Bozak B4000 Symphony - Restoration - Seeking Advice

cubby01 said:
<Chuckle>. Tell my wife that. :) She liked the sound by the way, even before any updates. She also asked if they could be stacked, you know,like if we had four of them....:D
Click to expand...

COOL. I think you and I have darned nearly the only wives that react like that. I drug home a pair of SpeakerLab K Horns one fine afternoon. She gave a guttural rumble and said, now those are speakers.

Biggles
 
A stock N-105 crosses at 200 and 1,500 Hz, using an 8 uF and 50 uF capacitor bank

Each bank uses a 25 Ω resistor to reduce the capacitor's attenuation.

I believe that crossover has a swapped wire. The connection should be on the other side of the resistor as a connection for orange from amplifier.

The midrange is reversed polarity, BTW.
 
cubby01 said:
<Chuckle>. Tell my wife that. :) She liked the sound by the way, even before any updates. She also asked if they could be stacked, you know,like if we had four of them....:D
Click to expand...

You can't do that. Comb interference. But what you could do is make a kick-ass quad setup. Or use one as the center-channel and one as the rear omni, and have the best surround setup you've ever heard.
 
Retrovert said:
A stock N-105 crosses at 200 and 1,500 Hz, using an 8 uF and 50 uF capacitor bank
Click to expand...

Well this is a stock N-105A network which I believe is for use with the B-800/B-800A midrange. There is a 5 uF and 45 uF (3x 15uf) cap bank. Definatly stock old CCC caps. If one looks back at both Rudedogg's b4000 w/ 800A in this post...

rudedogg said:
It's crazy! But it is not there. What i believe has been done is there are three 15 mfd caps strapped together to make 45 mfd. The 5mfd cap is seperate from the banks connection and with its resistor in tandem is used for the 200y tweets! It is working after 50 years... Wth???
Click to expand...

He mentions that 45uF mid / 5uF tweeter setup. As does Jerryfan and geg197, elsewhere. I thought I saw where Biggles had this same x-over but couldn't find the post. I've seen other examples of this same cap lineup for the N-105A on the web. I think it's that way to take advantage of the much wider range that the B-800 has over the B-209/b/bc line of mids.

Retrovert said:
I believe that crossover has a swapped wire. The connection should be on the other side of the resistor as a connection for orange from amplifier.
Click to expand...

That's what I thought at first too. But the second x-over is the same. Then I started looking around, and you can clearly see in Matt0404's initial post in this thread (7th Picture) where the orange wire internal in the box between the two 25ohm resistors. Same as mine.

And also there's this hand draw schematic from the FB group that, although a bit harder to follow, matches my drawing above. So it appears intentional. My understanding is the B-800 mid was a special order for the Symphony so that may be the key.

index.php


For now, I'm going to keep the original wiring and just replace the out of spec caps with like values. I would love to find an easy way to model this.
 
Hey Cubby, I absolutely love my Symphony set with just refreshed crossovers like yours. I did add bypass caps to each bank of larger films. They are by far my favorite speaker I have owned so far
 
Last edited:
Too much to go over via forum . I have 6 sets of b4000 from 62-73’s. Enough drivers and xo’s to build 2 more sets. Let me know what u need and yrs of your drivers. Hope u meant u have 800,800a, it 800yc mids range. My 73 had 209 which I quickly removed!
 
I have 6 sets of symphonies and restored countless sets. I have a few updated xo’s on the shelf. The mid doesn’t sound better to me with reversed polarity . Bob betts has nice article.
 
Hi everyone -
I just inherited a pair of Bozak Symphony speakers. 2 B199A 16ohm woofers in each, B800 midrange, 8 200Y tweeters. The crossover is in the picture - The 3 caps are 5uF each, wired parallel, and only connect to the tweeters. I notice the schematics in this thread are quite different.... I expected to see caps involved with the mid also. I am trying to find / figure out what version I should use to get these sounding right, etc. I'm sure the old caps are bad, they don't measure good on my capacitor tester. Also looks like I need to isolate the mids from the woofers. And upgrade the wires from the scrawny OEM ones. I also notice the back panel of the enclosure is far from sealed in the corners. The speakers I have scratchbuilt are air tight except where intended. Any directions to the correct thread, [or should I start a new one?] would be appreciated.Bozak Symphony Original crossover.jpg
 
I have an early set like your and 5 sets after. I talked to Bob Betts about the early ones with the 15uf in tweeter only. He said put and 8.2 and resistor in tweeter and 47 or 50 in mid with second resistor. I use 47 because 50 so hard to put in can. Then there’s a cross wire from top of tweeters to opposite side of mid. I have three xo on the shelf all done.
The earky 800a is undamped and makes the speaker brighter then ones after 64 which use 800a a damped version. The grill cloth is a big job on that style. Call me 4103366490 if questions and I’ll email or text you images. Make sure all driver work before starting.
 
Luuuuke! I am your crossover math. Ussssse the formulas!

Midrange-Tweeter Crossover Point
15 µF = 1,326 Hz ≈ 1,300 Hz
8.2 µ = 2,426 Hz ≈ 2,500 Hz

Woofer-Midrange Crossover Point
50 µF = 398 Hz ≈ 400 Hz
47 µF = 423 Hz ≈ 425 Hz

This difference may be slightly audible because the inductor is crossing 400 Hz and a gap in frequency response therefore occurs. Non-musician and (non Bud-Fried) humans tend to poorly hear that low, but it may be slightly audible.

Any arbitrary capacitor value may be obtained by adding in parallel additional capacitors of smaller value. This technique also improves the performance; look up "bypass capacitors" on AK. (Digital circuits use the same technique.)

This crossover point, like all other aspects of speaker design, is a tradeoff, because the B-800 suffers from increasing cone breakup with increasing frequency, but the tweeter, too, has distortion issues, increasing as frequency drops, plus very poor dispersion aka beaming. At the same time, humans acutely hear in the midrange — as Klipsch noted, we live in the midrange — and taking the crossover hit at 1,300 Hz is problematic. Hence the move to a higher frequency and accepting cone breakup as a compromise to avoid overlap and distortion from filter handoff.

Here's what I previously wrote about cone breakup and the B-800:

Consider the two-way systems with an 8" driver, and look at the crossover points which are traditionally set at ≈ 1,500 Hz, as smaller (tweeter) drivers typically becomes unusably beamy around 1,000 Hz, but sometimes as high as 2,000 Hz, but cone breakup of the larger driver is a bigger factor.

Here are some handwavy arguments with rough numbers, as I have not measured the exact diameters of the cone.

Consider the characteristics of an 8" B-800:
// 8 inch diameter driver has a cone — excluding surround — slightly less than 8 inches, guesstimate at 7 inch
Vs = speed of sound at STP = 344 meters per second
Dmax = Diameter maximum = 8" (cone plus surround) = 20.32 cm = 0.2032 meters (cone plus surround)
Dmin = Diameter minimum = 7" (cone only) = 17.78 = 0.1778 meters (cone only, excludes surround)
fmin = minimum frequency before beaming
fmax = maximum frequency before beaming​

fmin = Vs / Dmax
= 344 m/s / 0.2032 m
= 1,693 Hz

fmax = Vs / Dmin
= 344 m/s / 0.1778 m
= 1,934 Hz​

So best case is ≈ 1,900 Hz, worst case ≈ 1,700 Hz. Hence 1,500 Hz to 2,000 Hz as safe values. Either way, 2,500 is likely too high.

Compare with the 6" B-209, again guessing for 5" as actual cone diameter:
// 6 inch diameter driver has a cone — excluding surround — slightly less than 6 inches, guesstimate at 5 inch
Vs = speed of sound at STP = 344 meters per second
Dmax = Diameter maximum = 6" (cone plus surround) = 15.24 cm = 0.1524 meters (cone plus surround)
Dmin = Diameter minimum = 5" (cone only) = 12.7 = 0.127 meters (cone only, excludes surround)
fmin = minimum frequency before beaming
fmax= maximum frequency before beaming

fmin = Vs / Dmax
= 344 m/s / 0.1524 m
= 2,257 Hz

fmax = Vs / Dmin
= 344 m/s / 0.127 m
= 2,708 Hz​

So best case is ≈ 2,260 Hz, worst case ≈ 2,700 Hz. Hence 2,500 Hz as a safe value.

Any of the lower points are problematic, being smack dab in the region where humans acutely hear.
 
Can you post a schematic / diagram of the wiring? Or point me to a file on the board? With what I am seeing with so many different mods and different variants of these speakers, I want to be sure to start with the correct one for the drivers I have.

Once I have that info, I can do some tests, etc. I have access to an audio testing setup that can generate graphs of frequency response, THD, Level, etc. I can use swept sine waves or pink noise for frequency response.

The B800s I have have no foam, or any evidence of foam that crumbled. I unmounted one to peek.

On the back of the cabinets - Code 62 4 Serial numbers B24130 and B24131.
 
N-105 schematic attached.
 

Attachments

  • Bozak N-105 Crossover (B-4000 Symphony) (1964, Sept).jpg
    Bozak N-105 Crossover (B-4000 Symphony) (1964, Sept).jpg
    98.9 KB · Views: 32
The B-200Y tweeter places damping foam underneath the dustcap. Said foam over time dessicates and hardens.

The B-800A uses a special surround to mechanically limit response below 100 Hz to permit using the driver without a crossover. This is a bad, bad idea as I've elsewhere explained.
 
Ah, I hadn't come across the B-200Y having foam.... I wonder how this changes their characteristics...

My mids are marked B800, with no A, no visible evidence of degraded foam I could find.

I don't think it's likely anyone made changes to these speakers before I got them. I am the second owner.

Than you for the schematic / confirmation that it's the one I am looking for. Another one I had found showed slightly different capacitor values in the crossover. 5uF vs 8uf, and 45uF vs 50uF.
 
You may run the formulas and verify the frequency differences between various capacitor values.

Here is what I have previously posted about that, since most people are using calculators which mask understanding.

Formulas for crossover filters:
Given:
R is Nominal Speaker Impedance (Ω)
C is Capacitance (Farad)
L is Inductance (Henry)
f is Frequency (Hertz)

Inductor Formulas:
L = R / (2π × f) (Henry)
f = R / (2π × L) (Hertz)
R = 2π × L × f (Ω)

Capacitor Formulas:
C = 1 / (2π × f × R) (Farad)
f = 1 / (2π × C × R) (Hertz)
R = 1 / (2π × C × f) (Ω)

L is inductance in Henry.
To convert from Henry (H) to milliHenry (mH) multiply by 1,000.
To convert from milliHenry (mH) to Henry (H) divide by 1,000.​

C is capacitance in Farad.
To convert from Farad (F)to microFarad (uF) multiply by 1,000,000.
To convert from microFarad (uF) to Farad (F) divide by 1,000,000 or multiply by 0.000001.​

The easiest way to run these calculations with a calculator, be it physical or computer, is to compute the invariant portion, store in memory, and then tinker with values. For example, the inductor calculations can be performed as:
f = R / (2π × L) (Hertz)
R = 8 Ω
L = (unknown value) in milliHenries, not Henries
f = 8 / (2π × 0.001 × L) (Hertz)
Compute without L and then divide by the mH value (not scaling to Henries) as needed, essentially:
f ≈ 1,273 / L (Hertz)

Example:
L = 1.59 mH
f ≈ (memory recall) / 1.59 (Hertz)

L = 1.60 mH
f ≈ (memory recall) / 1.6 (Hertz)
​
So, mechanically, compute the entire equation without L, and store in memory. Then recall that value and divide by L as mH. For example, to compute the frequency for, say, 1.59 mH versus 1.6 mH, compute everything except L, store in memory, then do a memory-recall and divide by 1.59 and then a memory-recall and divide by 1.6, since the 0.001 factory is already entered. Makes it far less time consuming.

A similar trick can be done with capacitance.
f = 1 / (2π × C × R) (Hertz)
R = 8 Ω
C = (unknown value) in µF
f = 1 / (2π × 0.000001 × C × 8) (Hertz)
Compute without C and add as needed, essentially:
f ≈ 19,894 / C

Example:
C = 8.2 µF
f ≈ (memory recall) / 8.2 (Hertz)

C = 8 µF
f ≈ (memory recall) / 8 (Hertz)
​
So if you wanted to compute the difference between 8 µF and 8.2 µF (standard value) all that would be needed is to divide the pre-computed value by 8 or 8.2.

Greatly simplifies the calculations and saves considerable time.
 
A little more investigation of these speakers... I unmounted 1 tweeter. Looks like what I expected. Gently pushed the cone, wow, it doesn't move at all. Others I have worked with move at least a little. But they do play, I wired up a new set of stacked film capacitors to replace the original ones from '64. Just the 15uF 1 capacitor version for now, while the others for the full 3 way crossover ride the truck. I also noticed the tweeters are all just touching each other over the slot cut for them - not really sealing the box for the woofers... Some more listening, these speakers seem really light on the bass end, especially the low stuff. Oddly, with the back off it *seems* to have more bass, at least from behind. Other thing I noticed, is that all of the drivers are mounted to the back / inside of the front panel.

I learned from running frequency response tests on speakers at my job years ago, that this essentially makes a weird short horn on a driver. What I found back then was that even a 1/4" deep "tunnel" made by mounting behind the panel had a dramatic effect at 2-3KHZ, making an obvious and annoying bump in the frequency response of a driver about the diameter of these tweeters. I re mounted that speaker from the front, and the bump was gone. Ever since, I have always mounted drivers from the front.
 
Chamfering Baffle

The described effect regarding rear-mounting drivers is a well-known baffle issue. The solution is chamfering, i.e. rounding over the hole to smooth the transition and prevent disruption.

Troels Gravesen has presented actual measurements for the benefits of chamfering:

Sealing Cabinet

Infinite baffle requires a sealed cabinet to lose the backwave. Without the back the backwave will interfere with the frontwave. Leaving the back off will result in peculiar sound with peaks, but will not deepen the bass nor will it make it more accurate.

The cabinet need not be perfectly sealed, and should, in fact, not be perfectly sealed. Some minor leakage is required for any cabinet to properly function. In acoustic suspension, for example, the pressure is the restoring force, but that does not mean that a perfect seal is desirable.

Here's what I previously wrote on that subject of sealed cabinets, in addressing infinite baffle and acoustic suspension, and why a bit of air movement is required. It deals more with acoustic suspension, but the basic function is identical.

Perfectly Sealed Box

When any sealed box, be it infinite baffle or acoustic suspension, is perfectly sealed then the air pressure internal to the box differs from the air pressure external to the box. The interior cabinet air would be trapped at the time of manufacture, and its volume fixed. As a result, as the external pressure increased or decreased the cone would be moved in or out, changing its resting position. This is how a barometer functions.

Air Exchange

To avoid having the resting position vary with air pressure — air pressure varies with altitude and with storms, so it constantly changes, which is how a barometer works — every cabinet must therefore have a means to permit air exchange between the interior and exterior. The mechanism of exchange must be small to avoid having the backwave exit through that orifice and thus (a) (in infinite baffle and acoustic suspension) interfere with the front wave or (b) (in acoustic suspension) reduce the restoring force on the woofer by reducing the interior cabinet pressure.

At the same time, the orifice must be sized large enough such that air does not rapidly move through it or the orifice will squeak, like one of those rubber squeeze toys for dogs or children. A pinhole will squeak.

Dust Cap as Exchange Orifice

The dustcap is a nice way to exchange air from the interior of the cabinet to the exterior, and this also cools the voice coil preventing it from overheating and even burning out. It provides a nice surface area so the rate of air flow through any spot is small.

Cabinet Sealing

Despite the need for tiny amounts of air exchange, in both acoustic suspension and infinite baffle the cabinet best functions with a nearly perfect air seal.

This is why a gooey caulking like Mortite (tm) or other filled butylene rubber was historically used around the drivers and the back. It is why the surround must be resealed after some decades, because the butylene rubber in the surround has oxidized, hardened, flaked off, or otherwise degraded. The drivers must be perfectly sealed to the baffle, the back must be perfectly sealed to the cabinet, and the surround must be perfectly sealed to prevent air from moving through it. The only air exchange is through an opening kept small and diffuse, and that is the dust cap.

As has been demonstrated, the dust cap must be a bit porous to allow some air flow to equalize the air pressure and to cool the woofer voice coil which can otherwise become quite hot.
 
Ok, so the box not "perfectly" sealed makes sense - The ones I made were sealed, except that they are tuned port, so the air pressure problem is avoided by the planned "hole." I was never thinking of using them without the back in place, it was just an odd observation, which I figure is caused by the mid and tweeters being driven a bit on the hot side compared to the woofers. Possibly the bass being able to move the midrange cone messes with things in addition to just bothering the midrange's performance also?
 
You are correct about the issue of cross-modulation. Eliminating cross-modulation is why I suggest the flower-pot modification. Other threads explain that in detail. For example:
I suspect the issue you are experiencing in muffled bass is more a muffled midrange because of distortion arising from poorly performing capacitors. The sound will improve when the midrange high-pass filter is replaced. Same for the tweeters.

The midrange is not imbalanced with the woofers; both drivers are perfectly matched. You are, however, fully correct in observing that the tweeters are slightly too bright. This is why I suggest an L-pad for adjustment. (The L-pad is a critical modification to the Tobin mod, as the attenuator is best replaced with an L-pad for reasons I describe in gory detail.) Adjusting the tweeters, with new capacitors, will improve the sound.

You may also be interested in my four-way modification which cuts the old tweeters at 8k Hz and transitions to a soft-dome. The advantage is removing distortion and providing more air.
 
You must log in or register to reply here.
Back
Top Bottom