That may be true, some of us are.
If you mean in an every day use home system then yes. 4way active with a DBX 4800 for crossover, delay, gain etc. TEF as the measurement platform.
So the discussion around the WE stuff, in particular the 15A/555 and also, the Quad ESL 57 got me thinking about what they do well.
Allow me a slight detour...
As is well known, both the 15A/555 and the ESL 57 cover a significant amount of the frequency spectrum all from one compression driver/horn and (admittedly) three panels (80Hz to 6.5K and 50Hz to 16K respectively and approximately). No crossover or other drivers to introduce phase or transient-related problems. In fact, their energy storage - their ability to stop when the music does - is phenomenal (though I've only seen a waterfall graph for the ESL 57). Both of them use diaphragm/panels that are small, light and low in resonant frequency.
The other significant attribute that strikes me about both the 15A/555 and ESL 57 is that each requires a power supply - the 555's field coil needs an external tungar supply operating at 7V DC, and the Quad a mains transformer/voltage multiplier per side.
And it got me thinking - what other transducers are like this? And the answer, which I've used many times, is:
Small-diaphragm condenser microphones. (I'm omitting large diaphragm condensers due to their measured performance relative to their SDC counterparts.) The similarities are kinda interesting (to me, at least).
Firstly, they use small, incredibly light diaphragms and, secondly, a 48V "phantom" power supply (from a desk or a mic pre).
A single half-inch capsule mounted on an appropriately-designed body - for example, a Microtech Gefell MK221 omni capsule on a Joesphson C617 body - gives quite astounding measured performance. (Look DavidL! I'm writing about measured performance...!) 10Hz-20kHz plus or minus 1dB (that's eleven octaves - D#-1 to D#10 with 1dB of variation (that's ridiculously flat)); an equivalent noise level of 14 dB SPL (A-weighted); and a sensitivity of 66 mV/Pa. (I don't know of an equivalent Earthworks, Schoeps, Sanken or DPA mic that has as extraordinary measured performance, though I imagine they're not far off in various areas.) Really, though, to cover eleven octaves flat with such low self-noise and be as sensitive as that is truly an impressive feat of engineering. (For comparative purposes, an AKG C12 VR LDC measures 30Hz-20kHz plus or minus 3dB, has an equivalent noise level of 22 dB SPL (A-weighted) and a sensitivity of 10 mV/Pa.)
Okay, okay, a mic is not a speaker and vice-versa. I get it, I know. I merely mention this as an idle curiosity really, but I think the small, light, low resonant frequency stuff and especially, the power supply stuff are potentially key technologies inherent to both the WE 555 and the ESL 57 that we've largely abandoned today.
So, as a non-engineer, I wonder - is there anything from current state-of-the-art microphone technology that is or could be applicable to loudspeaker design today?
Allow me a slight detour...
As is well known, both the 15A/555 and the ESL 57 cover a significant amount of the frequency spectrum all from one compression driver/horn and (admittedly) three panels (80Hz to 6.5K and 50Hz to 16K respectively and approximately). No crossover or other drivers to introduce phase or transient-related problems. In fact, their energy storage - their ability to stop when the music does - is phenomenal (though I've only seen a waterfall graph for the ESL 57). Both of them use diaphragm/panels that are small, light and low in resonant frequency.
The other significant attribute that strikes me about both the 15A/555 and ESL 57 is that each requires a power supply - the 555's field coil needs an external tungar supply operating at 7V DC, and the Quad a mains transformer/voltage multiplier per side.
And it got me thinking - what other transducers are like this? And the answer, which I've used many times, is:
Small-diaphragm condenser microphones. (I'm omitting large diaphragm condensers due to their measured performance relative to their SDC counterparts.) The similarities are kinda interesting (to me, at least).
Firstly, they use small, incredibly light diaphragms and, secondly, a 48V "phantom" power supply (from a desk or a mic pre).
A single half-inch capsule mounted on an appropriately-designed body - for example, a Microtech Gefell MK221 omni capsule on a Joesphson C617 body - gives quite astounding measured performance. (Look DavidL! I'm writing about measured performance...!) 10Hz-20kHz plus or minus 1dB (that's eleven octaves - D#-1 to D#10 with 1dB of variation (that's ridiculously flat)); an equivalent noise level of 14 dB SPL (A-weighted); and a sensitivity of 66 mV/Pa. (I don't know of an equivalent Earthworks, Schoeps, Sanken or DPA mic that has as extraordinary measured performance, though I imagine they're not far off in various areas.) Really, though, to cover eleven octaves flat with such low self-noise and be as sensitive as that is truly an impressive feat of engineering. (For comparative purposes, an AKG C12 VR LDC measures 30Hz-20kHz plus or minus 3dB, has an equivalent noise level of 22 dB SPL (A-weighted) and a sensitivity of 10 mV/Pa.)
Okay, okay, a mic is not a speaker and vice-versa. I get it, I know. I merely mention this as an idle curiosity really, but I think the small, light, low resonant frequency stuff and especially, the power supply stuff are potentially key technologies inherent to both the WE 555 and the ESL 57 that we've largely abandoned today.
So, as a non-engineer, I wonder - is there anything from current state-of-the-art microphone technology that is or could be applicable to loudspeaker design today?
Actually IMO a microphone and speaker are the same.
You are also on to something with the field coil. I think it makes a nicer magnet for the cone to work against. It would be interesting to know the strengths involved.
I found the original 555 paper at the lucent website. It contains some parameters for those that want to model the compression driver. Also shows efficiency which they claim as 30% in production.
http://www.alcatel-lucent.com/bstj/vol07-1928/articles/bstj7-1-140.pdf
More papers here (Bell System Technical Journals, from 1922 on
Bell System Technical Journal, 1922-1983
David
http://www.alcatel-lucent.com/bstj/vol07-1928/articles/bstj7-1-140.pdf
More papers here (Bell System Technical Journals, from 1922 on
Bell System Technical Journal, 1922-1983
David
I was talking to a field coil guy on Sunday. Some interesting info about flux and finding the the correct point on the curve. If I understand correctly, saturating the magnetic circuit makes the F.C. sound like Alnico. I've often wondered what difference the flux could make - other than Qe, of course.
Well I`ll be dammed! Heard my first WEs today by sheer co-inky-dink
... 555 compression drivers mounted to 14a horns.
Very...very..very...impressive...driven by an old Leak Stereo 20 and energized by a couple of outboard Heathkit D.C. power supplies.
Yep, there`s something going on with these old-timers that is very magical indeed! I count myself one lucky camper.... Not the same size as the big WE horns, of course, but in a 15` X 15` room, more than sufficient.
Difficult to come back home and listen to my stacked six-pack of Minimus 7s despite their "D'Appolito Configuration"
... 555 compression drivers mounted to 14a horns. Very...very..very...impressive...driven by an old Leak Stereo 20 and energized by a couple of outboard Heathkit D.C. power supplies.
Yep, there`s something going on with these old-timers that is very magical indeed! I count myself one lucky camper...
. Not the same size as the big WE horns, of course, but in a 15` X 15` room, more than sufficient.Difficult to come back home and listen to my stacked six-pack of Minimus 7s despite their "D'Appolito Configuration"
I found a number of Western Electric response curves in their various papers. (see Lucent site referenced yesterday).
http://www.alcatel-lucent.com/bstj/vol08-1929/articles/bstj8-1-196.pdf
A Sound System for use in Motion Picture Theaters is mostly about the projector with sychronized record player as used in the initial Vitaphone system. The response curve of the horn looks like the damped room measurements of the second article. (see indoors vs outdoors, the curve using the rotating microphone). The same basic response curve with about 15dB rise in the free field response or 10dB in the in-room curve due to the high frequency deadening of the room and the high HF directivity. I have to assume this is a curve of the 15a.
http://www.alcatel-lucent.com/bstj/vol08-1929/articles/bstj8-1-135.pdf
The second article Acoustic Consideration in Steady State Loudspeaker Measurements is a good tutorial on the subject of response measurement and looks both at a horn and a 3.5 inch direct radiator measured out of doors, in a deadish (felt lined and curtains) room and finally via an averaging technique where the mic was moved in a circle to reduce standing waves. The outdoor curves include a set with off axis curves revealing the high directivity that a long exponential horn would have to have (before the evils of diffraction slots!).
Note that the 115 cycle horn appears to be the smaller folded horn as developed for the Victor Orthophonic Victrola. It has the same characteristic response as the first theater horn above, very similar artifacts at 400-500 Hz and the top Octaves, but gives up an LF Octave compared to the 15a. Both have the same rising or shelved up upper range, at least on axis. The polar curve of the larger horn would have to be the same but shifted down (same polar nearer 500Hz that this has at 1000).
Not very flat but reasonably smooth. In a cinema application the very high directivity would have to be considered when trying to achieve the best audience coverage or even in determining what the average response curve is. This shows why the articles indicate that they were used in groups of 4 for large theaters: a splayed pair aimed down at the audience and a second splayed pair set low and aimed upwards to the balcony. The Altec designs with multicell horns, of your choice of cell configurations, would be more helpful in that regard.
David S.
http://www.alcatel-lucent.com/bstj/vol08-1929/articles/bstj8-1-196.pdf
A Sound System for use in Motion Picture Theaters is mostly about the projector with sychronized record player as used in the initial Vitaphone system. The response curve of the horn looks like the damped room measurements of the second article. (see indoors vs outdoors, the curve using the rotating microphone). The same basic response curve with about 15dB rise in the free field response or 10dB in the in-room curve due to the high frequency deadening of the room and the high HF directivity. I have to assume this is a curve of the 15a.
http://www.alcatel-lucent.com/bstj/vol08-1929/articles/bstj8-1-135.pdf
The second article Acoustic Consideration in Steady State Loudspeaker Measurements is a good tutorial on the subject of response measurement and looks both at a horn and a 3.5 inch direct radiator measured out of doors, in a deadish (felt lined and curtains) room and finally via an averaging technique where the mic was moved in a circle to reduce standing waves. The outdoor curves include a set with off axis curves revealing the high directivity that a long exponential horn would have to have (before the evils of diffraction slots!).
Note that the 115 cycle horn appears to be the smaller folded horn as developed for the Victor Orthophonic Victrola. It has the same characteristic response as the first theater horn above, very similar artifacts at 400-500 Hz and the top Octaves, but gives up an LF Octave compared to the 15a. Both have the same rising or shelved up upper range, at least on axis. The polar curve of the larger horn would have to be the same but shifted down (same polar nearer 500Hz that this has at 1000).
Not very flat but reasonably smooth. In a cinema application the very high directivity would have to be considered when trying to achieve the best audience coverage or even in determining what the average response curve is. This shows why the articles indicate that they were used in groups of 4 for large theaters: a splayed pair aimed down at the audience and a second splayed pair set low and aimed upwards to the balcony. The Altec designs with multicell horns, of your choice of cell configurations, would be more helpful in that regard.
David S.
Attachments
Yes Bear. I know they aren't the big boys. I checked the number at the back of the enclosure. And yes they look like small wedges. The guy got them out of a smaller old theatre with maybe 100 seats. He was playing classical instrumental music through them during my listening so more mid and high program material than anything else...but impressive nonetheless...made me wonder what the 13a and 15a are capable of.
Simon,
do you get that sensation of "coherence" with any fullrange driver? Is driver size a factor?
Fairly flat power response out of the compression driver and constantly rising directivity of the horn. This is inevitable with a low cutoff exponential design working over many Octaves. Directivity for the treble is formed way back in the throat where the wall angles are very narrow.
A horn good to 60 to 115Hz has to have a very slow flare rate.
Most later horns in the preCD horn era worked well above the 3k mass break point. They tended to balance falling power response above 3k with rising directivity and give just flat axial response and strongly falling off-axis response.
The fairly flat power response from the compression driver is also different than modern practice. Most modern units push the resonances up as high as possible so the are flat to the mass break point and then roll off at 6dB per Octave from there to 15 or 20k. The 505 pushes some parameter downwards to get flatter response to about 4k, and then a steep rolloff from there. Response is flatter but less extended. This is like the comparison between a JBL 2440 and a 2441. 2440 is flat to 10k then dies. 2441 rolls off first order from 3k, but makes it to 16k.
I haven't seen too many horn/drivers that slope downwards on axis. Which are you thinking of?
A horn good to 60 to 115Hz has to have a very slow flare rate.
Most later horns in the preCD horn era worked well above the 3k mass break point. They tended to balance falling power response above 3k with rising directivity and give just flat axial response and strongly falling off-axis response.
The fairly flat power response from the compression driver is also different than modern practice. Most modern units push the resonances up as high as possible so the are flat to the mass break point and then roll off at 6dB per Octave from there to 15 or 20k. The 505 pushes some parameter downwards to get flatter response to about 4k, and then a steep rolloff from there. Response is flatter but less extended. This is like the comparison between a JBL 2440 and a 2441. 2440 is flat to 10k then dies. 2441 rolls off first order from 3k, but makes it to 16k.
I haven't seen too many horn/drivers that slope downwards on axis. Which are you thinking of?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.