Bozak B-305 restoration and upgrade

[drbiggles]

Hey,

Well, thank you to Retrovert for the more coherent answer regarding heat and resistors!

I get it. I've gone in and out of my Bozak B302's so many times now, I'm going to have to fill the original holes with epoxy and re-drill. But I've got tons more issues than you do, I kinda expected it.

Mounting them in another cabinet to set on top of the existing speaker is very doable and it's something that can be un-done easily. Here's my similar solution in a recent project. I bought a pair of bamboo utensil cups at a local hardware store for 8 bux a piece. A little filing, a little drilling and you're done. I did have to use a drill press to get the pilot holes perfectly straight though, not a lot of margin for error there.

Biggles

 

[moondogtn]

Biggles. That is very smooth but I don't know that I have that steady of a hand. Looks sharp though and something to think about for sure.

Btw, I still have one pair of 200x paper come tweets in this setup. That is the first crossover I setup to the 104T.

After completing both crossovers and hooking things up it seems the speaker with both sets of 200Y is not as pronounced in the high area? Don't get me wrong, it sounds great I am just curious. I think I read those paper cones were hotter but not sure. Any thoughts/experience with this? I will swap channels and mess with it some more tomorrow and post findings.

Bill
Bold was edited to correct to B200Y.

 

[Retrovert]

After completing both crossovers and hooking things up it seems the speaker with both sets 200z is not as pronounced in the high area? Don't get me wrong, it sounds great I am just curious. I think I read those paper cones were hotter but not sure.

It is the reverse: the aluminum cone B-200Y and B-200Z tweeters are louder than the paper-cone B-200X. That's why the Tobin mod uses an attenuator to bring them into line with the other drivers.

When Bozak was sold in 1977 the B-200Z quality control reportedly became quite poor. What is the datecode on yours? The B-200Z also had an internal foam damper ring which might have deteriorated. I have never heard a B-200Z.

 

[Retrovert]

Biggles. That is very smooth but I don't know that I have that steady of a hand. Looks sharp though and something to think about for sure.

You mean for drilling the holes?

Here's the trick for that. Carefully centerpunch the hole you want to drill. You can use anything sharp and fine. Then drill that centerpunch with a very fine bit; use the smallest that fits in your drillpress. It will be easier to center the bit with the centerpunched hole. Then drill with the larger one and it will tend to follow the path you drilled.

 

[moondogtn]

So, I think I confused you. I don't have 200z. Sorry. I have B200Y. The cab that has all four B200yb all have a stamp 664.
The other cabinet has a set that are stamped 6510. I also notices led a felt pad on the backs of those stamped 6510. I can post pics of you like.
Not sure what any of this means but I am sure you can shed some light on it as you seem to have much knowledge about these.
Thank you for helping.

I may try something like the cups but need to look around. I have a feeling that will make a big difference.

Bill

 

[Retrovert]

B-200Y vs. B-200Z

So, I think I confused you. I don't have 200z. ... I have B200Y.

Ummm. Ok. You wrote:

... I still have one pair of 200x paper come tweets in this setup.

and

... it seems the speaker with both sets 200z is not as pronounced in the high area?

Hence the confusion over which tweeters you have.

Ok, you've got the good tweeters so you need to use the attenuator to bring them in line with the rest of the drivers.

Decoding Bozak Datecodes

The cab that has all four B200yb all have a stamp 664.The other cabinet has a set that are stamped 6510. I also notices led a felt pad on the backs of those stamped 6510.

No great mystery there.

Bozak datecodes are YYM (single-digit month) or YYMM (two-digit month).

So your tweeters were made in April 1966 and October 1965.

Now you can decode Bozak manufacturing dates for fun and profit!

I can post pics of you like.

No need, we're good.

Thank you for helping.

You're welcome.

 

[drbiggles]

Hey,

I was actually quite impressed with how easy those bamboo utensil holders were to make. All I did was set the tweeter in the hole and mark where the notches were to go. Next I used a smaller hand-held wood saw to cut downwards to some depth that seemed appropriate. I grabbed a very small jewelers rasp and simply grated my way down. This species of bamboo just powdered away like butter. I don't believe one could drill the holes for the screws without a drill press of some kind. Supreme right angles and a slow/steady downward drilling action was key. I thought that since the bamboo powdered away so easily, I would have trouble getting the screws to hold well. I was wrong. The bamboo held the screws taught as though they were glued into place. Clearly, you wouldn't be able to use the same situation for the Bozak 2 tweeter array, but you get the idea.

I hope you were able to get your tweeter situation unwangled.

Biggles

 

[moondogtn]

I have verified crossover connections, wiring, and swapped channels. It still seems to be the speaker that contains 209BC and one set of B200Y and one set of B200X has more louder miss/highs versus the one that has B209B and two sets of B200Y. I'm not certain what exactly the issue is. It's probably slight enough that I'm the only one that will notice. I will have some others listen to it before pulling my hair out....more.

Bill

 

[Retrovert]

I have verified crossover connections, wiring, and swapped channels. It still seems to be the speaker that contains 209BC and one set of B200Y and one set of B200X has more louder miss/highs versus the one that has B209B and two sets of B200Y. I'm not certain what exactly the issue is. It's probably slight enough that I'm the only one that will notice. I will have some others listen to it before pulling my hair out....more.

Diagnostics require isolating the variables.

Given two crossovers, C1 and C2:
C1 -> Y + Y
C2 -> Y + X

Try swapping crossovers (C1 and C2) or just swap the tweeter connections (Y + X) for (Y + Y) to save some time.

That will isolate it to the crossover or the speaker combo.

If you can isolate it to a pair, you'll need to swap the Y tweeters between C1 and C2.

Change only ONE variable at a time and record your findings.

Do not despair, this is fixable. I went through this with Biggles. It just takes some time to diagnose and repair.

 

[moondogtn]

Retrovert, I will swap tomorrow for fault isolation but I think at least part of the issue is the midrange as I was testing the B209bc which isn't labeled and it is an 8 ohm driver which would explain why it sounds a bit louder. As I stated, speakers sound really smooth and great but something just seemed off a bit. I still plan on swapping to isolate issue tomorrow. Thanks again!

 

[Retrovert]

Retrovert, I will swap tomorrow for fault isolation but I think at least part of the issue is the midrange as I was testing the B209bc which isn't labeled and it is an 8 ohm driver which would explain why it sounds a bit louder. As I stated, speakers sound really smooth and great but something just seemed off a bit. I still plan on swapping to isolate issue tomorrow. Thanks again!

After reading your reply I looked at the earlier posting and noticed you have a midrange mismatch. Ah. That was not previously clear to me. I thought you were talking about differences in the HF (tweeter) band.

Yes, these drivers are different:
B-209B: alnico magnet
B-209Bc and B-209C: ceramic (BaFe) magnet. They are otherwise identical except the mounting, and the Bc is rear mount and the C is mountable front or rear.

The ceramic magnet may be slightly stronger than the alnico. Swap them to be sure that's the problem.

If this is the problem you can balance it with a large value potentiometer (so it only passes a tiny percentage of the current) across the driver to divert up to about 3 dB worth of sound. It will not appreciably change the impedance which, in any event, is an average and approximate value. I doubt that the difference in SPL would be more than a trivial amount, likely < 1 dB. 3 dB is about the detectable threshold, but many people can hear 1 dB difference and, as Klipsch noted, we do live in the midrange. Once you've got the right value, substitute a fixed resistor.

Again, change only one thing at a time.[/QUOTE]

 

[moondogtn]

One thing at I time is a must. I have been working on electronics and data systems for 16 years(not audio equipment) but it's always the little things such as this speaker system with a mixed set of variables.

I couldn't wait until tomorrow so I ran jumpers and swapped the tweeter outputs and it is absolutely the midrange.
I also see what you are saying about possibly adding a small resistor to Tone it down a bit. I will work on that tomorrow for sure now that I know I'm not just making this up.

 

[Retrovert]

One thing at I time is a must. I have been working on electronics and data systems for 16 years(not audio equipment) but it's always the little things such as this speaker system with a mixed set of variables.

Complex systems often have so many inter-dependent variables it is mind boggling at times that we can even begin to debug them.

I couldn't wait until tomorrow so I ran jumpers and swapped the tweeter outputs and it is absolutely the midrange.

Great! So you know the problem and the solution.

I also see what you are saying about possibly adding a small resistor to Tone it down a bit. I will work on that tomorrow for sure now that I know I'm not just making this up.

BIG resistor. Not small. Remember, you want the resistor to pass very little current. You are making a current divider and you want most of the current to pass through the midrange. Otherwise it goes through the bypass and you get no sound. The bypass should pass little current which means it is very high value compared to the midrange. You can work out the percentages for 3 dB as the minimum value. Remember as you go up in value it passes less current so the dB attenuation goes down. 0 ohms would be passing ALL of the current and 1 Mohm is passing nearly none of it.

 

[moondogtn]

So, I am guilty of dropping the ball to not update this thread.
Let me say it was something I missed by assuming. The midrange were mismatched impedances. One was 8 ohms and the other was 16. I noticed one was labeled 16 ohms and assumed the other was. I traded another fellow Bozak enthusiast and got another 16 ohm mid. 209b or 209 bc? I don’t recall at this point. Problem solved.
Apoogies for not updating sooner!

Bill

 

[Retrovert]

An impedance mismatch will throw everything off because the crossover points dramatically move.

Here's an quick explanation for those B-305 Bozak owners reading this and who either have purchased or might purchase the wrong midrange upgrade to move from paper-cones to aluminum cones. It is very common to have midrange mismatches in Bozak upgrades because sellers often intentionally sell 16 Ω drivers as 8 Ω, because the 16 Ω is not useful in most of the units aside from the 16 Ω B-305 which puts the 8 Ω woofers in series yielding 16 Ω, so it needs a 16 Ω midrange. (The Symphony goes the other way, using 16 Ω woofers in parallel to yield 8 Ω, so it uses an 8 Ω woofer.)

The crossover for the midrange is a traditional bandpass filter whose parameters are determined by an inductor (ceiling) and capacitor (floor):

Given:
R is the nominal speaker impedance
C is the capacitance (Farads)
L is the inductance (Henrys)​

Then:
f = 1 / (2π x R x C)
f = R / (2π x L)​

With all other parameters and constants in the equation remaining constant, since only the speaker impedance is changing, we need only consider the effect of increasing or decreasing R:

The capacitor crossover point varies as (delta f = 1 / delta R), so if R increases by 2 then f drops by 1/2, if R decreases by 2 then f increases by 2.
The inductor crossover point varies as (delta f = delta R) so if R increases by 2 then so does f , and if R decreases by 2 then so does f.
​

So you can see how a simple change to the midrange impedance can throw it all off. The ceiling drops and the floor increases so the bandpass region moves up or down, shifting a significant portion of the midrange.

Glad it all worked out. Hard to believe this was over two years ago!

Two more potential mods:

Have you added the L-pad to better balance the tweeters?
Have you added the flower pot cover for the midrange to prevents cross-modulation between woofer and midrange? Cleans up the sound.​

Thanks for updating us on the status!

 

[moondogtn]

I will have to find my thinking cap to perform those calculations but I believe you!

Just for clarification, I used the 104T schematic below and moved the leads closest to inductors, I believe it was tbe 104 taps?

I have not yet added the Lpad. For these, will I need a 16 ohm mono Lpad? Then, simply remove the 7.5 ohm resistor and install Lpad.
Of course, I will need to open up that can that I never should have reused in the first place. Trying to keep it original looking. Not worth it unless you just want to make it difficult.

I can see if I can find the flower pots locally to cover mids.

I actually just have the pink insulation in them. So, I will need to source the cotton batting.

Once these are done, I have a pair of 302s that I was planning on working on soon.

Bill

 

[Retrovert]

Attenuator
The exact attenuation value required for your setup will be empirically determined based upon the type of music, your room configuration, the placement of the tweeters, and your ears. Remember, as the attenuation increases the volume drops, and as it decreases the volume increases. Since the tweeters are closely balanced, but a little too loud for the other driver, you need a small amount of attenuation. Again, idiosyncratic and not predictable, so no standard value suits all. What I'm saying is that this is a by-ear measurement, not an absolute value, so you want to maximize flexibility.

Having the exact attenuation allows you to dial in what you prefer, and that may considerably vary. As I've elsewhere opined, since we can't ask Pat Tobin what his intentions were, the fixed value he selected was likely to ensure the cabinet did not require modification to add a variable attenuator and his customers could not use the wrong value. He likely selected this based upon his musical preferences and hearing, but at this point there really is no way to know.

The L-pad we use is basically configured as a rheostat (adjustable wirewound resistor) in the configuration we're discussing, not as a parallel impedance element. L-pads are easier to source in higher power than rheostats. Hence the suggestion.

You may purchase L-pads (or rheostats for that matter) on eBay or just about anywhere else. I suggest you purchase a NOS device, however, as the new ones from China often have poor quality wiper contacts which can fail. I've seen enough complaints about them. When I needed 16 Ω rheostats for another project I purchased Russian military surplus units. The appearance is very similar to the US made ones of the 1960s, except that the porcelain and glaze quality are greatly inferior.

Curtain and Lining
The curtain loses the backwave and delivers the clean bass which is the Bozak's signature. Suggest you switch from fiberglass (which is an acoustic brake) to thick cotton batting.

Tufflex is no longer available, but it wasn't fancy, merely a dense reconstituted cellulose very similar, but not totally identical at the core, to cotton fibers, so any dense cotton batting should identically perform. The issue between reconstituted cellulose (tradename rayon), is mostly in the area of stiffness, water resistance, absorption of oils, and flammability. None of which really matter for this application.

The side lining absorbs reflections, not slows them.

Discussions in other threads.

Flower Pot Midrange Cover
Flower pots at your local garden center. Remember to line inside and out with cotton batting. Explanations in other threads.

Cabinet Backs
Many times with all the on/off movement of the screws for the back the holes loosen. You must have a good seal for the infinite baffle to work, so the back must be tight. (It's a lot like a Karlson.) If your screw holes loosen you can fill them with epoxy, or epoxy putty, and then redrill. I've had good results doing this and so has Biggles. Again, described in other threads.