advertisement


Tannoys always full of surprises...

I'm not sure if this is acting as an all purpose Tannoy thread, or specific to cooky's work. If the former, I just thought I'd mention that there is absolutely no doubt in my house that 10" HPDs like a 4 Ohm tap (at least on a Primaluna Dialogue Premium).
 
The Hf screws should only be tight enough not to compress or deform the washers/ diaphragm edge. I tighten them to the bite then back off 1/8 turn as a starting point. A dab of nail varnish across the screw and the cap will prevent any movement. Super glue will do the same but I’d only use the gel variety. Setting the Dias up requires the patience of a saint but is the single most effective way to realise the true potential of DC tannoys .
 
Thanks for the advice. I just put a good blob of gel super glue across the vertical of the each screw/washer, making sure to make contact with the cap. Seems to have made an immediate improvement to the solidity and clarity of the image, hopefully it holds!
 
Super glue is strong in tension but relatively weak in shear so a bit applied to a screw or nut is strong enough to prevent movement but remains easy to break the bond with a bit of torque.
 
First of all hello to all you Tannoy lovers. I am hoping you may point me in the right direction with regard to some MG15's. I acquired the drive units last year, they came with some Mastrad crossovers. Turned out that the cones were incorrect, so they were re-coned by Lockwood. I subsequently got a hold of some York corner copies, very slavishly constructed to original plans ( originals are off the map price wise as I'm sure you all know ).

I stripped the old wadding out and replaced with a combination of damping materials. The sound was as good as I had hoped with the exception of an anomaly around the critical upper mid band. Thinking it may be the result of incorrect values in the crossover, after some research I ended up going with Troels re-design with some fancy caps and coils.

No doubt an improvement, but still has not resolved the spike in the response.

Driver coil values are fine. Treble is very good so not a coil issue, I think. Could alignment or any other issue account for this ? It seems over a narrow frequency range. High female singing or upper range of piano for instance.
 
First of all hello to all you Tannoy lovers. I am hoping you may point me in the right direction with regard to some MG15's. I acquired the drive units last year, they came with some Mastrad crossovers. Turned out that the cones were incorrect, so they were re-coned by Lockwood. I subsequently got a hold of some York corner copies, very slavishly constructed to original plans ( originals are off the map price wise as I'm sure you all know ).

I stripped the old wadding out and replaced with a combination of damping materials. The sound was as good as I had hoped with the exception of an anomaly around the critical upper mid band. Thinking it may be the result of incorrect values in the crossover, after some research I ended up going with Troels re-design with some fancy caps and coils.

No doubt an improvement, but still has not resolved the spike in the response.

Driver coil values are fine. Treble is very good so not a coil issue, I think. Could alignment or any other issue account for this ? It seems over a narrow frequency range. High female singing or upper range of piano for instance.

What is the nature of the anomaly? A spike? Distortion?
 
What are ‘Mastrad crossovers’? I missed that in the earlier PM contact and assumed you had the standard Tannoy crossovers with their controls. As far as I’m concerned all bets are off as soon as the crossovers are altered. I’ve heard a lot of third-party ones over the years and nothing I feel that gets the voicing ‘right’ the way the original ones do. One of many mistakes that seem to be made is not to account for the treble-loss that was designed-in to the originals for the cables and switches in the control panel. Remove them and keep the same values and things get crazy bright and forward, so this is the area I’d look to first. If you still have the proper Tannoy crossovers I’d certainly pop them back in and see if things are more to your taste. I use mine totally stock and love them.
 
Ok this is why I changed from the crossovers that came with them. Mastrad are the e bay Tannoy guys crossovers. They are now the Troels items with high end caps and the original autoformers. I could construct the original circuit but the Troels one is close especially in the treble.
The best way to describe the sound in that region is shouty but it’s over a narrow band. I was hoping the new crossover would fix it. I’m pretty convinced it’s the drive units, so may dismantle and see if there is any oxidation or misalignment.
 
If the crossovers aren’t too bright (i.e. they electrically compensate for not having the control panel, connecting cable etc) then it is almost certainly an alignment issue with the compression driver. The original versions that shipped with the speaker were in two pieces, had a thin coated metal back plate and could be jiggled in the opening a bit to achieve correct alignment. The more modern replacement have a black thermoplastic back, are thicker, and are far harder/borderline impossible to align (they are meant to be ‘plug and play’, but are anything but IME. As I understand it these are actually HPD units, not Red/Gold. Whichever type you have the tightness of the four bolts is absolutely critical and you will likely find just the slightest amount of pressure that holds the unit in place is the best sounding position. Just turn the bolts with you fingers until they stop, and that is usually about right tension-wise. If necessary hold the bolt in place with a bit of nail-varnish or paint when finished to stop it vibrating loose.

My Monitor Golds have the original two-part compression drivers and aligning them to the point I was really happy took me absolutely ages. I used a combination of pink and white noise signal and frequency sweeps. The crossover occurs around 1kHz, so you can hear a lot at that frequency, and if the compression driver is misaligned it tends to loose the frequencies immediately above that point, i.e. the presence region ends up notched. If you have the later thermoplastic type or not be aware they may sound slightly different in all four possible orientations. If you have the original type be very careful indeed as they are very fragile indeed. Fragile, rare and very expensive!

PS Here’s a pic of one of the old two-part compression drivers:

3949484793_bf9391dc31_c.jpg
 
I set the crossover hard wired to a neutral position, the treble is absolutely fine. It’s that region around the 1 kHz I would say, but a spike rather than trough. Given they had incorrect cones ( looked like ribbed JBL ) God knows what lurks in there but at least the coils seems ok.

Question - so the alignment is effectively down to the movement permitted on the threads/holes ? Would the magnetic field not centre the coils automatically?

Other than this the sound is fabulous. Working well with an Audio Research Ref 110 and passive, though I’m looking for an active pre.
 
Question - so the alignment is effectively down to the movement permitted on the threads/holes ? Would the magnetic field not centre the coils automatically?

To my mind the things in play are lateral-location (precise alignment of the coil in the magnet gap) and damping/over-damping (bolt tension).

One thing to understand here is Tannoys of this era are complex hand made/assembled devices. The predate ultra-precise modern production techniques so do have a degree of imperfection. As an example you will almost certainly find a different number or combination of card spacing washers under the compression drivers left to right, even different types (e.g. brown are a different thickness to white). These were chosen at the factory to position the compression diaphragm at exactly the right hight from the pepperpot, so whatever you do don’t mix them up!

Another thing to consider/experiment with is how tightly you have the drive units bolted to the baffle. If you are using a traditional thin-wall Tannoy-style cabinet, i.e. with the drive units mounted from the back to a rather thin baffle over tightening them sounds absolutely horrible and does really mess up the treble/make them over-bright/forward. I have mine with plastic washers under the bolts on the back and just very gently nipped-up so they can’t move, little beyond finger-tight. My guess is this allows the thick rubber front gasket to add some damping to the baffle rather than the metal basket directly coupling and energising it, but whatever the reason it sounds hugely, hugely worse tightened up too hard. This has proven true both with my previous DIY York cabs and my current Lockwoods. Certainly play with this.
 
I set the crossover hard wired to a neutral position, the treble is absolutely fine. It’s that region around the 1 kHz I would say, but a spike rather than trough. Given they had incorrect cones ( looked like ribbed JBL ) God knows what lurks in there but at least the coils seems ok.

Question - so the alignment is effectively down to the movement permitted on the threads/holes ? Would the magnetic field not centre the coils automatically?

Other than this the sound is fabulous. Working well with an Audio Research Ref 110 and passive, though I’m looking for an active pre.
If you were able to measure the frequency response then diagnosing and fixing the problem may become a little easier.

e.g. here is a comparison of the stock LSU crossover vs a particular aftermarket crossover from mainlytannoy. As you can see, this aftermarket crossover has an issue around 1.5kHz:

48658895752_429d883d93_o.jpg
 
Interesting that the problem occurs with the Troels, the Mainlytannoy crossovers and on both drivers.
Do both xovers include the notch filter? Are you sure they aren't 'silvers' type with the LF running full open and minus the notch?
As both crossovers are aftermarket first thing I'd try is inverting the HF polarity and listen if this tames the shout.
 
To my mind the things in play are lateral-location (precise alignment of the coil in the magnet gap) and damping/over-damping (bolt tension).

One thing to understand here is Tannoys of this era are complex hand made/assembled devices. The predate ultra-precise modern production techniques so do have a degree of imperfection. As an example you will almost certainly find a different number or combination of card spacing washers under the compression drivers left to right, even different types (e.g. brown are a different thickness to white). These were chosen at the factory to position the compression diaphragm at exactly the right hight from the pepperpot, so whatever you do don’t mix them up!

Another thing to consider/experiment with is how tightly you have the drive units bolted to the baffle. If you are using a traditional thin-wall Tannoy-style cabinet, i.e. with the drive units mounted from the back to a rather thin baffle over tightening them sounds absolutely horrible and does really mess up the treble/make them over-bright/forward. I have mine with plastic washers under the bolts on the back and just very gently nipped-up so they can’t move, little beyond finger-tight. My guess is this allows the thick rubber front gasket to add some damping to the baffle rather than the metal basket directly coupling and energising it, but whatever the reason it sounds hugely, hugely worse tightened up too hard. This has proven true both with my previous DIY York cabs and my current Lockwoods. Certainly play with this.

Interesting. The drivers are rear mounted and yes tight, so I will try that first. I was suspicious of the Mainly Tannoy crossovers as they are sold with same values for Gold and HPD, I know the values of the cap in the low pass section are different..
The crossovers are as on the Troels site, it’s worth a read and he does use measurements. I went for the most expensive option of the 3 kits. As I said I wired them for neutral balance. I could change but I know the issue is further down. Its obviously material dependent ie the excited frequency range is highlighted with recordings having instruments and voices in that band.
 
Interesting that the problem occurs with the Troels, the Mainlytannoy crossovers and on both drivers.
Do both xovers include the notch filter? Are you sure they aren't 'silvers' type with the LF running full open and minus the notch?
As both crossovers are aftermarket first thing I'd try is inverting the HF polarity and listen if this tames the shout.
Interesting that the problem occurs with the Troels, the Mainlytannoy crossovers and on both drivers.
Do both xovers include the notch filter? Are you sure they aren't 'silvers' type with the LF running full open and minus the notch?
As both crossovers are aftermarket first thing I'd try is inverting the HF polarity and listen if this tames the shout.
Yes they have the notch filter as originals. I was very careful with wiring. So certain it’s correct. It’s an upper midrange issue. The treble is very good. It’s a fairly simple crossover and the schematic for Troels redesign is very clear.
 
Yes they have the notch filter as originals. I was very careful with wiring. So certain it’s correct. It’s an upper midrange issue. The treble is very good. It’s a fairly simple crossover and the schematic for Troels redesign is very clear.
I've seen the drivers wired up incorrectly at the basket in the past. Another potential area to check is that all 4 spider nuts are done up and haven't worked loose.
 
I've seen the drivers wired up incorrectly at the basket in the past. Another potential area to check is that all 4 spider nuts are done up and haven't worked loose.
When I get the opportunity to do so I will take out drivers and dismantle. Odd that it’s such a narrow frequency on both drivers. Tony’s suggestion regarding tightness of bolts interesting. Maybe there is an resonance in basket transferring to cone ? Unless the red/black wires are soldered to HF units incorrectly I think the polarity is ok. Is there a simple way to set up measurement of drivers ? Thanks to all for replies- I’m determined to sort it as feel they are end game speakers
 
Looking at the eBay listing for a version of the Mainly Tannoy crossovers (link) it does appear that the level and roll off settings can be altered. If these crossovers in any way mimic the original ones I’d certainly try level -1 and roll-off -1 just to see if it is a simple level thing. Certainly every third party crossover I have heard to date has sounded too bright, and the advert for these actually claims it makes them ‘tighter and brighter’, which may well not be what you actually want.

PS The reason I use Tannoys is I love the way they sound, I don’t want them altered! (Mine are totally stock with level and roll-of in their ‘level’ position, though my amplification is a bit on the warm side, I’d likely alter the settings back a bit with modern solid state).
 
This site contains affiliate links for which pink fish media may be compensated.
Heres what Troels describes as the correct wiring shown in red.
49013897901_16fbe8f23d.jpg

But look at his final/finished graph,
49013899236_b3438c0e62.jpg

it's the same as the blue, ahem 'incorrect' wiring curve shown in the first..
 


advertisement


Back
Top