My Devon Dilemma - the boxes are too small

[Pete Alcock]

Newby here, so not sure of all the protocols yet, but here goes...

I love most of the sound of my Tannoy Devons, but because of the small-ish cabinet, the tuning of it seems to coincide with the lounge's room resonances, so room-boom is prominent rather that true low bass. The problem is, to correct this with the lovely HPD315 drivers I have, I'd need to build boxes as big, if not larger than Cheviots, and that's just not going to be domestically acceptable, if you know what I mean.

Anyway, my first dual-concentrics were a pair of Dorsets which I still have in storage (slightly larger boxes, 10" D-C with an ABR), and I was thinking that with a 10" driver, I could again make larger cabinets tuned low, in order to gain really satisfying bass, but they wouldn't need to be as big as an optimum cabinet for a 12.

However, I don't like the sound of the plastic coned, ferrite magnetted 2528 drivers from the Dorsets as much as the HPDs - somehow just less tingly and delicate. Pursuing this route, I thought I could have them re-coned with paper 2558 cones, build some larger cabinets and end up with something similar to Caernarvons. Only thing with that is that 2558 cones are as rare as hen's teeth, and having this done by Lockwood will cost £285, which is a lot for what is essentially an experiment.

Just wondered if anybody had any suggestions, possibly around improving the 2528 drivers' sound through crossover tweaks? Many thanks. Pete.

[dantheman91]

Welcome Pete.....

We can headbang together im also in Bournemouth :lol:

[spendorman]

My first Tannoy DC's were 12" Monitor Silvers in Chatsworth corner enclosures. I got these in the early 70's. Looked old fashioned to me, so I made some rectangular Chatsworth enclosures, not much difference in sound, but they looked more modern. I happened to have some Mordaunt Short MS 700 cabinets, these were quite a bit larger that either Chatsworth enclosures, put the 12" Silver DCs in these, and the resulting sound was really excellent.

About 5 years ago I managed to buy a pair of Devons. The cabinets were rough, the surrounds had been replaced quite well apart from not looking 100% cosmetically due to a little bit of excessive glue on the cones. I noticed some distortion in the treble of one unit, re-centred the tweeter voice coil and all was well.

I thought the bass a bit light, also slightly boomy. Adding a little extra damping inside the cabinets helped in tightening up the bass, but it's still a little light.

The original Chatsworth corner enclosures are still here. It would be nice to try the HPD315s in these larger cabinets, although a bit larger that these would probably be even better.

[walpurgis]

You can alter the port tuning on the Devon quite easily, to raise or lower the system resonance.

The 2528 driver will accept the SRM 10B paper cone. This is a conversion I'm planning, as I have a pair of Ascots knocking about. The drive units will hopefully end up in transmission line cabinets.

I don't see why a Cheviot sized box would be unacceptable for the HPD 315 units. Surely this would take up no more room than Devons on stands?

I assume you use stands? If not, that would most likely be the cause of your room matching issues.

[Pete Alcock]

I don't see why a Cheviot sized box would be unacceptable for the HPD 315 units. Surely this would take up no more room than Devons on stands?


Thanks Geoff. Yes, my Devons are on stands, but although the same depth, Cheviots are 2" wider. That's mainly what would make them "too imposing" within the lounge. I'll have a look at modelling what would happen if I changed the port length. I suspect I'm at the edge of the envelope as far as trying to gain more LF extension is concerned though, due to the small cabinet volume.

For your conversion (2528 to 2558) do you have paper cones already? If so where can I get some?

Cheers.

[spendorman]

The Chatsworth cabinet is not very wide.

Have you tried blocking up the port completely?

[Pete Alcock]

Hi Alex, thanks. I've not tried blocking the port up, but I suppose it would be easy to try. An infinite baffle would theoretically give a faster bass roll-off, but more ultimate LF extension...

I've just googled Chatsworth and they're the same width as Devons, so that's a possibility too.

[spendorman]

Actually, I think the infinite baffle will give a slower roll off, will go down lower. But less output at upper bass frequencies.

[walpurgis]

For your conversion (2528 to 2558) do you have paper cones already? If so where can I get some?

Cheers.

I have some. They were an ebay purchase. It took me quite a while to find a pair though.

Mine are mint, but second hand. They can be bought on ebay, at a price! You could contact the seller and ask if any used examples are available.

[walpurgis]

Actually, I think the infinite baffle will give a slower roll off, will go down lower. But less output at upper bass frequencies.

That's quite likely.

[spendorman]

Another thought, at the other end of the frequency scale, treble, the Tannoys are merciless in revealing roughness in the top end. A lot of class D amps are optimised for 4 Ohm speakers, that is a flat frequency response in to a 4 Ohm load. If used with a higher impedance load, top end treble is likely to be emphasised.

In this case the output filter capacitors should be changed. I sometimes run a Class D TPA3116, it's pretty good, but this treble emphasis can be heard. I should really change the filter capacitors as mentioned.

I generally run Rogers BBC LS3/6, which have a tapped auto transformer on their input. This would allow me to change the input impedance to something lower and around 4 Ohms if I wanted. I have not tried this as yet. My favourite amp with these speakers in a Radford STA25 III.

[danilo]

:eyebrows: Yup.. tannoys ARE merciless on low quality upstream gear.
Also respond marvelously to even the tiniest quality improvements.. A double edged sword.
I tried an Ta2020 'amp 6' for a short while.. before it died it proved a decent gizmo, for a wee chip device.
But tripath chips are Extinct these days
LM3886 chip amp device can be actually a pleasingly good amp.
IF of first world design and parts ;)
MY 3116 chip amp (bought because of much internet babble praise) proved Horrid to my ears.
Not surprised you are displeased with it
Screechy with ugly speaker threatening extraneous on off noises
Another Chi-Fi Ebay item..Immediately for the trashbin.

[spendorman]

TPA 3116 amps are very much variable depending on how they were built, mine was built by a class D expert (not me), and is totally silent, no switch on / off thump, no other extraneous noises, just a very slight upper treble lift as mentioned, because it's used in to higher impedance speakers than it was designed for. I am far from displeased with it. At some point I may change the output filer capacitors to suit higher impedance speakers.

I think Geoff heard my TPA3116 in to my home built LS3/5a clones. I don't think he complained too much.

[walpurgis]

I think Geoff heard my TPA3116 in to my home built LS3/5a clones. I don't think he complained too much.

Yes. I thought it sounded very good. Actually, it was the best I've heard LS3/5a type speakers.

[walpurgis]

:eyebrows: Yup.. tannoys ARE merciless on low quality upstream gear.
Also respond marvelously to even the tiniest quality improvements.. A double edged sword.

That is very true.

[Macca]

Actually, I think the infinite baffle will give a slower roll off, will go down lower. But less output at upper bass frequencies.

Depth of bass is dependant on the overall design of the speaker, If it is ported in a way designed to augment bass then FR will roll off steeply below the tuning point but that point could be very low indeed. Before that there should be no roll off so bass would be plus or minus 3db to the lowest frequency the speaker is able to produce. If a sealed cab it will roll off slowly but gradually reducing in level.

I think the problem here is the speaker is able to reproduce a bass frequency that excites the room, If they were producing that frequency at a low level this would not happen. I suggest that the first step is to find out at exactly what frequency this is happening at and see if it cannot be cured by simply recalculating port size and length so that the response rolls off steeply before it reaches this frequency. Changing the port size would seem the easiest fix. And easily reversible if you move to a larger room or sell them.

However if the offending frequency is quite high, and it may be, it largely depends on the room volume, this may be impossible; or possible, but make the speaker sound very bass shy.

A room calculator would be a start. Input the room dimensions and see what frequencies excite the room nodes. That should give a rough idea where the problem lies.

You could also potentially solve it with bass traps but they would be large and from what you say, domestically unacceptable.

[spendorman]

Yes. I thought it sounded very good. Actually, it was the best I've heard LS3/5a type speakers.

Praise indeed, thank you. I built those LS3/5a clones about 1976. They sound quite similar to my Chartwell 15 Ohm LS3/5a's, but with slightly more extended and firmer bass, due to marginally larger/ more rigid cabinet with slightly more damping.

It also shows that TPA3116 amps can be pretty good. Although, I'm not sure that I would build one, even though I've had a good part of my working life dealing with electronics.

[Reffc]

A few points of note:

1. It is unwise to consider varying alterations to a properly designed port in a vented enclosure. The port is calculated to work for a specific enclosure size, so altering the point of resonance will have a marked effect on system efficiency can can lead to damage to both driver and driving amplifier if got wrong. Better to plug it completely and see how that goes.

2. Tannoy didn't always get their port dimensions correct and the Cheviots used a port length resulting in EBS tuning, to a point way too low to be of any use at all, and allowing excessive cone excursions, especially with vinyl replay.


D... I suggest that the first step is to find out at exactly what frequency this is happening at... +1

however, it is not generally recommended that a port is retuned to try and solve the problem unless it is wrong to start with. The port is designed to work in conjunction with the enclosure's specific volume and the driver specific parameters. It is not a movable feast without consequence.


I would start by assessing (as Martin points out) room response at the listening position. Beg, borrow or steal a calibrated mic and run a set of pink noise bursts and assess room response from listening position. Failing this, just use your ears, and try moving speakers to reduce standing waves. It is important not to have them positioned so that they are equi-distant from rear or side walls as this will double output at certain frequencies with standing wave reinforcement.

12 inch drive units won't have any real disadvantages over 10 inch. Both will result in the same room excitement of the 40-60Hz room modes when the levels are matched.

The cabinets are no doubt adding to colouration. Most vintage Tannoy factory cabs benefit from internal damping and bracing. A shaped or corner style narrow unit, made taller to get the correct internal volume, would work well because such shapes are inherently stiffer than regular rectangular boxes as well as less prone to internal standing wave formation.

Chatsworth style cabs do have a little lift in the 40-60 Hz range partially due to internal standing waves, partly due to cabinet resonance and partly due to volume and how that works with the drive unit T&S parameters. As you can't do too much about the volume in this case, any improvements are limited to minimising internal waves, stiffening the cabinets and increasing mass to sink and damp resonances. To do this effectively, even where you may not wish to add a lot of panel mass, you need to break up the resonant frequency of the panels such that no two sub panels share the same harmonic frequencies. This means avoiding equi-spaced internal bracing. Use something like sound-clear internal damping poly-elastomer damping compounds for car panels (available from Ebay) to damp internal surfaces. Add a sloped base panel made from MDF set onto sloped battons and fill the annulus with sand or grout. The slope only needs to be 5 or 10 degrees. This will aid the avoidance of standing waves long the length of the cabinet. Finally, rip out the old mattress foam and use higher efficiency acoustic wedges to rear/top/base and sides.

If plugging ports, then system "Q" needs working out to maintain a Q of .707 for a Butterworth response (B2), and if this proves too boomy, the stuffing can be altered (lowered) to provide a gradually lower Q to a point of say 0.5 which is usually considered over-damped (although in theory it is critically damped) but in your case, such tuning could lower the bass output in the problem area enough to make it worthwhile experimenting with this. This of course assumes a closed box (plugged port and use a substantial air tight plug). You can approximate a Q of around .71 with most cabinets by stuffing the boxes with damping compounds until about a 70 to 75% by volume is achieved as a rule of thumb (not with vented speakers). This ideally needs measured though using something like DATS (dayton audio test system) which will provide a value for Q when measuring impedance with frequency response. Then work back, removing a little at a time until the response matches the best compromise for your circumstances.

Of course, it is better to design such things accounting for the correct internal volumes, and not always possible to effect all such improvements in smaller cabinets. The 315 requires a closed cabinet of some 75 to 100 litres or a vented cabinet from 100 to 155 litres net volume, tuned accordingly. Correct tuning if using a vented design is critical, not variable by choice without consequences. There's an article which may help explain things linked HERE
(http://www.referencefidelitycomponents.co.uk/design-and-performance-of-ported-loudspeaker-enclosures/)

There is no unforgiving HF response if the crossovers are properly implemented or the originals set at -1 (energy) and flat (rolloff) which is the true flat electrical setting (albeit, there is still a slight voltage lift from the autoformer). They can be very revealing but only if every other area is got right to provide that clarity. Some people mistakenly think that the revealing nature comes from using a lifted response but this is just boosting HF response to provide the detail and is painful ion the ears from the midrange upwards! The clarity comes from a balance of sorted cabinet design and sorted crossover. Then they can deliver in spades without any lift in response.

Happy to help further and best way might be to call for a chat and take it from there. You can get 315s to work in surprisingly small spaces but it is all down to getting the cabinet right.

There is one final trick that you can play once the port dimensions are checked and proved to be correct: Introduce a resistive plug of dense open-cell foam or similar of about an inch thick into the port. This modulates the speaker’s impedance characteristics, reducing and smoothing bass output as it creates a semi-aperiodic loading. Part-stuffing the ports or using bunches of straws as was popularised some years ago is NOT recommended.

[jandl100]

If you have a ported speaker and room-boom, then it is definitely worth experimenting with stuffing the port with material. I use speaker wadding in mine and the amount of wadding used can be tuned to quite significantly vary the bass output.

I actually started experimenting using socks, and that encouraged me to go further with the idea!

I see that Paul is against the idea ...


Part-stuffing the ports or using bunches of straws as was popularised some years ago is NOT recommended.

... but I find it can work surprisingly well!
I'm a let's-just-try-it empiricist and Paul likes to get the techie stuff right! :)

[Reffc]

If you have a ported speaker and room-boom, then it is definitely worth experimenting with stuffing the port with material. I use speaker wadding in mine and the amount of wadding used can be tuned to quite significantly vary the bass output.

I actually started experimenting using socks, and that encouraged me to go further with the idea!

I see that Paul is against the idea ...



... but I find it can work surprisingly well!
I'm a let's-just-try-it empiricist and Paul likes to get the techie stuff right! :)

I'm not against stuffing a port Jerry but part stuffing is definitely a BAD idea for the health of the loudspeaker and possibly even the driving amplifier. I am not against empirical trials but I'd like to think that an intelligent chap like yourself would appreciate that I speak from a standpoint of hard earned knowledge and experience, not from guessing! ;). I've given the reasoning and if you actually read the link that was posted, you'll perhaps appreciate why I gave it. It's a basic fact that tuning ports too high or low diminishes system efficiency and can cause damage. I can only give advice though and if someone chooses to ignore it, at least others reading it who might wish to learn more about it have the information :). It's a science, not a guessing game where ported enclosures are concerned. As you suggest, using different materials but COMPLETELY and NOT partly stuffing the port is worth the experimentation but if one is to do that, then to get the best effect, it's often found by either closing the port completely with a hard plug OR creating an aperiodic loading whilst maintaining the port, so a dense foam about an inch thick will be better than using a sock stuffed willy-nilly into a port and at least means that you'll have something left to wear under your shoes...

How a ported enclosure loads a room is dependent upon where the ports are positioned, their cross sectional area and the tuned frequency. Often, two different but same volume enclosures tune to the same frequency but one using ports located rear instead of front, or one using choked port tuning and too small a cross section, can ave appreciably different effects on the room response. Rear loading can lead to more boom problems for example.

[spikeyfaz]

There is no unforgiving HF response if the crossovers are properly implemented or the originals set at -1 (energy) and flat (rolloff) which is the true flat electrical setting (albeit, there is still a slight voltage lift from the autoformer). They can be very revealing but only if every other area is got right to provide that clarity. Some people mistakenly think that the revealing nature comes from using a lifted response but this is just boosting HF response to provide the detail and is painful ion the ears from the midrange upwards! The clarity comes from a balance of sorted cabinet design and sorted crossover. Then they can deliver in spades without any lift in response.





I've just re-set my own Devon energy and roll off controls as you've suggested and I now have speakers I can enjoy! Cheers for posting your advice. I'd always assumed the controls should have both been set at the neutral positions. No wonder I couldn't live with the treble response. Only taken me several years to suss this! Better late than never I suppose.

I was going to go down the valve amp route (hoping that would tame the treble) but no need now. However, my trusty old NAD 3130 is a bit past its sell by date so wondering what might constitute a good replacement.

Thanks again Paul (Reffc) for posting your advice here.

Mike

[spikeyfaz]

I've also got a spare (working) set of crossovers for the Devons. Is it worth getting these rebuilt to see if they improve the sound?

[spendorman]

3130 used as preamp

https://c7.staticflickr.com/6/5230/5687938702_b4e5b621d5_b.jpg (https://flic.kr/p/9ECb1E)NADford STA25 III (https://flic.kr/p/9ECb1E) by A60man (https://www.flickr.com/photos/41679262@N02/), on Flickr

https://c5.staticflickr.com/6/5027/5688335580_766e2e80bb_b.jpg (https://flic.kr/p/9EEcZo)NAD 303 (https://flic.kr/p/9EEcZo) by A60man (https://www.flickr.com/photos/41679262@N02/), on Flickr

https://c1.staticflickr.com/6/5244/5688134920_f2c8e3a769_b.jpg (https://flic.kr/p/9EDbkJ)QuadNAD (https://flic.kr/p/9EDbkJ) by A60man (https://www.flickr.com/photos/41679262@N02/), on Flickr

[spikeyfaz]

Hi Alex

I have thought about using the NAD as a pre, but probably not enough room for an additional amp. I have however got a little nuforce icon amp in the loft which I might try. It's a bit basic and needs a case but could work. I'd discounted it prior to discovering the correct speaker settings, so might be a winner!

[danilo]

I've also got a spare (working) set of crossovers for the Devons. Is it worth getting these rebuilt to see if they improve the sound?

IF original the electro caps are likely 'off'. Replace them.
This for Golds.. but HPDs are basically differing only in the Crossover packaging; http://www.44bx.com/tannoy/goldcrossovermods1.html

Replacing more bits will only thin your wallet... But Hey! it's Your wallet ;)

[spendorman]

I have not had my Devon HPD315 crossovers out, but was under the impression that they had non electrolytic capacitors. Is this so?

[Arkless Electronics]

A quick google reveals that these are "proper" Tannoys.... A mate briefly had some about that size with 10" plastic cones, still dual concentric though, I can't remember the model name but they were bloody awful! Talk about coloured...:eek:

[spendorman]

Devon's are certainly real Tannoys, HPD315 12" driver, early ones are basically development of the 12" Monitor Golds.

[danilo]

Actually the Harmon International 'Revision' of the Golds' design. The last Golds production run,
of 12's, were first/last run to receive the Lovely Butyl surrounds rather than the Crap foam ones Harmon immediately decreed...
throughout ALL of their recently purchased Speaker brands.
Last Golds V similar to what was relabelled as HPDs. Similar T & S #'s Similar weight cones even ;)
sans the added bits of folded paper support.
Royals (blue painted Cdn Market HPds) were the direct result of the (one time only) sale of 3000 pairs of Golds to the CBC.
Harmon/Tannoy guessed they were into a whole new market... Surprise :-)
Development is pure Brochure Babble... Devolution is more apt.

Small need to mention the widespread employee morale destruction the Takeover generated..resulting in HPDs becoming The worst built/assembled Tannoys on record.
All this history from a friend who spent decades with the Ontario based distributor of Tannoy in North America.
Apparently it wasn't a Happy time "Chez Tannoy". Certainly Not the official story either

Currently it's looking as a repeat performance..

Yes a pointless digression.. Bored with the Pollyanna Infomercial nonsense.

[walpurgis]

A quick google reveals that these are "proper" Tannoys.... A mate briefly had some about that size with 10" plastic cones, still dual concentric though, I can't remember the model name but they were bloody awful! Talk about coloured...:eek:

They may have been Ascots or another model from that particular range.

The plastic cone does not sound too bad actually, the 'tone' is unfamiliar to those used to regular pulp coned Tannoys, but the level of 'plastic cone sound' is moderate.

I have a pair of Ascots, they use a 10" version of the later ceramic magnet pepperpot dual concentric. The classic Tannoy imaging and mid/treble character is all there. They certainly sound far superior to early tulip waveguide types like the 9" driver DC1000 and 10" driver DC200, both with rather obvious plastic cone colourations. The DC200 is the worst sounding Tannoy dual I've heard.