward and must be delayed to align with the mid/woofer. To compensate for the forward tweeter center sometimes a delay is used like Phase Correction - Myth or Magic

Full credits to Tom Danley for the innovative concept of the Synergy Horn

Its too early to say whether any additional electronic delay will be needed in this particular implementation.

 

[goldyrathore]

How to get the FS Graph?

Which software/devices to use?

I want to do it for my DIY Sub

Thanks

I use op-amp based preamp from Linkwitz (System Test, scroll down to end of page) with a general purpose electret mic capsule.

HOLMimpulse s/w and XonarDX soundcard.

The measurements are taken very close to the port by running a frequency sweep across the band.

 

[kanwar]

I did a lot of trials with not so good results initially. I had almost given up with these drivers. After incremental tweaking, finally managed to get some good results.
Here is the 2" plot. As you can see there is pretty good response from 500hz to over 1k.



And here is the Sony 2.5" plot. This one goes from 200 hz to about 1k.


As mentioned earlier these results are after sealing the rear chambers. Considering that I ve not used any simulation so far (for the mids) and only relied on my bandpass sub experience from past, I think these are very nice results. The setup is temporary so not posting the images.

This gets me closer to tap the horn:clapping:

Goldy

Nice bandpass effect for Mids, Sony driver gives more flatter and wider bandwidth......:cheers:

 

[goldyrathore]

Nice bandpass effect for Mids, Sony driver gives more flatter and wider bandwidth......:cheers:

Yes indeed, I too was pleasantly surprised at the response of the Sony driver. There are some out of band peaks on the higher side but those are due to a bigger sealed rear chamber and are a result of the reflections. The 2" driver seal resulted in a very small rear chamber and displays no such reflections.

Some threads did mention that a lighter cone helps more than a heavier one but even though the Sony happens to have not so light cone (as compared to the other 2" driver) its response is flatter. Maybe things will change once they are put in the horn, I dont know yet. I am not using any simulation (Hornresp or Akabak) and so will have to tweak by trial and error.

 

[quad]

I use op-amp based preamp from Linkwitz (System Test, scroll down to end of page) with a general purpose electret mic capsule.

Goldy, what about the frequency response graph for the Mic capsule?
With the WM-61A, the response is well known and linear in the normal
audible range.

 

[goldyrathore]

Goldy, what about the frequency response graph for the Mic capsule?
With the WM-61A, the response is well known and linear in the normal
audible range.

From what I gather from Rod's site (Recording and Measurement Microphones), these capsules are pretty flat. For the time being this would suffice since I am focussing on the mids right now.

 

[kanwar]

Yes indeed, I too was pleasantly surprised at the response of the Sony driver. There are some out of band peaks on the higher side but those are due to a bigger sealed rear chamber and are a result of the reflections. The 2" driver seal resulted in a very small rear chamber and displays no such reflections.

Some threads did mention that a lighter cone helps more than a heavier one but even though the Sony happens to have not so light cone (as compared to the other 2" driver) its response is flatter. Maybe things will change once they are put in the horn, I dont know yet. I am not using any simulation (Hornresp or Akabak) and so will have to tweak by trial and error.

The response will not deviate much unless some serious changes are done. Having a natural bandpass is very tempting, one can use an active X-over to match the response further when combining it with whole system.

I simply love the projects which are pure science and fun

 

[goldyrathore]

Considering a bigger mid

In order to get a better low end response from the mids, I considered a bigger 3" Aiwa mid. Attached are the impedance sweeps of open backed (green) driver overlaid with a sealed back (yellow). The open back driver has Fs of about 135Hz and sealing it results in Fb of about 490Hz. The mids operate in bandpass mode and so for 250hz to 1000hz response the Fb should be at the center ie at about 500hz. This driver seems suitable.

The driver was sealed first with masking tape and then two layers of Araldite was applied over the tape to provide a hard surface. The impedance sweep is clean so far so it means the seal is doing fine.

 

[kanwar]

How many mid drivers you are going to use?

 

[goldyrathore]

4 per horn.

 

[kanwar]

What is the average sensitivity of each driver and what is the target sensitivity as combined ?

 

[goldyrathore]

What is the average sensitivity of each driver and what is the target sensitivity as combined ?

Driver sensitivity is 86 dB/watt. I dont have any specific target (for sensitivity ) in mind for this build since its a prototype. Level matching will be done by adjusting gain in the active setup. 4 drivers in series parallel should make it 92 dB/watt.

I do have a target for max SPL which is about 105 dB. 16 watts gets me 104 dB.

 

[goldyrathore]

Some photos of the mids holes and the mids mounted on the horn. The hole is tapered.
The white colored thing in the rear of the mid is wire mesh that is used to seal the mids along with Araldite.

 

[goldyrathore]

Firing up the mids

A brief explanation of the mids' operation. The mids operate in a bandpass config covering about 2 octaves. The outut from the mids travel to both the directions ie towards the mouth and also towards the throat which reflects back to the mouth. This reflection combines with the original output from the mid and causes a deep null at some higher frequency beyond which comb filtering occurs. The mids are supposed to be crossed below this null. The distance where the mids tap into the horn determines where in frequency this null is formed.
The more the distance from the throat the more lower in freq the null is.

Below is the measurement showing the notch. I had mounted the CD also but it was not connected. Output till about 1k is desired and the notch is well above it at about 1.7k. The blue curve is on axis and the maroon one is about 45 degrees off axis. The response is not calibrated and both measurements are taken at slightly different distances so take them with a pinch of salt. The response is similar to other Synergy horn projects. Do note that this is without using Hornresponse. So far so good.

Also note that the response is far from flat. This is just the nature of the beast. The response will be tamed by EQ.