That is completely false. It was never "contradictory". I think the word you are looking for in "controversial" and that it was. But it never contradicted anything.
Lynne
Here is my take on the VC overhang issue - its not clear cut either way.
First, if you know for a fact that your underhung VC will never come to the end of the gap then it will always sound fine. But if it does come close to the ends then its distortion will be very high order and sound terrible. So its a "fine until its not" situation.
With overhung we have to consider the physics. The field and hence the BL must go to zero for some excursion, so how to do that. We know that low orders of distortion are not very audible - like 2nd and third - so one should minimize any higher order nonlinearity in the BL curve even if this means maximizing the lower orders (remember the square of the order weighting). This says that a very gradual BL that is really never flat would be the best. (Flat regions only means that when it is not flat the slopes are greater and hence the orders are higher.) This is precisely what companies like B&C do. The speakers will have higher 2nd and 3rd harmonics, but virtually nothing above that - and here is the point - no matter how much excursion the speaker sees. The output will fall slightly, but the distortion will never get objectionable. Basically the overhang is an "always OK" situation IF the field is designed properly (like B&C).
The underhung is a huge waste of flux if in fact the distortion from the overhung coil is not a problem. If properly done, I'll take the overhung over the underhung. (Few drivers are "properly done").
One possible option is to have the power transformer rated, or gapped, to a much higher current than actually consumed by the amp. Of course, there is a price to pay for it, in dollars…
I use a server-grade "sinewave" UPS for my 58" Panasonic plasma TV and cable box, and another server-grade "sinewave" UPS for my Mac Pro and cable modem/router. According to the claims of the manufacturer, when AC power falls outside of spec, it switches over to the battery supply in the zero-crossing interval. They've done a good job so far of riding through transients, glitches, and power failures, and protecting the devices they're attached to.
But I wouldn't use these gizmos for linear audio; I'm not that sure the waveform is all that great. There are very expensive "power regenerators" that are basically 60 Hz oscillators that drive big Class AB transistor amps that put out 120V. There's no "ride-through" of line dropouts unless batteries and a lot of complex additional switchover electronics are involved.
I'd probably go the motor-generator route before buying one of these ... at least a motor-generator has 100% isolation of transients, several seconds of "ride-through" due to rotating mass, and runs down slowly when the power fails completely. It's basically a little power station that's all your own, completely electrically isolated from the incoming AC power. They do make a whirring noise, which I why I would exile it to the garage in a small box where it could do its thing.
Dr. Geddes, thanks for the B&C recommendation. Not all professional drivers have correct shading of the magnetic field in the gap; good to know that B&C does. I suspect this is an area where a lot of audiophile drivers fall down rather badly.
Last edited:
Not sure if you have measured yourself, but most noise are actually in that additional ground line. Due to the way most ground are implemented in equipment, this does become a serious problem. But power factor is also just as critical.
Heh, just last week I was looking at an old UPS/AVR I had used with my Apple II+ wondering whether it still worked or not.
Another benefit of a motor-generator is a completely isolated ground; that is, there's a separate "technical ground" rod that is physically separate from the main house ground.
All of this recapitulates old-school radio-station practice: isolated "technical grounds", and motor-generators that have enough ride-through so separate diesel gensets can spin up and deliver full power during an extended blackout.
The power around here is reliable enough that I haven't considered a backup generator. Even though Colorado gets ferocious lightning storms, the system is very reliable and flicker-free even when we're getting a light show that goes boom-boom-CRASH every few seconds. The longest outage was during the big floods, and even then it was only for a few hours.
WOW, in my poor memory, this is probably where I agree with you most. The only thing different is that I would try to optimize BL, excursion, and induction, and see what results instead of worry about whether it is underhung or overhung.
Aren't the generators acoustically noisy?
Don't forget the noisy stuff in your own home like computers and other electronic devices.
Last edited:
I think the ground noise issue is more how you load the line rather than the source or ground rod. If the AC is not balanced, things like unbalanced EMI filtering or unbalanced capacitor coupling of a transformer to a chassis just feeds that current back up the ground line to the source, instead of cancelling it at the load. I've always thought that negative reaction to EMI filters at IEC inputs might be due to this. Otherwise, why would elimination of unwanted noise be a "detriment" to audio quality?? I don't buy the inductor blame game in the filter that supposedly "slows the current down". Yeah, at the high frequencies you don't want through anyway. Why not let the power transformer do some line filtering and clean it up more on the balanced secondary side? Use dual winding/dual bridge supplies and don't feed the cap charging pulses through the star ground; i.e., don't use the transformer center tap as star ground. In fact, don't use a center tap at all, but dual isolated windings. Connect the primary polarity for the least capacitive coupling to chassis, and/or even balance this coupling with an external cap to cancel the ground current as much as possible? Minimize the ground noise to prevent conductive coupling to other circuits sharing the mains, and/or go all balanced.
Hi Lynn, interesting design. I am looking for a new speaker project to better the 3-way horn based speakers I have now. I know better is a subjective word, but the description of the 3-way and objective measures are in an article I wrote:
Advanced Acourate Digital XO Time Alignment Driver Linearization Walkthrough
Hopefully this isn't viewed as link spam as I am genuinely interested in a speaker design that can outperform what I currently have.
I have not heard the speaker components used in your design, hence my interest on getting your thoughts as you may have heard the same components that are in my current speakers and can compare and contrast.
Thanks for any insights.
Best regards, Mitch
Earl,
Although I agree with you that B&C woofers have a great BL curve, and that distortion should generally never be an audible problem with them at the power levels Lynn can achieve with the amps he prefers, I disagree that their speakers have virtually no harmonics higher than 2nd and 3rd.
In the screen shot below you can clearly see harmonics out past 11th order in the B&C 18SW115-4.
Not to say that this speaker would be a great choice for crossing around 700 Hz, but the BL curve does not eliminate upper harmonics.
That said, the JBL 2226 distortion is also so low at the levels Lynn has presented it in the LTO Apollo (or Athena) that I would doubt distortion has anything to do with his preference of the GPA 416B (16 ohm version) over it, impedance, inductance and frequency response variations are more far more likely.
Art
Attachments
Last edited:
Grounding the equipment is a safety measure. In my country it’s illegal to operate electric or electronic apparatus ungrounded, unless they have proper double isolation.
On top of that, noise is also on the ground connection, not only on it.
I have moved many of the Off Topic posts here:http://www.diyaudio.com/forums/everything-else/247259-power-cords-plugs-split-beyond-ariel.html
Let's try to keep this thread at lest somewhat related to speakers. Thank you.
I agree that many so called underhung motor designs are little more than an even hung design with a small margin of useful excursion. I have worked very hard not to do that and to have an extremely long gap with even flux density along the gap. The gap length is 1.25" while the coil height is limited to 0.375" and this in on a 6 1/2" cone driver. So you could say that this is an ultralinear motor design if there ever was one. I think that the Mark Audio motor design is using a similar concept though I have not heard the device. Aura is another that uses this type of design though the rest of the speaker is not of the best quality and they use some very low value Neo in those commercial speakers they produce.
I agree with Earl that this will lower the efficiency compared to a focused gap motor design but I would gladly give up some efficiency for the linearity. Now that being the case you also have to look at the magnetic material itself and what alloy or grade that is being used. This is seldom if ever mentioned in the literature, only the B/L factor is generally given. I happen to be using the highest grade of Neo that is available at the moment. If they come out with a higher value material I will move in that direction. One of the problems with this material at this time is that the very high value Neo magnet materials is that as the value is going up the curie temp is also falling so much thought has to go into making sure you do not reach that temperature or the magnets will soon loose charge. As in anything else in speaker design it is a balancing act to say the least. Not many will use this type of Neo material as the cost is substantially higher with higher grade material.
You also have to look at how much energy is wasted by having a major portion of the coil outside of the gap. that portions is wasted energy. I have yet to see any magnetic focused gap design that is balanced on both sides of the gap with equal flux lines which mean that it is a non-linear motor by design. The coil length inside the gap is not working at the same rate as the external coil section, there is no way that I know of that can totally correct for this factor.
I agree with Earl that this will lower the efficiency compared to a focused gap motor design but I would gladly give up some efficiency for the linearity. Now that being the case you also have to look at the magnetic material itself and what alloy or grade that is being used. This is seldom if ever mentioned in the literature, only the B/L factor is generally given. I happen to be using the highest grade of Neo that is available at the moment. If they come out with a higher value material I will move in that direction. One of the problems with this material at this time is that the very high value Neo magnet materials is that as the value is going up the curie temp is also falling so much thought has to go into making sure you do not reach that temperature or the magnets will soon loose charge. As in anything else in speaker design it is a balancing act to say the least. Not many will use this type of Neo material as the cost is substantially higher with higher grade material.
You also have to look at how much energy is wasted by having a major portion of the coil outside of the gap. that portions is wasted energy. I have yet to see any magnetic focused gap design that is balanced on both sides of the gap with equal flux lines which mean that it is a non-linear motor by design. The coil length inside the gap is not working at the same rate as the external coil section, there is no way that I know of that can totally correct for this factor.
I did not write my review as an advertisement for B&C, but to state why I think that the prevailing philosophy of motor BL design is not really correct. B&C just happens to do it right that's all. I still hear people talk about how the BL curve should be flat and that is just not the best way to do it.
Last edited:
Art - since I never said that they did your comment is moot. I said that one should minimize the higher orders by letting the lower ones rise. I alluded to B&C appearing to do this.
You really need to read what I say better.
I have no idea if the speaker that you posted in done right or not. Some B&C drivers posted in Voice Coil did show a very smooth rounded BL curve, quite symmetric - clearly done intentionally. I have no idea how prevalent this is in their products, just that I know that they have done it for some. Most people believe what Kindhornman states - I don't, which is why I posted what I did.
Last edited: