Review: JM Labs Diva Utopia Be Speaker

Category: Speakers

I am an avid 2-channel audiophile for 32 years and counting. I’ve been modifying my own gear for almost 10 years now, having started with the usual tricks of re-cap, Schottky, choke, resistors, and progressed on to fixing the flaws of Chinese amps, and recently moved on to building my own amps. I recently took on modifying and building amps for fellow audiophiles. I’ve had little experience with speakers and crossovers, none other than recently upgrading the crossover for my Quad 989s and a supertweeter for my Diatone 610s.

“What? You buying a pair of home-theater speakers?” a friend comments as I made my decision to pickup a pair of JM Lab Diva Utopia speakers.
“Love the beryllium tweeters. The size is right.” I said.
“We heard them at the dealer before, don’t you remember?” the friend exclaimed. “No decay, wooden bass and dead sound. You are a sucker for punishment!”

Currently I own Quad 989s and Avalon Eidolons (the latter are my perfect reference but
both pairs have a tendency to overwhelm my room). The Eidolons are current on loan (boy has it been a long while) to a fellow modder friend.

12 years ago I almost bought a pair of Mezzo Utopias. Too bright, I passed.
Then a few years ago while helping a friend pick up new speakers we auditioned the Micro, Diva, and Alto. The Diva was actually the worst of all three, it had a wooden and dead sound. The friend ended up buying the Altos (priciest, and as expected, it’s the best of the 3).

The Alto Utopia has a deep but a bit flabby bass, but with the right upstream gear it has transparency that rivals the best of electrostatics. Very transparent, but even more so very critical of upstream components. Anyway, I kept thinking there has to be a way the Diva Utopia can sound good. The form factor is what I want in my listening area.

Recently I had the opportunity to re-cap (more like resolder instead of recommend and install) and replace resistors (with Mills NIWW of the right rating) on a pair of Micro Utopias that distorts and saturates when stressed. The owner had gotten 4 V-Cap OIMP 6.8uFs to be installed in the HF portion of the crossover. “You are taking a huge leap of faith here.” I said, “the tweeters are voiced with the stock caps and as far as I remember the sound is not bad. You are betting big money buying these expensive caps.”
“The V-Caps TFTFs worked well for me on my heavily modded Jolida.” He said.
“Well I know they’re good. I’m using V-caps in my Altavista NP220. But these OIMPs are completely different. These are lower voltage OIL caps.”

Few hours on the bench with a soldering iron and the end result was stunning. I was floored and the owner and I were ecstatic when we listened to the speakers with the resoldered crossover. The Micros successfully pulled the trickery of convincing us they are not mini-monitors. The sound stage was huge, the dynamics awesome (still a boom due to its slotted design interacting with my room). But talk about detail, decay, and dynamic headroom of anything above 100 Hz. They are so stunning it revived my interest in the Diva Utopias enough to look for a pair for myself.

“Same configuration, same crossover point for mid and tweeter.” I said to myself, “they can’t be bad.”

A few weeks later I found a pair on Audiogon and the owner was local. Got a deal and the owner told me he was driving them with a BAT VK75SE. He said he didn’t get them to sound right until he borrowed another VK75SE to drive dual mono.

“Wait. This can’t be. Aren’t they supposed to be 90 dB/W speakers?!?!”
Anyway, transaction done, and I hauled them home.

Excerpting from on AA forum thread: “There is a 5-6 dB suckout between 80 and 230 Hz …lacks the soundstaging magic of the 907's … voiced a little hotter on the top …soundstage sounds narrower and flatter than the 907's…”

Sadly, other than the “little hotter on the top” this writer’s right…Major bass suckout. But I think the top is not hot. It’s hashy. Lack’s air compared to other Utopia’s. Nothing wrong with this particular pair. They’re pretty much the same as the pair I heard at the dealers.

Within an hour’s listening. I pulled out the Torx wrench from my toobox and popped off the back plate where the binding posts are to see the crossover. “2 crossover boards? No. one is a crossover board, the other is the Zobel circuit to stabilize the impedance of the 2 woofers.”

Taking a closer look at the main crossover board. I realized why the Diva’s don’t sound even as good as the Micros. Focal JMLab committed the SAME CRIME Quad commits: Using a 180uF bipolar lytic as a coupler for the midrange driver. This lytic is supposed to cutoff low frequency to let the woofers takeover. But as a coupler it is now filtering anything between 100Hz (yea right!) and 2500Hz. For the tweeter there’s a Solen/AXON 3.3uF 250V cap that’s known for its raspiness. This can’t be right. Aren’t these supposed to be high end speakers? I am almost inclined to call up my friend who has the stock Altos and open them up to check on the caps.

Based on TJN’s measurements. The bass suckout can be explained by the LF cutoff being tuned too high. OK, since I am doing away with the lytic I might as well get a bigger cap to lower the cutoff point. Does anyone make a 220uF film cap? It’s Solen of course. And heck I have to stick with the raspy 250V ones. Well, at least they are much much better than lytics based on my experience fixing my Quad 989s. The 220uF should lower the 3dB drop point of the midrange a bit to narrow the midbass suckout.

For the tweeter, I do have a pair of Jantzen Audio Superior caps of the same value. I popped out the AXON 3.3uF for the tweeter and replaced them with the Jantzens. Sound stage opened up. Less of the hashiness. But a sense of dryness came with the Jantzens. It’s a little bit like that of Infinicaps used in high voltage bypass in amps. The sense of clarity and openness comes at a price. Now I hear what some people would refer to as Beryllium ringing. Tried to band-aid that with Russian K40 PIOs. The mid high became very colored and resolution actually suffered.

The mids and highs are definitely not good enough compared to the modified Micros. Time to bit the bullet and get the same caps in the modified Micro. Got a pair of OIMP 3.3uFs and Chris was nice enough to match them for me.

Took me no time installing them. The leads on the input side remain a bit long because I didnt want to solder point to point to the resistor yet. But the output side goes straight to the HF output of the crossover board bypassing the traces.

Even with only 10 minutes run-in I can already recognize the sound. That’s exactly what impressed me the most with the Micros that were modified. A reviewer / manufacturer, most of whose opinion I agree with and respect a lot (in fact I bought many of his surplus caps used in his shootout tests), once wrote and that the oimps sound uninvolving. I beg to differ this time. In both the Micro and the Divas the OIMPs are anything but ininvolving. Do the V-cap OIMPs have a sound? Yes, it’s the V-cap house (Haus... pun intended?!) sound. Fast yet always under control, extremely airy with no analytical coolness, and never any overshooting or saturation. This, in my opinion, is exactly what modern tweeters need. Subjectively I feel that these are like Jensen copper oil caps with much less color and much more honest and accurate headroom. Some caps are de-focused, most caps glare. Most caps sound beautiful when playing simple stuff but goes non-linear or thins out when dealing with complex passages. The V-Caps (both TFTF and OIMP) in my applications are completely free from that. The smear-free delineation of strings in symphonic passages while maintaining texture is what separates V-Caps apart from all others.

Too bad there is no way I can open the crossovers of the Eidolons otherwise I bet $$$ the V-Cap OIMPs can get rid of the slight grain of the Accuton tweeters.

My next step? Let them run-in for a while. Perhaps get another pair of V-cap OIMPs to bypass the Solen 220uFs (risky bet, some caps cannot be bypassed, it smears). The woofer integration is a different story. It will take me some time to really address the problem of the still remaining mid-bass hole.

For fellow owners of Utopia Be’s. I strongly urge you to check the crossover to see if they are Solen PA / AXON True-caps. Get a pair of V-Cap OIMPs of the same value, won’t take more than 15 minutes to install (unless you are also changing out the midrange high-pass lytic) and you will be able to bring out the absolute absolute best of the Be tweeters.

Reference material:
Sibelius Violin Concerto, Midori / Mehta / Israel Philharmonic
At the Blue Note, Keith Jarrett trio
Firebird, Shaw, Atlanta Symphony, Telarc
DeeDee Bridgewater, Jai Deux Amour

Assoc. Equipment:
Sonic Frontiers SFCD1
A) Home made Aikido preamp with Janus regulator
B) Home made preamp referencing MFA Lumi
A) Counterpoint NP220PG
B) VAC Phi 70

Associated gear
Sonic Frontiers SF-CD1
Sonic Frontiers Processor 3
A Aikido Preamp with Janus regulator
A VAC Phi 70 Monos
B DIY preamp referencing MFA Lumi
B Alta Vista NP220PG

Similar products
Avalon Eidolon
JM Labs Alto Utopia Be
Wilson Watt Puppy 7
Quad 989
Stunning appearance, underwhelming performance. A friend had these and I was surprised to hear how much they cost. Another friend said, "They have no midrange." Well, of course they reproduce the midrange, but we agreed that something was wrong and this seemed to best identify it. Maybe my friend would still have them if he had done what you did to yours.
Why do you have to do so much to make them sing?

Seems like alot of work.
It's not a lot of work.
NOthing at all compared to modding a Ming Da MC-2A3 preamp.

Just 2 V-Cap OIMPs and 2 Solen 220uF 250Vs.
The Solens are hard to install, creative tie-wrapping required due to size. Once figured out the mod for the pair can be done in 30 minutes or even less.

The dead midrange was the $2 SCR lytic cap.
This is the biggest difference between Diva and Micro.
The Micro doesnt have it, thats why even in stock form the Micro has better midrange.
Damn the speakers cost $14k new, how on earth can they get away with using this??? I am not using any premium film cap for that and its already night and day.

The blunt highs were because of the 3.3uF Tru-cap.
The V-Caps gave it the airiness and decay.

Like I said I bought them just because I know somethings wrong with the design and can be fixed easily. The drivers are worldclass and I am a sucker for looks.
I ditched a pair of Divas and thanks to this post I know why I did it. I had owned Mini-Utopias (pre Be tweeters) before and loved them, so I was confident the Divas would be better. But it was a mixed bag. The Devas were less colored and more transparent but all the problems you mention were there, especially what you aptly call the "mid-bass suckout." No changes in speaker positioning or associated equipment cured the problems. I have a basement listening room with limited ceiling height, and I decided the Devas just couldn't develop their sound in that room. I didn't imagine that JM Labs would actually produce an inferior sounding Utopia speaker. I "traded down" to much less expensive Gallo Reference 3s and they were better in every respect.

Beside the Gallos, I also have a mint condition pair of Quad 63s. I'd appreciate your thought on whether a cap upgrade can make a worthwhile improvement in their performance and, if so, how to go about it.
Ahhh Quad ESL63, same circuit as my 989 with 1 less bass panel. I am selling my 989s now so go figure how good I have made the Divas. Look at my listing and you will see what I did to it.
Update: The mid bass dip and strange overtone actually appears courtesy of the series notch filter in the crossover board that connects to the 2 woofers (not the main Xover board where the speaker cables connect to, it's the board that's secured onto the inside of the front baffle.
There lies the woofer low pass crossover and right before it gets to the drivers are in series a 440uF cap and an inductor and 3-paralleled-47Ohm resistors which makes the series notch filter which I had originally mis-identified as a Zobel.
The series notch filter's purpose in theory was to smoothen and extend bass response, but sadly it does the exact reverse in practice.
I removed it by disconnecting 1 lead of the 440uF and the bass comes alive and there is no hint of a mid bass suckout anymore.
The Divas now sound a few classes better in dynamics and even transparency.
Did Focal put that in for the sake of "spec"manship?
Or was it to make these speakers possible to drive with OTLs? (Series notch fixes impedance spikes like Zobel, and calms resonances due to impedance peaks in theory)
Even with extensive listening I dont hear a bass hump or a tendency to resonate. I will leave it disconnected until I find a compelling reason to put it back.

You should check the waterfall response for resonances - the well known problem with very light rigid drivers, such as the w cones, is ringing. Look for resonance around 1100 Hz from the woofers and around 4 kHz from the midrange. If you do not have test equipment then listen for a nasal sound (for the lower ringing peak) and an etched sound for the upper ringing peak.

I am not a fan of rigid light weight driver designs (as you can tell) - to me they always sound impressive but lack a natural timbre - although the Be tweeter is world class, as the resonance peak is well outside the audible range.
I should add that notch filters are usually necessary to cure resonances on these type designs but I expect you understood my meaning.

You should check the waterfall response for resonances - the well known problem with very light rigid drivers, such as the w cones, is ringing. Look for resonance around 1100 Hz from the woofers and around 4 kHz from the midrange. If you do not have test equipment then listen for a nasal sound (for the lower ringing peak) and an etched sound for the upper ringing peak.

I am not a fan of rigid light weight driver designs (as you can tell) - to me they always sound impressive but lack a natural timbre - although the Be tweeter is world class, as the resonance peak is well outside the audible range.
Thanks for the advice Shadorne. Dont feel there's any anomaly at 1k or so right now especially after I pulled the series notch filter. There used be a bass suck out and wooden "middiness" (in addition to low freq muddiness) to these speakers which a friend has always described as nasal but its now gone. After reading up on the purpose of notch filters it seems that it works specifically at calming a resonance peak at a certain frequency which supposedly eliminates a nasal character due to resonance.

I have a feeling the Focal came up with the values based on modeling and calculation but did not really check whether the suppression is dead on when everything is on the box along with that tuned port. A destructive cancellation might have turned into a whole can of worms and exacerbated the problem instead.

Not touching the filter for the midrange for sure as all sounds fine from the mids to highs eversince the cap swap.
What I meant to say was:

A good intention to create destructive cancellation to calm the resonance might have been off-target and exacerbated the problem (more muddiness and more nasal sound) instead.

I don't know. I think JM Labs make awesome speakers with some pretty serious engineering. If there is a notch filter there then my guess is that it is there for a good reason (either driver ringing or to reduce an unwanted resonance in the driver/box system).

Could they have made a mistake on one pair of speakers by someone inadvertently installing the wrong caps or resistors (got mixed up in the wrong box and was missed by QC) - sure it is possible as mistakes do happen even to the very best.

It is also possible that you find the cure worse than the disease - and prefer the system with the ringing problem and without the deleterious phase effects of a sharp analog notch filter.
I dont think its only my specific pair of speakers.
When I auditioned the Divas a few years ago at a dealer the "wooden" bass and the nasal sound was all there,
and I am sure most owners and ex-owners would agree with me.
Funny even the salesguy said in a lot of respects the Electra Be sounds better than the Diva.
I say with confidence that the cure is worse than the disease in this case.
I would expect woofer ringing to be audible as a high bass low mid hump or nasal sound, but contrary to theory without the notch filter the preexisting nasal character actually went away.
It's been exactly a year since I've owned the speakers.
A fellow audiogoner bravely took the step of following my crossover mods and is quite pleased with the results.

Recent communications with him sparked my interest in further pursuing the improvement of the still somewhat lethargic, constricted, and occasionally "one-note" bass as most evident in Holly Cole "I can see clearly now" and Eagles Live "Hotel California."

I removed the woofer crossover board and carefully took measurements of the components. It's a 4th order crossover with L1=1.2mH C1=390uF L2=530uH C2=220uF.
What disappoints me is the DC impedance of the inductors.
1.1 Ohm and 1.0 Ohm respectively.

The 2 woofers are wired in parallel. 6 Ohms nominal per driver that makes it 3 ohms. With the inductors of such high impedance, no wonder the bass is lethargic. Kind of like having to drop 1/3 of the power before getting to the drivers! Better inductors of similar values should be 0.12 Ohms (Check Michael Percy Audio or Partsconnexion)

A friend suggested that I do some tspice simulation of the circuit. Ouch, this crossover is extremely notchy and phasey (the phase shift as the frequency increases is not smooth at all.) The woofers are 4th order, the midrange is second order for sure. A rather strange combination.

A friend suggested I use tspice and come up with values for just a Butterworth 4th order and run simulation to verify the drop is smooth. I compared the 2 plots and decided to be adventurous. Ordered the parts from Partsconnexion and THEN I realize how much bigger and heavier those inductors are. Even the strap on 68uF Mundorf bipolar lytic is 3 x the size of the incumbent 390uF (I need 460uF for C1).

This is a much much more involved mod than the cap replacement of the mid and tweeter obviously.
Took me many hours figuring out how to mount them creatively and somewhat non-invasively without having to build it externally. But even with only 1 speaker done I was able to tell the difference putting my ears close to the woofer. There is a world of difference in clarity. Another hour gets the 2nd channel done... drum roll...

There is a lot more dynamic contrast in the bass notes. The decay characteristics of one drum beat vs another is clearly evident. It doesn't go any lower than before, but has the "fast bass" (a term used by Avalon's Niel Patel) that never existed with these speakers. Listening to no less than 10 of my favorite bass-challenging recordings I don't hear any note that stands out or causes a nasal sound.

I'm not quite done yet. Ultimately I might just build an external crossover box for it so I can replace ALL bipolar lytics in the crossover and see if I can take it up another notch.
Thank you for the writeup! That was a lot of fun to read.
I admire your courage and expertise; but this story causes me to lose any confidence I might have had in JM Labs integrity. This is not a small company, and there is no excuse for sub-par crossover design. I just don't understand.
The Diva Utopia had home theater "lifestyle" speaker in mind according to the brochures.
Bad compromise. Just the cost of the inductors I put in was over $450 so figure the usual 5-7x multiple it would have increased the price of this already expensive speaker by quite a bit.
sorry to review this thread, Johnsonwu id like to discuss this mod more with you, do you have a email contact?

Interesting to see there is no "contact member" option on Audiogon. AT least I cannot find it. You can just post questions here I guess. more than happy to help.
Hi there, please drop me a email? dastrix at gmail dot com

Thankyou :)
I own a pair of Electra 936s. I like the treble and mids after using a good tube amp to drive them. The bass however isn't always quite as punchy as I would like it to be. Have you ever worked on Electras? I realize they are much less costly speakers at the outset and ultra expensive caps and other higher quality crossover components may seem like overkill.
What did strike me was just how much better the speakers sounded when I changed amps. I qas using Aragon Palladium 1k 400wpc monoblocks with an ARC SP6-B modded pre. I then bought the little 6SN7 based AES AE-3 by Cary a very simple tube preamp, I am a tube collector so absolutely wonderful tubes were rolled in. Then when my friend became the distributor for Opera Consonance, USA, I bought a pair of Cyber 800s using the stock EH big bottle 6CA7s but rolled the 5687s to Tung Sols and the 6922s to Mullard CV-2493 (really NOS).
Man I cannot tell you how many compliments I got on how much speakers improved, which I didn't touch. I kept telling them it was the amps but no one would believe it, until a friend modded the pre amp minimally with hexfreds.

Just a thought how much better do you things would be if you used trulu SOTA top of the line caps and other components changing values only when absolutely necessary?
It's not the caps when it comes to bass. It's the coil inductors.
Changing them requires a lot of testing and measuring. Not to mention size issues esp if you use the highest end air core foils.
So theoretically a different presumably larger inductor coil, help the bass in a crossover. You did say however there was no simple way of knowing apriori what inductors to use to get that better bass. So I guess I have to give up on that idea, although changing just one (large) element, albeit expensive, sounds tempting. Any ideas on what might work?
It's a combination of analysing the current components and figuring out what kind of short-cuts the orig. mfr took to minimize cost/size and whether/how they relate to the current shortcoming in the sound first, then validate that in your current listening environment and decide what mods to apply.
The approach of replacing a coil with the same inductance with a huge air core foil inductor (which will likely reduce series resistance) will give you improvments but likely not get you the max benefit vs a well thought out mod. But you never know until you try it out. Just be careful not to damage the old components in case you want to revert.
Hi Johnson , wow ! The difference with vocals is just mind blowing. The imaging is sensation , so much more space . So more holographic - its for about 1hr on them and I haven't really sat down , just back ground playing. Highly impressive - thank you !!!!

I took the suggested route from johnson and installed VCap Oils and Mills Resistors for the BE Tweeter and Mid range on my Focal 927 BEs, they are superb now!
Is it possible to do the same mod on 1008 Be too? Thanks!
Even simpler and cheaper go mod than the three way ones. It will take it to micro utopia level.
Thanks for the reply.
I have never done any mod myself. Any advice on how to do it on 1008 Be? Anyone that I can contact?
Do you have access to anyone who knows how to operate a soldering iron of 25w or more and knows how not to overheat a component?
Hi Johnsonwu,
None of my friend are into audio and think buying speakers like Electra is a nut job. I too have not been in this for long. Anyway, a quick google search gives me someone that I think I can ask for help, maybe with a service fee. It's called True Sound in Campbell, CA ( On its website, it says the owner has done Quad 63 mod. I will try to contact him and see if doing mod on 1008Be is something that would interest him.

By the way, is there anyway I can PM you for detail info? Again, thanks for the info and answering my questions.
Nick Gowan is a scholar and a gentleman and I am about 15 miles away from him. As if he would like to do it. Maybe I can help chip in my 2 cents.
Hi Johnsonwu, it seems we live close to each other as well. I'm from San Jose. If you don't mind, here's my email khinz.htoo@gmail.

Called and left a message for Nick today. The lady on the phone said he's busy.
Hi Johnsonwu,

I have enjoyed reading your post. I have the Divas have been looking listenning to some really high end speakers from TAD the other week and thought of whether I could get better sound from the Divas.

I'm in the UK and have found someone who is willing to modify my speakers for me. I am thinking about ordering the various parts required from Parts Connexion as they are cheaper than sourcing in the UK. Just wanted to ask if all the parts that you used are within the thread. I would be grateful for any assistance. I am looking at may be getting external crossovers built for the Divas. Have you explored this route since you last posted.

Hi Jazzao1, The coupling caps for the tweeter is a good start which will get you 70% of the way.
Getting the lytic replaced with a Solen AXON 250V 180uF bypassed with a 2uF Mundorf silver oil will get you to 100% of where the mids and highs should be.
Both mods take less than 10 minutes per speaker.
The woofer crossover mod is very involved.
Honestly I would not recommend it unless you have some skilled DIYer with some special drills handy (or build and ext crossover box outright)
First thanks to all contributors here for all the insights and detail, especially Johnsonwu obviously.

I have just obtained some Diva Utopias, with, incidentally, new tweeters due to previous child damage!

In case anyone is interested, they replace my Monitor Audio Studio 60’s, which I never thought I would part with. In comparison, the Diva’s just sound more effortless, though they do not – yet – image anything like as well as the Studio 60’s.

Hence my interest in this thread… I’d just like to be crystal clear about the suggested mods here, can I ask Johnsonwu a few questions?:-

1/ “Tweeter” – I think this is simply remove “Solen/AXON 3.3uF 250V” and replace with V-Cap 3.3uF OIMP.
- Question – 250V again for the OIMP?

2/ “Bass suckout” – Not sure here…. Early in the thread, the suggestion is – I think – Remove SCR 180uF lytic and replace with Solen 220uF 250V Film Cap
However the last reply suggests these should be 180uF… and bypassed with 2uF Mundorf silver oil.
- Can you clarify what to use here?
- Also for the Solen do you mean “Fastcap PB series – metallised polypropylene caps”?

Also there is the suggestion to disconnect the 440uF cap on the woofer crossover board – is this still recommended?

3/ Woofer inductor mod - -I’ll take the advice to pass on this for now….

Question for Jazzaol (Alan)– I’m in the UK also – have you had the mods done yet?

Thanks everyone
Regards, Will