New project: 15PR400+HF108+BigOldRadial+HypexFA123

hi guys,
After reading a lot of information on this great forum I've started a project to build some big speakers for my upstairs listening room.
I want to experience the effortless effect of big speakers cruising along on living room listening levels.

Requirements:
- 15" woofer in a closed box
- Big horn with small compression driver to reach as low and high as practical
- Active system because I love the flexibility of DSP

Choices:
- the woofer was simple, Faital Pro 15PR400 gets almost universal praise for use in a two way on this forum
- I found some big old cast bi-radial horns on our local 2nd hand website, brand is unclear, could be Vitavox. The throat size is a very strange 41mm or 1 3/4", I'm having an adapter 3D printed after getting advice from this forum (https://www.diyaudio.com/forums/mul...ng-throat-adapter-hf108-vitavox-biradial.html)
- Compression driver will be Faital Pro HF108, mainly after reading the good results forum member Legis is getting with them in his project: https://www.diyaudio.com/forums/multi-way/357825-shadow-colossus-build-thread.html (Faital Pro is relatively affordable here in Europe)
- DSP and amplification will be one off the well known Hypex Fusion plate amps, I choose the FA123, plan is to bridge 2 channels to get 250W + 100W (8 ohms)
- Dimensions were chosen starting from the horn which is 62cm wide, height is 80cm because it looked nice, depth is just 31.5cm. This results in the HF108 sticking out of the enclosure which enables experiments with a bigger rear chamber
- A Linkwitz transform will be used to eq the low cutoff.

I will use REW and a calibrated ECM8000 microphone to measure the response of both drivers and use these to decide were to put the crossover frequency.
I'm hoping to find a point where the directivity of of both drivers is more or less the same, while still not pushing the tweeter too low.

I hope to start measuring in a few weeks, I will post the results.
For now I've attached some photo's of the project.

kind regards, Peter
 

Attachments

  • IMG_0226s.jpg
    IMG_0226s.jpg
    741.6 KB · Views: 1,011
  • IMG_0227s.jpg
    IMG_0227s.jpg
    713.3 KB · Views: 1,021
  • IMG_0241s.jpg
    IMG_0241s.jpg
    351.5 KB · Views: 1,806
  • IMG_0244s.jpg
    IMG_0244s.jpg
    426.6 KB · Views: 1,172
  • IMG_0272s.jpg
    IMG_0272s.jpg
    336.8 KB · Views: 1,008
  • Like
Reactions: 1 user
Nice project Peter!

I've always loved the classic look of a large radial horn sitting over a 15" woofer.

Can you show a couple more pics of the backside of that quasi Vitavox horn ( I'm still really curious about its origins ).

:)

hi EarlK,
I've attached some photos of the horn, I can add other views if you like.
One thing is clear, they need damping, they ring like a bell.
I spend quite some time with google to find out what brand they are, only found one reference on faceook, I will post these in the next message to keep the images separated.
regards, Peter
 

Attachments

  • IMG_0225s.jpg
    IMG_0225s.jpg
    237.7 KB · Views: 180
  • IMG_0277.jpg
    IMG_0277.jpg
    90.1 KB · Views: 164
  • IMG_0276.jpg
    IMG_0276.jpg
    102.7 KB · Views: 171
  • IMG_0275.jpg
    IMG_0275.jpg
    75.6 KB · Views: 161
  • IMG_0274.jpg
    IMG_0274.jpg
    77.7 KB · Views: 218
Last edited:
Google found this on facebook, I don't have a facebook account so I cannot contact the guy. Therefore I don't know why he thinks it's a Vitavox horn...
I think they look very similar to my horns.
 

Attachments

  • Vitavox3.jpg
    Vitavox3.jpg
    300.4 KB · Views: 221
  • Vitavox2.jpg
    Vitavox2.jpg
    367.2 KB · Views: 191
  • Vitavox1.jpg
    Vitavox1.jpg
    374.7 KB · Views: 202
  • Capture.JPG
    Capture.JPG
    41.2 KB · Views: 219
I use a pair of 15PR400 from 75 to about 300-500Hz, depending on my mood. Plan is to lower it to 200Hz by getting a bigger midrange horn. Dizzie Rascal - Bonkers. FaitalPro 15PR400 8ohm, 80-270Hz, Xindak 100W - YouTube

They are used in a 92L ported box. Two pipes with 10cm diameter and 16cm length. Looks good in WinISD and sounds good too. Compared to my previous speakers the JBL K140 / E140, housed in the same cabinet, the 15PR400 are a bit timid, but nice and clean. Motor is a bit weak also. Nothing to get excited about, but does nothing wrong either. I compared them to a Martin Sound SB115 midbass horn and those are less correct, but gives a great sense of presence. Something that is lacking in most box speakers.
All in all, 15PR400 is a problems solver that fixes midbass toba hood level, and after that you can experiement and see if you find something more exciting.
 
Nice project, also looking forward to your results.

The volume of the box is about 110 liter, according Winisd this will give me a Fsc of 64Hz with a Q=0.58.
With a Linkwitz transform is simple to bring this to 30Hz with a Q=0.707 and still go plenty loud with 250W (and stay within Xmax)

I'm modelling the same'ish box in WinISD and while I'm getting the same results, Xmax is reached at about 8 watts input signal, resulting in a max SPL of around 104dB. I'll probably never reach that in my living room, but I'm still curious about your simulation and you mentioning 250W
 
hi Cooljazz,
Thanks for you interest en remarks.
I'm clearly not an expert in WinISD, the cone excursion I saw was at 1W, and indeed Xmax is reached at about 8W.
I hope to measure response and create a first filter this weekend weekend, I'll report my findings.
kind regards, Peter
 
Update!
The speaker building project is finished, measurements have been performed outside, first filters have been made in Hypex Filter Designer.
First listening tests are very positive, the sound is very 'big' and effortless without a trace of harshness. Overall sound in warm, I can listen for ours and only want to hear more (and louder).
Bass is massive with the cone moving just a few mm.
Crossovers were made following the Grimm Audio LS1 white paper, in short, equalizing the raw drivers to a flat response to 1 or 2 octaves outside their pass band and then apply a 4th order LR filter
Crossover is at 750Hz at this moment, and as seen from the measurements, crossover is not perfect, I think I have the delay of the woofer not correct, I don't completely understand HFD regarding this. I measured the individual drivers with HFD as adviced, this should make sure the measured timing is correct. To be continued.
Response seems reasonable constant up to 30 degrees, beyond that response begins to drop, to me the takeover from woofer to tweeter looks quite good with this crossover frequency. I will have to work on the delay issue
For a next project I think I would try the XT1464 horn, it will measure better, but this old thing does look cool :)

I really love the Hypex Fusion amps, they have great possibilities and great sound, but as all DAC's and active speakers with a digital input, they are extremely sensitive to the quality of the digital input signal.

thanks for looking, kind regards, Peter
 

Attachments

  • IMG_0374s.jpg
    IMG_0374s.jpg
    375.2 KB · Views: 644
  • IMG_0375s.jpg
    IMG_0375s.jpg
    304.9 KB · Views: 524
  • IMG_0376s.jpg
    IMG_0376s.jpg
    424.4 KB · Views: 552
  • IMG_0378s.jpg
    IMG_0378s.jpg
    266.6 KB · Views: 683
  • total 0-90 degrees.jpg
    total 0-90 degrees.jpg
    91.4 KB · Views: 711
  • Like
Reactions: 1 users
Wow, that is smooth.

Using it from 75-200Hz with an 8" taking over the lowest midrange for more oomph! Or more relaxed 15" and compression driver.

Quick drawing of how it will be.

Is yours sealed or ported?
 

Attachments

  • 222973140_1450939275267521_5140292205874503169_n.jpg
    222973140_1450939275267521_5140292205874503169_n.jpg
    54 KB · Views: 396
  • thumbnail_Screenshot_2021-08-13-23-53-44-35.jpg
    thumbnail_Screenshot_2021-08-13-23-53-44-35.jpg
    169.1 KB · Views: 339
  • 15PR400 sealed vs ported.png
    15PR400 sealed vs ported.png
    45.4 KB · Views: 311
  • 15pr400 winisd.png
    15pr400 winisd.png
    89.5 KB · Views: 303
Last edited:
crossover is not perfect, I think I have the delay of the woofer not correct, I don't completely understand HFD regarding this. I measured the individual drivers with HFD as adviced, this should make sure the measured timing is correct.

If you flattened the driver responses past 1 octaves from the xo point and applied the LR slopes, then reverse one driver's polarity and adjust the timing to the point where the reverse null is the deepest on your desired axis. Reverse back the polarity to normal of the previously flipped driver, done.

Of course, this is just one method, but it works.
 
  • Like
Reactions: 1 user
For science.

I took the liberty to compare my 92L ported (tuned to 47Hz), and your sealed at 110L. I will try it and stuff the holes and see if the midbass is clearer.
 

Attachments

  • 15pr400 winisd 92ported vs 110 sealed.png
    15pr400 winisd 92ported vs 110 sealed.png
    45.9 KB · Views: 206
Last edited:
If you flattened the driver responses past 1 octaves from the xo point and applied the LR slopes, then reverse one driver's polarity and adjust the timing to the point where the reverse null is the deepest on your desired axis. Reverse back the polarity to normal of the previously flipped driver, done.

Of course, this is just one method, but it works.

hi Deni,
Thanks for your response, that's what I did, resulting in a delay of about 1500us (about 500mm) which is way more than expected, in HFD the sum is perfect.
The response of the total system was measured with REW and that's not perfect.
I also measured the individual drivers without filters with REW and when I subtract the 'Start time (seconds)' of both i get a calculated delay of 666us which seems much more logical (about 200mm).
Something goes wrong in HFD.
I have to put the 666us in the speakers and then measure the overall response again, hope I can measure this inside as I won't bring the speakers down the stairs again.
regards, Peter
 
Hi petervv,

can you please clarify the crossover and its implication on power?

Per your post # 12, the crossover frequency is 750 Hz. If you equalized the HF108 1 octave away from that, the frequency is 375 Hz, and another octave is 187.5 Hz.

Looking at the HF108 data sheet, it appears that you had to apply ~ 20 dB gain at 375 Hz and ~ 30+ dB at 187.5 Hz.

Given that the Minimum Crossover frequency is 1.3 kHz, and you are using 100W amplifier per your post #1, even with a crossover slope of 24 dB/octave, are you not worried about destroying the HF108?

Or, am I missing something?

Kindest regards,

M