Jump to content

Ricci's Skhorn Subwoofer & Files


Ricci

Recommended Posts

Samps it's really no different than a vented cab. Look at the group delay measurements for the cab or the spectrogram. Consider these during design in your simulation software. In general the deeper the frequency the less group delay or ringing is of a concern IMHO. Below 25Hz it is much more difficult to hear than at 100Hz. 100Hz is less of a concern than 300Hz, etc. Pay more attention to the upper / XO range of the design you are working on. You don't want sudden spikes in delayed energy or ringing in narrow frequency bands either. As long as GD is smooth over the operational bandwidth and remains below 1 cycle of delay I don't worry about it. For an example of a bad sub issue that was audible look at the DTS10 at 55Hz. Also even if the cab imparts significant overall time of flight delay such as with a large FLH the mains can be delayed relative to the subs to match more closely over the XO region. A small room such as most HT's will greatly affect this far more than a few ms delay inherent in the design.

Link to comment
Share on other sites

Thanks.   I haven't decided on whether to build the half skhorn or try to modify my existing ported cab.  My theory is that I already have the ported section and it performs exactly how I want it to.  Seems simple enough to just add the horn section onto it.  Biggest difference would be that the mouth would not be adjacent to the port.   I could actually build the mouth as big as I want if it would help any.  

Link to comment
Share on other sites

  • 3 weeks later...

635 Liters net reflex enclosure, 1680 sq  cm port, 40 cm length measured in the same place, same mic, same calibration, same windowing. I am very surprised that the Bandpass SKHorn XL loads the driver that low. The enclosure is very big and I don't like how it sounds on music, it lacks the real kick. But the deep end is very powerful and clean.

comparisson BPH vs reflex.jpg

Link to comment
Share on other sites

On 7/27/2018 at 7:13 PM, Ricci said:

That's what I'm talking about... So you would still be a little ahead overall vs 6 Xoc1 TH18 and still no comparison down near 30Hz.

What do you think of the sound subjectively?

Add 6 dB for 2 meter results.

 

CEA 2010 4 meters
1*F1218 ohms MC2 E45 bridge
25 Hz 102
30 Hz 109
31.5 Hz 111.1
35 hz 114.8
40 Hz 121.1
50 Hz 126.3
63 Hz 128.5
80 Hz 130.2

1*F221 4 ohms E45 BRIDGE
25 HZ 108.7 dB 
30 HZ 115.4
31.5 HZ 117.5
35 HZ 122.9
40 HZ 129.2
50 HZ 133.4
63 HZ 135.5
80 HZ 135.8
100 HZ 136.9

1*TH18 8 ohms 4 METERS MC2 E45 BRIDGE

25 HZ 104.4
30 HZ 117.3
31.5 HZ 119.1
35 HZ 122.2
40 HZ 123.3
50 HZ 127
63 HZ 128.3
80 HZ 128.8
100 HZ 130.6

2*TH118 8 ohms 4 METERS MC2 E45 BRIDGE

25 HZ 110.8
30 HZ 124.3
31.5 HZ 125.9
35 HZ 129 AMP LIM
40 HZ 130.3 clip
50 HZ 133.2
63 HZ 134.2
80 HZ 134.4
100 HZ 136.7

 

The results I got with one and 2 F221 this summer were not ok, something was absorbing a lot of power under 60 Hz. I got +6 dB now from them today in comparison at 25 and 30 Hz. But still no match for a single SKHorn XL. It seems that it is very close to 4 TH18 at 30 Hz and 3 of them under 100 Hz. 

 

 

Link to comment
Share on other sites

On ‎10‎/‎20‎/‎2018 at 9:47 AM, radulescu_paul_mircea said:

635 Liters net reflex enclosure, 1680 sq  cm port, 40 cm length measured in the same place, same mic, same calibration, same windowing. I am very surprised that the Bandpass SKHorn XL loads the driver that low. The enclosure is very big and I don't like how it sounds on music, it lacks the real kick. But the deep end is very powerful and clean.

comparisson BPH vs reflex.jpg

Lacks kick? That's odd to me. Not the way I'd describe mine at all. What was the processing used while listening? In room or outdoor? Compared to what?

Which RCF driver is in the measurement? Is it measured in the same cab as the 21Ipal reflex? I've noticed something over the years in vented cabs with that type of response shape. In general it looks good but will tend to exhibit greater losses and compression near tuning and less low bass output with higher distortion. Indicative that the airspace is undersized and the vent is not loaded sharply if that makes sense.

21 hours ago, radulescu_paul_mircea said:

The results I got with one and 2 F221 this summer were not ok, something was absorbing a lot of power under 60 Hz. I got +6 dB now from them today in comparison at 25 and 30 Hz. But still no match for a single SKHorn XL. It seems that it is very close to 4 TH18 at 30 Hz and 3 of them under 100 Hz. 

Thanks for posting the extra measurements. I'm always glad to see data.

What do you think the cause was with the lower values on the previous tests? Perhaps a limiter setting left on? Same amp used?

Link to comment
Share on other sites

@Ricci I messed up the phrasing on that. I meant that the RCF loaded enclosure with 29 hz tuning lacked punch and depth. The enclosures with RCF and 21Ipal are the same.

After I changed it to 33 Hz, increasing the surface with 30% and decreasing the length with the same percentage, it improved quite a bit, but still no comparison to SKHorn XL, which indeed is absolutely devastating in that category! New tuning in green.

It makes sense what you said about loading. TBH I don't think the main problem lies with the port losses, but with the magnetic strength and damping. Off course, the performances of a reflex enclosure is in fact simple to deduce from measurements and Sims and also is pretty simple to design properly. How it will sound though based on that,you could never really know until.you build it. I found something interesting regarding group delay and linearity. If it is my feeling that a lower group delay is good, but it makes things sound thin. A higher and more even group delay will exacerbate the lower frequencies and it will make them jump into attention. 

FB_IMG_1540211129524.jpg

Link to comment
Share on other sites

Another important fact are the distortions. For the same input voltage, 28.3V , between all 5 designs, 2*TH, 1*F221, 2*BR29hz RCF, 2* BR29Hz Ipal, 2*BR33Hz RCF and 2* SKHorn XL, the last one is the best one, with under 2.5 % distortion from 31-110 Hz. Compared to F221 that has up to 7%, or BR33hz RCF that goes to 5.8% at 40 Hz.. the proportion of the orders of harmonics are also very different and the ones with the drivers close to the surface have their own sound signature. I believe what I can hear is probably some very small air turbulences across the drivers diaphragms and suspensions and other things because of the fact that the acoustic impedance mismatch will always have a part of the air molecules wiggling over the surface, slipping away to the edges. Also any noise from the driver itself is not masked, but that is just it. 

Link to comment
Share on other sites

Ah...That makes more sense. I was baffled thinking that your XL lacked punch!

I also worded my response badly when talking about the RCF driver reflex. I agree with damping and motor strength as the main contributing factors. I've found that reflex systems with a flattened knee at the vent tuning (undefined or less defined rolloff or tuning point. RCF measurement) seem to perform worse overall than reflex designs with a very defined sharp knee (rolloff below vent tuning, 21Ipal measurement).

THD is one component, but there are also non harmonic mechanical and operational noises that seem to give a signature to the sound of direct radiator cabs. Especially when the speaker is being pushed hard. I believe that this second factor is an important component of why people like the sound of horns and some BP subs and say that they sound different. This was a big consideration for me when I designed the Skhorn. It allows horns and BP's to be pushed further before these driver noises to start coloring the sound.

How do you like the MC2 amplifier? They are rare over here in the US.

Link to comment
Share on other sites

I've commented before that I believe punch depends substantially on the linear response of the system in the upper bass as well as the sub.  I'd go even further and say that mid and high frequencies can matter a lot too.  Even fairly tiny resonances in the treble can perceptually overwhelm the bass, causing several dB reduction in subjective output.  Why this happens is a good question for scientists to answer some day.  In any case, punch depends *crucially* on linear response, and a major fault *anywhere in the spectrum* can be harmful.

Note also that native sub responses are mostly minimum phase, which means that measurements like group delay and waterfall are not really distinct from frequency magnitude response.  What you'll typically see is higher group delay and waterfall decay in and around regions where frequency response slope is changing a lot, such as around the vent tune or sealed box resonance but also around higher frequency resonances that appear in the driver, port and/or horn.  As such, I believe it's helpful to focus on (native / anechoic) frequency magnitude response shape when talking about linear response in general.

Of course linear response is not the whole story with subs, especially when they are driven hard.  In general, non-linear response is very complicated and strongly dependent on the signal.  THD and power compression measurements are the easiest ways of characterizing non-linear response, but unfortunately they are not very useful.  What we mostly care about is response to music which mostly involves transients and complex multi-tone signals.  THD and power compression only offer small hints as to what might happen there, and there are many mechanisms of non-linearity that neither THD nor power compression adequately reveal.  These include mechanisms that cause inter-modulation distortion (IMD) and noise, which are probably far more important perceptually speaking.

In terms of direct negative impact on sound and tactile quality, THD is probably least important.  Indeed, THD (both odd and even harmonic) is frequently added to music *on purpose* during mastering to give it a *subjectively improved* (and louder) sound.  This is done especially to create high frequency content an alternative to doing EQ boost, because the EQ boost often brings out more noise and more HF harshness caused by recording deficiencies.  The spectral balance of the synthesized harmonics can be very smooth, which will sound very good when reproduced on accurate speakers.  (It can sound harsh or bright on inferior speakers though).  I'd guess that more than half of music masters and probably most movie soundtracks get this treatment.  THD added to sub bass can potentially increase tactile sensation also.  In terms of equal SPL, the body has a lot more tactile sensitivity in the ~100-500 Hz range.  The reason why we feel subs as much as we do is because of the comparatively higher SPL.

3 hours ago, Ricci said:

THD is one component, but there are also non harmonic mechanical and operational noises that seem to give a signature to the sound of direct radiator cabs. Especially when the speaker is being pushed hard. I believe that this second factor is an important component of why people like the sound of horns and some BP subs and say that they sound different. This was a big consideration for me when I designed the Skhorn. It allows horns and BP's to be pushed further before these driver noises to start coloring the sound. 

I think the bolded part along with IMD, which I presume is often caused substantially by non-linear inductance effects, are key when it comes to distortion.  Almost all music involves transients (wide-bandwidth) or multi-tone signals, and the mechanisms that influence IMD but not necessarily THD will impact reproduction of both of these types of signals profoundly.  Of course, IMD depends a lot on driver characteristics too but the cabinet reshapes not only the linear response but also distortion components of the driver.  A cabinet that naturally suppresses higher order overtones (harmonic or otherwise) and noise is likely to help the sub "get out of the way" of the rest of the system in the upper frequencies that are crucial to punch. 

Probably any sound that is produced that is not harmonically related to the signal, including noise and IMD is very degrading to sound quality, especially when it reaches into the higher frequencies where the ear is much more sensitive.  One experiment I think would be interesting would be to try to strategically adding some absorption into one or more passages to try to further smooth and suppress the output of mid / high frequencies.  Obviously there's a trade off as far as losing "CEA output" and possibly adding some power compression, but there may be a "sweet spot" which gives an even cleaner sound and perhaps a tad more "usable output" if one's trying to avoid any hint of audible strain.

3 hours ago, Ricci said:

I also worded my response badly when talking about the RCF driver reflex. I agree with damping and motor strength as the main contributing factors. I've found that reflex systems with a flattened knee at the vent tuning (undefined or less defined rolloff or tuning point. RCF measurement) seem to perform worse overall than reflex designs with a very defined sharp knee (rolloff below vent tuning, 21Ipal measurement).

Are you saying that the systems with a sharper, more defined knee give better punch?  Or that they are better in general?  I'm a bit confused because sharper knees are typically associated with more internal resonance, less "control", less damping, and/or less motor strength, which seems a bit counter-intuitive.  In terms of tactile sensation, the subjective emphasis of frequencies around that knee could help or hurt depending on your baseline. It's really complicated and I don't really have a formula for determining how a particular linear response will feel.

I am wondering @radulescu_paul_mircea , have you tried EQing the lower part of the response of one cabinet to look like another?  I'm pretty sure doing so will change the sound quite a bit, but I'm curious as to how much it closes the gap in terms of "kick" or whatever.  Another thing to consider is that a sub with more extension may substantially activate room modes that were largely untouched before, and so supressing just the new room mode(s) with EQ could help clean up the sound and get the "kick" back.

Link to comment
Share on other sites

I have to run but Paul and myself are discussing 2 different things. One of which is punch, which I had thought he said that the Skhorn lacked in that department, which made me think he was crazy town...He meant the RCF 21 reflex sub. The other subject was the shape of the response knee at vent tuning, something I started about the RCF 21 reflex by making an observation about the response shape. 

Link to comment
Share on other sites

CEA 2010 measured at 2 meters
F121 MC2 E45 bridge normalised to 2 meters 
25 Hz 108
30 Hz 115
31.5 Hz 117.1
35 hz 120.8
40 Hz 127.1
50 Hz 132.3
63 Hz 134.5
80 Hz 136.2

F221 measured at E45 BRIDGE normalised to 2 meters 
25 HZ 114.7 dB 
30 HZ 121.4
31.5 HZ 123.5
35 HZ 128.9
40 HZ 135.2
50 HZ 139.4
63 HZ 141.5
80 HZ 141.8
100 HZ 142.9

1*TH18 xoc1 measured at 4 METERS MC2 E45 BRIDGE normalised to 2 meters 

25 HZ 110.4
30 HZ 123.3
31.5 HZ 125.1
35 HZ 128.2
40 HZ 129.3
50 HZ 133
63 HZ 134.3
80 HZ 134.8
100 HZ 136.6

2*TH18 xoc1 measured at 4 METERS MC2 E45 BRIDGE normalised to 2 meters 

25 HZ 116.8
30 HZ 130.3
31.5 HZ 131.9
35 HZ 135 AMP LIM
40 HZ 136.3 clip
50 HZ 139.2
63 HZ 140.2
80 HZ 140.4
100 HZ 142.7 

Edited by radulescu_paul_mircea
Link to comment
Share on other sites

Paul if you keep this up you might as well start adding to the Data-Bass. I can get you a login and password you know...You seem to have better access to the high end pro speakers than I do. I feel like this stuff is going to get lost in here and is getting a bit off topic. I can move it out to another thread...

Why the switch of amps for the RCF test? I assume this is outdoor GP testing? Are you using the bursts inside of REW?

  • Like 1
Link to comment
Share on other sites

13 hours ago, Ricci said:

Paul if you keep this up you might as well start adding to the Data-Bass. I can get you a login and password you know...You seem to have better access to the high end pro speakers than I do. I feel like this stuff is going to get lost in here and is getting a bit off topic. I can move it out to another thread...

Why the switch of amps for the RCF test? I assume this is outdoor GP testing? Are you using the bursts inside of REW?

All the measurements were made outdoor ground plane. One space where I measured the TH and F121 , a wall of the building was 14 meters away but the rest of the space is a lake and a field. I truncated the impulse response to eliminate as much of the load as possible. 17 meters the reflection came 100 ms after so I had very good resolution and clean response down to 10 Hz.

The space where I measured Stasys XV2 that I will not publish, the BR with RCF , the SKhorn XL and others is a very large space where the building behind and on one side was at 25 meters and the other 2 sides were empty for more than 100 meters. 

I switched from MC2 to clone and X4 because I own a clone, a K20 and I have 2 X4 for an install that I could use but the MC2 was installed in a fixed place where Funktion and TH are installed.  

I am using REW for generating and measuring the output, Focusrite soundcard, earthworks M30 calibrated mic , a class 1spl calibrator , a Behringer calibrated mic for comparison purposes.

Unfortunately one of the RCF drivers kicked the backplate with the coil former and it isn't in the perfect shape right now.

I will soon test the SKhorn XL with 21ID , each driver on a K20 channel to see the numbers. Probably next week.

Link to comment
Share on other sites

Thanks for details on the equipment and testing areas. I'd suggest bridging the K20 or any bridgeable amplifier for high power bass testing. It definitely performs better for longer term signals. It may not make much difference for burst testing though.

Sorry to hear about the RCF. I've slammed my fair share of drivers but the B&C and 18Sound stuff seems exceptionally resilient to mechanical bottoming. Perhaps RCF is slightly less so. Was this the 451 or 551 driver?

Link to comment
Share on other sites

I did the measurements again to directly compare the 21ID Vs 21IPal

Measured at 4 meters, normalised to 2 meters 

CEA2010 2 m K20 bridge 

BPH221 ID 4 ohms 
20 Hz 104.6
25 Hz 126.2
28 Hz 132.2
30 Hz 133.4
31.5 Hz 133.1
35 Hz 133.1
40 Hz 137.0
50 Hz 141.2
63 Hz 140.1
70 Hz 142.0
80 Hz 143.2
100 Hz 144.5

BPH221 Ipal 2 ohms 2 meters
20 Hz 105.8
25 Hz 126.9
28 Hz 134.5
30 Hz 136.6
31.5 Hz 137 5
35 Hz 135.4
40 Hz 138.6
50 Hz 140.7
63 Hz 142.8
70 Hz 144.0
80 Hz 144.6
100 Hz 145.1

I love these tests. The amp asked for huge amounts of power at 28 Hz where the port is tuned. I've seen 720 V peak to peak and 110 amps on the output with series IPal before the 20A breaker dropped. 

I can say that the drivers in series are a problem, a big one IMO. Very different excursion at 35 Hz between them, that's why the output is lower there 
 

Link to comment
Share on other sites

1 hour ago, radulescu_paul_mircea said:

I did the measurements again to directly compare the 21ID Vs 21IPal

I can say that the drivers in series are a problem, a big one IMO. Very different excursion at 35 Hz between them, that's why the output is lower there 
 

How do you know this? Do you have cameras in the cab or some sort of viewing window?  I don't see how you would confirm this? I'm doubtful.

 

Mine do not exhibit issues with this. The drivers have separated vented chambers partly because of this reason and these control the majority of the driver excursion in this design until above 65Hz where there isn't much excursion. The horn section only loads down to about 65Hz. I know it seems like the pressures in the throat would be insane and amplify any small difference in driver behavior but it simply isn't a big factor compared to the motor force of these drivers on the voice coil and the individual vented chambers. This wouldn't be too much different than a pair of individual vented subs both operated facing into a corner or in a vehicle hatch which people do all the time.

35Hz is getting close to maximum driver excursion in band. Output is about what would be expected...40Hz is higher frequency, 31.5Hz is loading on the vent. 135dB at 35Hz in this setting is going to be well past xmax.

 

Link to comment
Share on other sites

2 hours ago, Ricci said:

How do you know this? Do you have cameras in the cab or some sort of viewing window?  I don't see how you would confirm this? I'm doubtful.

 

I put my camera inside many times. it does appear to have a problem there. Also the behavior is different when drivers are in parallel, the response between 32-39 hz is 1 db higher in sensitivity, but this I still need to verify again. when the hatches are open and i pass a sine wave anywhere under 50 hz, the drivers excursion is visible and the difference can be very big, like double somewhere under Fs. 

I'll instal the Ipal modules on them loaded with ID and I have decided shorten the ports to go to 32 Hz instead of 28 Hz as it is now. I wanted them to be 30 hz, I thought they ended up at 29 but 28 is definitely too low, especially for the 18Sound ones. I need to cut 16 cm out of them to be able to get close to where I want them to be but I have a vertical brace in my way. from the experience with the BR with RCF and from sims, going to 32 Hz will increase the output by about 3 db under 45 hz, the output at 30 will be the same and under that it doesn't matter anymore, because the place where they will be installed doesn't need anything that low. 

Again, I'll have to say about the sound of it, it is absolutely fenomental. The impact that 2 of these have is awesome. I'm in the process of designing and building another set of 2 slightly modified but the same dimensions. I can't wait to hear them together.

I took a measuerement between front and back. Unknown voltage, 2 subs one on top of the other, equitemporal distance, so the distance are matche in impulse delay, not in ruler. very good directivity!

 

 

2 subs front back.jpg

BR VS BPH.jpg

Link to comment
Share on other sites

If you look at the measurement I posted since I built these 2, youll see they all are a bit different, especially between 32 and 45 hz. The first ones are with the drivers in parallel, the last ones are with drivers in series. A different Q behavior... 

Look also at the response of one enclosure vs 2 of them, the ports get better with 2, a higher Q and at least 1 hz lower extension.

Off course some of them differ because the space where I measured them is different and no matter what post analisys I use, some reflections will get into the final impulse response, otherwhize I loose resolution and accuracy down low. these differences account for about +- 0.5 to 1 dB sometimes.

 

 

Link to comment
Share on other sites

Paul,

I've been busy.

Why do you think the series connection is causing some sort of performance impact? Is it because of the info I posted about the Skhorn/ MAUL design process and uneven driver loading and excursion profile with the MAUL? In that case the word "series" had nothing to do with electrical wiring. The word "series" is in regards to the acoustic loading of the drivers. The drivers were physically in series acoustically inside of the horn/slot. This causes uneven cone loading not the electrical wiring. The Skhorn is symmetric so this does not apply.

Series connection vs parallel connection of the woofers should not amount to anything significant unless the 2 drivers are way off spec from each other, in which case you will still see this with parallel wiring.

Any objective quantification of it would have to be very strictly controlled as well. Same cab, same drivers, measurement equipment, same environment, same day, etc. If you have some of this I'd like to see it. Impedance curve taken back to back with the wiring resistance compensated would be ideal.

Link to comment
Share on other sites

I am referring to electrical series not acoustical. I confirmed my theory by Akabak first and then I measured the voltage at the drivers connectors and they differ by up to 6 V when driven with 80 V. So one driver gets 37 V and another 43 V at the same time measured at 38 Hz where driver excursion is maximum. 

I'm going to test them as you said Sunday. Outdoors with drivers in series and parallel impedance and magnitude.

 

Link to comment
Share on other sites

40 minutes ago, radulescu_paul_mircea said:

I am referring to electrical series not acoustical. I confirmed my theory by Akabak first and then I measured the voltage at the drivers connectors and they differ by up to 6 V when driven with 80 V. So one driver gets 37 V and another 43 V at the same time measured at 38 Hz where driver excursion is maximum. 

I'm going to test them as you said Sunday. Outdoors with drivers in series and parallel impedance and magnitude.

 

Ok.

What were the driver specs used? Different for each driver I assume? That is a large difference in voltage. This would likely be caused by differences in the drivers performance rather than the wiring type unless there is a big difference in the amount or type of wire used between each driver. It makes me wonder how well matched the drivers are. If it isn't too much trouble flip the polarity or wiring on the drivers when you do this and see if it stays with a driver.

Perhaps I'll investigate on mine next time I have a chance.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...