Romy the Cat wrote:

The brickish finish does not make me cringe, in fact I do like the idea. I do not feel that it would be some kind of non-truths. There is no truths or non-truths in it. The only a few first feet of the horn will be visible and then it will go to dark anyhow.  My view of the brickish finish is combined with my view of how I might finish the ender back wall. I like brick finish and I absolutely adore it as sound detractor surface behind the listening sit. So, my long–terms décor objective are to have the mouth of the horn to work beneficial and integrateivly in the room. So, I was thinking to finish the back wall right behind the listening chair with half-brick. Then to finish the rest of the wall with something like Eldorado Brick Veneer that has .5” death. Then the hors internals with brick wallpaper. If to arrange to match the brick, Veneer and wallpaper of the same color then it might look very good, considering that I have also in the room some brick islands. I did not decide it finally yet but I know that if I want to lay vinyl wallpaper in the horns then I need to do it while the horn on the ground….

Both mouths are ready and glued and this is very good sign.  The mouths are largest and the most complex pieces that hold 2 cords – it will be one section. I am talking about mouths being ready but they look like nothing. Sure, they are  but they are VERY far from beeing truely ready – they are just the plywood skeletons. However, they already have shapes and now we need to juts add to the shapes the frames, jointing harnesses, second layer of plywood or damper and so on…

This is very interesting subject.

I expect that tghe new Midbass horns will have somewhere 108dB sensitively anechoicly.  The driver is 99dB and the hyperbolic curve has much more EQ then Tractrix and that is more important the EQ that work much closer to the horn rate (if the horn is properly designed). So, I hope the horns will gain approximately 8dB-9dB. Then they get something from the room at Midbass region. Would it be another 8 db up or doe no one can say now. so, if everything  goes as I planed then the new Midbass channels will have sensitivity as the rest of Macondo. Now, how to drive the midbass?

My initial sentiment was to use the current Melquiades bass Channels. It will have most possible enough gain and power – a full 6C33C at 50W of plate dissipation.  I have this amp in my 6-CH assembly and then why do not use it? Sure I would need to modify the crossover and it is VERY comfortable to do. Take a look at the Melquiades bass (channel A on the top of the circuit)

http://www.romythecat.com/Site_Images/6-Chennal_Melquiades_DSET_Amplifier_Rev3.jpg

Currently it has just one cap to ground. Ironically it is very sophisticated crossover. A cap to ground can write low-pass filter only after a resistor. The resistor 30K in this configuration serves a combined duty:

1)     Maintain the input impedance for the bass channel

2)     Serves as base resistance for the filter

3)     Generate bias offset for the driver tube

Currently it runs 78Hz first order. It is not a big deal to move it up. I hope that I will be able to use the first order in there with Midbass Channel but it most likely will not be so. If I need to go in there with higher order filter than it will be so. Then there is a subject of the high pass filter. I know that people do not do it and no one talk about it but I also know that people are deaf Morons – I am not. As I told many times the un-used bass mast be removed from LF horn. What is the point to make the driver cone to excurt more than the horn’s mouth can pass? So, most likely the channel will have some kind of 25-35Hz filter, just under the horn’s natural roll off. In Melquiades’ case it will be very easy by changing capacitor between the stages – a perfect spot for high-path filter.

Still, there is something that make me feel “not perfect” for this channel when I think about the Melquiades bass channels. This is primary due to the current output transformer that I use in Melquiades for bass. The Melquiades bass channels were design to use with open-bottom type LF. It has enormous transformer with ultra low DCR and huge inductance. It has the most insane bass transformer you ever see for 6C33C. It has 7.5Hz frequency response. Mind you there are zillion idiots out there who claim that their not DSET output transformers work at 2Hz or 10Hz. This is all BS as none of them does measurement at full power. My persistent position is that any frequency response with OPT MUST be quoted only at full power. So, the Melquiades bass channels have now 7.5Hz  at 22W and gaped up to 450mA – you can imagine the construction of this transformer.  This transformer was one of the most important contributor to the truly pronominal bass I got in the old room.

Now, the funny thing is that this pronominal transformer might not be the best bet to drive my midbass horns. Here are the things that I have problems with: fixes ratio, use of M6 steel. I would like to have more flexibility with loading the new Midbass Channel and I would like to have much faster core. I guess I would need to get another much smaller transformer for the Melquiades bass channels….

Paul S wrote:

Romy, is it a goal to make the horn strong enough to raise one end or turn it without having it fall apart from its own weight? From your earlier descriptions, I expected to see at least one continuous piece of plywood bridging the full length of the horn, to tie it all together. I also expected to see glue blocks, perhaps some temporary clamping blocks (to facilitate clamps that pull sections together until the glue dries, maybe some band clamps, continuous cleats rather than small temps or small permanent cleats that will interfere with continuous braces/cleats in the future, if that is still planned. At least, I would expect to see "biscuits" where edges are joined, both edge-to-edge and L joints, to get and keep everything located/lined up and to strengthen those joints. Another trick I do not see is any sort of control templates, to ensure that sections destined to mate together will do so directly, since you likely will not be able to "rack" the glued-up sections together once the glue has set. Any diviation in any plane makes for big fights trying to line up and effectively glue thick plywood sections like these.  Maybe as a hedge you could glue up each section while it is lined up temporarily with its ultimate partner/partners, to ensure subsequent compatibility. The weakest parts of this horn will likely be the full edge-to-edge joints between the pre-built sections. IMO these warrant the "braces" you spoke of earlier, and, again, they shoud be installed as part of the skin, then glued, clamped and screwed together along the edges rather than added as afterthoughts. Because of the frequent changes in the "cord" of the horn it will be a challenge to come up with continuous "braces" or cleats to cleanly bridge all the "natural", directional breaks in the horn. But IMO this is something that should be undertaken sooner rather than later. One idea is to make the "braces" themselves out of plywood, so they'd combine the functions of braces and cleats and still function as part of the horn shell, as well. Also, maybe, some "bent" cleats could be fashioned from framing hardware and fastend with short, self-tapping "truss head" screws at the "bent" joints.  Just understand that any cleats are a supplement to rather than a replacement for well-fitted, glued (and clamped) joints.

Paul S wrote:

Funny, because your "sections 3 - 4" explanation was the perfect response to the first thing/question that popped into my head when I first looked at the picture. Yes, theoretically, one might "reduce" any flowing dimension as long as the area of the mouth none-the-less increases according to the recipe. Still, this is the first time I remember seeing a "decrease" rather than an increase along any extending axis of any horn (other than moronic nonsense).  Since you explained it unprompted going in, I ass-u-me this feature is planned. Is this something you have done or seen/heard done before; or is this a little twist of your own?

Well, the horn has a few functional curves.

1)    First Section is a transition from circular throat to rectangular mouth of the First Section. This is a complex transition and it will be done in an interesting original way. You will see as time goes by.

2)    Second section is a simple section - it is a symmetrical trapezoid that is pretty much a conical horn

3)    3rd section is a complex transition from symmetrical trapezoid to very much non-symmetrical trapezoid

4)    The rest sections are in a way the same expending non-symmetrical trapezoids

My carpenter assures me that the mouth of the 3rd section is right and I think he is correct as the 3rd section is relatively swiftly expends in one dimension and relatively rapidly collapses in another. I do understand it but I also feel that my carpenter might go a bit too far in his approximation. This is exactly why we test-assemble the horn before to solidify it.

I am debating with my carpenter and the most important with myself if the connection between 3 and 4 sections need to be redone. Intellectually I think it is fine as is but my carpenter teaches me that an agley looking airplane does not fly well…  I think if we slow a bit the horizontal expansion of the 3rd section and smooth the indention of the 3rd section in vertical plain then we will have more natural curve with the same mouth aria on the exit.

Paul S wrote:

From the photo (parallax?), it looks like the rate of flare turns a little "odd" in a couple of places, not just at the cited 3 - 4 junction. And I suppose that even if this is the case, yet the "expanding" mouth area "explains" any similar case equally.

Again, it may be parallax (or, it may be late...), but the following do not appear to me to be "smooth progressions", whether simple or exponential:

1) Section 2, vertically

2) Section 4, horizontally

3) Section 4, vertically

Of course I don't know what the thing is supposed to look like, in the first place, but this is where large, plotted paper (or cardboard) templates are useful as gauges. Maybe plot and gauge the inside of the curve, so differences would be easier to spot (if it matters...).


It's very easy to cross a line with a hand-held power saw, and a snapped line can be fairly wide (relatively speaking, here), to begin with. If this proves to be an issue, Japanese-type ink lines are quite narrow and visually sharp. If/when the line is critical, one can always "leave the line", then plane or sand back to it with a disk sander (experts only...), for greater control.

Best regards,
Paul S

Paul S wrote:

Again, it may be parallax (or, it may be late...), but the following do not appear to me to be "smooth progressions", whether simple or exponential: 1) Section 2, vertically 2) Section 4, horizontally 3) Section 4, vertically Of course I don't know what the thing is supposed to look like, in the first place, but this is where large, plotted paper (or cardboard) templates are useful as gauges. Maybe plot and gauge the inside of the curve, so differences would be easier to spot (if it matters...). It's very easy to cross a line with a hand-held power saw, and a snapped line can be fairly wide (relatively speaking, here), to begin with. If this proves to be an issue, Japanese-type ink lines are quite narrow and visually sharp. If/when the line is critical, one can always "leave the line", then plane or sand back to it with a disk sander (experts only...), for greater control.

Thinking about the integration scenarios of my midbass horn I come to a realization that it highly possible that the horn in THIS room will do something very unique. I have wrote before that I was successful to set channels delayed but the delay was masking but the fact that echo frequency channels was delays in respect to a neighboring one and offset by frequency range. This way delays was not as single hit but speeded across the bandwidth and was much less annoying, if annoying at all. What I think is possible is that in my new room with my new midbass horn I might have a model where the exchange between frequency and delay will be super gradual and absolutely natural. I do not know how it will affect sound but it shall be very interesting. Let look at the proposed mechanic of the process:

Let pretend that wide bandwidth come from main speakers and from midbass horns. On the picture is the side view of my room with more or less right proportion. The LF do not propagate like light but they obey the pressure wave pattern, so the pressure fill all space at the same time and pressure rises at the shortest distance. So, at LF the L1 distance and L2 distance are identical – the horns are in perfect time alignment.

Now the frequency rises. The more it raises then more and more sound losing pressure wave pattern distribution and get directivity, like light. The sound at HF shots out of the horn but the horn very much not pointed to listener but sits way above. So, the HF are shooting in the opening of cathedral ceiling and then descend to listening spot as reflections (whatever were not killed by my sails). This HF sound will be delayed in regard to the reference arriving time but here is an interesting kink: the HF will be delay proportionally to bandwidth and he most important this dependency will be self-regulated. So, in my drawing the L3 will be never equal to L4, however the LF will be longer then L3 very smoothly in respect to the acting frequencies.

I guess it might be very interesting effect. The L4 will be of close way attenuated but I think they will be in play somehow. How effective will it be the time will show but if the impact of FH from my midbass horns will be not the same as amount FH from the horns. I am very confident in it.