Subwoofers and LS3/5As

[Tony L]

Of my many amps, the Grant works best with the Spendor LS3/5As. However it doesn't fill the lower-mid gap. It smooths over the slight graininess I hear in both the LS3/5A and the JR149.

I love the sound of the LS3/5As with either Leaks (TL12+ or Stereo 20) and on the 303 downstairs. They sound a bit boring and dead on the Pass Aleph 3 though, which the JR149s seem to prefer as it is just so smooth and clear over the mid (the 149s are a little mid-forward in my installation).

Thinking aloud:
I wonder if the LS3/5a might suffer from thermal effects. For the same SPL as a JR149, the LS3/5A will need to dissipate rather more heat inside the cabinet. The JR has a metal cabinet which must be more efficient at removing heat than the wooden sealed box of the 3/5A.
Drive unit parameters change significantly with temperature. I wonder if the sound of the LS3/5A changes quite a bit over a period of use?

I don’t notice any ‘warm up character’ at all with either speaker beyond the initial burn-in phase with my 149s after restoring the crossovers and replacing the drivers where they certainly relaxed over the first month or so. I feel I play them both well within their performance envelope though, I’d be surprised if even the fastest transient peaks were over about 90db or so, my median C-weighted level around 75-80db at the listening seat. There is a resistor in both the LS3/5A and JR149 crossover which is known to get very hot if the speaker is asked to do more than it wants. A spare pair of 149 crossover boards actually had a burned area on the PCB, and I believe similar occurs with some LS3/5As. I’m sure the B110 will warm up a bit after a few hours use, but I don’t personally hear any character changes.

 

[ToTo Man]

LS3/5A was designed for voice monitoring in an outside broadcast van.

This intrigues me because, if monitoring speech, I'd have thought it would have been better to monitor it on speakers that did not have a lower-midrange recession, otherwise you might end up mixing the vocals too bloated/thick, which IMO would be much worse for speech intelligibility over a typical 1970s consumer TV set with built-in speaker than vocals that were mixed too clear/thin?

I'd love to see a photo of an LS3/5A being used in a broadcast van as it might give some clues as to the kind of response they got from it. E.g. reflections from the desk or extending the baffle width with books may have brought up the level of the lower mids?

My S3/5R2 are at the other end of the scale to the Q7 I auditioned, I find them to be a little murky/bloated in the lower mids (IIRC J.A. detected a resonance in the 5R2 cabinet around these frequencies). The JR149's lower mids sits in between the Q7 and 5R2 and, to my ears, is the most balanced of the three in this area. Perhaps if I heard them in a different room my impressions would change...

 

[S-Man]

I'd love to see a photo of an LS3/5A being used in a broadcast van as it might give some clues as to the kind of response they got from it. E.g. reflections from the desk or extending the baffle width with books may have brought up the level of the lower mids?

Seem to be very thin on the ground. These look a bit bigger and later than LS3/5A:

However, I suspect that they would have been (effectively) soffit/flush mounted.
So I expect the LS3/5A was designed with no baffle step compensation (was this even known then??).

Is this common knowledge?... LS"3" = outside broadcast and "5" = studio, at the Beeb.

 

[Tony L]

I’d say that was fairly typical of broadcast monitor use, i.e. nearfield and sitting on the broadcast equivalent of a mixer meter-bridge. Can’t find any picture of an outside broadcast van, but I’d expect that to be pretty compromised. I guess a live OB van was only going to be dealing with vocal balancing and detecting mic clipping etc, just basic QC so positioning would be less critical.

PS As I understand it the LS3/5A was a voice monitor, so well used by say R4 etc in studio locations. The bigger stuff (LS5/8s etc) being use for live music recording/monitoring, e.g. Proms, Peel sessions etc.

 

[S-Man]

I will have to ask my pal who owns the JRs and 3/5As. He used to work for the BBC as an audio technician in the 80s, he certainly got calls to particular studios to sort out problems often with the instruction "fix it, we're on air in 20 minutes!". He was the one who told me about the 3 and 5 designation. It would be interesting to know if he worked on the kit in the OB vans.

 

[Tony L]

I suspect the key thing to differentiate in the BBC environment is the difference between live recording, editing, and broadcast monitors. They are all very distinct roles, and you really don’t need a pair of LS5/8s to present Gardners Question Time, Woman’s Hour or whatever, nor to cut together a spoken-word news report. Likewise to play records, CDs etc as there is nothing to EQ. The 5/8s etc will have been kept for serious music recording, creating radio plays and TV shows with music and sound FX etc. Then there are the various LS3/6 (BC1) size boxes lying somewhere between the two extremes (I suspect music DJs etc got these). I have been in BBC local radio studios decades ago and I’ve certainly seen LS3/5As there along with 3/6 size boxes.

PS They seem to use little two-way active Dynaudios nearly everywhere now!

 

[S-Man]

I would prefer a pair of LS0/0s for The Archers

 

[Tony L]

Likewise. They really shouldn’t use any mics for that.

 

[S-Man]

Not sure if it's the sunshine or something in the water, but I summoned up the motivation to move a load of kit into the back room. I suppose I just really wanted to know if a pair of bass extenders that are flat to 10Hz would cause major problems in a small room. I also wanted to hear the LS3/5As in the nearfield in a smallish room 3.5 X 3.3M + an entrance way. The speakers are 145cm apart and I'm sitting 180cm from the line between them. The back of the box is 37cm from the front wall, my lugs are about 110cm from the back wall. The 3/5As are driven, with or without the bass extenders, by my KT88 amp. I tried them briefly with with my impure class A amp, but it was not a happy match:
https://pinkfishmedia.net/forum/threads/contaminated-class-a-amp.264761/

The first thing that I noticed is that the LS3/5As sound inferior in this room. The upper mid shoutiness is more apparent. In fairness the room does have less soft furnishings than the main room. My 2nd conclusion is that the bass extenders improve things a lot in the small room and surprisingly the very extended bass doesn't cause any "subwoofer" noises.

 

[Tony L]

The first thing that I noticed is that the LS3/5As sound inferior in this room. The upper mid shoutiness is more apparent.

What LS3/5As have you got? “Upper bass shoutiness” are three words I’d personally find next to impossible to use in a sentence with mine. The mid is genuinely stunning. Very like ESL57s tonally: clear, clean, open, natural.

If they are a vintage pair I’d crack them open and have a look around. Check the cabinet seal, look for stressed crossover components, and if possible REW sweep them each individually on the same stand as any sample deviation likely implies neither are performing to spec.

 

[ToTo Man]

Upper-mid shoutiness could be a sign of B110s that have gone off-spec, this is something Jerry discussed in his RMAF presentation, IIRC. I've owned multiple pairs of mk1 JR149 over the years and some have been more 'shouty' than others, but I can't say for sure if this was due to off-spec B110s. The two pairs of mk1 JR149 I currently own are quite similar to each other but, without resorting to EQ, I prefer the tonality of the 'green' pair:

 

[S-Man]

What LS3/5As have you got? “Upper bass shoutiness” are three words I’d personally find next to impossible to use in a sentence with mine. The mid is genuinely stunning. Very like ESL57s tonally: clear, clean, open, natural.

If they are a vintage pair I’d crack them open and have a look around. Check the cabinet seal, look for stressed crossover components, and if possible REW sweep them each individually on the same stand as any sample deviation likely implies neither are performing to spec.

Upper mid shoutiness not upper bass shoutiness.

I accept what you say but remember that I preferred these particular speakers to Stirling V2s. The owner has another pair of Spendor LS3/5As, 3 pairs of JR149s and a pair of ELS63s and we have done numerous comparisons.
One or two of the pairs of JRs are undergoing some work. We plan to measure and audition them. I will see if we can get the LS3/5As measured as well.

My final experiment was to try my Murphy CAOW1s in the back room (with the bass extenders). They were nearly as good as in the larger room, whereas the LS3/5As were not.

 

[S-Man]

How does one identify an out of spec B110? (Other than Tony's suggestion).
Can it be done in situ or does it need to be tested on its own outside the box? The Spendors have biwire terminals so the B110 and its crossover can be driven on their own.

@ToTo Man - Most of differences in your plots are >2KHz. Isn't it more likely that the T27 is responsible (or the sheer difficulty of making consistent in-room measurements)?

 

[Tony L]

Short of comparing them to a fresh pair of Falcons all you can really do is compare them to each other. Speakers very rarely age identically, they are usually in different lighting conditions etc so glues age differently. I alway pan some white and pink noise left and right between speakers just sitting next to each other on the floor. It is easy to pick up differences by ear, if they exist it is then time to measure.

PS According to Jerry from Falcon just about every original B110 and T27 will fail to meet spec now. The glues used just weren’t stable over a 40+ year timescale. I had a good phone chat with him when I was diagnosing my 149s (which initially sounded like two different speakers despite all drivers “working”). Admittedly there was also a factory crossover error that emerged much later, but those drivers were way out.

 

[ToTo Man]

How does one identify an out of spec B110? (Other than Tony's suggestion).
Can it be done in situ or does it need to be tested on its own outside the box? The Spendors have biwire terminals so the B110 and its crossover can be driven on their own.

@ToTo Man - Most of differences in your plots are >2KHz. Isn't it more likely that the T27 is responsible (or the sheer difficulty of making consistent in-room measurements)?

Identifying an out-of-spec B110 is beyond my pay grade I'm afraid, but my best guess would be to disconnect it from the crossover and connect directly to a signal generator, either leaving it installed in the LS3/5A enclosure or removing it and putting it in an open baffle, and comparing it to a brand new Falcon B110.

Here's a measurement I did on a pair of Type SP1003 B110s. I did not use an enclosure or an open baffle so the measurement is of limited use from an objective standpoint, but you can see that the B110 has more output above 3kHz than the Audax HB13 and Monacor SP-135TC I also measured. The JR149 crossover frequency is 3kHz, and in a conventional crossover filter the raw response of a driver still has an effect on the system's frequency response until it's -20dB down in level, so unless the JR149's "anti-quack" filter is particularly aggressive I expect that the B110's output up to 5kHz will be affecting the JR's measured frequency response. It could, however, also be the tweeter that's contributing to the difference, as you rightly suggest.

Note that these B110 drivers came from a pair of KEF Concerto I stripped many years ago, which I've been holding onto in case I ever use them in a project (i.e. this was before Falcon began manufacturing new T27 and B110 units). I haven't tried installing them in my JRs so I don't know whether they are any less or more 'off-spec' than the ones in my JRs. If I was going to the hassle of swapping out drivers in my JRs I'd probably just order new ones from Falcon, at least then I'd know they were 100% on-spec.

 

[S-Man]

Thanks both for the suggestions. I suppose we do have a total of 6 JRs and 4 3/5As to play with, so hopefully can find the outliers.
I think my pal has some reasonably sophisticated test gear as well.

 

[Tony L]

Are your LS3/5As 15 or 11 Ohm? They each have a different B110 version, the earlier 15 Ohm version having the same driver as the JR149 (albeit BBC selected at one end of the tolerance bell-curve).

 

[S-Man]

I don't know. They measure 6.2 Ohms DCR. I expect they are the later 11 Ohm type.

... and they are on loan.

They have spent most of their life in the carton. My pal bought 2 pairs when they were being sold off cheap about 20 years ago.
Of course he should have bought all their stock along with bucket loads of GEC KT88s - but I think he thought they were a bit steep at the time .

 

[Brian S]

I really like the way they do piano! Stick on some well recorded solo Chopin or whatever and they sound superb. Same with well recorded close-mic’d stuff, e.g. try Herbie Hancock’s River.

I was just listening Elisabeth Leonskaja playing Mozart sonatas through my Spendor D1s just under 3mtrs away and thinking, would I gain from anything bigger (Spendor 3/1) in this configuration. Excellent.

 

[TheDecameron]

I was just listening Elisabeth Leonskaja playing Mozart sonatas through my Spendor D1s just under 3mtrs away and thinking, would I gain from anything bigger (Spendor 3/1) in this configuration. Excellent.

I have a pair as well and was struck by how better resolved they were than any of the s3/5r, HLP3ESR and 3/5as I had, which is not to say any of the latter are bad, they’re not but the D1 tweeters and cabinet design bring out even more. You might take a punt on or borrow a BK sub to try with them.