TANNOY 3128 speaker 1012 and HPD315A crossover. What's the difference?

[tonymorris]

First may I say this my 1st post on here (or any forum) so I don't know if this is the correct protocol, Sorry.

I have been asked to build a crossover for the TANNOY 3128 driver. Some say the 3128 has similar properties to the earlier HPD315A driver.
I notice the 3128 had a crossover designated 1012 when used in the Cheviot cabinet.

So, ... the question ... Can anyone tell me the difference between a 1012 crossover and the HPD315A crossover PLEEEESE?
Also when the 3128 was used in the Bradley a totally different and much simpler design was used. If anyone has info on this I would be interesred too.

Thanks to all who take time to answer. I bet Paul Coupe would know but I have no idea how to get in touch!!!

Cheers
Tony M

[Spectral Morn]

Hi Tony

Welcome to AoS.

I have moved this post from the Welcome section because it will get more attention here than there as its a question, need for help rather than a general introduction post.

I can't help you but I am sure there are folks here who can.

[RothwellAudio]

You'll find Paul's contact details here
http://www.referencefidelitycomponents.co.uk/contact/

[Reffc]

The 3128 used a time compensated crossover for the ABF type motor, typically, the time compensated 1020 crossover, as used in the Balmoral. It is not the same as the earlier AlNiCo HPD crossover which cannot be substituted. It requires time compensation due to the shorter magnet body and horn dimension meaning that, unlike earlier HPD and MG drive units, the horn phase response is not time aligned.

Tannoy adopted a version of a lattice all pass time delay circuit originally attributed to Harry F. Olson which was used to delay the HF response for the 3128. It was the same all pass lattice system that they used in many other of their later designs, including some of the K series and later drive units.

The problem is, that it doesn't integrate the HF well, and can be audible in circuit as a result.

Adopting the HPD crossover won't work because both the acoustic and impedance response of the later drive unit is not the same as the earlier HPD. You really need to use the 1020 time compensated crossover.

[cooky]

Something that has intrigued me over the years is the short horn K series Pro Monitors came with the all pass (sync-source) network on a separate little circuit board but the Prestige/Time Compensated series didn't include one at all. Perhaps better electronics minds than me can point to what Tannoy have done in the networks.
Below is the network for the 'Time Compensated' Mk1 RHR

[/url]

[Reffc]

Something that has intrigued me over the years is the short horn K series Pro Monitors came with the all pass (sync-source) network on a separate little circuit board but the Prestige/Time Compensated series didn't include one at all. Perhaps better electronics minds than me can point to what Tannoy have done in the networks.
Below is the network for the 'Time Compensated' Mk1 RHR

[/url]

The horn design changed a little since the RHR.

I understand that the reasons for dropping the TC circuit (not shown in the above schematic) was partly because it didn't work well (partially as the time delay may not have been accurate enough for the actual time domain difference at crossover) and secondly because, being cynical about it, costs were cut and crossover designs for what is quite a complex electro-acoustic design were simplified. Not because this made them any better (sadly, the reverse is true IME with the prestige line) but because it was cheaper to manufacture. The autoformers which used to be used to impedance match the nominally 10 Ohm (at DC) horn from around 18 Ohms at crossover to closer match the LF were also dropped for the prestige line plus earlier models. There are other ways though to balance a circuit without them (I use an HF compensated circuit for my own designs and shun the autoformer although it would be easy enough to get them made).

Tannoy, instead played with component quality, marketing in particular the use of VDH cables and "high quality" cost-no-object type poly capacitors. It is a shame then that they simply connected said "quality VDH" cables using cheap gold plated push fit spade connectors and plastic automotive-type connector blocks (to the drive units). It is just inconceivable with speakers costing what they do, that such cheap and nasty methods of connecting cables were used. Caps, despite the hype, are bog standard re-branded for Tannoy polys. It was a triumph of marketing over common sense good design. The accountants though were probably responsible for directing the engineers, who have to play to the budgetary fiddle.

[cooky]

I built some Arundels in 1982 using 3828 drivers, I recently found an old receipt for a pair of 'Arundel 'time compensated' 1019 crossovers' from Tannoy for £184/pr!-they didn't supply an all pass network even then.

They looked very like Arden/Berkeley etc crossovers and had the auto formers(and the Molex connector).
There's a pic of the network here;
http://www.canuckaudiomart.com/detai...toredupgrated/

[tonymorris]

I would like to offer my sincere thanks to everyone, particularly Paul C at RFC and also Cooky, of course for responding to my plea.

I have learned, over the years, to take the plethora of Tannoy Crossover diagrams on the Internet with heavy scepticism. So I don't use them. Being able to undertake real time analysis of electrical output under real speaker load conditions confirms my doubts.

There is much to be praised for Mr. Coupes' contributions to this forum. I thank him and appreciate his learned analyses. Thank you sir.

To conclude this discussion, my client has a set of the small Bradley with speakers with 3128 drivers but no crossovers???. I have passed your comments to him and reasserted the concerns as detailed here. I await a reply.

Thanks again all

[cooky]

As the reality is that the Prestige /Balmoral range didn't use any extra time compensation circuitry you could use the Cheviot 2 crossover(it used the 12" K series driver).

if you wish add the sync-source circuit to taste.

http://www.hilberink.nl/codehans/tannoy34.htm

[Reffc]

As the reality is that the Prestige /Balmoral range didn't use any extra time compensation circuitry you could use the Cheviot 2 crossover(it used the 12" K series driver).

if you wish add the sync-source circuit to taste.

http://www.hilberink.nl/codehans/tannoy34.htm

It would seem safe to assume that there were minimal differences between the 1012 X/O used in the Mk2 end-of-run Cheviots (themselves using the 3128) and the 1020 time compensated X/O used in later Balmoral/Bradly and Chertsey lines.

The time comp circuit in the latter series were separate from the X/O board itself, hence why it is sometimes assumed that the Balmoral for example, didn't use time compensation (it did, and Tannoy produced measurements for the effects on phase which were a published part of the specifications). There are always pairs that have been altered though and this circuit was often removed by owners. That's a different debate though as the answer, failing any other option, is to use the 1012 circuit if you have that Tony. You can build the TC circuit and have it fitted to the HF input via a switch (ie switch in or out).

Off topic, I know, but seeing as Tony has his answer, it's interesting to mull over some development aspects of the DC drivers in general:

Tannoy marketed the newer ABF motor as "superior" to the earlier HPDs but it wasn't quite as well received in some quarters. The whole thing needed refinement, but as previously mentioned, the TC circuit was dropped in later Prestige lines, which bizarrely used the superior Tulip Waveguide only in the entry level models and the pepperpot in the more expensive ones (a nod to the fact that Tannoy were targeting the nostalgia buyer or tannoyista fetishist for sales rather than pushing what was technically superior!). I had this conversation with their Technical Director a few years ago, and he admitted that the main market (Far East) were less inclined to run with technical advances in DC design, such as the Tulip Waveguide and more efficient motors, and more inclined to buy a newer nostalgia model in the belief that it was superior, hence the DC was never really finally developed the way it should have been. The irony here is that the very fans that want it, only want the older designs in the belief that they're all superior, which is not the case. They exhibit higher HF distortion as the HF pepperpot designs break up resonance occurs lower down whilst the TWG can extend cleanly to 30KHz and is smoother within the audible band. Sadly, marketing drives what is developed, not common sense or best technical solution.