SDA-SRS crossover points?

madmax

Does anyone know how the four tweeters in the SDA-SRS are crossed over? For example which one is crossed over the highest and which one starts at the lowest freq?
madmax

RuSsMaN

I think F1Nut posted a 'firing order' sometime back. I'll do a search see if I can find it.

Cheers,
Russ

madmax

I remember that too. That is where I learned that the second one from the bottom is the highest. I've searched for it but can't find it. What I would really like to know is the range of each tweet. For ex: top one from 2500 to 5K, second one from... to..
madmax

F1nut

Ok, I'm going to edit my "running on memory" post as Raife has chimed in with his expertise and I agree that this is the right order.

Top = 3rd
2 = 4th
3 = 1st
Bottom = 2nd

I have no idea what the range of each is, maybe Ken will chime in on that one. Then again, maybe not.

DarqueKnight

I do not know what the frequency ranges of the individual tweeters are. Looking at the schematic and doing some quick impedance calculations (at 2000 Hz and 10,000 Hz), the second list posted by F1Nut makes more sense:

Top = 3rd
2 = 4th
3 = 1st
Bottom = 2nd

The top tweeter is in series with a resistance of 15 ohms and an inductance of 0.4uH.

The second tweeter is in series with a resistance of 22.5 ohms.

The third tweeter is in series with a parallel combination of a 4.4 uF capacitor, a 22.5 ohm resistor, and a 750 pF capacitor.

The fourth tweeter is in series with a 7.5 ohm resistor and a 0.7 uH inductor.

I do not think that tweeter crossover points will fall under the domain of "proprietary information". However, Polk's engineering department is not as forthcoming with information as they were during the "golden age" of SDA manufacturing back in the 1980's and 1990's. I tried to obtain frequency response charts for the SL2000 tweeter, the SL2000T replacement tweeter, and the SL3000 replacement tweeter . My request was politely declined for proprietary information reasons. When I reminded them that they published the response curves for the SL3000 tweeter in a September 1989 Technical Report, they said they remembered that, but they don't release detailed technical information anymore.

If anyone has the time, they could always simulate the circuit in software and calculate the response for the individual tweeters.

madmax

The 2nd tweeter only has a 22.5 ohm resistor? So that means it is full range with a reduced amplitude. That is very strange.
madmax

DarqueKnight

Originally posted by madmax
That is very strange.

Not really. The entire tweeter array is in parallel with:

(1) a series combination of a 0.4 uH inductor and 2.7 ohm resistor. The 2.7 ohm resistor is bypassed with a 92 uF capacitor,

and

(2) a 12 uF capacitor that is bypassed with a 750 pF capacitor.

madmax

I was trying to draw it out but I don't think I got it right. No wonder the inside of the old SDA's had that smell to them.
madmax

DarqueKnight

I sent an email to Ken asking about the operating ranges of the SRS and SRS 1.2TL tweeters.

This would be nice info to have. I could simulate the circuit in PSPICE, run a frequency sweep from 1 to 20 KHz, and plot the four frequency responses. However, its easier to just ask Ken.:)

DarqueKnight

O.K., my curiosity got the best of me. While we are waiting on Ken to get back to us on this interesting technical question, I simulated the SRS high frequency crossover circuit in PSPICE.

This simulation result supports our earlier discussion regarding the firing order of the four SRS tweeters. An arbitrary 14 volts was used as the source signal amplitude. The signal was swept from 2000 Hz (the SRS high frequency crossover point) to 20 kHz.

Tweeter 3 is by far the dominant transducer from 3.5 kHz to 20 kHz.

Tweeter 4 is dominant from 2 kHz to 3.25 kHz. Its response falls off sharply after 7 kHz.

Tweeter 1 has a tenuous hold on 3rd place and starts going to sleep at 7 kHz.

Tweeter 2 hangs in from 2 kHz to 8 kHz, then offers mostly moral support after that point.

According to a Polk brochure,

"The SRS uses Polk's Progressive Point Source technology to maintain a constant vertical directivity of mid and high frequencies, which prevents undesirable beaming. As frequencies increase, the tweeter array adjusts its radiation area and eventually becomes an ideal point source at the very highest of frequencies, eliminating high frequency interactions and reflections between drivers."

DarqueKnight

Here is another response plot sweeping from 1 Hz to 26 kHz.

madmax

That is very useful information in understanding how they work together. Just wondering, can you do a summation of the four curves into one output curve?
madmax

DarqueKnight

Here is a composite curve summing the response of the individual tweeters. The reader is reminded that this is a simulated output voltage response rather than an anechoic chamber response.

I was surprised to find a peak in the response at 13 kHz. It is well known that SDAs from generation 1 through 4 had a tweeter resonance at 13 kHz. I always assumed that this was a solely due to the properties of the SL2000 tweeter. This simulation result implies that the rise in response at 13 kHz was a characteristic of the high frequency crossover circuit. The performance parameters of the SL2000 tweeter probably just enhanced the crossover circuit's emphasis at 13 kHz.

madmax

This does not take into account the curve of the tweeter itself, does it? This is just what is being fed to the tweeter or am I wrong.
madmax

DarqueKnight

Here is a room response plot of the SDA SRS 2 that was published in the November 1986 issue of High Fidelity magazine (pg. 37).

The tweeter room response plot tracks closely with the simulation results shown in my previous posts. The poor quality of the graph is due to the source (library microfilm). The shaded left portion of the plot was difficult to copy without washing out the rest of the plot.

DarqueKnight

Here is a room response plot of the SDA 1 that was published in the January 1983 issue of High Fidelity magazine (pg. 41).

madmax

Let me explain where I'm trying to get to here. When I first received the wrong tweeters (the '3000 replacement with lower resistance) I unknowingly put them in and although it didn't sound quite right with all 4 wrong tweeters there were several things which seemed much better. The very high frequencies seemed much more extended and the upper midrange was lowered slightly which made the speaker very dynamic sounding and also seemed to extend the bass dramaticly. The sound was alive and most recordings sounded much better. There were, however some things I couldnt stand. The highs around 15K were VERY loud (kind of a lisp sound) and the midrange (trumpets for example) was a little too depressed. I'm thinking a few resistor changes are in order to get the best of both worlds. I wonder what you would see if you plugged in the '3000 tweeter coil resistance values in.
madmax

DarqueKnight

Originally posted by madmax
This does not take into account the curve of the tweeter itself, does it? This is just what is being fed to the tweeter or am I wrong.

You are correct. I could not accurately model the tweeters themselves because I did not have their electrical parameters. Polk does not release information on their driver parameters and I do not have, nor do I have convenient access to, one of those machines that will measure and derive driver parameters.

In the simulation, the tweeters were modeled as 7.5 ohm resistances. The curves represent the signal that is being input to the tweeters from the crossover circuit.

DarqueKnight

Originally posted by madmax
I wonder what you would see if you plugged in the '3000 tweeter coil resistance values in.

The simulation model tweeter resistances were changed to the 5.6 ohm value of the SL3000 tweeter:

DarqueKnight

Simulated composite high frequency response of the SRS with SL3000 tweeters.

madmax

Hmmm,
The curve is the same basic shape but is a higher amplitude. I guess that loud peak I hear is the 13Khz peak. Now why in the world would this make the bass so much more impressive sounding?? It is very noticeable! It is full, deep and very forceful. I would have expected it to seem weaker with the higher amplitude of the tweeters.

I'm downloading pspice right now. Could you send me the file you made for the xover? OK if not. I would like to play with it.

Thanks,
madmax

DarqueKnight

I will email the files. They are very small in size. In case you have not used PSPICE before, an excellent tutorial resource page can be found here:PSPICE Tutorials

madmax

Thanks!

I just received them. I will try to get it going over the weekend.
madmax

[Deleted User]

Sirs,
I've sent your request to Stu Lumsden and will forward his reply.
Take care, Ken

F1nut

You're both geeks , but I'm loving this.

DarqueKnight

Yeah, you know you're a geek wheeeeen.......you're looking at circuit schematics and response plots when you're on vacation!:)

DarqueKnight

Originally posted by Kenneth Swauger
Sirs,
I've sent your request to Stu Lumsden and will forward his reply.
Take care, Ken

Thanks Ken.

trubluluc

I must have nodded off or something.

-Luc

madmax

Raife has achieved official SDA Knighthood because of his extensive knowledge of SDA's.
madmax

I-SIG

This is good stuff, guys. Keep it coming! Thanks to Ken as well for doing some checking in-house.

Wes

[Deleted User]

Folks,
Here is Matthew's reply:
"I should check the old x/o drawings but I recall that the order was from highest to lowest 2, 3, 4, 1 with 1 being the tweeter closest to the floor. The crossover points are not standard slopes. The idea was to arrange the high end roll-off of each tweeter to achieve constant vertical directivity vs frequency. Vertical directivity is determined by the ratio of the wavelength to the length of the line source. The four tweeters create a line source which, if they all covered the same frequency range, would become more directive at higher frequencies (shorter wavelengths). By progressively rolling off tweeters, the line source becomes effectively shorter at higher frequencies. By carefully choosing the roll-off characteristics of each tweeter the output of all the tweeters adds up to flat response and pretty much constant vertical dispersion across the range covered by the tweeter array."
Regards, Ken