SynTripP: 2-way 2-part Virtual Single Point Source Horn

  • Like
Reactions: rrobot and Jazzfan
Hi All,

User @woolwool asked for my Fusion 360 files, so I thought I would share them for everyone.

I worked in metric measurements, so that might be useful for the non-North Americans.

Use these files at your own risk. Note my concerns / caveats with my drawings in post 1398.

The files :

- Nested inner horn parts

- The base jig which was created I think by @Jennygirl

- Throat adapter plate top and bottom parts. Again by Jennygirl IIRC.

Syntripp outer panels were cut on the tablesaw so you're on your own for those.

1739258484041.png
1739258390676.png




1739258688055.png


1739258720284.png
 

Attachments

Thanks for the upload. So many discrepancies in those metric plans.. I've started to build according to those and have noticed what might be errors. It seems they have been designed to give a lower overall inner horn enclosure as the dimensions are generaly shorter than when converting from the imperial plans. I think i might scrap and start again using your measurements in fusion (which convert roughly correct the imperial plans when multiplied by 2.54).

For example:
  1. The width of the inner horn top and bottom is only 587.38mm in the finbot metric plans, but is approx 622mm in the imperial plans (this is not insignificant)
  2. The length of the inner side horns is only 391.04mm vs approx 414mm in the imperial plans

Before i scrap and start again, does anyone have any insight into whether the finbot metric plans are correct if you follow the dimensions (assuming i correct for symmetry in the ports and routing the speaker diameter out of the sides which is incorrect in the plans)? I originally thought it was because the thickness of ply between 1/2 inch (12.7mm) vs the metric sizes we get (12mm = 0.7mm thinner) was factored in? But i feel like it can't as that's not significant when one side is more than 30mm shorter in length.. (see point 1 above).
 
If you refer to Art's original parts list, the top / bottom horn panel (Part D) is listed as 23.125" That's 587.38.

1740259512898.png



I converted Art's imperial measurements to millimetres and then inputted into Fusion.

1740260544825.png


When I tell Fusion to label the dimensions in inches, it matches Art's drawing.

1740260491203.png


This doesn't match the Finbot drawing (has the width as 587.375 but the height is 291.2 which does not match Art's dimensions).

1740260704164.png



I went back and looked at what I did. I followed plans from user Sam Boenhoff. These plans were drawn out in imperial and I converted to metric.

Sam didn't include the gap for the grille cover. I believe this is why his plans have a deeper and wider horn than the original plans.

But I already see one difference between my top / bottom horn plan and the drawing I used from Sam - he has the width as 622mm and I have it as 629mm I cant remember now whether that was intentional or it could have been a typo.

I appreciate this doesn't really answer your questions, but for now at least I think any new builder's options are either

1 - start from Art's original plans and find a way to overcome the tricky bits that have thrown off all the other builders who started from scratch,

2 - use someone else's plans and expect to go 'off piste' a bit from the original plans.

I am planning to go and take masurements of my single completed Syntripp tomorrow using REW.. Hopefully by the end of tomorrow I will have some frequency response charts I can post here. It will be my first time measuring so fingers crossed I get meaningful results. But I'm already expecting a bit of a different plot to previous builders because (based on Art's recommendation) I built my SyntripP without the ports in the top and bottom horn panels.

I will report back tomorrow and should be able to provide a little reassurance (or not!) that the Sam Boenhoff plans will produce a cabinet that works.
 
That is very useful information, thanks for sharing mate - the context is enough to sleuth and reverse engineer..

I hadn't fully visualised the end product to realise that the inner horn (outside) battons accounted for that extra width/height (i assume this is what you are referring to as the material that can be used to drill into for a grille). In my case, i'll likely not use them as mounting points, but instead to just fill the box to give it the same volume.

In my case, the additional depth of the top bottom (291.2 - an extra 5 mm or so) is something that can likely be fixed by adjusting the angle it sits at or worst case, can be sanded away once assembled. Will see how it all comes together. Regardless - my margin of error is probably in the +-3mm range so realistically as close as i'll get with my available skills and tools and i accept that i'll have to do some creative woodworking to get it all together regardess of what plan i follow!

I suspect your width is a miscalculation (i also noticed that) but is probably not too significant -> i also doubt this will have a huge difference in sound quality - especially given that our margin for error probably supercedes a lot of these discrepancies.. Would be curious to see if it impacts your analysis measurements - keep us posted

And finally, thanks for getting back to me! It has been a thoroughly enjoyable process so far and i've learnt a lot in the re-engineering from the various posts and content. So if that was your intent Art, then thank you, i've finally had to employ my physics degree to good use!

Cheers
 
First try at speaker measurement last night. The results were... inconclusive. My results do not seem to be smooth like many of the frequency / SPL charts that I have seen posted in this thread, but I am looking at this very much through layman's eyes so perhaps I'm misreading (or not measuring correctly).

I am measuring in a large woodworking workshop (I assumed a large space would be good) but surfaces are all hard and reflective and a lot of stuff rattles as soon as SPLs start to go up. I have a feeling this is negatively affecting my measurements, especially with the mic placed further away from the speaker.

The playback chain was REW running on my laptop, going into an A&H ZED14 via USB. Mixer output into T Racks DSP 408 into DAP QI4600 amp into SyntripP

The recording chain was a Behringer ECM8000 into a channel on the Zed14 and back into the laptop.

I aimed to get SPL to be around 75 - 80dB at the microphone using an SPL meter on my phone.

I am aware that the ECM8000 is not ideal because there's some doubt over how accurately it responds in the HF range. I can upgrade the measurement mic and buy a proper SPL meter if necessary.

1740554744429.jpeg



My SyntripP is built with 2 x 10CL51 and B&C DE900TN. I do not yet have the secondary horn installed. Also important to note : I built this SyntripP without the ports in the horn, as per Art's suggestion.

I plan to use my SyntripP with XOC1 TH18 and Cubo kick so I HPF the mid drivers at 160Hz LK24 and LPF at 850Hz LK24. The HF is HPF at 950Hz.



First measurements with the mic at 20cm from the horn mouth


HF. What is happening around 14k?

1740556050905.png


Mid

1740556073075.png


Both. A big drop around the crossover. And a bump at 4k

1740556125524.png



Second round with mic at 2m

I moved the mic to 2m from the horn mouth and measured again. I applied a 6dB cut at 4kHZ because there was a big bump there. That seemed to smooth the bump nicely.


1740556658143.png


Mid

1740556773362.png


Both. There seems to be a bump at 400Hz and 600Hz (or a dip at 500Hz?).

1740556888904.png




Could anyone help on these beginner questions :

1 - Are the 20cm measurements or the 2m measurements a more accurate representation of how the SyntripP is behaving? I see differing opinions on this.

2 - The goal is to get the frequency response as flat as possible on the graph right? What I found with the measurements is that the gain on the mid and HF should both be set at 0dB on my DSP to make the SPL from each driver roughly the same on the chart . But when listening to music rather than measuring, the HF is way louder than the mid. I find myself reducing the HF gain down to -9 or -10dB vs mid in order to listen, otherwise the HF is painful. Is this normal / expected?

3 - Are the charts from REW useful without smoothing? If smoothing makes the chart more readable, which level of smoothing should I use? it seems that 1/3 or 1/6 produce a beautiful looking line, but hide a lot of large variations in frequency response.

4 - What is happening here? Should I attempt to reduce 400Hz and 600Hz with very high Q EQ?

1740558048586.png




5 - I have put some wadding in the top, bottom and rear sections of the back of the cabinets, but I was unsure whether this will make much difference in this SyntripP because I built the SyntripP without ports in the horn mouth. Does the back of the cabinet still need a lot of damping? The wadding I have is substantially thinner than the 3" fibreglass which Art suggests, so I will get some fibreglass and improve this if the consensus is that it will improve the performance.

1740559370574.jpeg


6 - Possibly related to point 5....When playing the mid section only, there is distortion of some kind. The best way I can describe it is almost a kind of static, like an out of tune radio. but it tends to come in short bursts when there's a loud transient in the music. I removed the back panel during testing and you can barely hear the distortion at the back; it's much more audible at the front. Nothing in the cabinet is loose or rattling. I tested the 10CL51s one at a time and the noise is present in both of them, so I do not believe that the speakers are damaged or faulty (yes, possibly both are faulty but it seems unlikely).

My next test will be to remove the 10CL51s from the cabinet completely and test them in open air to check the noise is coming from the cabinet and not the speakers.

I appreciate that this is a vague description of a problem, but does anyone have any suggestions on how to approach this? Methods to rule out certain possible causes?


Thanks in advance to anyone who can provide some input 🙂
 
Hi, read VituixCAD measurement manual for REW, link is on top of this page: https://kimmosaunisto.net/

Your data looks weird, or different than many show, because it contains sound of the room.

For crossover work we need only sound of the speaker, not arbitrary reflections from the room that always change depending on room and positioning, hence the measurement data needs to be avoid of sound of the room and include only direct sound of the speaker.

This is achieved by windowing the measured impulse, the impulse is cut off before any room reflections arrive to the mic after the direct sound. Typically this is only ~4 milliseconds time in a normal room speaker middle of room height. This is called semi-anechoic measurement, because we cannot get true anechoic measurements in domestic conditions and best we can strive for is maximally long window, speaker and mic as far away from any boundaries as possible, including the floor / ground.

If you can raise the speaker and mic to ~3m height and measure outside, you can use window as long as +10ms and get good data down to ~100Hz or so. Low bass wavelength is so long one has to do ground plane measurement outside far away from any buildings, or nearfield measurement. These are then merged with the windowed far-field measurement to get somewhat accurate data of the speaker for the whole bandwidth without using actual anechoic chamber.

In addition you also want the mic to be further away from the speaker while measuring, about 3x largest dimension of the face to get mostly into speaker far field so that the data represents sound at the actual listening distance and not include anomalies that are at speaker near field.

In general, you can make only as good of a crossover as your data is, so take your time, expect all sorts of errors and expect to spend some time to wrap your head around all these practical issues to get good data that represent acoustic radiation of the speaker with minimal error.
 
Last edited:
Thank you @tmuikku

I will review the manual you suggested. I notice that it recommends against USB measurement mics like the UMIK-1. But is it worth ditching my ECM8000 and getting an analogue measurement mic with a calibration file?

At this point, I do not have the option to perform measurements outside. So I have to work within that limitation for now.
 
Hello everyone,

Just about to get started with the impulse testing of my first SyntripP and I thought I would first address point 6 in my previous post, which I assumed was something rattling in the cabinet rattling.

Well, I am now quite certain the problem is the drivers. I removed both drivers and tested them out of the cabinet, and the noise is still present.

I tested a brand new 10CL51, also fresh out of the box, also in free air, and there's no noise.

So it seems that somehow I've damaged the drivers, which is a massive PITA because I thought I had my limiters set up OK, and I did not push the SyntripP that hard (or at least I didn't think I did).

Would really appreciate the opinions of the more experienced guys on this :

The problem is a buzzing sound on both 10CL51 drivers when they play the bass notes. I made two videos - one with drivers mounted and one with a driver playing in free air. It's a little hard to hear through the phone recording of course, but I hope the videos make it clear enough.

You can only hear this noise when the mid section is played in isolation. As soon as the HF is unmuted on the DSP, the buzzing sound from the mid is inaudible.

If I move the HPF on the mid section up to about 160-170Hz, the sound also becomes pretty much inaudible.

Obviously the situation I really want to avoid is putting the new pair of drivers from my as-yes-unbuilt second SyntripP into the first box and then do the same thing again.

Has this been caused by driving the 10CL51s too hard? Are they toast?





 
Hello, my bet is on cable leads touch the cone. If that's the case wires are too long which makes it poor quality control issue, something you should give feedback to the seller and directly to the company. One of my two Faital 15FH520 did/does the same. Some other woofers I have route the leads through spider so the same issue isn't there.

You can test this theory, play the system open air and use finger or stick to prevent the leads to touch the cone and hear if it helps. You might be able to just pinch the leads together using two fingers to move them just teeny bit further from the cone and stop the noise. Fix would be to tie them together permanently with piece of string.
 
Last edited:
  • Thank You
Reactions: FartyMcfly
I encountered a similar issue on some old JBL drivers a few years back. It was due to a poor reconing job, but the symptoms sound identical to yours. I tried to fix it with various methods but ended up having to recone the driver and position the leads further from the spider.

Worth trying to bend the leads slightly though and maybe a small amount of hot glue to hold them. I couldn't get the glue nozzle in close enough to have any success there.

1000015292.jpg
 
What is happening around 14k?
The DE900TN has a +8dB narrow peak in that frequency range.
Screen Shot 2025-03-21 at 12.36.48 PM.png

I'd suspect minor measurement angle changes around that frequency will result in erratic response deviations when measured in the near field.
Mid

Both. A big drop around the crossover. And a bump at 4k
The 10" response should drop at around -6dB per octave below the upper horn gain.
Yours is dropping at over -24 dB per octave, which suggests either a HP filter is applied, or one of the drivers is wired with reversed polarity.
Both. There seems to be a bump at 400Hz and 600Hz (or a dip at 500Hz?).
Also could be indicative of a polarity reverse.
2 - The goal is to get the frequency response as flat as possible on the graph right? I find myself reducing the HF gain down to -9 or -10dB vs mid in order to listen, otherwise the HF is painful. Is this normal / expected?
Flat, or a gently sloping downward HF response is the usual goal.
The HF should have higher sensitivity than the 2x10", but again it sounds like one of the 10" is reversed polarity.
3 - Are the charts from REW useful without smoothing? If smoothing makes the chart more readable, which level of smoothing should I use? it seems that 1/3 or 1/6 produce a beautiful looking line, but hide a lot of large variations in frequency response.
Smoothing makes it easier to see the "big picture" to determine the low Q filters that are needed to shape the general response.
Small peaks visible with no smoothing may change off axis, so be careful using any high Q corrective filters, make sure they are required globally.

4 - What is happening here? Should I attempt to reduce 400Hz and 600Hz with very high Q EQ?
Again, a -35dB drop between 100Hz and the first upper peak is not right.
Measurement distance won't result in that kind of error.
5 - I have put some wadding in the top, bottom and rear sections of the back of the cabinets, but I was unsure whether this will make much difference in this SyntripP because I built the SyntripP without ports in the horn mouth. Does the back of the cabinet still need a lot of damping? The wadding I have is substantially thinner than the 3" fibreglass which Art suggests, so I will get some fibreglass and improve this if the consensus is that it will improve the performance.
The sides and back either side of the HF driver should be covered, but the difference is not going to be huge, especially with that thin polyfill, it's only doing a fraction of the damping fiberglass of the same depth would achieve.
That said, with the sealed box, you are only hearing back reflections through the cone, so the effect of damping will not be as apparent as through ports.
6 - Possibly related to point 5....When playing the mid section only, there is distortion of some kind. The best way I can describe it is almost a kind of static, like an out of tune radio.🙂
The tab for the tinsel leads is obviously pushed way in, probably when you pushed on the female spade connectors:
Screen Shot 2025-03-21 at 12.08.44 PM.png

Bend the tab back to the normal angle and the buzzing should stop.
It is a good idea to cable tie the leads to a speaker spoke so the connectors are not carrying the wire weight, and can't shift into contact with the cones.

Check the polarity of the woofers, easy to see the cone movement direction applying a 9 volt battery voltage across the parallel "+/-" connection.

And after determining correct polarity for the woofers, you will want to wire the HF polarity for the greatest output in the acoustic crossover region with no time correction.

Art
 
Last edited:
Looking at the photos the non-buzzing driver the tab that holds binding posts seem to be bent out some, which streches the leads further from the cone. You could likely bend the tab to fix the issue on the buzzing drivers.


The tab for the tinsel leads is obviously pushed way in, probably when you pushed on the female spade connectors:
Screen Shot 2025-03-21 at 12.08.44 PM.png

Bend the tab back to the normal angle and the buzzing should stop.
It is a good idea to cable tie the leads to a speaker spoke so the connectors are not carrying the wire weight, and can't shift into contact with the cones.

Thank you both of you. Wow, a proper beginner error. Yes, it appears that the tinsel leads were making contact with the cones on both drivers after I'd bent the binding post while mounting the spade connectors. This is the kind of mistake you only make once eh.

Well, I tested the drivers in free air and the rattling sound was gone. Problem solved!

Then I put them back in the enclosure, aaaaaaaaaaand the buzz is back. Two steps forward, one step back etc.

To cut a long story short, I had two problems

1 - tinsel leads touching - now solved.

One issue remains :

2 - after a load of testing and moving things round, I can say with certainty that the HF buzz / static sound from the front of the SyntripP enclosure is actually coming from the HF compression driver. The buzz goes away if I electrically disconnect the driver by unplugging it from the amp. When the HF driver is connected, low frequency content in the music is making the HF driver buzz even though the HF channel is muted on the DSP.

I have the mid drivers connected to amp output 1, HF to amp output 2. DSP mid output is connected to amp in 1, DSP HF output is connected to amp in 2 but muted. Amp is set for stereo operation.

So it seems like there's some bleed / crosstalk from output channel 1 (mid) of my SyntripP amp making its way onto output 2 of the amp (HF). Because the mid channel has signal going down to 80Hz, it looks like it's causing the compression driver to make a noise which it shouldn't!

My SyntripP amp is actually four channel as I hoped to power two SyntripPs from one amp, so I also tried moving the compression driver to a different amp output and the problem remained.

I then temporarily moved the entire test setup to my sub / kick amp , and the problem went away completely. So I am quite certain that the problem lies with the SyntripP amp - it is not a problem with the DSP or the wiring etc etc.

So it seems that my SyntripP amp (it's a cheap thing - DAP Audio QI-4600) is not up to this job. I will contact the supplier tomorrow and see what they say.

When playing music with mid and HF channels active, the SyntripP sounds really nice and the buzz is not audible. But could this crosstalk (or whatever it is) damage the compression driver? I guess I need to measure channel output 2 (the HF channel) of the amp when there's no input on channel 2 to see what it's actually doing right?
 
  • Like
Reactions: tmuikku
But could this crosstalk (or whatever it is) damage the compression driver?
Probably not, unless its level increases as input level is advanced.
I guess I need to measure channel output 2 (the HF channel) of the amp when there's no input on channel 2 to see what it's actually doing right?
Yes.
When playing music with mid and HF channels active, the SyntripP sounds really nice
Can't understand how it would sound nice with the mid response dropping at over -24 dB per octave.
Did you check polarity of the 10" drivers?
 
Did you check polarity of the 10" drivers?
Yeah, I tested both drivers with a 9V battery while they were wired in parallel as you suggested. Both drivers move out or move in together when the 9V battery is connected, so I can confirm that both drivers are wired with the same polarity. For what it's worth, the drivers move out when +Ve is applied to the red terminal, and in when -Ve is applied to the black terminal.

Can't understand how it would sound nice with the mid response dropping at over -24 dB per octave.
I'm dubious about my test results based on everything that tmuikki mentioned.

I would like to resolve this issue with the amp, then measure the SyntripP again with impulse testing in the way @tmuikku suggested and then report back with (I hope) more meaningful data.

One question for you Art : How does one work out or otherwise discover the exit angle of the compression driver? I searched for this info for my comp driver (DE900TN) but it did not seem to be in any datasheet or officia repository I could find. It was on a to-do list from months ago to check this because I'm currently using a thraot adapter plate design based on the original Celestion comp driver.
 
One question for you Art : How does one work out or otherwise discover the exit angle of the compression driver?
Shine a bright light into the throat, look across, use a protractor or folded piece of paper to match the angle you see.
The DE900TN throat angle is probably ~6.5 degrees.
The angle could be called 83.5 (complementary), or if measured as an interior angle (like the nominal dispersion of a horn) 13 degrees.
Interior:ComplementaryAngles.png


Art