I just got done doing one of Martin's great recap kits for my crossovers but my midranges are still distorting. I called the B&O service headquarters in Chicago and they said the coils are shot. They also said any 3" 8ohm midrange will do as a replacement, but I'm having trouble finding one that will fit the cabinet just right. The hole is about 3.5 in and most 3" midranges require a 4.5inch cutout. Any suggestions for a replacement that would fit? I don't want to get used originals because both drivers are shot and I'm pretty used ones will fail sooner than later.

B&O service are talking rubbish these filler drivers are very specific and no modern equivelent will do.

First try taking the drivers apart, but do so very carefuly so as not to damage the hair-thin enamled wires to the outer ring. I suspect you will find the voice coil has come loose from the dome so a little speaker cement will stick them back on again. Just make sure they sit in their groove and are dead straight! Then there should be no rubbing when installing them back to the magnet assembly!

If this fails Martin is the man to ask for spares!

They said this specific speaker - which he had the documents for and recognized without looking up - was one of the few that was exactly 3" and could be replaced by anything the same size. He also knew that the coils on these specific models can fix themselves after a time and go bad again - which is exactly whats happening as one speaker had this problem, cleared up, and now the other one is doing it. So I would tend to trust them. I don't mind taking the driver apart, but I didn't really see a way to do it last I looked at it.

I haven't been able to find any modern replacements for the SEAS drivers but admitted, I really haven't
had the need for it either.
Of the few faulty drivers that I've seen, I've been able to repair most and replace the rest with good used drivers.

If you have any luck finding a suitable modern replacement, do let us know.

Martin

Well if you have refurbished replacements I wouldn't mind buying them from you. I'd even be willing to send you the ones I have if they are of any use to you. I don't particularly want to do a lot of foot work involving this.

Mark I repaired mine some time ago and they are fine! If I remember correctly there are three clips round the outside which will all the dome to be pulled from the magnet assembly, perhaps along with a tilted voice coil so make sure it doesn't get stuck!

Just hang on a minute....

Thought I had a pic!

I had forgotton but the dome had come away from the surround on mine too!

I remember there was a similar situation where even a recone specialist said the dome was not repairable but the OP ended up taking apart and I think he did it himself - his coil was partially burnt too!

So Martin hooked me up with new midranges that are on their way. The repair fellow at A Brown Soun told me the reason why the mids died was apparently a design flaw in the speakers themselves. There's apparently a small resistor that is sending low end frequencies to the midrange driver on purpose. The old drivers couldn't handle the strain anymore. He's going to try to replace the resistor with a larger one or remove it all together to seperate the frequencies as much as possible. Anyone else ever seen something like this?

I've been reading nearly everything on the forum about speakers for the last 2 years,  and nobody ever mentioned anything about this. If you are sure that no repairs or alternations was made to your speakers before, and I'm sure B&O would never design a crossover that the drivers couldn't handle, check with the manual onsite to get the right resistor, but I'm also sure that Martin has some opinions about it.

We have the service manual with the schematics of the crossover. This guy is pretty well known in the states, so I trust him so far. Considering I have 2-3 weeks before the drivers get here I have time to discuss it here and see what the forum experts think.

Yes its interesting, let the experts speak.

Assuming that Soren's point regarding previous mods is not true, I would be asking "And the capacitors and inductors in that part of the crossover are doing nothing?"

This driver fills in a gap between bass and treble, if I remember rightly there is an overlap but the peak in response is fairly narrow, the majority of which lies within the gap left by the other two units - I very much doubt B&O would have put the mid drivers at risk!

I would say altering the values in these crossovers, more than any other speakers is something you should not do without some sound theory behind it, I should imagine the results could be quite disturbing!

Anyway it will be interesting to see if Martin has experienced problems of this nature I am betting he has not!

Just one more thing, my speakers were well loved over time and used for a lot of party's until they were stored away by the previous owner. Look at the coils in the above picture... Would you say there is any sign of fatique or that they have been over-heating???

Personally I think you would be better working on these speakers yourself!

I never experienced problems of this sort.
I have had a few burned midranges from the speakers being generally over powered and
one or two where the coil had losened from the diaphragm (which is usually fixable) so I cannot
help thinking that unless the tech guys are very unexperienced with speakers and measuring
on crossover filters (doubt that) something is very wrong with the filter boards.

Martin

The coil apparently did come loose on one of them. The same problem isn't true of the other. Yet both are suffering from the same midrange blow out. He does have his particular way of testing the frequencies. These guys were recommended on a number of forums because musicians send them their stuff from around the counrty. He's just not familiar with this unit, but he says hes never seen this level of overlap so he doesn't want to put in two new mids just to see them blow again.

Is there anything else that I should tell him to look at if this isn't a built in deficiency? Which bits are the filter boards?

The filter boards are where the electronic frequency filtering takes place.
In your speakers, it's a printed circuit board with all the components needed
to separate low, medium and high frequencies and send them to their respective drivers.
With the circuit drawing in front of them and a bit of calculating, the tech guys should be able
to tell if something is seriously amiss.

In your speakers, the board looks like this:

Martin

I see, so its just the crossover board. Let me ask you, do you know what the frequency overlap is supposed to be exactly so we can know if its just the normal activity of the board or if its out of order?

The crossover frequencies is found in the service manual but, as said before,
they can also be calculated from the circuit diagram and measured in a test setup
to confirm correct operation.

Martin

Will the following description mean anything to the techs...?

Jon:

Well, I wouldn't consider myself to be anything approaching "quite knowledgable", but thank you!

As far as I understand it to be, Bang and Olufsen decided on using the filler-driver concept as a way to compensate for the phase shift introduced by a crossover.

Because of the resultant phase shift with a crossover, the driver's polarities usually are reversed so that their acoustic output sums flat at the crossover point. This, however, obviously means the speaker system can never achieve a minimum phase response (zero phase shift being the goal).

What B&O did with the 3-way Uniphase speakers is to design a crossover with symmetrical 2nd order low-pass and high-pass acoustic slopes, but with the drivers connected in phase. Then, in order to counteract the resulting cancellation in response, they used a midrange driver to "fill in" the response - that midrange driver's passband being centered in the middle of the high and low frequency driver's crossover point. Above and below it's passband, the midrange driver's acoustic response rolled off at a shallower rate. This meant that the overall frequency response of the speaker summed flat, yet the drivers were all in phase with each other.

As well, B&O "time-aligned" (physically lined up) the acoustic centers of each respective driver. This is why you see some of the drivers a bit more recessed on the baffle than others. So, the Uniphase designs are "transient-perfect" as well, in that they can (mostly) accurately reproduce a square wave. Notice that "time-aligned" designs which are not also either minimum phase (zero phase shift), or linear phase (the entire passband of the speaker phase-shifted by an equal amount), are not transient perfect.

I believe the 2-way Uniphase speakers have the response of the high-pass and low-pass sections overlapped, and then the hump in response is tuned out.

There are WAY more knowledgeable people than I who discuss this at length. You can probably Google things like "transient perfect" and "filler driver" and find alot of useful, in-depth information. Search for a man named "John Kreskovsky" - he's written a few papers about this subject.

The merits of this design are highly debated, with many people saying you'll never hear the difference. I'm the kind of person that doesn't always trust my ear, but I do believe what I can measure to be true. Within certain limitations, I believe in the merits of these designs.

Many people claim that due to the reflections and phase anomolies introduced just by your listening room, a transient perfect design is a moot point, and I can see where they're coming from.

Also, by nature, these designs have poor vertical polar response, illustrated perfectly by the big difference in sound quality you get when you listen standing up versus sitting down. Even if you could somehow get around the room influences, the Uniphase speakers' transient perfect quality (and really, it would be the same with ANY similar design, not just B&O's) would be evident only through a very narrow listening window. Move up or down off axis out of that listening window, and that phase coherency is lost.

These speakers were made back when B&O was more concerned with technical perfection, IMO. It's not an easy thing to do; what they did, that is. It resulted in a more complex and expensive design, but one they considered worth it. Again, many doubt the validity of these designs in the first place, and don't even bother to acheive these design goals. And I've heard many fine speakers that are nowhere near transient perfect in nature. But there are a number of manufacturers that do strive for this, so I guess it just depends on what you choose to believe.

I hope I made sense the way I explained that, and didn't mess anything up.

Jon

Not sure how unique this approach was?

Wow, Olly. That's interesting. Where did you find that?

Mark,

What Jon wrote is all true, he has a lot of knowledge and it's wonderfully put but it doesn't say anything
about the crossover frequencies and it doesn't talk about your problem area.
Besides, I never considered this series of speakers as being truely "Uniphase" and none of the drivers
have reversed polarities in the S120.
I see the Beovox S120 as more or less a standard, even if somewhat beefy, livingroom type three-way system
and a bass-reflex system at that and B&O used the name "Perspective" for some of the models in the range
rather than "Uniphase".

The tech guys may not have a lot of experience with this exact model but they will know the general
workings of a speaker and its circuits and they will know how to calculate. They are trained people
and they should not rely on you to provide them with all the tech info.
If they think there is a fault, they will have to locate and correct it.
My experience is that the midrange is playing quite a lot in the larger B&O speakers like these. Considerably
more than f.e. the midrange in the good old S45 (true Uniphase) does. The diameter of the midrange coil
alone suggests that this is a driver that will speak up loud.
At normal livingroom listening level, you can easily feel the vibrations of the membrane when playing.
That may trick a non-experienced repairer into believing that the midrange is fed too much
signal (or too low frequency with too high energy). I don't have the filters here so I cannot tell
if that is indeed the case, I just know that I was quite surprised myself.

Two original midrange drivers are on their way to you. I suggest you bring them to the repairers and ask
them to repair the speakers. All components in the crossovers are fairly standard types and they should
have no problems locating suitable replacements, should they need any.
If they have any specific doubts or questions, let them ask us and we'll do what we can to answer.

Martin