Cerwin Vega L-7 Bookshelf Speaker Crossover-Enclosure-Tweeter Upgrade!

Last month I was at my parents house when my dad says, “Hey do you want a pair of speakers?” I thought heck yeah, I’m not one to turn down a free pair of speakers. So he shows me a near perfect-condition pair of vintage Cerwin Vega L-7 bookshelf speakers just sitting on the floor of his bedroom. He said he hasn’t used them in years, they weren’t even connected to anything, and were just sitting around collecting dust. So I absolutely said I would give them a new home and took them off his hands. Almost 30 years ago I had borrowed these same speakers my sophomore year of college and seeing these speakers again instantly brought back memories of blasting Cake and Depeche Mode for all my neighbors to hear. These were decent speakers, even though they don’t have the traditional Cerwin Vega look. Without the badge on the grill, you’d never know who made them. But I always thought they were a good speaker. And now we get a chance to see what this speaker is made of and, of course, see what modifications we can perform to improve their sound and make them even better.

Speaker Breakdown

An external visual of the speaker reveals a 7″ paper cone woofer with rubber surround and a 1/2″ mylar dome tweeter. The cabinet is ported in the rear with a spring-loaded terminal cup. The cabinet is wrapped in a black wood grain veneer, typical of just about any inexpensive speaker available today. The L-7 is fairly lightweight and tapping on the cabinet sides and top reveal a very hollow “knock” indicating little to no internal bracing or damping. Opening up the speaker is quite easy (there are no decorative rings trying to dress up the stamped steal basket of the woofer) which reveals a hollow box with no bracing and basically zero actual acoustic damping, except for a single piece of what looks like a 3/8″ wool fiber pad that had completely dried out and was as stiff as cardboard. At this point we can see that the enclosure is made from 5/8″ particle board and not the more traditional MDF. The crossover is as bare as they come with only 4 components – a single 0.5 mH iron core inductor on the woofer (1st order), a 2.2 uF cap and inductor on the tweeter (2nd order) and a protection device to prevent damage to the tweeter in case of overload. This is extremely bare bones and definitely designed to hit a low price point.

An inspection of the woofer reveals a pretty decent design, although nothing fancy, it appears solid, beefy size magnet, it’s got a cheap stamped steel frame, but the paper cone, spider and surround all look really good. I read somewhere that the drivers were actually made in Germany and not by Cerwin Vega. There is certainly potential here. But the tweeters on the other hand are pretty much garbage. They were very common in cheap speakers back in the day and you can still buy a version of this 1/2″ mylar dome tweeter from Parts Express for about $4. Overall it’s not a great tweeter and will probably be the first thing to go in the upgrade. I measured the speaker with my ECM8000 mic and REW on axis at 0.5m in my room to get an idea of the speaker’s overall tonality and the plots below give you an idea of what we are working with. Cone break-up modes dominate the midrange region right around the crossover point. The sorry tweeter just can’t play low enough to push the crossover point any lower which results in a crossover point of about 5.3 kHz. The tweeter response also can’t keep up with the volume output of the woofer, so it rolls off leaving the midrange to sound even more forward. Overall just not great and clearly designed to be inexpensive and not designed to sound as good as it could. When I was looking up this speaker online I did find that there are two versions of the L-7 with the other version using what’s look like a much better tweeter. Probably a good choice, since besides the hollow particle board box, the tweeter is definitely a major limiting factor to making this a great speaker.

Cabinet Upgrades

So let’s start talking about what we can do to improve this little Cerwin Vega L-7 bookshelf speaker, without completely redesigning it, by keeping what’s good and ditching everything else. The first four things that have to go are the spring-loaded terminal cups, the crossovers, the crappy dried up wool pad and those terrible tweeters. What we’re left with is a ported particle board box and a nice paper-cone 7″ woofer. So the first thing I tackled was the cabinets. With everything removed, we basically just had a hollow box with lots of options for bracing. The tricky thing with bracing an enclosure that is already built, is that is not traditionally how bracing is incorporated into an enclosure. It’s usually built into the enclosure at the same time as the construction of the enclosure, which allows for solid pieces of bracing to extend side to side and front to back with no breaks in material. Since we have a single 6″ circular opening in the cabinet, all the bracing has to be able to fit through that opening. What I came up with was a basic window brace for the weakest part of the cabinet, the baffle just between the woofer and tweeter. To do this I cut 7 pieces of 3/4″ x 1″ MDF into lengths that perfectly fit the cabinet side panels front to back, a middle brace going from side to side and another middle brace between each front to back. Basically a normal window brace with 7 separate pieces all measured and cut to fit snuggly in place and then glued (and clamped if possible) with liberal amounts of wood glue. Normally if I were to design a window brace for a new speaker, it would be cut from a single piece of material, like you would see any typical window-style brace. In the end this should be just as effective though, even if it is more tedious to assemble and requires precise cuts to get everything to fit tightly. Some cuts I had made too small, so the piece would just fall out, in which case I had to scrap it and try again.

Next I added a small block of 1/2″ MDF to the top panel, about half the area of the top panel and ran another brace from the top to the center window brace. This thickens and stiffens the top panel of the cabinet and ties it to the window brace which is then cross-braced to the front and back and each side of the cabinet. The next brace I added was just below the terminal cup hole on the rear panel. This stiffens the back panel about a 1/3 of the way up. The next pair of braces I added was to each side about the height of the middle of the woofer opening. Since the woofer sits about 3.5″ deep into the cabinet, we can’t do a full window brace here, so I was left with just a cross-brace tying the two sides to each other, along with the bracings running front to back to stiffen the side panels. A final brace connects this cross brace to the window brace just above it, thus connecting all the bracing to each other and subsequently all of the panels to which they are attached. The bottom of the cabinet I added a pair of smaller 2″ x 6″ by 1/2″ blocks of MDF just to stiffen the bottom panel, but did not run any cross bracing since this location has the port, and was going to be the final location for the crossover. So I needed that area free and clear without bracing. At this point we had stiffened up this little cabinet tremendously. Knocking on each panel revealed a nice solid “thud” completely absent of that hollow sound it had before. All 6 sides were now braced (or made double thick) and tied side to side and front to back, or some combination of both. Additionally the braces were tied to themselves wherever possible to ensure the highest rigidity possible. We lost some total cabinet volume, which means we we’re going to lose some low-end extension, but the overall tonality of the speaker in the bass and lower-bass to midrange region should improve significantly.

Now that the box is properly braced, it’s still full of hard surfaces that reflect sound, which means we need to dampen the enclosure with a nice, soft, sound absorbing material. For this I used an 80/20 mix of cotton/poly batting which you can buy from either Walmart or JoAnnes. I folded the batting over itself to create 4 layers and then cut strips at 3 and 4 inches wide to fit into the enclosure along the sides and back in each of the subsections created by the bracing. I had some pink fiberglass insulation left over from one of my sub builds that I also stuffed into the upper section of the cabinet. I also added about 1/2 ounce of 100% polyfil to fill that top chamber above the window brace. Lastly I covered all of the braces with a single layer of the cotton/poly batting to soften any hard edges that remained. All of the batting was glued into place using 3M spray glue. See the pictures below for a view of how this looks. This method completely damped 5 of the 6 panels leaving only the bottom open/exposed for the crossover. Which is also why I added the two 1/2″ MDF blocks to the bottom, to stiffen it up, since I wouldn’t be adding any additional damping material in this location.

Picking Out a New Tweeter

For the new tweeters I set some design constraints for myself, just to limit the total number of options from over a hundred to just a dozen or so. Cost was one of the primary factors and so I opted for about a $20 upper limit here. Next I liked the look of a smaller-diameter tweeter with that big 7″ woofer, plus didn’t want to cut a new hole in the cabinet. There are lots of tweeters that have a large 104mm face that could have worked with a quick jigsaw but I sort of just wanted something that was a drop-in fit. I know it seems like a lame constraint, but even as such, that left a good number of small-diameter tweeters to choose from. The tweeter opening was about 2″ in diameter with the screw holes being about 2.5″ apart at the widest, so I was looking for a tweeter that required a hole that was smaller than 2″ but had a face that was larger than 2.5″. I didn’t want the original 4 screw holes showing through as that would have looked dumb. I also figured a regular soft dome tweeter would be ideal as they tend to be cheaper and have a nice workable frequency response. There are about 4 Dayton Audio tweeters that fit this criteria and about 4 Peerless tweeters. I checked out the FR plots of all of these tweeters in hopes of finding a standout and this is basically how I laded on the Peerless DX20BF00-04 3/4″ Silk Dome Tweeter. It has a low fs, a semi-horn loaded design for improved efficiency, plus the smaller 3/4″ dome offers a ruler-flat frequency response all the way to 20 kHz and beyond. Not to mention I just thought it looked super cool. With an outside diameter of 2.6 inches, it was just large enough to hide the 4 screw holes from the original tweeter. I also liked the fact that it used only 3 screws, so there was no concern about whether or not the holes would just barely-sort-of-but-not-really line up with the existing screw holes causing more problems. You be the judge but from a pure visual perspective, I think the tweeter turned out great and totally matches the look and feel of the L-7 speaker in keeping in tune with the original aesthetic.

Designing the Crossover

So with the cabinet fully upgraded, braced and damped, and the tweeter all picked out, I bought a bunch of crossover parts (caps, inductors and resistors) to provide options for designing and fine-tuning the crossover. I had some parts on-hand and had already started testing different options, and with the help of Passive Crossover Designer, I had a basic idea of what value of parts I would need, but crossover parts still ended up being a fairly large portion of the budget. With the speakers installed into the cabinet, I ran a pair of wires for the tweeter and the woofer out the port where I could test and tweak the crossover to my hearts content. And this is where the real fun began. After over hundreds of measurements I came up with 3 basic crossover options and then after listening to a bunch of music with each one, I settled on the one that I liked that sounded best to my ears, even though, oddly enough, its on-axis frequency response was arguably not as flat as the other two. Where this final crossover really shined was in its off-axis response, both vertical and horizontal, proving that on-axis response alone does not mean everything. Ultimately good tonality comes from a balanced design that looks good both on-axis and off-axis as opposed to favoring one over the other. In fact, you could argue that off-axis response is more important than just a straight on-axis response and indicates more about how a speaker will really sound in your room. As you’ll see in the plots of the final crossover, off-axis response horizontally is excellent, and even the vertical response is quite good, considering this is a basic vertical 2-way design and not a coaxial one.

I started with the woofer and measured various 2nd order options ranging from a 0.5 mH inductor and 5.2 uF to 15 uF caps all the way up to 1.8 mH inductor paired with 5.2 uF to 15 uF caps. Each of these options provides a slightly different roll-off pushing around the crossover frequency and changing the overall presence in the midrange. The woofer actually has a really nice natural roll-off just above the cone break-up modes, and with just the right values, you can bring the break-up modes down to be flat to the lower midrange of the driver and then roll off sharply where the crossover corner lines up with the woofer’s natural corner. This is how you can start with an electrical 2nd order filter but end up with a 4th or even 5th order acoustic response. Not all woofers play this nicely though. But this woofer with its natural paper cone fibers does an excellent job of staying well damped by controlling those break-up modes and then rolling off nicely. So from a purely on-axis response perspective, you can cross this woofer anywhere from 2 kHz as high as 5 kHz and it won’t sound terrible. Directivity will begin to suffer however as you go higher and higher in frequency so the lower we can stay under 5 kHz the better we will be in that regard. Note the original crossover frequency of this speaker was way up at 5.6 kHz, which was mainly driven by the cheap mylar tweeter. No 7″ woofer should ever be required to play that high. With the new peerless 3/4″ dome tweeter we are able to push the crossover frequency down to around 3.4 kHz where it blends quite nicely with the woofer. Still a bit high for a 7″, but for this design, it works.

I had played with one option that pushed the tweeter even lower, which made the response flatter, but this is where some of my listening tests ruled that out. This tweeter despite having a low fs, isn’t really designed to play well below about 3 kHz, in my opinion. Above 3 kHz is where this tweeter really shines. Trying to get it to produce some of that 2-3 kHz response, taking the load off the woofer, just sounded bad. However, the woofer plays quite nicely in that 2-3kHz range, especially with the newly braced and damped cabinet, we get some really nice vocals that sound rich and full. Which is just one of the areas that you don’t really get to see or hear in frequency response plots alone. My recommendation is to always fine tune a crossover to music, especially vocal-heavy music, because your ears will tell you when you’ve got a nice, natural and smooth sounding vocal response, that just disappears into the speaker, as apposed to a honky, nasally screech beating your ear drums to death.

All said with the woofer we end up with a very simple 1.5 mH inductor paired with a 7.5 uF capacitor. This gives us just a little bit of baffle step compensation, I’d estimate it at about 3 dB and puts the 6 dB down point at around 3.4 kHz. Leaving out any kind of series resistor inline with the parallel cap keeps the filter corner peak nice and high which sums nicely with the woofer’s natural response creating a very clean roll-off overall. This filter corner and slope is primarily responsible for the subtle dip in the overall frequency response between 2 and 4 kHz. You can bring this up by going with a smaller inductor and or smaller capacitor, but I actually found the subtler midrange response very easy to listen to, especially with vocals, sitting them just a bit softer in the mix and not screaming at you. This is easily a preference kind of thing, but seriously, bringing up that midrange to sit actually flat, sounded harsher, more gritty and not as mellow as I liked, so I left it more subdued. Also, don’t try and push the tweeter down to pick up this region, it will sound worse.

Which brings us to the tweeter section of the crossover. This tweeter had me stumped for a long time. I just could not get it to cross with the woofer anywhere and be happy at first. I pushed the woofer up and the tweeter up and it sounded bad, I pushed the woofer down and tweeter down and it sounded even worse. Plus nothing looked flat through the crossover region. The drivers just did not want to sum. In phase, out of phase, off-axis, it was just a lumpy mess. Worse than that, nothing sounded good. That’s when I found the sweet spot, not too high and not too low, where I finally started to get something that sounded great, but still didn’t look great in the plots. I ended up with a crossover point of 3.4 kHz for the tweeter with a 3rd order slope. This was comprised of a 0.5 ohm resistor, a 6.2 uF cap, 0.15 mH inductor and a 15 uF cap in a standard 18 dB/octave topology. This made the response above 5 kHz just ruler flat all the way to 20 kHz, but left that slight depression in the at 2-4 kHz range, like I talked about above. Maybe I just hate midrange, but having the response come back up right around 5 kHz just worked. The tweeter was happy, the woofer was happy, and more importantly my ears were happy!

At first glance it might just look like both the woofer and tweeter just need to move another 500 Hz towards each other and that dip would go away and you’d be right, but I promise it will not sound as neutral and as clean, and as flat, as it actually sounds. Also, once you go off axis either up or down or left or right, the response flattens out much more so in effect it is much more subtle that it appears in the one direct on-axis plot. The dip is really only present dead on axis and exactly between the woofer and tweeter. I flipped the polarity of the tweeter initially thinking it needed to be wired out of phase, but that made the response worse, especially in all of the off-axis measurements, so I knew it wasn’t a polarity issue. The flattest on-axis response is with the mic placed about 2″ below center, or just about the edge of the surround of the woofer, where we see no suppression at all between 2 and 4 kHz with just a mild drop at about 4.5 kHz. This is probably more or less the actual acoustic center of the speaker as apposed to being between the frame edges of the woofer and tweeter. Since this spot is centered between the center of the woofer and the center of the tweeter thus ignoring their actual dimensions.

In the end, all of these measurements only represent the sound of the speaker at a single point in free space. It’s really meant to just provide a visual indication that the crossover is doing what it should, and that each driver is behaving in a predictable manner. And that the basic FR shape which defines the tonality of the speaker appears to make sense. There are some other options for the tweeter that you can tune to taste with the series resistor. I chose the 0.5 ohm as to me it sounded the best, it left the speaker a touch brighter than having it be completely flat because I sort of wanted it to have what I would consider a classic Cerwin Vega sound, and so I left them arguably bright. You can dial back the brightness of the tweeter without changing anything in the crossover and it will not change the overall frequency response at the crossover, but it will just adjust the volume above about 4 kHz. Want it brighter? Lose the 0.5 ohm all together! Want it softer, double or triple it to 1 or 1.5 ohms. Even going to 2 ohms is reasonable, although I imagine the speaker is going to start to sound lifeless and boring. And that’s definitely not the Cerwin Vega style. So you don’t want to get too crazy. But with these final values shown here, I went ahead and soldered up the crossovers onto two separate little boards and installed them into the cabinet. I ended up just hot-gluing them down into the cabinet as that seemed simple enough and didn’t require me trying to figure out how to fit a screwdriver in there to screw them down, as that would have been the only other option.

Final Thoughts

So there you have it, the complete breakdown of the Cerwin Vega L-7 bookshelf speaker and 4 improvements we were able to make to it that absolutely are guaranteed to make this speaker sound like a million bucks: new 5-way binding posts, a fully damped and braced cabinet, a new crossover and a new soft dome tweeter. Take a look at all the pictures for a visual tour of this project. I had a lot of fun playing with these speakers, measuring them, bracing them and just getting to listen to their sound as we shaped it into something that sounds simply beautiful. Not to mention once I got both speakers done, I could actually listen to music in stereo and get a real feel for what kind of stereo imaging we get and how well the speakers fill the space. I listened to a bunch of different songs and just sat back and smiled ear to ear as everything I threw at them sounded great. And just like my Dad let me borrow them in their stock form some 30 years ago, I have the privilege of giving them to my son as his first pair of speakers where 3 generations of Marx’s will have been able to enjoy these little beauties. We’re going to set them up in his room this weekend as part of his first real stereo and will be connected to a nice older Denon receiver with a Pioneer tape deck and a blue-tooth receiver. He is super excited. Also because though I didn’t mention it, he helped build and test almost every part of this speaker upgrade with me. He hung out with in the garage while I worked on them and sat and listened to every Bwoooooop! of REW. He was very interested in the whole process. Even though he didn’t know I was planning on giving them to him. He was so excited when he found out they were going to be his speakers. I’m sure he will get some good years out them still. Love you Ben. You will always be my little music buddy.

Parts List

Peerless DX20BF00-04 3/4″ Silk Dome Tweeter 4 Ohm
5-way Binding Post Terminal Cup
1.5 mH Iron Core Inductor
7.5 uF Poly Capacitor
0.5 ohm Audio Grade Precision Resistor
6.2 uF Poly Capacitor
0.15 mH Air core inductor
15 uF Poly Capacitor
3/4″ x 1″ MDF blocks
1/2″ MDF blocks
80/20 Cotton/Poly batting
3M spray glue
Various Lengths 16 AWG wire


Build Pics

Final Completed Shots

About Dan

For a complete biography, stop by www.danmarx.org and click around for a while.
This entry was posted in DIY Audio and tagged , , , , , , , , , , , , , , , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *