Do you mean sandwich as in isobaric? If you do, I believe you are wasting four neo 8's and also giving rise to bizarre distortions that wouldn't otherwise be there in a normal configuration. You would be much better off stacking them all vertically..
Last edited:
Yes, isobaric configuration. I don't think I am wasting anything. On the contrary: the sandwich means double magnetic force, double moving mass but single air mass. Result: increased efficiency especially in the lower frequencies, lower cutoff. Distortions -- where should they come from?
Stacking vertically only increases the size of the speaker which I consider undesirable because it creates more reflections, intermodulation and room reactions. I had four Neo8 stacked before and was unhappy with it.
Incidentally, Loewe, a German producer of TV/home cinema gear is offering a relatively small electrostat (150Hz-20kHz) satellite fitted with two closely sandwiched ESL panels, as far as I could ascertain. They tout this as a breakthrough.
Stacking vertically only increases the size of the speaker which I consider undesirable because it creates more reflections, intermodulation and room reactions. I had four Neo8 stacked before and was unhappy with it.
Incidentally, Loewe, a German producer of TV/home cinema gear is offering a relatively small electrostat (150Hz-20kHz) satellite fitted with two closely sandwiched ESL panels, as far as I could ascertain. They tout this as a breakthrough.
It is almost certain that there will be comb-filter type cancellations with Neo 8s used this way. It'd be interesting to run a fr plot on this configuration. I haven't seen the Loewe electrostat, but it depends on the diaphragm spacing, and whether they are physically connected as to what frequency you would start to see cancellations.
One of the things I can see as a potential problem with this isobaric configuration is that the balanced "symmetric" motor configuration is no longer balanced and symmetric. The diaphragm of each transducer will see a stronger magnetic force on one side than on the other. This may add HD that wasn't there to begin with. The "neutral" air mass between the drivers may not be as neutral as you think do to possible turbulence within the elongated motor and frame structure. As long as it's crossed over low enough, I don't see comb filtering as a potential issue..
Last edited:
The distance between the diaphragms of two sandwiched Neo8 icorresponds to the thickness of the baffle, in my case 1/2 " or 12 mm. Comb filter effects would therefore only arrrive at beyond cutoff frequencies, as Remlab pointed out. On the other hand I believe the magnetic forces acting between the two Neo8 are suifficiently damped by the distance not to pull the diaphragms out of symmetry. The sheet of urethane foam, i.e. the 'stuffing' of the sandwich, does reduce internal reflections and prevents part of the treble emitted by the speaker facing the inside of the enclosure from radiating through the front speaker.
As regards Loewe they are very cagey in describing their 'stat. I asked the sales people in one of their fancy downtown stores what the principle of their stats was, and they were totally uninformed. I'll look up Loewe catalog and find the precise (nebulous) term they use and they are so proud of. Since the stats are very thin, only about one inch thick I can only imagine that they use three stators (and two diaphragms between them); the middle stator showing opposite polarity in both directions.
As regards Loewe they are very cagey in describing their 'stat. I asked the sales people in one of their fancy downtown stores what the principle of their stats was, and they were totally uninformed. I'll look up Loewe catalog and find the precise (nebulous) term they use and they are so proud of. Since the stats are very thin, only about one inch thick I can only imagine that they use three stators (and two diaphragms between them); the middle stator showing opposite polarity in both directions.
Just found this thread. Some really great and inspiring experimentation being done. A note about harmonic distortion - digitally filtering out the high frequency peak won't touch the distortion of lower fundamentals, as the peak is actually still there in the speaker. Acoustical filtering, such as a mesh grill, is needed to lower distortion, although getting the desired frequency response then becomes an issue.
Spartacus
Panel speakers are different from other speakers in that not only do they have hd above the fundamental,they also have pronounced subharmonic distortions that exist below the fundamental due to semi chaotic resonances produced as a byproduct of a flimsy diaphragm stretched like a drum over the frame. Electrostatics also have this problem. This is different than conventional drivers that primarily have their hd above the fundamental. The logic you have to apply to this is somewhat counterintuitive and confusing. I do need to point out though that this discussion is ultimately academic in nature if the distortion is below the threshold of audibility anyway...
Seth
Panel speakers are different from other speakers in that not only do they have hd above the fundamental,they also have pronounced subharmonic distortions that exist below the fundamental due to semi chaotic resonances produced as a byproduct of a flimsy diaphragm stretched like a drum over the frame. Electrostatics also have this problem. This is different than conventional drivers that primarily have their hd above the fundamental. The logic you have to apply to this is somewhat counterintuitive and confusing. I do need to point out though that this discussion is ultimately academic in nature if the distortion is below the threshold of audibility anyway...
Seth
Last edited:
Thanks for explaining, and after a few moments staring at the graphs, I think I follow what you are saying. However the graphs still look like conventional frequency response induced HD, just like with most other drivers (although much lower in amplitude). How do you tell the difference, without actually filtering the peak and then measuring?
Pinco
Your isobaric alignment is starting to make a lot of sense. You have obviously thought through every aspect of this design..
Seth
Spartacus
I have to admit that this was inferred(initially). I noticed that the 2nd order HD curve at 5khz (ZAPH- Neo10) looked almost exactly like the the frequency response curve at 9khz. It is almost a pure 2nd harmonic occurring at almost half of the 9khz peak(Since subharmonics are semichaotic, they don't always occur at exact divisions of the fundamental). At that point, I cheated and asked Igor Levitsky(The Neo 10's designer) if my suspicions were correct..
OK. So once the response is flattened, distortion will be reduced, rather like with a waveguide loaded tweeter. No wonder it sounds so good!
In this Pro20d finished speaker(downloads), you will notice that the aforementioned 2nd order hd peak completely disappears from the measurements(Compare it to the raw Neo 10 data in the same review) as predicted. If it was a 2nd harmonic of a 3khz fundamental, it would still be almost at full level in the finished speaker..
http://www.audax-speaker.de
Seth
http://www.audax-speaker.de
Seth
Last edited:
This is discussion is ultimately irrelevant(Educational, but still irrelevant in a subjective sense) because even unfiltered, the harmonic and subharmonic distortion is so low in level in this driver or for that matter, any of the drivers in the BG Neo line, that within reason, you can't even come close to hearing it in any of them..
Last edited:
In the final speaker, there seems to be a peak in both K2 and K3 at 3KHz. i wonder what that corresponds too.
As to the audibility of distortion, that's obviously a topic that's been done to death, but personally I think in an ideal speaker it should be be approaching the same level as other components in the playback chain.
In the finished speaker, everything above 100hz falls below .3%, except a tiny area at 300hz that hits .5% . For all intents and purposes, that's pretty much as good as it gets..
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.