Re: Just did it!
Most of my experience is here. I can also add that another fellow took a number of drivers & coated them with quite a few different things and puzzlecoat worked best.
The baskets and the magnet basket should be ductsealed. They are cheap drives and one place this shows up is on the "paper" thin baskets. I guess you could go crazy and pot them in epoxy 1st.
I don't think they actually paid much attention to that
dave
Bas Horneman said:
Most of my experience is here. I can also add that another fellow took a number of drivers & coated them with quite a few different things and puzzlecoat worked best.
The baskets and the magnet basket should be ductsealed. They are cheap drives and one place this shows up is on the "paper" thin baskets. I guess you could go crazy and pot them in epoxy 1st.
SY said:
I don't think they actually paid much attention to that
dave
Yeah, there's a quote I always loved, "Highly designed, lightly engineered."
OK, for Bas, a quick explanation of the gibberish. At a distance that's large compared to the length of a line (4-5x), the line appears as a point source. At closer distances, it appears as a line source that has a dramatic loss of higher frequencies above and below its ends. The consequences to the polar pattern are obvious; you'll have a narrow vertical window where the tweeter and the line source woofer are in balance, and the reverberant portion and early reflections will have a very different spectral balance.
Worse yet, the falloff of intensity versus distance is different for point sources and line sources. This means that whatever balance you select between tweeter and woofer will be different as you move closer or further from the speaker. And again, this will be reflected (bad pun) in the reverberant response. This has plagued every commercial implementation of line source ESLs/ribbons crossed to a cone woofer in a box that I've ever heard.
I'm about to add a 5th Law to SY's Four Laws: Lines with lines, points with points, and never the twain shall meet.
Thanks all again.
Here is a post by Doc B. who designed the S8's (From AA)
"I wouldn't call it a subwoofer. Without any EQ or the effects of corner boost the S8 woofer array will probably run down to somewhere around 55Hz or maybe a bit higher, as we have noticed the Fs has been climbing up a bit over the years, averaging about 65-70 Hz out of the box, rather than the published 54 Hz, which I did measure 5 years ago when they were first introduced.
It is interesting to note that the current 1870 comes with a straight cone rather than the curved cone of the earlier ones. Maybe this relates to the change in Fs somehow, though PJ speculates that it's usually an increase in rear suspension stiffness that causes an upward shift in Fs. Also interesting to note that we didn't measure much difference in Fs between straight coned raw drivers and straight coned drivers with several hundred hours on them.
The other parameters seem to remain close to those published in the catalog. I think there is some data published on the full range driver forum that shows a measured sensitivity of 84dB for these drivers, but we just tested a group of 5 or six from various batches with the Praxis software and our result was that they were pretty consistently within a few tenths of a dB of the published 87 dB, as was a 55-1855 and a 55-1290.
I have in the past measured bass response of the S8 as pretty even down to the mid 40's with careful room placement, and I have been using a subwoofer coming in at around 40Hz for the past couple years. One thing we know is that when measuring the lower midrange to midbass response of an an array like this you need to get data from a lot of different points in space for consistent results.
I'd guess that you are correct that you can probably get by without the cone dope if you are only running to 500Hz. The peak around 8-10 Khz will probably be 20dB or more down with a first order slope. A second order slope would probably make it insignificant. My hunch is that an even higher order crossover will be best."
Here is a post by Doc B. who designed the S8's (From AA)
"I wouldn't call it a subwoofer. Without any EQ or the effects of corner boost the S8 woofer array will probably run down to somewhere around 55Hz or maybe a bit higher, as we have noticed the Fs has been climbing up a bit over the years, averaging about 65-70 Hz out of the box, rather than the published 54 Hz, which I did measure 5 years ago when they were first introduced.
It is interesting to note that the current 1870 comes with a straight cone rather than the curved cone of the earlier ones. Maybe this relates to the change in Fs somehow, though PJ speculates that it's usually an increase in rear suspension stiffness that causes an upward shift in Fs. Also interesting to note that we didn't measure much difference in Fs between straight coned raw drivers and straight coned drivers with several hundred hours on them.
The other parameters seem to remain close to those published in the catalog. I think there is some data published on the full range driver forum that shows a measured sensitivity of 84dB for these drivers, but we just tested a group of 5 or six from various batches with the Praxis software and our result was that they were pretty consistently within a few tenths of a dB of the published 87 dB, as was a 55-1855 and a 55-1290.
I have in the past measured bass response of the S8 as pretty even down to the mid 40's with careful room placement, and I have been using a subwoofer coming in at around 40Hz for the past couple years. One thing we know is that when measuring the lower midrange to midbass response of an an array like this you need to get data from a lot of different points in space for consistent results.
I'd guess that you are correct that you can probably get by without the cone dope if you are only running to 500Hz. The peak around 8-10 Khz will probably be 20dB or more down with a first order slope. A second order slope would probably make it insignificant. My hunch is that an even higher order crossover will be best."
SY's First Law: "Is it plugged in?"
I've already posted SY's Second Law on a few occasions- "Stop speculating and do the #$@*ing experiment!"
SY's Third Law: "An analogy is not an equality."
The Fourth Law was particular to a company/technology that I worked with for about 10 years and has no general application. My proposed replacement Fourth Law is, "When a major breakthrough is made that will change our thinking about the Universe, it is unlikely to be made by a guy selling HiFi gadgets out of his garage."
Sorry, nothing too original. But I've had a pretty good track record of turning out bright, practical engineers from my product development groups, and they all say that these laws were the most useful thing they ever had beaten into them.
Hi Sy,
Easy enough with the EQ. The rest is gonna be built on circuit board by Gary Kaufmann. Any ideas? ..or should I just put it all together with clip leads first...and build the final version on the pcb...Guess the last option heh
The preamp is transformer coupled with 600ohm output impedance...(Just for info..don't know if that is useful to know or not)
Cheers,
Bas
Easy enough with the EQ. The rest is gonna be built on circuit board by Gary Kaufmann. Any ideas? ..or should I just put it all together with clip leads first...and build the final version on the pcb...Guess the last option heh
Why should I buffer?...just curious...I'd hate to put even more circuitry between pre-amp and power amp.
The preamp is transformer coupled with 600ohm output impedance...(Just for info..don't know if that is useful to know or not)
Cheers,
Bas
Attachments
If you're certain about your low source impedance, then the buffer may be superfluous. I'd still be inclined to use it- a sonically transparent buffer is no big deal. And, of course, you'll need to be sure that the input of the next amp in the chain is sufficiently high that you don't have to buffer the output.
One way that I've used to construct circuits that will be tuned is to build the board, but solder turrets or flea clips or some other sort of terminal in the holes for the frequency-determining components. That way, the components can be soldered and unsoldered from the top of the board with little danger. Some EQ designs even put those components on a plug-in header, but I'm always nervous about adding plugs and connections in the signal path for reliability reasons.
One way that I've used to construct circuits that will be tuned is to build the board, but solder turrets or flea clips or some other sort of terminal in the holes for the frequency-determining components. That way, the components can be soldered and unsoldered from the top of the board with little danger. Some EQ designs even put those components on a plug-in header, but I'm always nervous about adding plugs and connections in the signal path for reliability reasons.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.