Here are 3 simple modifications I made to a pair of Jamo Concert Series 93 II bookshelf speakers that give them a much more balanced and more neutral sound with better bass extension and better vocal response. I’ll do a complete teardown of the speakers, sketch out the schematic for the crossover and go through each of the modifications one by and show how each modification produces a measurable improvement in the sound. If you’re up to the task, feel free to take on this project with your own Jamos, or if you’re just curious about what’s inside these little speakers, then read on!
So I picked up my Jamo C 93 IIs on sale from Crutchfield as a open box deal and did no-rush shipping so I think I only paid a little over $220 for the pair. I’m not sure what prompted the purchase, they aren’t one of the more popular speakers on the market but for the most part people who have bought them do like them. I thought they looked really cool and also thought just maybe they would lend themselves to some simple upgrades to make them sound better. Plus, coming from someone who’s spent most of his life designing and building speakers, there is something super intriguing to me about buying commercial speakers and figuring our what makes them work. The immediacy of the process is amazing too, to go from the day of purchase to fully-built speakers on my doorstep in 4 days is something I could get used to. I mean I love spending hot weekends in my garage, breathing MDF dust just as much as the next guy, but this takes the simplicity of purchasing/owning speakers to a whole new level.
Which is why I couldn’t leave well enough alone and within the first week of getting these speakers, decided to do a complete teardown and started working out what modifications might be possible. My initial intention wasn’t really to change anything, but to just inspect what we were working with and see if there was some performance left on the table, so to speak, due to the commercial drive to mass-produce cheap speakers. Some sound elements of a speaker are design choices, not technically limited by cost, but perhaps driven by the need to give a speaker that showroom sound, a little extra bass, a touch more treble, anything to give it the edge and make the consumer say, yeah this one sounds better. But once you get it home, the added bass and treble end up being too much for day to day listening, and without the A/B comparison to any other speaker, the speaker ends up becoming is own worst enemy always trying to out-do itself. Let’s see if we can fix that.
First Modification – The Crossover
The first thing I started to dive into was the crossover. See here a complete schematic of what we are working with. I was overall impressed with the complexity of the crossover. It contains a total of 11 elements comprising iron and air-core inductors, electrolytic and poly caps and sandcast resistors. I would say the quality of the components is just acceptable at this price-point and none of my modifications made any attempt to improve the quality of the parts. That of course is still an option down the road, though it comes at significant additional cost. But if you want to purchase better parts of the same values, like resistors, the cost is minimal, move up to better caps, that will cost a little more, and better inductors would cost even more. Initially I am just going to adjust the crossover a touch without swapping out any of the major parts.
The low-pass section is a quite nice, it’s a just simple 2nd order filter but it has some decent baffle loss compensation built into it. It’s got a big 1.3 mH inductor and large shunt 33 uF electrolytic cap. Shelving starts at about 200 Hz and with the inductor alone would be about 3 dB down at 400 Hz and 6 dB down at 800 Hz. Introducing the 33 uF cap brings up the response at 800 Hz by about 3 dB and adds a corner to the response at about 1300 Hz, which then transitions to the overall filter to a 12 dB/octave slope at this point. The series RLC notch filter (8.2 ohms + 1.2 mH + 22 uF) is centered at about 1000 Hz and provides a 4 dB notch that’s just a few hundred Hz wide. This cleans up some peaking in the driver response and also helps out with the final baffle step compensation network. This puts the crossover frequency to the tweeter at right around 1.9 kHz, which is a good spot for most woofers, considering it’s about 6″ in diameter and the tweeter is a full 1″ soft dome. Pushing the crossover lower can put stresses on the tweeter and can make vocals sound nasally while pushing it higher can lead to poor directivity and more interference from cone break-up modes. Jamo does not provide a specified crossover frequency in their cutsheet nor do they mention the filter slopes.
I played around a little bit with just the low-pass section, adjusting the notch filter and changing the shunt cap value, but after all my fiddling I decided that there wasn’t anything drastically wrong with where the response was sitting and decided just to leave it alone. I think Jamo did a decent job here getting the baffle step right, setting the crossover point just low enough, the notch filter prevents some peaking around 1 kHz and enables the woofer to blends nicely to the tweeter without any massive suckouts. The final response ends up being more 3rd order acoustic once you add in the natural roll-off of the driver. So I was happy with not changing anything here. Maybe a future upgrade might be swap the 22 uF and 33 uF electrolytics for some poly caps just because they’re pretty cheap and make for a better-looking crossover and you never know, maybe to some people it will even sound better. You could also swap the iron-core inductor for a lower-gauge air core.
So let’s break down the tweeter section: we’ve got a single 1.5 ohm series sandcast resistor feeding a 7.5 uF poly cap and 0.18 mH air core inductor creating a nice 2nd order filter. This is almost textbook, but with a smaller inductor value which increases the Q and creates a sharper roll-off with a tighter corner at the cut-off frequency. Pretty standard adjustment that I tend to do as well. So the only red flag I see at this point with the tweeter is that itty, bitty, teeny, weeny 1.5 ohm resistor. You know that soft dome tweeter is more efficient that the woofer, we know the woofer has nearly 4 dB of baffle step compensation, so there’s no way a single 1.5 ohm resistor is going to provide enough attenuation to level match the woofer and the tweeter. That tweeter is going to be bright! And that is clearly by design, because it’s so easy to make that resistor anything the designer wants at zero cost. Maybe the guy hand-tuning the final design at Jamo was in a heavily damped room, maybe the speakers were tuned off-axis, maybe they really wanted to compete in the showroom by giving that upper end some sparkle. I mean when you’re up against the likes of Klipsch and B&W at Best Buy, seriously, these won’t win an A/B standoff with your average listener if the treble is deadlocked with the mids and bass. Even if that’s where it should be. Or maybe that’s just the Jamo signature. As you can see here I adjusted everything from 0 ohms to 8 ohms which provides a 15 dB spread in tweeter level and eventually settled on 3.3 ohms as the ideal spot from both a measurement perspective and from my own listening even though the treble is still just above flat. I like a touch more brightness over flat if I can choose it. So let’s double the 1.5 ohm series attenuating resistor to 3.3 ohms and move onto the last crossover tweak.
The final part of the tweeter crossover is another series RLC notch filter similar to the woofer. This one is comprised of a 3.3 ohm resistor, 0.32 mH inductor and 2.7 uF poly cap. This provides a basic 4 dB notch or so at around 5.3 kHz that’s about a 2 kHz wide. See plots here that show the tweeter-only response with and without this notch filter. Again Jamo did a good job recognizing that the tweeter response was peaking badly in this region and bringing the response down was absolutely necessary to getting towards a flatter response. Only now with the tweeter adjusted down a touch with the new 3.3 ohm resistor, this section now looked a little depressed, the notch was too deep, so I increased the 3.3 ohm resistor to 6.0 ohms to bring it back up a little. And that did the trick. I tried making more aggressive adjustments to the cap values and resistor values but nothing looked as good as this so that’s where I left it called the crossover tweaks done. The tweeter still has this rising response above 10 kHz but realistically it’s probably got the perfect slope to give a touch of airiness to most music so I made no attempt to try and flatten anything above 10 kHz and just accepted it.
So ultimately what we end up with instead of a continuously rising treble response that starts at ~2 kHz and increases by almost 8 dB as we approach 20 kHz, we end up a nearly-flat response from 200 Hz to 10 kHz and then have only a shallow +3 dB rise up to 20 kHz. For me this is perfect. If you can get this region of a speaker to be ±1.5 dB then you’ve got something worth listening to. A little extra above 10 kHz, I mean, no one is going to mind. Especially if it’s only 3 dB. And the best part is this mod actually only requires you to buy a pair of 6.0 ohm resistors because the 3.3 ohm resistor used in the RLC network can be reused as the 3.3 ohm resistor in attenuation circuit. I almost convinced myself that we needed a solid 4 ohms in that location, but after a few long listens, I was happier with the 3.3 ohm series resistor which in the end just made the mod even easier. See plots here for more details. And that’s the crossover modification. Worth every penny. And seriously, this mod literally costs pennies. A 6.0 ohm 10W resistor from Parts Express costs $1.49.
Second Modification – More Acoustic Damping
So this is where honestly I don’t understand why manufacturers skimp because it can make such a huge difference in sound quality. I guess it’s a cost thing, maybe it’s a manufacturing thing, because it clearly looks like this part of the assembly is done by hand. The random splattering of hot glue holding in 4 tiny pieces of poly-fil and that’s it. But nothing sounds worse than a poorly damped enclosure sounding like a big hollow box. Vocals sound muddied and instruments carry on a tonality that’s just off. Cabinet resonances probably make up the bulk of why small bookshelf speakers just sound bad. The enclosure is made from only 1/2 inch MDF instead of the typical 3/4 inch which would have been better. The front baffle is about 1.5 inches, which is quite nice actually. But those 1/2 inch side panels, top, bottom and back just are not doing this speaker any favors. There is one small window brace in the center of the box which thank goodness helps stiffen the side/top/bottom panels a little bit. I debated attempting to add additional bracing, like dowel across the side panels, but in the end just decided to add more acoustic damping material. This was easier and in the end provided a good enough result that I was happy enough with it. Bracing techniques really should be built into the design at the beginning when they can be done properly. Slapping a fat dowel between the side panels might have worked to some degree, if I didn’t overcut it, but that would have complicated the acoustic damping installation with such a tightly cramped space, and I figured the boxes were small enough as it was, the one window brace was going to have to do.
The damping fix I came up with for these enclosures was to add multiple layers of 100% cotton batting available from Joanns/Michaels. This is not 100% polyester, this is 100% cotton batting, it’s slightly more expensive per pound, but has better damping properties in my opinion than poly as it is slightly more dense and the cotton fibers just have a better feel to it than the polyester. I bought a bag that contained a single piece that was 45″ x 60″. This I was able to cut into (12) 3-1/2 inch by 60 inch pieces. I folded each piece to produce (6) double-layer pieces for each speaker. So each speaker ended up with (2) 6-layer thick strips that fit into the front and back halves of the enclosure. This added roughly 1″ of cotton damping to the top, bottom and sides of the enclosure. Then I added back in the original 4 strips of poly-fil that came with the speaker. Now this thing was packed! I took this opportunity to stick a mic into the enclosure and take some measurements with no damping, the original 4 pieces of poly-fil and the final configuration of 6 layers of cotton batting and the poly-fil and you can see that the combined effect of both the poly and the cotton batting provides the best acoustic attenuation in the enclosure. Because ultimately that is all we are doing, is providing some level of acoustic attenuation inside the cabinet which in turn will help prevent those thin cabinet walls from resonating as much and becoming audible to the listener. The additional stuffing also helps prevent reflections in the cabinet from making their way back out the woofer and will help with pipe resonances of the port. This much stuffing does come at a small cost though to what I’d call upper mid-bass presence. Bass will be deeper and sound smoother, but will also be less pronounced and softer. Totally worthwhile the trade though once you catch your first listen of a male vocal with a deep baritone. It will sound warm and inviting, effortless and less honky. So note to manufacturers, seriously, more stuffing, less poly, more cotton. With the crossovers installed I also added another 4 oz. or so of regular poly-fil over the entire crossover and back of the enclosure just so there was effectively no bare wood showing anywhere inside the cabinet.
Third Modification – Port Tuning
I measured the T/S params of the woofer using REW so I could model the driver in the enclosure to see if the box volume and tuning were appropriate for this driver. Unfortunately the box is a touch small sitting at only 6.5 liters (0.23 cu.ft.) when the optimum volume should have been more like 11 liters. This is actually typical for most small drivers and even a lot of speakers I build I can’t afford to make the box as big so as to provide a standard 4th order design. The tuning frequency is right at 72 Hz which puts a 2 dB peak in the response at about 106 Hz. The -3 dB point is about 68 Hz. Excursion is kept in check above 60 Hz and about 50 watts. So while there’s nothing I can do about the volume apart from building completely new enclosures, and realistically the enclosure volume is perfectly adequate for a driver this size, there is something we can do with the tuning frequency. The port is about 4 inches long and does not quite extend all the way to the front of the cabinet, which meant there was some room to tune this box lower with a longer port. I opted to extend the port a full 2 inches which brought the end of the port to just about 2.5 inches from the front baffle/tweeter. General rule is to make sure the port has at least its diameter of breathing room in front to not impact air flow or tuning. This brought the tuning frequency down from 72 Hz to 56 Hz. This completely eliminated the 106 Hz peak and improved the extension below 60 Hz to more than 3 dB. While the modeled f3 didn’t technically change (due to less bass above 70 Hz but more bass below it) the overall response is flatter and has more extension, allowing the driver to play lower and louder than before.
Check out some of the pics for how I did this modification. Basically I bought a 2 foot piece of 2″ black PVC pipe from The Home Depot and cut it on my miter saw to larger than 2″ to give me some pipe to work with while I chamfered the outside edge of the port to match the inside flare of Jamo port. I used a 1/4″ round-over bit on my router and my angle grinder and did the best I could to get them to match up and then I cut if off at 2 inches. I then chamfered the inside of the port end to provide some flare, though it’s small, it takes off that hard square edge of the port end. Then I just hot-glued it in place right on the end of the existing port. It worked out just fine and only left a very small notch on the interior where the two ports meet up but causes no significant audible additional port turbulence. As it stands port air speed is just about 25 m/s (which is just under the general rule of 8% the speed of sound) so port noise is kept at a minimum. Plus since the port is rear-firing that further prevents any audible port noises or pipe resonances from making their way into the room at any significant volume. Check out the impedance plots of the before and after below. And they say this is a 6-ohm nominal speaker, it’s not, it easily falls into the nominal 4-ohm category. Even dropping to 3.6 ohms at about 200 Hz. Still, should be no problem to drive for any modern amplifier.
Conclusion and Parting Thoughts
So with all three modifications complete, I got a chance to measure the speakers with REW as well as give them a good listen. Up until now I had been taking measurements and sneaking listens of bits of pieces of the changes, but this was when I actually got to sit down and listen to the speakers with all 3 mods completed, the drivers fully installed, the decorative rings re-attached, the crossover fully put back together and not hanging out the back of the enclosure. So let me just say that these Jamo Concert 93 IIs absolutely sound like completely different speakers. And I don’t mean different just to be different, they sound better, ten times better. I’m not playing around here, I could not believe what I was hearing upon my first real listen after having just listened to these speakers in their original form not a week earlier. Now granted a week is a long time between A/B’ing any gear so any memory I had of these speakers had surely mostly faded, but I do remember what I thought the first listen straight out the box, “Hmm, I guess this is why I build my own speakers”. I mean they sounded fine, but they also sounded like every other commercial-thrown-together speaker you can get at this price point. But now, oh wow, now all of those pitfalls, the boxy, shallow, upper bass, the overextended treble, the lack of really deep bass, the muddied vocals, it was gone. It was all gone. The speakers sounded smooth, the bass was deep and rich, vocals sounded natural and clear, the treble was balanced with the mids, the overall tonality was exceptional, neutral and balanced. Upon my first listen in contrast to my original impressions, I smiled and said, “Ah, now that’s what a speaker’s supposed to sound like!” Call it confirmation bias, call it what you will, the measurements below give some indication of the improved frequency response, but they don’t really show the whole story, how much better the bass sounds, how much better the vocals sound and just the general way these speakers effortlessly convey music into the room. Definitely recommend.
So these three updates will set you back about $18 and a single weekend to complete. Between the 6.0 ohm resistors ($3) the cotton batting ($7) and the 2″ PVC pipe ($8) you will need a bunch of tools and a lot of patience getting the drivers out of the enclosure. I’ll explain it in the pictures below but you have to be super careful taking off the little decorative rings around the woofer and tweeter, after that, it’s a walk in the park. You need minimal soldering skills to solder the two resistors, some decent hot-gluing skills to hot glue in all the cotton batting in place and the port, but otherwise this is a super straightforward modification. I’m guessing this voids any kind of warranty on the speakers, so if you’re the type who likes it loud and figures the manufacture is responsible for replacing your blown drivers, then maybe this mod isn’t for you. Sure the drivers can suffer other damage that might be related to a manufacturing defect that maybe would have been covered otherwise, but I guess I’ll cross that bridge when I get to it. For now I am extremely happy with the Jamo Concert 93 IIs with these three modifications and can recommend trying them out if you’re feeling courageous enough or if you’ve felt like these speakers are great but could be just a little bit better, then I say go for it! And let me know what you think of this speaker update in the comments below. Thanks for reading, see ya!