New/Old project finally under way, and looking for driver data - old PE buyout driver

[mdocod]

This is a project that was actually sort of started back about 10 years ago when I picked up these 6.5" woofers and some "silkies" to build some MTMs. They wound up getting packed away, and the boxes were never built, sort of forgot about them until my interest in speaker building has been rekindled in the last year or so. Knowing that the silkies aren't exactly up to par with other offerings these days in a similar price point, and knowing that the woofers are not likely very special either, I'm not going to fret a thousand little details trying to make a perfect speaker here. If they turn out sounding pretty decent and looking alright when finished, I might sell them, otherwise, they will be our new garage speakers to replace some 1980s panasonics.



I have the originally published T/S specs for the woofers, so the boxes are built to suit that to my liking. What I am interested in, is if anyone knows if there is any response and impedance files out there for this woofer, or posted response charts? I'd like to simulate a crossover. I originally had crossover building parts purchased, but know more now than I knew then and have access to more software now than was even conceivable back then.

Here's some picts of the woofer...





Any help finding charts for that driver would be really great.

-----

I'm liking the way the boxes are turning out so far:





Got a new toy today:



I will never own a corded jig saw again in my life. I'm actually very surprised at the performance and perceived heft and build quality on this little jig. Hard to beat for $50 if you already have ryobi batts/chargers. The laser guides is surprisingly useful and bright.

Regards,
Eric

[chrisn]

Looks like a fun project. Any idea what brand they are? The back looks kind of like a Peerless of India, the front looks like a Peerless (not of india) CSX, at least the cone. Is the surround foam? You might need to use a WT3 or Arta (with a jig) and measure the impedance in box. Do you know which generation of silkies they are? I've heard some measured different than others, while some were relatively intechangeable. I have 3 and they all measure a little different.

I never plan on buying another jig saw either lol.

Good Luck

[mdocod]

Hi chrisn,

I don't think they were sold as any sort of "brand" when PE carried them. Yes on the foam surrounds. At the moment, I'm not really looking to buy test equipment, was hoping to see if anyone recognised the drivers and knew a source for some online charted data. I'm actually surprised I haven't been able to dig anything up. I'm sure thousands were probably sold. They were dirt cheap with easy to work with T/S characteristics, and were sold for many months, maybe even a year before they ran out.

The Silkies are as old as the buyout woofers, about a decade, so I'm guessing they are early generation. I have a spec sheet for them here (paper) and it looks very very similar to the spec sheet for the latest gen so I'm not too concerned about the tweeters.

----

On a side note, I have "completed" the speakers less painting etc for some test running.... I went ahead and used the originally planned network. The network I put together back then, was based on only knowing "calculated" component values, so the network was built to precision match what an online calculator spit out. It's a 4th order electrical with a zobel on the woofer and a series notch on the tweeter, centred around 1.6 or 1.8K IIRC [correction: 2400hz]. All the caps less the correction filter caps are dayton poly, inductors are all air.... It's an immensely overkill and stupid waste of x-over parts more than likely, (if I knew then what I know now sort of thing)...

The results are a lot better than expected. My Wife and I sat out in the garage with the rotel's knob set around noon for about an hour tonight, listening to various music. We both commented that they seem to sound "better" than the speakers in the living room by a long shot [edit in: further listening resulted in an increase in dissatisfaction]. There's no BSC, so they sound a bit thin on the bottom, however, the laid back bottom end still sounds very smooth, "poised," "effortless," etc. Setting the 2dB LF boost on the Rotel solved most of the bottom end [but still not enough after further listening].

Regards,
Eric

[chrisn]

You could alwys position them close to the walls/ceilings to compensate some for the lack of BSC. Being what looks like a poly cone, it shouldn't have a strong breakup, and you might be able to tweak it by ear. What are the crossover values and locations? Do you know the PE part number for it? They usually keep the product page up for old drivers. A search of the numbers on the driver turned up nothing useful. Peerless of India may have some info on it, if they built it. Any idea what speaker it was part of when new?

[mdocod]

You could always position them close to the walls/ceilings to compensate some for the lack of BSC. Being what looks like a poly cone, it shouldn't have a strong breakup, and you might be able to tweak it by ear.

Yep

What are the crossover values and locations?

If you plug in 1800hz [edit in correction: ~2400hz] to the 4th order calculator here: http://ccs.exl.info/calc_cr.html <mine should be very close to that. The cap values are achieved in most places with 2-3 caps in parallel to add up to the value called for by the calculator. The inductors are singles closest to the value the calculator calls for.

[trust me when I say, I am aware of the fact that this is a very outdated, wasteful, and less than ideal method to x-over design, I would not do this today]

It's possible though, that because of the impedance correction, and the fact that the x-over point that is likely far away from either drivers natural rolloff or breakups, that there could be a "lucky" result. Listening certainly suggests that the result isn't terrible, though my ears on not terribly sensitive to linear distortion.

Do you know the PE part number for it?

Seems to have been taken down from the site.

They usually keep the product page up for old drivers.

Apparently not this old

A search of the numbers on the driver turned up nothing useful.

There are some things that pop up if you do the right searches...

Peerless of India may have some info on it, if they built it. Any idea what speaker it was part of when new?

They could be from a PSB Century 300i speaker. There is an ebay seller and a "product finder" search engine that show up results for the driver. Seller claims they were made in Canada, and is selling them for about 4X what PE sold them for back during the buyout.

I may look into trying to get data from someone down this path, though I suspect the chances there highly unlikely. We'll see though.... [I have already asked the ebay seller if he has data for them, he does not]

Eric

[chrisn]

Is there anything you object to, soundwise? If you can't find the info to sim it, maybe somebody near you could measure them. If you can find the crossover originally used for this driver it may provide some insight. Is the current XO soldered together? You could mix and match parts from it.

They do look the driver PSB used.

[mdocod]

Hi Chrisn,

Thank you for sticking with me on this. As it turns out, we're on the same page for most of this. Mostly just keep in mind that I was hoping for this thread to be a sort of "feeler" to see if driver data was readily available in the community. Perhaps with exposure time, someone with driver data might notice (crossing fingers).

The x-over is not soldered, just a twisted mess on a breadboard. -the plan is exactly as you are suggesting, if I can find some driver data or have these tested, I would cannibalise the current networks, and probably be able to come up with a simulated response I like with those components. I'm betting I have enough selection in components in these networks to come up with numerous working alternatives that would be an improvement.

I'm getting the impression that no matter what I do, these speakers will never really match my taste. Granted, when listening from a certain position, to particular genres, they sound spectacular. As I listen more, I question whether an x-over fix will solve some of the limitations. The bass extension and capability isn't going to cut it as they are, and if I have to implement a sub, then I don't want the trade-offs of an MTM when a TM or TMM2.5 would be better [IMO]. Beings that these are 4 ohm drivers, I'm not aware of a way to implement them as a 2.5 way that would result in a practical load for most amplification. At this point, I'm leaning towards tinkering with the x-overs and implementing some BSC, making them look pretty, and selling.

Regards,
Eric

[awwnuts]

They are Bravox woofers from Energy speakers. That's all I know.

[chrisn]

Have you simulated the tweeter? You could always try an existing XO that uses your tweeter on a similar baffle, and tweak the woofers to match. An inductor 2mh or higher and a cap 10uf or more along with the zobel should get you something listenable. Nothing wrong with selling them, it encourages building more

[mdocod]

I spent a few hours studying response charts around the net for as many 6.5" polycone drivers with the same basket and non-vented pole design as mine, and found that most actually shared reasonable similar specifications in other ways, and were all sharing pretty similar response behaviours. I picked a middle of the road driver response in the category and used it for simulation.

The following simulation has the main gist of baffle effects and listening position and driver offsets accounted for and assumes the tweeter is at ear level with the front baffle vertical. No room interactions are accounted for. Upon further investigation of my prior x-over, this is actually targeted at 2400hz x-over, way higher than I would want, not sure what I was thinking, or not thinking back then, obviously I had a lot to learn.



It's not as bad as I thought it would look. I suspect the current listening position hurts them more than anything [edit in: nope, it was the speakers]. I'll have to bring them inside to our bedroom or something and have another listen. I'll certainly be tweaking the x-over in time, but I'm glad to see it's not as far off as I feared.

[edit in: the simulation in this post has been updated to be more accurate (reveals more error and takes into account the lower sensitivity of the woofers than originally factored), please ignore the uselessly quoted simulation below]

Eric

[thekorvers]

I spent a few hours studying response charts around the net for as many 6.5" polycone drivers with the same basket and non-vented pole design as mine, and found that most actually shared reasonable similar specifications in other ways, and were all sharing pretty similar response behaviours. I picked a middle of the road driver response in the category and used it for simulation.

The following simulation has the main gist of baffle effects and listening position and driver offsets accounted for and assumes the tweeter is at ear level with the front baffle vertical. No room interactions are accounted for. Upon further investigation of my prior x-over, this is actually targeted at 2400hz x-over, way higher than I would want, not sure what I was thinking, or not thinking back then, obviously I had a lot to learn.



It's not as bad as I thought it would look. I suspect the current listening position hurts them more than anything. I'll have to bring them inside to our bedroom or something and have another listen. I'll certainly be tweaking the x-over in time, but I'm glad to see it's not as far off as I feared.

Eric

That is an awful lot of complexity (and expense) for some garage speakers for which you do not have the driver data. Why don't you keep it simple?

.

[mdocod]

Hi thekorvers,

Why are you quoting on a direct response with a picture in the quote to make matters worse? That's poor forum behaviour FYI.

The x-overs were assembled as pictured 10 years ago. The free simulation software scene looked a little different back then, and I knew a lot less back then. I have made it clear that I would not design X-overs as pictured above several times in this thread. I don't need someone to remind me of something I have already made clear. I'd appreciate it if you read the thread before commenting. The simulation above was performed to get an idea of what I might have.

Regards,
Eric

[thekorvers]

Hi thekorvers,

Why are you quoting on a direct response with a picture in the quote to make matters worse? That's poor forum behaviour FYI.

The x-overs were assembled as pictured 10 years ago. The free simulation software scene looked a little different back then, and I knew a lot less back then. I have made it clear that I would not design X-overs as pictured above several times in this thread. I don't need someone to remind me of something I have already made clear. I'd appreciate it if you read the thread before commenting. The simulation above was performed to get an idea of what I might have.

Regards,
Eric

Thank you so much for your friendly response. In my opinion that is poor forum behavior!

And for your information I did read the complete thread before responding and you have not made anything clear.

.

[mdocod]

Hi thekorvers,

Here's an example of a good application for quoting someone... In this particular case, we'll call it "review" for someone who claims to have read the thread.

It's an immensely overkill and stupid waste of x-over parts more than likely, (if I knew then what I know now sort of thing)...

not sure what I was thinking, or not thinking back then, obviously I had a lot to learn.

The point of admitting to the crossover being "wrong" was so that nobody would have to waste a breath telling me about how wrong it is. I believe that your contribution to this thread can be summed up as quotes that people have to scroll through that are unnecessary, and salt in a wound that was already identified.

Cheers!
Eric

[chrisn]

You have a lot of parts to tweak with. Some more BSC and padding on the tweeter should get it closer. How does the tweeter sim with the notch and large coil removed? I would also try the 1.5mh inductor in place of the 1mh inductor on the woofer, and maybe a small resistor on the tweeter.

[rpb]

Have you simulated the tweeter? You could always try an existing XO that uses your tweeter on a similar baffle, and tweak the woofers to match. An inductor 2mh or higher and a cap 10uf or more along with the zobel should get you something listenable. Nothing wrong with selling them, it encourages building more

+1

I'd take a wild guess at 2 mH and 12 uf. That's usually my first trial and error measurement when making a x-over.

[chrisn]

Yep, I keep a few values and some multi-tap inductors, but most designs end up with fairly similar values. Even if its not perfect, its pretty close, usually.

[mdocod]

I'll post what I come up with when I have a chance guys. I appreciate the input. I have a plan for how to switch it up just haven't had the time to simulate the plan.

Eric

[mdocod]

Ok, I've come up with the following:



Going to go rebuild the x-overs and have a listen.

Going from 18 parts down to 10 parts. Not ultra-simplified, but better all around IMO.

[edit in]: Finished up about an hour ago and have been planted in front of them since listening to some trancy sort of stuff. The "disappearing act" is very well executed now. The concept of "painting a sound-stage" is very real with these. I'm very happy with them at this point and would feel pretty comfortable gluing down and soldering up the x-over, permanently installing it, and either enjoying or selling. I'm pretty amazed at the improvement.

Eric

[chrisn]

Nice work and reuse, glad you were able to fix them with parts on hand