Rerurn to Romy the Cat's Site

Audio Discussions
Topic: A drivers own inductance

Page 1 of 1 (17 items)


Posted by Romy the Cat on 07-20-2007

 Merlin wrote:
I have … yet to hear a bass unit that can provide the low end quality whilst still allowing a first order slope. Unless we are happy to accept copious quantities of 2nd Harmonic of course. I would love to know of one.

From my point of view a first order on bass channels is uncontestable solution. The imaging, the ability to be integrated with room, the capacity of its second harmonics to cocoon the MF channels, the ability to be integrated with delay channel (if you advanced enough to use them), the imaging, the lessen sensitivity to phase mis-alignment, the imaging, the imaging … did I mention the imaging?

However, the first order on bass could not implemented everywhere just because you want it. Here is where the art of design come to existence: the playback system should be thought in a certain way where system’s ability handle the first order will logically flow from many other design decisions. Sure we all know that there are very-very many aspects how a playback system might be “thought”, and in some cases a first order on bass is not achievable, but still it is from my point of view is one of many ultimate objectives of what I call “well-thought design” or “Design for Sound”.

Below, I enumerated a few not necessary directly related points that people who would like to think about it the subject might use as starting points. What I will say is not a recipe but rather some moments that I discovered for myself while I was in pursuit my “first order discovery”:

1) With most of the bass drivers it is very difficult to roll them off properly with a coil between amps and the drivers as the bass drivers have own inductance that make the slope “immune” to increase of inductance in the filter’s coil.

2) With a full-range amp driving MF and bass practically all of the bids about first order bass are off.

3) With a full-range loudspeakers where MF and bass sections are not physically decoupled most bids about first order bass are off as well.

4) SS amplification is very-very complicated for first order bass as SS amps have very vulgarly sounding second harmonics.

5) Depends of the driver and the enclosure dumping of the driver with amplifier output is superbly critical. There are no objective or measurable mechanisms that would be indicative for proper damping. Each driver in a given enclosure, each amps and each interface are different and the only assessment available is use of the Proper Listening Evaluation Techniques ™.

6) The lowers order is “better”: the 1.5 order is ALWAYS better then 2nd order. A fraction of the order is made by implementation of two series first orders with a certain amount of octave between them.

7) It is better to have 2nd-Harmonic-affluent acoustic system and 2nd-Harmonic-leen electronics then wise-versa. It is always easer and less devastating to sound to reduce harmonic tail then to increase it.

8) The design of your bass section should be able to handle a room 2-3 time larger them yours. If in your room, at very high volume, you the bass section is stressed out then you have problems and you need to resolve the problem of power handling before thinking about the first order.

9) Drivers with low-mass cones and moderate flux are more preferable then high mass driver with high flux. Harmonics are written partially when the driver is rolling back and act as a generator… What car corners better: high mass or low mass if then have the same suspension, mass to HP ratio and the same mass allocation?

10) 30Hz and 120Hz are VERY different frequencies and in most of the cases they should be handled by different channels.

11) Electricity, electricity and one more time electricity! The subjective harmonics influence of your bass is HUGELY depending from quality of Electricity you have.

12) In many cases a bass driver is superbly sensitive to presence of any resistance between amp and voice coil, and this resistance is NOT a part of the amplifier output impedance. So, if you use a passive filter then keep the gage as high as you can and keep DCR as low as you can. Well, if you are in the world of  air core coils (how it should be) and cross at 50hz then you looking at a large barrel of cupper in your room… that most likely will pick your local AM stations without any amplification. Read one more time the paragraph #1, #2, #5 and… How much more incentives you need to go multi-amping?

13) In some instances the contemporary recordings (or 99% of the re-mastered LPs from 60s) are made in a way that first order on bass will not sound properly – it is not the problem of your system but the problem of the Morons with diploma of electricians who believe that they are recording engineers.

14) Use Absolute Phase to check quality of your fist order crossover on your bass section.

15) In many cases a first order and a narrow-bandwidth lack of linearity at crossover point in better then a flat response and second order.

I could go on but I think it is enough for now. With a correct selection of the bass driver and it’s enclosure (this is 60% of success) to implement a proper low-pass first order on bass channels is not so difficult but once it is done and a person learn about auditable advantages  then the person would never go for any sharper filters.

Rgs, Romy the Cat

Posted by stuck.wilson on 07-20-2007
hm..

lots of interesting information to chew on, and by and large, i've had similar experiences with first order crossovers on bass drivers .. this being my experience with a full range amp, and a two way system-- which is a far different critter than your own, for certain.

regardless though--   i do have a question about implementing the '1.5 order' crossover... how would one implement such a thing?   having seen the benefits of first orders in my own application tonally AND dynamically, but having to settle for an asymmetrical 2nd order (first order on the horns) to deal with acoustic phase (WAF, as well as percariously perched horns with first orders, and cats who don't take the hint not to look for the higher vantage points... including my speakers!)-- the idea of something besides a 2nd order sounds appealing..  what would be the phase 'offset', so to speak?  is there a difference from the traditional 45 degree shift?   is it just, say, a 2 mH and a 1.5mH inductor, where before it would be a 3.5?  sorry for the silly question-- but that's definitely not something i've seen outside my own mixing of inductors because i didn't have the right values!

also-  the higher DCR of the filter configured this way-- especially in the mids-- can do some damage to efficiency numbers and transients, wouldn't it?  i have noticed a significant slowing down of transients with air cores that i don't have as much of an issue with inductors like madisound sledgehammers (which are laminate iron cores).  any thoughts?  i know there's definitely a traditional bias against iron cores-- but i've definitely appreciated the 'lessening of the reins' below 1k!

i realize it's not your optimal-- but in such a compromised sitch as a 2 way,  with the 'narrow bandwidth lack of linearity' at 1k, would you think that could be a rectifiable situation-- or would y' think that you just deal with the compromises of the flat response and second order?

many thanks!

d.

Posted by Merlin on 07-21-2007
Romy,

if you see my post you will notice that I mentioned I did not know of any drive units that could successfully be employed. Can you list some?

Here is a typical high quality 15" cone - indeed one of the finest on the market today. It is interesting to see the behavior above 500hz.

Posted by Merlin on 07-21-2007
Sorry wrong attachment but as your software does not allow for editting, I can't be bothered to link to the correct one. Suffice to say - plenty of second harmonic and nasty break up modes.

Posted by Romy the Cat on 07-21-2007

 Merlin wrote:
Romy,

if you see my post you will notice that I mentioned I did not know of any drive units that could successfully be employed. Can you list some?

Here is a typical high quality 15" cone - indeed one of the finest on the market today. It is interesting to see the behavior above 500hz.

Merlin, I did not have 1500 I had 1400 drivers, still they essentially according to the subject of given discussion behave the same. You are a right and most of 15 inch drivers run all the  away up. The problem with them is not excessive high frequency but the fact is that a above certain frequency they produce very bad quality high frequency. For instance 15 inch Altec/JBL type of the drivers might be used up to 500 Hz, as they use in the commercial installations of those manufacturers, however from my point of view those drivers do not have good sound above 150-200 cycles. The drivers should be rolled off by other means, using extra 0.5 crossover, or narrowing down the horn’s mouth if the are horn-loaded or by any other means, that BT might be a cause off as your own problems…

What I usually see is not the excessive second harmonics but rather unpleasant quality of reproduction of higher harmonics. If cross a 15 inch driver at I would say 100 cycles, then I find mostly survivable what that driver does at 200-300 Hz, but the sound of the driver at 500-1000Hz is what bothers me. I know, I know it is a quite few dBs down but I still found it out auditable.

Anyhow, I can not directly answer your question as I do not know any good 15 inch drivers which have natural roll off at 200 cycles, as they rather should be MADE to roll off. The selection of the methods to roll them off should be viewed at each given case, and yes, in some cases it is impossible.

 Merlin wrote:
Sorry wrong attachment but as your software does not allow for editting, I can't be bothered to link to the correct one. Suffice to say - plenty of second harmonic and nasty break up modes.

I have fixed it. The posters, as now should be allowed to I edit on posts with the “new” forums. You still might repost the file you intended to post. Alternately, since your file is a graphic file, you can embed the image in the body of your post, hosting the image of my server, and using the following instructions:

http://www.goodsoundclub.com/Forums/ShowPost.aspx?PostID=1120

Rgs, The Cat

Posted by Romy the Cat on 07-21-2007

 stuck.wilson wrote:
having seen the benefits of first orders in my own application tonally AND dynamically

It's never was my experience. A never seen any tonal distinction between first and second order and I don't think that I experienced are any differences dynamically.

 stuck.wilson wrote:
regardless though--   i do have a question about implementing the '1.5 order' crossover... how would one implement such a thing?

The 1.5 order is very easy.  The “1.5 order” is not the order itself but rather it is a hype-name that I use. However, it is very simple to implement and very effective. The 0.5 order crossover based upon a phenomena that I discover that if I have at first-order crossover and then, for instants, after one octave another first-order crossover, then I of course have a second order crossover, however this type of  “distributed” first-order crossover does less impact to Sound (imaging) then when I use full second-order crossover (I used religiously Bessel curve on second orders). That distributed crossover is a very interesting fruit, as it is allows a lot of flexibility. For instants you have the bass section that you would like to low-pass at 60 cycles with first-order. That frequency you'll found is necessary to integrate your base with your upperbass channel. The integration of bass is fine but you have an excessive HF “tail” from your low-frequency driver. If you go at 60 cycles with sharper order then you'll my gain some “issues” in imaging, so here is where you might employ the “distributed” first order. You'll keep your first order crossover at 60Hz and put in play and other first order crossover at, let say, 200Hz. The channel begins to write a full second order at 200 cycles, and does all typical for a second order phase spinning. However, the key in it is that subjectively, listening with that playback, this type of the arrangement screw up imaging less. The additional beauty is that you can move the second first-order crossover from (for instants) 100 cycles to 500 cycles, finding how much you would like to truncate the bass driver’s tail. There is even more very important benefit in this approach: the second first-order crossover might be not the actual crossover but rather roll-off from other means: enclosure, amplification, magnetics or whatever. The interesting part in that case is that if in place of the second first-order crossover is used a pure second-order crossover then the entire channel does not sound (imaging-wise) as a third order crossover.

 stuck.wilson wrote:
also-  the higher DCR of the filter configured this way-- especially in the mids-- can do some damage to efficiency numbers and transients, wouldn't it?  i have noticed a significant slowing down of transients with air cores that i don't have as much of an issue with inductors like madisound sledgehammers (which are laminate iron cores).  any thoughts?  i know there's definitely a traditional bias against iron cores-- but i've definitely appreciated the 'lessening of the reins' below 1k!

Yes, the iron-core coils are OK as have fewer turns and consequentially lower DCR. In my past is I have used Erse Super Q Inductors with quite pleasant results. Steel being a snob I have paid a lot the money, perhaps stupidly, to wind 8-10 gauge high-inductively air-core inductors. It was a huge, it was heavy, it was outragesly expensive. If I had at a little brain during that time then I would buy for the money I spend for those inductors a separate amplifier and went bi-amping.

The caT

Posted by stuck.wilson on 07-21-2007

what a great idea-- i never really thought of a sort of 'low pass bypass' inductor of sorts?!   i'd always assumed that inductance was similar to capacitance in that regard-- the more you add, the more you get-- but hadn't thought about the potential for chaining them concurrently as a positive attribute.  it takes a lot of time experimenting to put it all together...

would you still end up with the net result of a second order (90 degrees) if you chain inductors?  that seems to follow what you're saying-- as you're basically 'delaying' twice.  hmm.. if that's the case-- that's a really interesting proposition.  two chained iron cores could still yield a lower DCR than an air core of similar size (although with different results), and the extra level of control could be really helpful.  PLUS having the same net phase shift..  i can keep the wife, cats, and the horns... i like it...    i'll have to look into that!  with a first order crossover around 1100hz--     the DCR might be naturally lower than it would be with the crossover of a larger inductance second order filter anyway..

i ALMOST went the same route as you with the big custom wound inductors-- especially running a modestly powered system with passives and trying to keep a 99db bass bin even close to a 118db compression driver--  but the outrageous expense of copper nowadays definitely sent me the direction of steel cores...  i even looked into spinning them myself-- but still the cost of high grade magnet wire of gauges like that is pricy!   and for the low price of the erse or sledgehammers.. that was a myth i just had to try busting for myself... it seems a lot more like an 'economic' bias than a sonic one from the sounds of it here!

excellent!  thanks for that!

d.

Posted by Merlin on 07-21-2007
Romy,

I agree with everything you say with regards to 15's - your experiences mirror mine. Obviously Keith's Alphas apparently have a 1st order slope above 600hz on a back loaded 15" paper cone. I would find that unacceptable as you would too.

Posted by Romy the Cat on 07-21-2007

Stuck,

You can NOT stack coils in series to get a “distributed half order crossover”. To make it “distributed” the first and the second first-order filters should be decupled. For instance the first filter might be at line-level and the second at the driver level. Or the first filter is at line-level and the second via a transformer or via plate filtering or feedback or whatever… You might also get the result loading the first filter to a proxy load and then write a second filter against the voice coil. As I said in order to have it working properly the entire channel should be thought from it's ground up…

So, it is like the note that I hang in my office says:

The caT

Posted by Romy the Cat on 10-03-2007

I spoke this morning with Yoshi and he made a phenomenal comment. He was complaining something about his bass. He uses a pair of big 15’ JBL in a box with an active second order crossover. So, when I asked him to describe what he feels is wrong Yoshi said replayed with something geniuses. He said:

“The bass decay notes are rather characteristics of driver/enclosure then the characteristics of the acoustic space of the recording…”

Wow! That is the phrase that I was looking for a long time!

Rgs, The Cat

Posted by Merlin on 10-12-2007
Sounds like he needs someone to build him a decent cabinet.

Posted by malinowski on 10-12-2007
Is Yoshi's box ported or sealed? If it's ported.. of course he's going to be listening to more.. a lot more.. of the box... everything below.. 60hz? is box music.
if it's not ported then this is a more interesting discussion. when I first read romy's response to yoshi's description that

“The bass decay notes are rather characteristics of driver/enclosure then the characteristics of the acoustic space of the recording…”

I assumed this was a good sealed box and Romy and Yoshi were referring to a characteristic assosicated with active crossovers retaing a perfect response curve at the expense of losing the sense of space..

so which is it..? interesting. thanks.

Posted by yoshi on 10-13-2007
The box is about 5 cu/ft apperiodic with 3 x Scan Speak vent on the rear panel.  Damped with cotton blanket (filling 60% of inside) and braced rather well.  The driver is JBL 2220 crossed at 80Hz with 12db slope.  All other channels are 6db slope.

It is difficult to have larger boxes in my room, so I'm thinking of trying different drivers that can be crossed smoothly to the mid with 6db.  I just got my own RTA set up, but have to leave to Japan tomorrow for about 3 weeks. 

Yoshi

Posted by jessie.dazzle on 10-14-2007
Romy wrote:

"...With most of the bass drivers it is very difficult to roll them off properly with a coil between amps and the drivers as the bass drivers have own inductance that make the slope “immune” to increase of inductance in the filter’s coil..."

In this case, would there be an advantage in using a driver that has lower (much lower) nominal impedance?

jd*

Posted by Romy the Cat on 10-14-2007

 jessie.dazzle wrote:
Romy wrote: "...With most of the bass drivers it is very difficult to roll them off properly with a coil between amps and the drivers as the bass drivers have own inductance that make the slope “immune” to increase of inductance in the filter’s coil..."

In this case, would there be an advantage in using a driver that has lower (much lower) nominal impedance?

Yes, from prospective of filter theories is it might be correct but it to look from the perspectives of practicality then I so not see it as a solution. In case of SET a lower impedance of load require a corresponding higher transformer ratio and this ratio would be most likely not suitable for the other none-bass driver as it will eat harmonics to much. So, we slightly cure bass’s inductance problems but we loose in correct loading for other drivers…
Here is where the ultimate tool of any single-ended amplifier kicks-in: DSET. In case of a dedicated SET all problems just got evaporated as the bass-channels are completely independent and as a result they can care any loading they wish. Not to mention that DSET most likely implies a line-level crossover that removes the evil inductance between amp and driver and, that is not a minor, remove DCR of the filtering coil out of the amp loading.

Rgs, Romy the Cat

Posted by Merlin on 10-14-2007
Yes but you are still going to need a higher order slope than 6db to prevent unnecessary non linear distortion from the bass driver and then the insertion losses inherent in a PLL crossover become an issue. Higher order slopes and active operation works well for me.

With regards to the JBL bass enclosure my first suggestion would be to dump the cotton blanket and replace it with a 1" or 2" lining of Wrap On or similar fibreglass wadding.

Posted by morricab on 11-07-2007
Hi all,
In my experience it is VERY hard to get anything resembling realistic bass from a vented box, particularly with a Q of around 0.9 or 1.  There is too much stored energy in the system and too much overhang. 

The best bass I have heard in descending order is:
Horn bass
dipole bass
Sealed box (around 0.6-0.7 Q)

The other problem in discussion, namely using first order crossovers, requires some examination of the two drivers at the crossover point.  If the drivers are two different materials, eg. aluminum and kevlar, to use a high order crossover results in a sonic disaster.  Why?  First you have a mismatch between the inherent sound of the drivers.  ALL drivers (except perhaps an ion flame or a diamond cone) have a characteristic sound from the flexing and breakup that occurs from the material not being infinitely stiff.  Then you are making an abrupt transition (with a high order filter) between one driver to the next and this makes a discontinuity in the coloration of an instrument that spans those two drivers.  The worst I have heard in this regard were some B&W speakers which had carbon fiber woofers, kevlar mid, and aluminum tweeter with 6th order filters.  No coherence to the sound whatsoever.  This is complicated by phase shift, which while perhaps inaudible in a short demo becomes obvious when listening for longer.  One hears softened transients that blunt any attempt at a lifelike sound.  A low phase shift with a 1st order filter makes it easier to get good transient response.  Any distortion increase will be more than offset by the benefits as long as a suitably robust driver is chosen.  Especially in the bass a little bit of extra harmonic distortion will largely be inaudible but a better integrated bass will give a much more coherent and upbeat sound (or PRaT).  To me higher order vented systems often sound quite sluggish.

Page 1 of 1 (17 items)