Thanks for your response! When looking at hificompass.com and comparing the step response of the RS180P-4 with a few other mid / bass drivers, it seems to mee the step response is less favourable compared to most others. Would this be an aspect to take into consideration when selecting drivers?
(source: https://hificompass.com/en/speakers/measurements/sbacoustics/sb-acoustics-sb17nrx2c35-4)
(source: https://hificompass.com/en/speakers/measurements/dayton-audio/dayton-audio-rs180-4)
Whats the issue you see in the step comparison above?
The Dayton looks OK to me
Last edited:
The Dayton seems to drop a little less the first 2 ms. Not sure if this tells anything about soundquality?Whats the issue you see in the step comparison above?
The Dayton looks OK to me
I dont know if the step response are that useful for a woofer doing bass? The step response looks to be a wide frequency impulse and includes the cone break up and ringing at > 8kHz for the Dayton.The steps all seem to return to input baseline at about 2ms. The waterfall plot being multidimensional separates out the high freq ringing but still doesn't show much about dampening of the low freq for woofers.
In the woofer measurements I look at the FR to see the SPL for the chosen freqs and so make sure the low freq doesn't fall off too fast and then look at the HD plots down low. They HD plots are all bad <100Hz but some worse.
For bass the IMD plots look really useful. None for the Dayton but IMD on other woofers shows where the frequency response is messed with IMD aharmonic tones eg at -22dB and wild HD3 at -6dB. Makes DACs look benign.
In the woofer measurements I look at the FR to see the SPL for the chosen freqs and so make sure the low freq doesn't fall off too fast and then look at the HD plots down low. They HD plots are all bad <100Hz but some worse.
For bass the IMD plots look really useful. None for the Dayton but IMD on other woofers shows where the frequency response is messed with IMD aharmonic tones eg at -22dB and wild HD3 at -6dB. Makes DACs look benign.
Yep. But the main question remains: is the step response of any value for a woofer? Maybe its best just ignored?
Hi,
it depends
Filtering with a lowpass filter most of the HF-ripple will vanish anyway.
Now it depends on the amount of peaking, the frequency of the peak(s), the projected xover frequency and the steepness of the filter flank.
The farer the xover frequency and the peak are apart, and the less peaking the unfiltered driver shows, and the steeper the filter flanks are, the less influence will show up in the filtered driver response.
If peak- and xover frequency are coming close one might need to add dedicated notch-filters to equalize the response .... the longer decay of the breakups will of course remain existent.
So , for the typical low xover frequencies in our applications here - with xover and peak differing by two ore more octaves- breakups beyond 1kHz are typically of very little concern.
The low frequency filtered step response will be slow but clean.
jauu
Calvin
it depends
Filtering with a lowpass filter most of the HF-ripple will vanish anyway.
Now it depends on the amount of peaking, the frequency of the peak(s), the projected xover frequency and the steepness of the filter flank.
The farer the xover frequency and the peak are apart, and the less peaking the unfiltered driver shows, and the steeper the filter flanks are, the less influence will show up in the filtered driver response.
If peak- and xover frequency are coming close one might need to add dedicated notch-filters to equalize the response .... the longer decay of the breakups will of course remain existent.
So , for the typical low xover frequencies in our applications here - with xover and peak differing by two ore more octaves- breakups beyond 1kHz are typically of very little concern.
The low frequency filtered step response will be slow but clean.
jauu
Calvin
Hi,
regarding the Q which driver to use, either the Dayton or the SB, I´d say it´s just a matter of Your personal preferences.
Both look very well executed and don´t differ much.
Built into dipole cabinets andd filterered and eq´d actively the differneces between both will probabely vanishingly small.
Budget-wise the SB18 has a clear edge.
The Dayton´s more glossy looks almost cost twice.
You can almost double the number of drivers with the accompanying increase in max. SPL, reduced excursion and THD level using the SB18.
Regarding the number of drivers I found 6, resp. 8 drivers sufficient to match well with my small, resp. large panel´s max. SPL when HP-filtered at around 50Hz.
If not utilizing a dedicated sub and eq´ing down to 35Hz 8 drivers match my small panel perfectly (6 might just be a tad too low at extreme SPL-levels approaching 110dB@4m).
There´s certainly no need to utilize 12 drivers SPL-wise.
Only if the ESL panel would be very tall, I might think about to use more than 8 drivers, but just to mount them more closely to each other to reduce vertical comb-filter ascpects.
jauu
Calvin
regarding the Q which driver to use, either the Dayton or the SB, I´d say it´s just a matter of Your personal preferences.
Both look very well executed and don´t differ much.
Built into dipole cabinets andd filterered and eq´d actively the differneces between both will probabely vanishingly small.
Budget-wise the SB18 has a clear edge.
The Dayton´s more glossy looks almost cost twice.
You can almost double the number of drivers with the accompanying increase in max. SPL, reduced excursion and THD level using the SB18.
Regarding the number of drivers I found 6, resp. 8 drivers sufficient to match well with my small, resp. large panel´s max. SPL when HP-filtered at around 50Hz.
If not utilizing a dedicated sub and eq´ing down to 35Hz 8 drivers match my small panel perfectly (6 might just be a tad too low at extreme SPL-levels approaching 110dB@4m).
There´s certainly no need to utilize 12 drivers SPL-wise.
Only if the ESL panel would be very tall, I might think about to use more than 8 drivers, but just to mount them more closely to each other to reduce vertical comb-filter ascpects.
jauu
Calvin
Hi,
oh sorry, mixed up all theses SB16whatthef* SB17strangenomenclature and SB18idontcare numbers.
Can´t they simply name them SB16 intro, medium and top?
Who but the king of nerds could possibly memorize all those crazy numbering and lettering?
Anyway, I just compared the two models from the Link, the Dayton and the SB18whatever.
jauu
Calvin
oh sorry, mixed up all theses SB16whatthef* SB17strangenomenclature and SB18idontcare numbers.
Can´t they simply name them SB16 intro, medium and top?
Who but the king of nerds could possibly memorize all those crazy numbering and lettering?
Anyway, I just compared the two models from the Link, the Dayton and the SB18whatever.
jauu
Calvin
Your panels look very professional! Excellent job, it looks like you have a very good handle on what you want to create. After reading though my h if the thread there seems to be the same issues that most everyone gets caught on with the bass.
Now I haven't made any panels (yet) had I known about electrostats 25 years ago it may be a different story since I was working in steel fab shops back then and had access to machines and material.
However, the last few years I have been working on this exact issue with panels made by JansZen and Infinity (RTR/JansZen). I knew instantly that the panels had great potential but we're unfairly matched with terribly lacking woofers. Typically the one gripe by everyone was how they integrated the two became a huge reason for many to dismiss them with prejudice.
Through the years (thanks to early car stereos addition of the subwoofer) woofers have been retooled to a point where they can be made to fit in every type of application requirement. But the timing for electrostats seems to have passed from mainstream thoughts, other than a few that found a way to make some adaptable designs that fit the changing times and found a place in high end AV systems. This has opened up a new DIY area that revives the panels use in an effort to fix that design area that couldn't be solved when they were new. Applying newer (no need to be brand new design due to large excursion dominating design now) subwoofers, really the best part is, at this moment in time the subs that were huge in the early 2000s, are now dirt cheap since they have smaller excursion and can play up the ladder of frequencies very flatly.
I'm not saying to match up old car subs with electrostats, even though you'll see this being done on very high $ stuff, the technology was trickled down on home woofers that could dig deeper and have higher impact from smaller and larger cones alike. Depending on amps used will determine the choice in drivers and number of them. Of course the more used, the more problems found in response. Using an isobaric design has been tried and must not have worked since I see none now using it. Depending on how low you need to get will still come down to size unless you really are set on making a marketable design and have to find a way to deliver the lowest tones at high pressures.
I looked into many gutted cabinets woofer drivers I had collected and a long list of car subs I have repaired and sold to see how they looked on typical home testing set ups using calibrated mics. This is by no means any kind of official test that has all the info printed in text and was more if an account of what was typically going on with woofers and how they fit my room and my ideas for using the panels.
I currently run 14 panels per side with 2 woofers, the woofers that I tested we're made after 2000 and most were able to mix fairly well with the panels, they needed to be able to show a decent response up to 1kHz (just a reference point I picked to be important) what I found was smaller drivers created large fluctuations between 100 to 500Hz, big spikes and sharp dips with 2x 8" woofers (not subs) and failed to play low enough to work for me. I wanted to stay away from 8" car subs since I felt they were never really well designed and we're made just to fill in that niche back then. Things have changed now but most will limit upper bass quality with design moved to support lower than 160Hz.
I also know from research than 2 of the exact same drivers can be a challenge to keep from reacting with each other. I did not explore this after the 8s all showed huge issues and moved to bigger drivers that had typical woofer excursion of home woofers. I tested a huge amount but found the 12" infinity kappa woofers (all black poly resin cones with cast baskets to do a remarkable job at plotting a good response thought the needed range but they did not have the output needed to match even half the panels.
I sought out a 10" driver to add, after doing more tests by themselves then while added to the 12s, I was lucky and had some drivers laying around for years that I forgot about and didn't know who made them. They turned out to be Monitor audio subs, when added they reinforced the bass perfectly.
Now the interesting part that applies to why you find mixing cone drivers to prove to be a problem more than a solution. I suspect it's based in the newer theories of magical DSP solutions being all the rage. Ik iw I fell for it myself and I was trying them one after the other only to find they all had limited applications and poor results due to the lack of memory and CPU power. It turns into an either or situation more than the specs imply. Even though they advertise these unlimited abilities but once you get them you find you're limited on what they can do. I found all of them had a profound affect on realism once the sound passed through their circuits. Simply destroying the perfection that I knew was there before. Hell running with I crossover sounded way better!
That's when it hit me, rather than working with these super high db slopes that we've been told are going to be better to use, but this is only because of the digital environment allowing this type of steep slope, you'd never see them used with analog components. I stated thinking a different direction, I got rid of all my DSP junk and tried some analog active crossovers and analog EQs. I found all crossovers are not made the same and no crossovers are either! I also found a 2 way crossover couldn't remove that obvious crossover point. This is what you have found as well?
Now knowing the issue and trying many points to cross at, I was convinced that I need more flexibility with my crossover. I ditched the DBX 2-way for a JBL 3-wsy and I had problem after problem with poor test results that I blamed on that crossover. I researched it and checked all jumpers inside and found it to be a complete waste of time and money. It has the worst possible way to configure it and was basically nads to correct fir their poorly designed pro audio speakers. I couldn't get anything to work in my favor and ditched it as well. So you see thinking 1 way and accepting those results without moving to another product can be a bad idea, I wasn't going to admit defeat yet and found a good brand new active crossover that has the ability to do what I thought I needed to do to get things the way I wanted them.
The Behringer super x pro 3 way active crossover was next in line for the task and I also found a issues with my previous analog eqs, basically they needed to be dual 31 bands to be able to work at the point required. Most if not all home pieces were simply junk and had no place in my system. I tried a DBX dual 31 band and found it was cheaply made and had issues with sliders even though it was fairly new. After inspecting it I opted for a replacement with only 1 extra requirement being it had to have volume controls that would go to zero out out so I could use it as a preamp.
I found every preamp I tried to be nothing more than an unneeded point noise and unbalanced output (more sound from left or right) I couldn't find that perfect centered headphone effect I had without but had no find tuning of volume unless I adjusted each amp constantly by a tiny fraction. Each track seemed to have a tiny imbalance and with 2 volume controls I was able to find it in a few seconds. This was a very real issue and I could not be without that control.
Now the juicy part! After about 2 years since I started working with electrostats I finally found the answer (for me and I'll bet most others will agree) the slopes on the crossovers were the issue. If I could find a way to allow more of the sound to extend across the crossover point I could blend it better. Only the best active crossover will allow tweaking the slopes but what if I used the mid point as a buffer that allows the high and lows to overlap? I talking real overlap, remember how I said it sounded better with none than with a DSP (but that not hard to imagine even when the woofer had a huge 3kHz peak of 9db)? What I needed was to make sure the woofer stopped before that.
Basically I now believe it can be accomplished with a 2-way active analog crossover, it took using the 3-way to figure it out and the 3-way provides perfect spot on results that can find the correct amount of lap and at the frequencies that work best. I had started originally looking at performance plots I tested and picking out where the panels rolled off and used that as the point just above it. That didn't work and I ended up with my lows crossed at 800Hz or so and my panels could run full range if I wanted too. But the are terminated at 660Hz. If the panels were able to run flat down to a lower frequency, I would have lowered it even more but they have some ups and downs at lower frequencies so I cut it off there.
All in all, a 2 way can be used as long as the panels don't create too much cancellation in low tones. The lap I have is most likely covering an octave and completely removes any noticable timber mismatching. That doesn't mean any woofers can be used, I would think that some testing is needed to find response in room before considering moving forward and dedicating them without any comparing. I cannot comment on how well a large excursion might work, my thoughts would be possibly if the panels can run flat to below 100Hz and the subs are getting huge amount of power to move them. I personally found they aren't needed and I have clean clear bass to 24Hz and have very little invested in the drivers sin e they were bought used. The other issue that I did not have to worry about but most might find is the resistance of the drivers becoming too low for many amps. I chose some old fan less QSC that are 1 ohm rated and use them for both high and lows.
Knowing the impedance curves may help but there's nothing that's more accurate than testing them in the real world. I know that computer programs can help but we all know everything changes once in room. I'm not making speakers to sell so this is what worked for me in my home.
The other thought I had was about your thoughts on trying to increase bass without a cone driver and the very first thing that popped in my head was to overlap panels. Then I read further, I see you considered that, very smart as I was confident you have considered everything far beyond what I could muster up. Then I read about the wings!?! I dismissed that as a possible solution but it gave me more ideas, right ir wrong, here they are.
I visualized a square panel that's wider than the one you have now and mounting it behind the other, making sure to be able to move it so you can test it. Set up a mic at center and play the frequency you want to boost. Move the bug panel around both side to side and up and down to see if you can find any boost. Since it's documented that panels reproduce different tones from different areas across each panel it might take a specific shape and size as well as placement according to your tall panel. I also could be entirely incorrect but I believe in trying over believing what's read without trying it for myself. I've found most if what's been proven never seems to work out on my stuff and my results are far from expected more often than not there's far too many variables in each situation to finalize or rule anything out entirely unless I try it myself.
I hope there's something in here of value, as I read through, I couldn't believe I landed on what I just had spent years seeking the answer! I was seeking something to support what I found and I was hoping that there was something that you found that I could compare notes with, it seems as though you have it stuck in your mind to be a closed subject judging by your initial reaction to others suggesting the cones, immediately supplying proof to stop any further discussion. What I didn't see was any tests involving no crossovers at all which is how I started and how I came to my conclusion. The other issue was using DSP which I would never recommend since I heard a very deteriorated signal with all the little fringe details stripped away. The digital twisting and stretching just doesn't work. I compare it to working with pictures in digital correction programs. You can only pull at it so much before detail is lost and the realistic properties turn into a cartoon look, to gain that realism again turns into quite a difficult task and unless you're a pro it will probably be very obvious the image has been altered.
I wish you continued success on your creations I'm very impressed and would love to see your outcome when you feel they are complete and performing as planned!
Hi,
some of the probs You mention I also experienced and found a way around, which is a concept that is talked about in this thread also.
This concept is ´matching of distribution character and xover flanks´.
If your panel for e.g. exhibits a dipolar cylindrical waveform shape, then the accompanying midwoofer should match that character at least +-octave around the turnover frequency.
To keep horizontal comb filter effects small the turnover frequency shall not be too high but this inevitably asks for rather small midbass drivers and a tall and thin cabinet shape in close proximity to the panel.
Also design the xover such that the flanks appear symmetrical.
You may either use analog xovers as well as DSP digital ones.
The results are mostly suboptimal anyway and the few really good xovers are either DIY or elaborate digital ones.
The typical OPAmp based analog xovers utilizing theoretically correct textbook filters only simply suck.
The same applies to digital filters utilizing cheap Codecs for A-D/D-A conversion and OPAmps for post-filtering and buffering, even if they allow for equing also.
After having sorted out those xovers I was faced with the challenge to build an dedicated analog xover, since there was nothing adequate on the market.
I used simple JFET-Buffers mostly in unity-gain-Sallen-Key topology, which I modified to my needs.
This means that I threw away all that academic overhead of ideal filter shapes, OPAmp building blocks, and tons of the mathematical rubbish around, but designed the filters as it is done for passive xovers.
(rem.: its of course not all mathematical rubbish, but it is for this application, simply because our transducers behave by far not as ideal as a resistor or a OPAmp output)
In passive technology a filter is designed with the speaker´s requirements in mind, which almost always asks for some equalizing means.
Then -in most cases- the values of simple standard filters with low parts number count are varied to achieve that special equalizer function at the same.
There´s absolutely no reason to not do it the same way in active technology.
In the end my filter for the panels consisted of a high-Q highpass and two Notch-filters and a bit of linear amplification, but required the signal only to pass two JFETs instead of maybe eight or more OPAmps.
This filter sounded much more authentic, and lifelike than any other analog filter before and also much better than any DSP filter as long as the room acoustic didn´t spoil it all.
It was just when we built our own DACs and DSP-Core, resp. applied for Accourate Software filtering that digital not only could keep up with the analog filter, but actually surpassed it by a considerable amount due to its vast equing and room correction abilities.
(rem.: Accourate runs as background application on our music servers, so no extra dedicated PC is required)
Imho its not the DSP cores that spoil the sound of digital, but those cheap A-D/D-A Codecs and surrounding analog stuff.
So if You´re chasing for the top, You need to invest a lot of effort - and money probabely also- in multiples of real high-class DACs, and -if You won´t restrict to an all digital signal chain- a similarly high-class ADC.
The concept of dual or more ´parallel´ diaphragms mounted mechanically in series has been known for long.
You can find threads about it here at DIYaudio also.
SONY and others wrote patents around it and the concept has been proven functional.
ML implemented it in the bass panel of their CLS (using 3 stators) and I used it successfully in my large panels also.
One can gain up to almost +6dB in SPL at low and medium frequencies, thereby countering a great deal of the loss in amplitude response due to acoustic phase cancellation.
Not only does it get rid of that anemic sound of many FR-ESLs in the lower-mids upper-bass range, but increases the dynamic range by more than 6dB due to less amount of electronic equing.
jauu
Calvin
some of the probs You mention I also experienced and found a way around, which is a concept that is talked about in this thread also.
This concept is ´matching of distribution character and xover flanks´.
If your panel for e.g. exhibits a dipolar cylindrical waveform shape, then the accompanying midwoofer should match that character at least +-octave around the turnover frequency.
To keep horizontal comb filter effects small the turnover frequency shall not be too high but this inevitably asks for rather small midbass drivers and a tall and thin cabinet shape in close proximity to the panel.
Also design the xover such that the flanks appear symmetrical.
You may either use analog xovers as well as DSP digital ones.
The results are mostly suboptimal anyway and the few really good xovers are either DIY or elaborate digital ones.
The typical OPAmp based analog xovers utilizing theoretically correct textbook filters only simply suck.
The same applies to digital filters utilizing cheap Codecs for A-D/D-A conversion and OPAmps for post-filtering and buffering, even if they allow for equing also.
After having sorted out those xovers I was faced with the challenge to build an dedicated analog xover, since there was nothing adequate on the market.
I used simple JFET-Buffers mostly in unity-gain-Sallen-Key topology, which I modified to my needs.
This means that I threw away all that academic overhead of ideal filter shapes, OPAmp building blocks, and tons of the mathematical rubbish around, but designed the filters as it is done for passive xovers.
(rem.: its of course not all mathematical rubbish, but it is for this application, simply because our transducers behave by far not as ideal as a resistor or a OPAmp output)
In passive technology a filter is designed with the speaker´s requirements in mind, which almost always asks for some equalizing means.
Then -in most cases- the values of simple standard filters with low parts number count are varied to achieve that special equalizer function at the same.
There´s absolutely no reason to not do it the same way in active technology.
In the end my filter for the panels consisted of a high-Q highpass and two Notch-filters and a bit of linear amplification, but required the signal only to pass two JFETs instead of maybe eight or more OPAmps.
This filter sounded much more authentic, and lifelike than any other analog filter before and also much better than any DSP filter as long as the room acoustic didn´t spoil it all.
It was just when we built our own DACs and DSP-Core, resp. applied for Accourate Software filtering that digital not only could keep up with the analog filter, but actually surpassed it by a considerable amount due to its vast equing and room correction abilities.
(rem.: Accourate runs as background application on our music servers, so no extra dedicated PC is required)
Imho its not the DSP cores that spoil the sound of digital, but those cheap A-D/D-A Codecs and surrounding analog stuff.
So if You´re chasing for the top, You need to invest a lot of effort - and money probabely also- in multiples of real high-class DACs, and -if You won´t restrict to an all digital signal chain- a similarly high-class ADC.
The concept of dual or more ´parallel´ diaphragms mounted mechanically in series has been known for long.
You can find threads about it here at DIYaudio also.
SONY and others wrote patents around it and the concept has been proven functional.
ML implemented it in the bass panel of their CLS (using 3 stators) and I used it successfully in my large panels also.
One can gain up to almost +6dB in SPL at low and medium frequencies, thereby countering a great deal of the loss in amplitude response due to acoustic phase cancellation.
Not only does it get rid of that anemic sound of many FR-ESLs in the lower-mids upper-bass range, but increases the dynamic range by more than 6dB due to less amount of electronic equing.
jauu
Calvin
Thanks for your response. Yes the product numbers are confusing.. I guess you were talking about the SB16PFC25-8 which costs around 33 euros:
https://www.soundimports.eu/nl/sb-acoustics-sb16pfc25-8.html
I have to check if I can order enough of them as they are planned to be discontinued...
@zinda: thanks for your extensive response, suggestions and compliments!
I will keep you updated about my progress regaring LF extension for my electrostatic loudpspeakers. In the future I will try the overlapping of multiple (2) electrostatic panels and publish results here. But as a quick solution / test, I would like to use "regular" magnetic drivers as they are easier to implement and give good results (I have 1 mono OB line array woofer made of 4 Peerless 12 inch drivers).
I will keep you updated about my progress regaring LF extension for my electrostatic loudpspeakers. In the future I will try the overlapping of multiple (2) electrostatic panels and publish results here. But as a quick solution / test, I would like to use "regular" magnetic drivers as they are easier to implement and give good results (I have 1 mono OB line array woofer made of 4 Peerless 12 inch drivers).
Hi all,
I'm still questioning which choises to make: currently I own 4 Peerless SLS-315 12" subwoofer drivers which are mounted in 1 (test) baffle. I use this woofer (currently 1 channel) with my 2 channel DIY electrostatic loudspeakers. The cross-over frequency is 150 Hz. My electrostatic panels have a resonant frequency at around 60 Hz. The panels are driven by custom made (not by myself) transfomers which are, in my opinion, of high quality. As these are "full range" transformers, HF output is limited because of LC resonance of panel + transformers at around 14 kHz. But as my hearing is less sensitive in this area this is no issue for me.
In the near future I would like to test toroidal power tranformers as they are cheap and widely available. I would like to build a hybrid esl with an OB line array woofer. I'm considering the Dayton HS-180-4 drivers suggested by Calvin for this prototype. I'm curious if there will be any audible difference in sound quality between the current Peerless 12 inch woofers and the smaller 7 inch Dayton drivers. According to Calvin there is an audible difference. I would like to test this. Another advantage of the smaller Dayton mid-woofer is it's frequency range: I can chose higher cross-over frequencies, if wanted. I would like to keep cross over frequency as low as possible, but it's still nice to have options.
This is where my first question is about: what is the highest usable frequency of the Peerless SLS-315 drivers in OB configuration? Is this after the first notch filter area, in my case around 300-350 Hz? If I use steep filters I would neet 1 octave for cross-over area, so maximum cross over frequency would be 350 / 2 = 175 Hz. Is this reasoning correct? If correct, this would be a good reason to buy the Dayton 17 inch drivers, as cross over frequency of the hybrid panel will be, if I'm correct, somewhere between 200 and 300 Hz, as smaller toriod power transformers will saturate earlier than my full range big transformers which I currently use for my panels.
The image above shows measured frequency response of the OB panel with Peerless 12 inch drivers at 3 cm from the panel, without equalization.
I'm still questioning which choises to make: currently I own 4 Peerless SLS-315 12" subwoofer drivers which are mounted in 1 (test) baffle. I use this woofer (currently 1 channel) with my 2 channel DIY electrostatic loudspeakers. The cross-over frequency is 150 Hz. My electrostatic panels have a resonant frequency at around 60 Hz. The panels are driven by custom made (not by myself) transfomers which are, in my opinion, of high quality. As these are "full range" transformers, HF output is limited because of LC resonance of panel + transformers at around 14 kHz. But as my hearing is less sensitive in this area this is no issue for me.
In the near future I would like to test toroidal power tranformers as they are cheap and widely available. I would like to build a hybrid esl with an OB line array woofer. I'm considering the Dayton HS-180-4 drivers suggested by Calvin for this prototype. I'm curious if there will be any audible difference in sound quality between the current Peerless 12 inch woofers and the smaller 7 inch Dayton drivers. According to Calvin there is an audible difference. I would like to test this. Another advantage of the smaller Dayton mid-woofer is it's frequency range: I can chose higher cross-over frequencies, if wanted. I would like to keep cross over frequency as low as possible, but it's still nice to have options.
This is where my first question is about: what is the highest usable frequency of the Peerless SLS-315 drivers in OB configuration? Is this after the first notch filter area, in my case around 300-350 Hz? If I use steep filters I would neet 1 octave for cross-over area, so maximum cross over frequency would be 350 / 2 = 175 Hz. Is this reasoning correct? If correct, this would be a good reason to buy the Dayton 17 inch drivers, as cross over frequency of the hybrid panel will be, if I'm correct, somewhere between 200 and 300 Hz, as smaller toriod power transformers will saturate earlier than my full range big transformers which I currently use for my panels.
The image above shows measured frequency response of the OB panel with Peerless 12 inch drivers at 3 cm from the panel, without equalization.
Imho you should try to keep the crossover an octave or more below the first peak on the woofers, and also at least an octave above the esl resonance. An octave is enough provided you put notches on both. That puts the crossover frequency in the 120-150hz range, for a typical h-baffle ob setup like Linkwitz uses.
I use exactly that setup using 3rd order LR crossover.
I use exactly that setup using 3rd order LR crossover.
Hi,
You bet! Sensitively useable from about 50-60Hz up, the array of the smaller drivers can perform with such ´speed´ and attack that larger drivers won´t.
Also, due to the smaller dimensions required, the xover frequency can be set a bit higher ... up to ~250Hz.
This in turn allows to build a panel with smaller d/s and higher mechanical diaphragm tension .... which in turn increases the max SPL and dynamic range.
If You can place dedicated subwoofers, I´d suggest to use the 12"ers in CBs as true subs playing no higher than 50-60Hz.
I´m sure You´ll be very pleased about what You´ll get to hear from such a ´big´ system.
jauu
Calvin
You bet! Sensitively useable from about 50-60Hz up, the array of the smaller drivers can perform with such ´speed´ and attack that larger drivers won´t.
Also, due to the smaller dimensions required, the xover frequency can be set a bit higher ... up to ~250Hz.
This in turn allows to build a panel with smaller d/s and higher mechanical diaphragm tension .... which in turn increases the max SPL and dynamic range.
If You can place dedicated subwoofers, I´d suggest to use the 12"ers in CBs as true subs playing no higher than 50-60Hz.
I´m sure You´ll be very pleased about what You´ll get to hear from such a ´big´ system.
jauu
Calvin
Thats appealing. Is there any useful way of understanding how smaller drivers can do ´speed´ and attack better? Useful would mean for me guiding which woofer to select based on specs. Thanks.
Hi,
I had made two prototypes, a larger one with big panel and an array of 8 22cm bass drivers, and a smaller one with the smaller panel and an array of 8 17cm bass drivers.
The larger system's bass performance without doubt was very good, with sufficient 'weight' and a nice kickbass.
The smaller system certainly had less weight down below, but a more pronounced kickbass.
More pronounced in the sense that it sounds more snappy, agile, 'fast'.
Though both system are the same by concept, they differ of course in size and the amplitude response of their basses and panels, as well as the xover frequencies.
It might well be that the sonic differences became alot smaller if both systems were equalized to the same amplitude response, but I'm quite sure that the smaller driver bass array would still sound snappier.
I couldn't name a single one of the TSP parameters responsible that could tell me how that driver will sound in the end.
As I mentioned before the small diver array might leave You wishing sometimes for more 'weight down under', hence a dedicated subwoofer.
As always ... a improvement in one point leads to a worsening in a other point.
jauu
Calvin
I had made two prototypes, a larger one with big panel and an array of 8 22cm bass drivers, and a smaller one with the smaller panel and an array of 8 17cm bass drivers.
The larger system's bass performance without doubt was very good, with sufficient 'weight' and a nice kickbass.
The smaller system certainly had less weight down below, but a more pronounced kickbass.
More pronounced in the sense that it sounds more snappy, agile, 'fast'.
Though both system are the same by concept, they differ of course in size and the amplitude response of their basses and panels, as well as the xover frequencies.
It might well be that the sonic differences became alot smaller if both systems were equalized to the same amplitude response, but I'm quite sure that the smaller driver bass array would still sound snappier.
I couldn't name a single one of the TSP parameters responsible that could tell me how that driver will sound in the end.
As I mentioned before the small diver array might leave You wishing sometimes for more 'weight down under', hence a dedicated subwoofer.
As always ... a improvement in one point leads to a worsening in a other point.
jauu
Calvin