New "The SOFF Loudspeaker" public demonstration ASoN 29th August...

[Joe Rasmussen]

Hi Guys

Hope some people won't get their nose out of joint by posting this thread. I just want to point out this is about DIY rather than some commercial interest. I am commercially involved with Hi-Fi, but unlike most such concerns I also have a strong interest in DIY and in experimentation, especially when it's about pushing the boundaries.

And that is what The SOFF Loudspeaker is about. It will be demonstrated for the first this Sunday 29th of August 2010 at the monthly ASoN meeting held a 78 Dalhousie Street, Haberfield NSW, 2.45PM. Parking is not hard to find and I am sure that speaking as a member of ASoN (Audiophile Society of NSW) I can assure you that all non-member visitors are always welcome.

The basis of this development follows years of thoughtful introspection. The speaker in mind was an attempt at producing a virtually pure resistive load on the amplifier and to work best with non-Voltage amplifiers. Basically all solid state amplifiers are Voltage and low output impedance. Most modern tube amplifiers are also sub one Ohm output impedance that makes them Voltage amplifiers. But classic pre-50's tube amplifiers and most if not alll Single-Ended Triode Amplifiers are not Voltage amps. Zero feedback Push-Pull tube amplifiers are also non-Voltage like my JLTi amps. These have limitations driving speakers with non-linear impedances. This make it possible for speaker designers as a whole to get away with making speaker with non-linear impedance curves. The SOFF speaker takes a very different approach in deliberately aiming at a loudspeaker with a linear impedance and electric phase load, from 20 Hertz to 20KHz.

The ASoN article in their monthly Newsletter was published earlier today to members. It has been reproduced on my website here.

http://customanalogue.com/the_soff_speaker.htm

I hope you enjoy reading it and also perhaps have your thoughts on what is a topic that has been an obsession of sorts for many years. The gratification I've gotten from the achievement in finally making this speaker possible is a really nice feeling.

Cheers, Joe

PS: Map of the ASoN location:

http://customanalogue.com

Edited August 23, 2010 by Joe Rasmussen

[Once was an audiophile]

this is intresting joe i will be following this.

your always playing around with something mr rasmussen

[planet10]

Very interested in how this turns out Joe. I too am working on a current drive speaker.

Those B200 can be made to sound quite a bit better -- we were listening to such a pair at this weekwnds diyFEST.

dave

[Antripodean]

Hi Joe

Congratulations! I look forward to hearing an audition in Melbourne I can't wait to hear them; especially compared with the Elsinores

I was wondering if this application can be applied to other drivers and if so, what type of driver is appropriate?

[Joe Rasmussen]

Very interested in how this turns out Joe. I too am working on a current drive speaker.

Hi Dave

In which case you must be working with conjugal filter/elements as well. Assume you are familiar with Nelson Pass' article I linked to.

I believe it is important to view a flat sub 100 Hertz impedance for optimum local bass damping and a flat impedance elsewhere to give the driver a flat response . An option is also to shape the impedance plot to compensate for diffraction loss, something many full-range users don't (Visaton as you may have noted does recommend compensation), by shaping the impedance. Since you are using current source/drive you can go as low impedance as you want. Many will not quite understand that you can drive a current amp into a zero impedance and not damage the amp. So you can aim at any impedance required to do the job. Just don't use the speaker with a Voltage amp.

Those B200 can be made to sound quite a bit better -- we were listening to such a pair at this weekends diyFEST.

dave

Have little doubt that is the case. But I'm actually stoked by the results. But using it 2-way with a waveguide tweeter I am definitely getting a result I like. But I suspect you wouldn't use it this way.

Peculiarly I have another driver here that has the same cut-out requirement and the same six screws exactly match as well, so screw out and screw other driver in. Sorry, but the short-coming I do have here is that the B200 is dynamically limited at the bottom end, using my Elsinores as a comparison, albeit an unfair one. But the quality of the bass is sublime in its own right within its comfortable power envelope. But the alternative driver I have in mind will be much less limited in that respect.

Best thing, I enjoy the challenge and this project has legs to run on. So will see where the next step will take it. With the new driver I have to do all the measurements again and computer model as before and see if I can get that same modelled linear flat impedance, only then build the passive network, measure that to see that it matches what modelling predicted. Finally listen to it. But the double cone mass maybe it won't cut the mustard and be as immediate like the B200. But who knows, it will go louder - so must try.

Cheers, Joe

[Joe Rasmussen]

Hi Joe

Congratulations! I look forward to hearing an audition in Melbourne I can't wait to hear them; especially compared with the Elsinores

I was wondering if this application can be applied to other drivers and if so, what type of driver is appropriate?

I have already made some comments re alternative driver to Dave. But it is not easy to answer.

Re Melbourne, I am simply available and would love to go, but the good people at Melbourne Audio Club must be the ones to have the final say if I am welcome or if the have an opening. Otherwise will be late Q1 2011 that could be arranged?

Cheers, Joe

[Guest Muon]

That was very interesting read in the OP, so may I ask Joe, as you mention in the OP "Most modern tube amplifiers are also sub one Ohm output impedance that makes them Voltage amplifiers." so with the Yaqin before or after mods would that amp be considered a Voltage amp or more a Current amp, given that the circuit is of an old design as I understand it?

Sorry if this a bit of newbie question, but my brain has be hurting of late trying to understand different circuits

Cheers

[Joe Rasmussen]

That was very interesting read in the OP, so may I ask Joe, as you mention in the OP "Most modern tube amplifiers are also sub one Ohm output impedance that makes them Voltage amplifiers." so with the Yaqin before or after mods would that amp be considered a Voltage amp or more a Current amp, given that the circuit is of an old design as I understand it?

Fair enough question.

Basically the standard Yaqin is based on 50's technology, Ultra-Linear, Pentode, feedback etc. While I did not measure the output Z before or after, I know from experience that before it is essentially a Voltage amp. One of the primary clues to a Voltage amp is global negative feedback, in particular from the amp's output terminals. Negative feedback forces an amp to behave like a Voltage amp because it forces the output Z to become low.

But after, as a JLTi EL34 amp, it becomes a totally different beast. No feedback does the opposite and keep the Z high, and as it will be typically in the 3-6 Ohm range, it will be a Unity Coupling amp.

Does a Unity Coupling amp behave more like a Current than a Voltage amp? In my view yes. The reason is that Unity Coupling so-called damping factor is ONE - unity. That just about means you have reduced damping factor to an insignificant low level. It will behave more along the line of a Current model.

Cheers, Joe

[Guest Muon]

Excellent explanation, thanks, Joe

[Zaphod Beeblebrox]

That was very interesting read in the OP, so may I ask Joe, as you mention in the OP "Most modern tube amplifiers are also sub one Ohm output impedance that makes them Voltage amplifiers." so with the Yaqin before or after mods would that amp be considered a Voltage amp or more a Current amp, given that the circuit is of an old design as I understand it?

Sorry if this a bit of newbie question, but my brain has be hurting of late trying to understand different circuits

Cheers

As a general rule, there are several things that dictate the output impedance of a valve amp. These are (in approximate order of importance):

* The type of output valve. Triodes exhibit significantly lower output impedance figures than Pentodes and Tetrodes.

* The number of output valves. Push pull topology will exhibit lower output impedance than single ended. Parallel push pull will exhibit even lower output impedance.

* The prescence of global NFB. Global NFB will reduce the output impedance figures over a zero global NFB or low global NFB amp.

* The quality of the output transformer. High grade, multi-interleaved output transformers (like the ones Earl builds) will exhibit lower output impedance figures than inferior transformers, typically found in cheap, Chinese amps.

Naturally, there will be variations on these truisms, but they provide a close idea of what you can expect. Very few manuafacturers 'try' to build current amplifiers, as 99.99999999% of all loudspeakers are designed to be operated with a Voltage source, not a current source. Here in the real world, all amplifiers lie somewhere in between a pure Voltage source and a current source. Obviously, the tendency is for those amps to exhibit a Voltage source type of operation, as this allows correct operation with most loudspeakers.

An example of a low output impedance valve amp, might be:

Parallel push pull, Triode amp, using a high grade multi-interleaved output transformer, using lots of global NFB.

I've not been able to locate any data on Yaqin amps (they appear to be strangely silent on the important data), but I believe Earl may have some measurements for them. You should ask him.

[Guest Muon]

Thanks for expanding on that point ZB

Just trying to understand so as to understand these new speakers in the OP for the most part, and It's a bonus to find out about a bit of kit I'm using atm.

[Audiobugged]

Couldn't make this though it did sound most fascinating. Would love to hear any thoughts on this speaker?

AB

[k-k-k-kenny]

So? How did they sound?

[theophile]

The model meant for placement with it's back to a wall could be named the 2 Pi SOFF.

[Joe Rasmussen]

So? How did they sound?

I will try to honestly explain the pros and also the cons.

First of all, the provisional SOFF nomenclature has been changed to The Linear Current Loudspeakers - or TLCL if you like.

The usual loudspeakers most of us use, it is the current that causes the speaker's Voice Coil to move. It makes inherent sense to use current. Only the edge of the Voice Coil sees the voltage, but this voltage is not in the magnetic gap. Send voltage down a line (a conductor if you will) and that voltage will be divided up along its path by various elements until it reaches zero. Send current down the same line and the current stays the same all the way through. That means that the very force that activates the speaker is constant. Voltage is never a constant.

So we would expect superior linearity from a current driven loudspeaker. So why do we not see it done regularly? Because the amplifier and speaker needs to be matched in a serious way and not in the lazy way that the voltage road allows. The marriage is more difficult but also worth it. For example, in so-called active speakers where the amp is inside there is a real opportunity to do it - and yet they don't. Here the designer has the chance to do both amp and speaker, and yet the opportunity is lost. Some enterprising manufacturer should realise this, what a great opportunity.

Noiw back to TLCL and how it sounds.

While the presentation at ASoN went well, it is after all a hall. But it managed quite well and we set up the equipment and the speakers is a different location that we normally do. I think that worked for the better.

But the test is really what it sounds like in a conventional room for which it was intended.

In the last month I have had a number of people turn up for a listen and it has been extremely well received. In fact nothing short of amazing. The speaker has an extraordinary expansive midrange and the treble has very low, and I mean audibly low, distortion. The waveguide gives you the option to trade for greater bandwidth OR lower distortion if you don't lower the crossover frequency. I have gone for the latter option and it really does pay off. Never been a great fan of dome tweeters and you have to go to extreme lengths of damping them in the crossover using a null technique that means it goes very high order before it goes below 1KHz (now also used by other designers). This is the best I have heard a dome tweeter sound and right up there with anything I've heard.

The expansive midrange coming up to meet the waveguide tweeter makes the speaker wonderfully clear and fluid and yet easy on the ear despite the very high level of detail. I suspect the speaker will not suffer fools with lower quality ancilliaries - which I haven't desired to try.

The bass is problematic, not that it doesn't match the rest of the speaker. It does. The Visaton driver is on the delicate side (compared to the much more rugged Peerless HDS drivers I've been using) and below 50 Hertz they run out of steam if pushed just a bit too hard. But low bass is reproduced up to that limitation. The bass has no Q like normal speakers do - so it is the localised damping that does the job instead. But the bass quality matches the rest of the speaker - in the sense of clarity is maintained.

They are not capable of the power of the Elsinore speakers (rarely does a DIY speaker match those). But if you are a lover of vocals this is a speaker you will love. Also jazz, small scale acoustic works like guitars and small ensemble string and or wind instruments. It does better that the usual small two-way speakers when it comes to more powerful music (except don't push low bass to hard), but with a full scale orchestral sound coming up to a climax, they cannot do what the Elsinores do. They can produce the scale required - but most speakers cannot do that anyway.

The new TLCL, I think it has shown the way that I need to continue in the future. I am developing a waveguide, with the help of a friend to help make them, for the Elsinores. Other things will also be looked at.

But it does also come down, and to some this wil be one of the cons, that it should not be used with Voltage amplifiers if you want to get the most out of them. They will work with Voltage amps (think Solid State mostly) but will not give their best. Other speakers out there designed for current operation will not work at all with Voltage amps as they don't have the ultra-flat impedance that the TLCL does.

The best description I have had was one listener who in the end tried to find one singular word to describe what he heard. He came up with holographic.

I'll take that.

Call me on 02-96074650 or 0412-203382 if anybody wants to come around for a listen in the South-West of Sydney where the M5 and M7 meet.

Cheers, Joe

PS: For anybody reading this who is not familiar with the Elsinores: www.customanalogue.com/elsinore/elsinore_index.htm

PPS: Website that discusses using current to drive speakers: www.current-drive.info

[planet10]

The usual loudspeakers most of us use, it is the current that causes the speaker's Voice Coil to move. It makes inherent sense to use current. Only the edge of the Voice Coil sees the voltage, but this voltage is not in the magnetic gap. Send voltage down a line (a conductor if you will) and that voltage will be divided up along its path by various elements until it reaches zero. Send current down the same line and the current stays the same all the way through. That means that the very force that activates the speaker is constant. Voltage is never a constant.

http://www.diyaudio.com/forums/vendors-bazaar/157787-secret-tube-amplifiers-revealed-much-more.html

dave

[Joe Rasmussen]

http://www.diyaudio.com/forums/vendors-bazaar/157787-secret-tube-amplifiers-revealed-much-more.html

dave

http://www.diyaudio.com/forums/vendo...much-more.html

Yeah, Dave. For what is a vendor 'announcement' it has certainly hotted up. Esa may be OK preaching to the converted, but he is potentially alianating the very ones he (and we) needs to convert.

I don't agree with the hybrid model that he proposes, so I am even more Left (or is it Right) field that he is? I may chip in with a comment there about that - and why I favour an all out current model - and no voltage drive below 100 Hertz being necessary.

But there is more to it...

For example, one fellow designer here in Sydney, and a good friend, has a commercial design (on sale here) that has linearised the currrent below 100 Hertz (single peak near 20 Hertz, but the design does not extend that low and little output there) even knowing that the speaker is going to be used with mostly SS. By doing this, making the current linear, then even the benefits arrive with voltage delivery amps. So under those circumstances voltage amps will also linearise the current - and hence similar bebefits. So there may be common ground; make speakers linear current and you will get benefits with all kinds of amps, not just current amps. BTW, his design approach also has superior group delay at LF.

Anything that can stabilise the current load will always lead to more linear drive whether or not the amp is current or voltage? That is how I see it. Make the speaker look resistive!!!!

This puts the speaker designer in charge rather than the amp designer? It's a thought that needs more thinking about!

Amazon has sent me an email, so the book is on its way.

Cheers, Joe

[planet10]

Anything that can stabilise the current load will always lead to more linear drive whether or not the amp is current or voltage? That is how I see it. Make the speaker look resistive!!!!

As i've been noodling my current speaker i have been having the same thots.

dave