New DIY Mid High (90deg) - AKA PM90

[David Chrapla]

Re: New DIY Mid High

The standard BMS I think will still cost about 750 each. The HE version a simple 1000 each. Something closer to 3-400 each for the CD is doable for me. I will not be thinking that any of these lesser options will compare to the HE BMS CD. I am just trying to get the highs accomplished by a lesser product. Then down the road I can save more money and buy the BMS AND better DSP.

I was looking at the 12PE32 for another design anyways. AND your design seems to be the best horn loaded design I could find for the size. I just cant afford the mid/high combo. I also was looking at the XT horn because I could not find a 90 degree horn such as we had discussed.

Was also curious about the Radian 950/850 and the 18 Sound 4015. BUT the 4015 costs about 550 dollars. So that is getting us closer to the price of the regular BMS. So I may have to try your 1480's and see how they go with the 12PE32's.

This will be a compromise regardless because of the CD. BUT I am just trying to find options that will work with my budget unfortunately.

 

[Mark Wilkinson]

Re: New DIY Mid High

Well, as excited as I've been about this project, I have to admit that I defected from DIY yesterday.
I saw that JTR is having a "moving sale" and I bought a pair of his 3TX...
This fills the 'speaker on stick' hole for me, and although I don't expect the 3TX to match what we've been reading about here !!!, the sale discount and the 3TX's light weight pushed me over the edge.
Best of luck to all building this, thx to those helping with layout and sourcing components, and great appreciation to Peter ! Mark

 

[Bennett Prescott]

Re: New DIY Mid High

Hi David,

Availability in Australia is always difficult, but there is no reason that our DE980TN (or the 2" throat version, the DE985TN) can't be used with an 800Hz crossover point. It should keep up well with a pair of 12", but if you're worried we also make the DE1080TN/DE1085TN in 1.5"/2.0" throat. The DE1080 is a larger diameter than the BMS coaxial (154mm vs 133mm) but considerably shallower so I imagine it could fit, our DE980TN is almost exactly the same diameter.

 

[Peter Morris]

Re: New DIY Mid High

Hi David,

Availability in Australia is always difficult, but there is no reason that our DE980TN (or the 2" throat version, the DE985TN) can't be used with an 800Hz crossover point. It should keep up well with a pair of 12", but if you're worried we also make the DE1080TN/DE1085TN in 1.5"/2.0" throat. The DE1080 is a larger diameter than the BMS coaxial (154mm vs 133mm) but considerably shallower so I imagine it could fit, our DE980TN is almost exactly the same diameter.

[SUB]Hi Bennett,

To get the most out of this design you really need to crossover at 700Hz or less to ensure good directivity, in addition the bent horn struggles to get up to 800Hz. It will work but it’s not optimum.

The recommended crossover for the 980 is 1200Hz and 1000Hz for the 1080. I looked at the 1080 however there are very few HF horns with a 1.5” entry, and none that I could find had the physical dimensions and cut off frequency needed. There were a lot more 2” entry horns but once again I could not find anything in the appropriate size with a cut of frequency that was low enough and that also had good pattern control.

If you are going to crossover at 800Hz or more I think another design approach would be better.

The driver that looked interesting from the B&C in terms of going low and sounding nice was the was the DCX50.

http://bcspeakers.s3.amazonaws.com/2013/04/30/09/59/06/40/DCX50.pdf

Although it’s relatively new it does not look like it was very popular and has been discontinued. It was also only available in 2” and I was not sue about its output capacities (??) [/SUB]

 

[Peter Morris]

Re: New DIY Mid High

Well, as excited as I've been about this project, I have to admit that I defected from DIY yesterday.
I saw that JTR is having a "moving sale" and I bought a pair of his 3TX...
This fills the 'speaker on stick' hole for me, and although I don't expect the 3TX to match what we've been reading about here !!!, the sale discount and the 3TX's light weight pushed me over the edge.
Best of luck to all building this, thx to those helping with layout and sourcing components, and great appreciation to Peter ! Mark

I expect the 3TX will be similar to my double 10 and horn. It uses similar components and is a similar size and weight. This combination has been such a useful box for me and I'm sure you will be very happy with them ~~:smile:. In comparison the DIY12 should make about 10dB more output; but its heavier and bigger and requires more processing and amplifiers ... slightly different design compromise.

Love to know what you think after you have done a few shows with them.

Cheers

Peter

 

[David Chrapla]

Re: New DIY Mid High

I expect the 3TX will be similar to my double 10 and horn. It uses similar components and is a similar size and weight. This combination has been such a useful fox for me and I'm sure you will be very happy with them ~~:smile:. In comparison the DIY12 should make about 10dB more output but its heavier and bigger and requires more processing and amplifiers ... slightly different design compromise.

Love to know what you think after you have done a few shows with them.

Cheers

Peter

I really didnt see many CD's from B&C that looked interesting to me. I like some of the 1" varieties but that was about it. I saw that 10" design and was curious how that sounded also. BUT in saying that they do use the HE coaxial so maybe not an option yet for me. Nut I am not giving up.

 

[Peter Morris]

Re: New DIY Mid High

I really didnt see many CD's from B&C that looked interesting to me. I like some of the 1" varieties but that was about it. I saw that 10" design and was curious how that sounded also. BUT in saying that they do use the HE coaxial so maybe not an option yet for me. Nut I am not giving up.

The DE980TN is a new driver from B&C. Although I have not heard it I believe its very very good as is the 1080. For application where the crossover was around 1K2 it would be on top of my list to check out.

In 2015 the bar is quite high - new drivers and horns from most manufactures are all excellent. The DIY uses the latest compression driver from BMS (4594HE), a new 12" from RCF, and their latest horn - the HF950.

I believe the 3TX uses a BMS 4594 as does my double 10.

 

[Ivan Beaver]

Re: New DIY Mid High

I believe the 3TX uses a BMS 4594 .

I heard the 3TX used the 4593 ( a less expensive coax driver)-but cannot confirm.

 

[Bennett Prescott]

Re: New DIY Mid High

If you are going to crossover at 800Hz or more I think another design approach would be better.

The driver that looked interesting from the B&C in terms of going low and sounding nice was the was the DCX50.

http://bcspeakers.s3.amazonaws.com/2013/04/30/09/59/06/40/DCX50.pdf

Although it’s relatively new it does not look like it was very popular and has been discontinued. It was also only available in 2” and I was not sue about its output capacities (??)

Hi Peter,

Just a suggestion. I know it can do it, and you're trying to save money. This could let you do that, possibly without compromising performance.

The DCX50 hans't been discontinued, just taken out of our show catalog. It is a very good sounding driver, and the MF section is well proven at very high output. I don't know as much about the capabilities of the HF section, and in fact most users opt for the MF-only DCM50. That sells well, but mostly in emergency paging and alert systems where it deals with siren tones and voice all day.

 

[Peter Morris]

Re: New DIY Mid High

I heard the 3TX used the 4593 ( a less expensive coax driver)-but cannot confirm.

...I think you heard correctly ... ~~:smile:

 

[Peter Morris]

Re: New DIY Mid High

Hi Peter,

Just a suggestion. I know it can do it, and you're trying to save money. This could let you do that, possibly without compromising performance.

The DCX50 hans't been discontinued, just taken out of our show catalog. It is a very good sounding driver, and the MF section is well proven at very high output. I don't know as much about the capabilities of the HF section, and in fact most users opt for the MF-only DCM50. That sells well, but mostly in emergency paging and alert systems where it deals with siren tones and voice all day.

Hi Bennett,

I have not had a chance to try to 980, however I believe it’s one of the best drivers B&C have made. I do have some B&C DE 750s, B&C DE900’s,18sound NDS1480Ns and ND1480s lying around. The low mid SPL capability of this box is quite substantial; I know none of these will keep up if I use a 700Hz crossover.

For my application I was trying to get the most pristine VHF that I could; like you get with the best 1” driver. I was trying to build something that exceeded the performance in terms of SPL and sound quality of another stick mounted speaker available and the cost of the HF driver was not really an issue. I wanted to use the newest and best drivers, components and processing available.

Having said that it looks like price is an issue for many people … so a cheaper option would be great, but I am a little concerned about crossing a 3” titanium diaphragm /suspension compression driver at 700Hz …but It would be great if the 980 was up to it.

The only problem is, someone would need to do the signal processing …

 

[Roy Novio]

Re: New DIY Mid High

Hi Peter, very interested in this design my RCF distributor here in the US offered me great price on the MB12N351 but the 950 Horn is not available in the US I am willing to use 18Sound horn but building the cab deeper would that compromise the sound since now the horn and rear chamber bigger? Thanks!8O~8-O~:shock:

 

[Peter Morris]

Re: New DIY Mid High

Hi Peter, very interested in this design my RCF distributor here in the US offered me great price on the MB12N351 but the 950 Horn is not available in the US I am willing to use 18Sound horn but building the cab deeper would that compromise the sound since now the horn and rear chamber bigger? Thanks!8O~8-O~:shock:

The only issue with 18sound horn is the 60 degree pattern. If that’s what you want I suspect the 18sound version will work even better than the RCF. The 18sound horn will actually take up more the of box volume than the RCF Horn so as a rough GUESS the extra volume resulting from a slightly deeper box will be offset by the extra volume the horn takes. I had a play around with modelling different volumes when I made my prototypes and it’s not that critical.

I’m also planning to build 4 of the 60 degree version. Now my summer season is over I will have a lot more time.

The one thing to stress is this box was designed with DSP correction as part of the original design.

Following on with Bennett suggestions trying to save us some money these are some of the options.

· Use a passive crossover on the HF / VHF
· Use the standard 4594 or even the 4593
· Use the B&C 12NDL76 … this may actually give you more LF capability as B&C driver has a greater Xmax than the RCF. The RCF will probably have lower distortion and has more power handling.
· Use a standard compression driver as Bennett suggested and cross it at 800Hz. Some compression drivers may be able to go down to 700Hz.

The RCF Horn is available from Team Audio in France. http://www.teamaudio.fr/en/

The horn is very important, it has to have a cut off frequency of 400Hz or less. The whole design concept revolves around getting the HF section to go lower than normal – 600 to 700Hz. By doing that, you can use a bent horn on the LF. If you can bend the LF horn the box can be smaller and the horn length can be increased to provide horn loading down to a lower frequency. To get just a bit more LF the box is ported.

The spacing for the two MF horns exists is also critical, the further they are apart the lower the crossover needed. The whole design is a balancing act between the box and horn physical dimensions and the crossover frequency.

Because the dimensions of the LF horn are as small as can be, it causes some ripples in its frequency response, hence the need DSP correction.

 

[David Chrapla]

Re: New DIY Mid High

Wow, thanks again for the link Peter. They have the prices on their page now which is a PLUS. I had to email them before to get any prices which took a while to get a reply. And thats also great because I have always wanted to try some RCF speakers out. Although shipping might put my wants back on hold for a while.

 

[Peter Morris]

Re: New DIY Mid High

IIR processing, 700Hz crossover - frequency and phase response.

I will put the settings up as soon as I have a chance.

Peter

UPDATE ... added impulse, group delay and an unsmoothed frequency response @ 5 dB per div ...and ...

Here is my first "best guess" at some IIR settings. They have been made on a Lake LM26 in IIR mode. Your DSP may define the BW or Q slightly differently. They are the settings used for the above plot and should be somewhere near the mark. They may need a little bit of adjustment to suit your taste.

Also note that the Mid and HF have the same gain ... i.e. allowing for the impedance difference the LF is about 6 dB less sensitive than the HF.

GAIN - DELAY - POLARITY

VHF (16 ohm) - 1.5 dB 1.62 ms delay +ve polarity
HF (16 ohm) 0.0 dB 1.49 ms delay +ve polarity
MID (2 x 8 ohm = 4 ohm) 0.0dB 0.00 ms delay +ve polarity


VHF PEQ

Freq 6K85
BW Oct 0.55
Gain dB +2.2

Freq 9K62
BW Oct 0.38
Gain dB - 3.5



HF PEQ - Crossover 6K33 24dB LR

Freq 890
BW Oct 0.22
Gain dB -2.5

Freq 1K15
BW Oct 0.67
Gain dB -3.2

Freq 2K1
BW Oct 0.27
Gain dB -4.1

Freq 2K88
BW Oct 0.44
Gain dB -7.0

Freq 6K98
BW Oct 0.77
Gain dB + 6.0


MID - Crossover 700Hz 24dB LR

Freq 144 LF Shelving Filter
BW Oct 2.42
GaindB +4.5

Freq 223
BW Oct 0.33
Gain dB -1.7

Freq 376
BW Oct 0.4
Gain dB -7.0

Freq 394
BW Oct 0.21
Gain dB - 2.1

Freq 564
BW Oct 0.33
Gain dB -2.4

Q - Bandwidth converter

https://soundforums.net/attachment.php?attachmentid=4371&d=1343451453

I have include some Time / Amplitude / Magnitude plots. The last one is the FIR setting for comparison so you can see the difference compared to the IIR settings.

 

[Don J Davis]

Re: New DIY Mid High

Peter has given me the go ahead to post the project drawings. Bracing is something to be considered during the build. Add more as you see fit especially in the dbl 12 chamber.

I have opted for my own build to make the entire HF950 baffle removable via threaded inserts, using a gasket to seal it. This will allow better access to the chamber. As Peter stated, you don’t need to do that, the drivers can be installed through the HF950 cut out.

I have left off the pole cup position and side handles. Best to locate those yourself as you build.

There are no considerations in the design for flying the box. If you choose to modify it to do so, do it at your own risk.

If this box is like most DIY boxes I’ve made, some field adjustments will need to be made during construction. Please post any errors found, adjustments or improvements that need to be made and I’ll correct the plans.

If anyone wants the dxf/autocad file PM me.

Have fun, hope to see some progress photos soon!

 

[Roy Novio]

Re: New DIY Mid High

Nice work Don! OK who's willing to do CNC flat pack??? I'm very interested, thanks!:lol:

 

[Neale Watson]

Re: New DIY Mid High

... Bracing is something to be considered during the build. Add more as you see fit especially in the dbl 12 chamber ...

Peter - I'm curious how much sound emanates from the rear of the box (via transmission, vibration, whatever) while it is working hard. Apart from bracing, is there enough simple density in the rear panel to prevent significant transmission?

 

[Peter Morris]

Re: New DIY Mid High

Peter - I'm curious how much sound emanates from the rear of the box (via transmission, vibration, whatever) while it is working hard. Apart from bracing, is there enough simple density in the rear panel to prevent significant transmission?

Sound from the back of the box in simple terms is mainly caused by the defection of the box walls under pressure and exacerbated by resonate issues.

For its weigh this is a very ridge box, and I have not noticed any issues with sound from the rear of the box.

By the way of simple example to demonstrate where you need to concentrate your box design efforts … and because the sums for supported round plates are simpler than supported square plates; the defection at the centre of a circular plate under pressure supported on the edges like this box = 0.696pr^4/Et^3

Where:
t =the thickness of the plate
r = the radius of the plate
p= the pressure
E = Young’s modulus – a constant for the material

Even though we are not using round plates the principle is the same - If you look at this equation what you see is defection is proportional to the radius of the plate to the power of 4 and the thickness of the material to the power of 3.

Sooo if you reduce the unsupported width of the panels and increase the thickness of the material there is a dramatic effect on the wall deflection and sound being transmitted through the box walls.

The other things you need to do is minimize resonances and standing waves. Parallel walls can cause issues, and once again this box has a lot of angles and not many parallel walls.

In terms of resonance the trick with a wooden box is to try and keep any resonant modes out of the operating frequency range of the box, in this case between 100 and 700Hz.

You need to make the box as strong as possible so that it does not flex, but not use so much wood and bracing that it becomes too heavy .... I'm not sure you need the rear centre brace as shown but its defiantly not doing any harm. There also needs to be some bracing on the side panels. It would be best if you used a cross brace from side to side, but then you would not be able to get the 12” speakers in and out.

I gave all of this quite a bit of thought even custom making the pole mount to join the bottom of the box to the 12" enclosure, making it as strong and ridge as possible without adding weight.

 

[Neale Watson]

Re: New DIY Mid High

Thanks Peter .. I'm sure there are more than me who are learning a whole bunch .. courtesy you sharing your thinking.