Collaborative Tapped horn project

OK, FYI/FWIW, a comparison of JLH's specs (solid) Vs published (shadow) and in each sim the conical flare was changed to parabolic for correctness, though of course barely any visible deviation/measurable/audible differences.
 

Attachments

  • B&C Speakers 8PE21 dual TH (JLH).PNG
    B&C Speakers 8PE21 dual TH (JLH).PNG
    5.6 KB · Views: 57
No, the shadow if the published specs are most correct. That said, it just 'dawned' on me (the caffeine finally 'kicked' in) that back then HR apparently wasn't programmed to sum driver specs, so he modded them to get the most correct response, i.e. don't use his specs under the assumption that they are measured.

Attached 'updated'/'corrected' sim; note the 3 ohm series resistance and power based on a 16 ohm load, making me wonder if it was tube driven..............
 

Attachments

  • 8PE21_2TH_JLHupd.txt
    2.7 KB · Views: 66
Last edited:
  • Like
Reactions: 1 user
Member
Joined 2008
Paid Member
That dual 8PE21 horn is really a nice design - pretty small for what it can do. I shortened it to 0.8 of original length and got this (black, grey is the original):
1671628368991.png


The motivation is that I was after 35 - 40 Hz at the low end in a compact box. With the smaller one I get almost 112 dB/m at 35 V still below XVar.
The net volume is 90 l, but I yet have to figure out how to place the drivers and how to fold it. The instructions mentioned in the thread say they are placed like in the SPUD (i.e. one reversed if I understand it correctly) and the folding should be like Volvotreters W6 horn - I checked the plans but cannot imagine fitting a reversed driver there.

My modified HR record is also attached.

I was however surprised by one thing. There is a difference between serial and parallel wiring for the two drivers - grey is series, black is parallel. Until now I thought this would not matter if connected series or parallel.

1671629345353.png
 

Attachments

  • JLHx0_8.txt
    2.7 KB · Views: 52
Hmm, ATM my understanding is that we can't mount them as simmed, but if true, then at this late date I'm guessing we couldn't sim it correctly with HR's original TH designer, so FWIW, etc., here's an example of my way for expanding, contracting termination (low pass filter):
 

Attachments

  • Expanding TH LPF.PNG
    Expanding TH LPF.PNG
    1.6 KB · Views: 47
  • Contracting TH LPF.PNG
    Contracting TH LPF.PNG
    1.6 KB · Views: 47
  • Like
Reactions: 1 user
Member
Joined 2008
Paid Member
I did play a bit with simulations yesterday, swapping BC 8PE21 and Fane Sovereign 8-225 and I have a few observations/questions.

1. Is it correct, that the lower the Qts (or Qes?) of the driver, the smaller volume/CSA the horn needs to be flat.

2. If the volume/CSA is too low, the low end suffers with a peak

3. If the volume/CSA is too large, I get peaks on both ends of the pass band

4. I can shape the response somehow by adding offset of one or other end.

I will try to modify the design so that it is actually buildable and uses only one 8PE21. It seems it can be squeezed into a really small box, so it could be used in pairs.
 
Member
Joined 2008
Paid Member
I was not able to come up with anything meaningful. I still like the idea of using two 8PE21s in a tapped horn since the box needs relatively low volume compared to other drivers. Both speakers would have to be placed along the horn path (not side by side), so I wonder how to simulate that with HornResp.

I can think of two options.

a) use the center line between the woofers as the driver position
b) simulate both positions and add up the responses

Which of these would be closer to reality? Or is there any better way?