New DIY Mid High (90deg) - AKA PM90

Oct 25, 2018
103
11
18
59
Bideford, Devon. UK
Thanks Kevin, Jeff, and Peter for your most generous and valuable advice.

As is obvious, phase correction is completely new to me so I appreciate your advice immensely. All the systems mentioned in Jeff's link are sadly outside my budget, with the exception of miniDSP. My system will be 5-way and I am planning to use - please don't laugh - a Behringer DEQ2496 feeding two DCX2496 via AES. Can phase be corrected at the input to the DEQ using a miniDSP or is phase correction intrinsic to the crossover filters?

Thanks all, Carl.
 

Rich Riotto

New member
Aug 14, 2019
2
0
1
50
USA
Hey everyone, to Peter and all those involved in helping this design along, THANK YOU for all of your work and the willingness to share your ideas with this community. I'm moving toward taking on the challenge of building the PM90 boxes, and have a couple of questions if anyone has the knowledge to share:

- I have access to a Powersoft K3 with DSP. Will the FIR tools available in this amp get close to the Lake LM26 processing Peter accomplished?
- I have access to RCF drivers at a discount. Are there any RCF compression drivers currently available that can closely match the BMS 4594 HE in Peter's original design?
- To anyone that actually built these, what was your as-built weight?

Apologies if any of this has already been posted and I missed it...very long thread!

Best,
Rich
 

Peter Morris

Senior
May 8, 2011
1,056
140
63
Australia
Hi Rich,

Yes people have built them with RCF ND950's and HF950 horns. The FIR fliters used in the Powersoft amps will do almost every thing the Lake will do. The only thing they cant do is a 24 dB / oct LR slope linear phase filter, but they can do a brick wall type filter that may work ????

If I was building now I would probably look at using the new B&C DCX464 compression driver.
 

Jim dee

Freshman
Jul 12, 2015
32
3
8
New Zealand.
FINALLY got mine finished. I must say I am very impressed with the U-pol raptor I used to coat them. Essentially very very easy to apply(spray) although as you can see in one photo the trick is to keep the same distance as you spray or you will end up with a varied texture across the surface which unfortunately I did but they still look smart enough. It was my first time using it... Ill get it right next time!

I powered one up briefly and sounds very very promising!

Peter, many many thanks!
 

Attachments

  • 01.jpg
    01.jpg
    3.5 MB · Views: 176
  • 02.jpg
    02.jpg
    3.3 MB · Views: 165
  • 03.jpg
    03.jpg
    3.7 MB · Views: 151
  • 04.jpg
    04.jpg
    4.9 MB · Views: 144
  • 05.jpg
    05.jpg
    4.6 MB · Views: 138
  • 06.jpg
    06.jpg
    4.9 MB · Views: 142
  • Like
Reactions: Jaroslav Mikyska

Max Warasila

Graduate
Feb 20, 2013
1,194
62
48
Richmond, VA
Hello Peter, a friend wants to build a PM90 box and asks if they would be suitable for the box and could use 18sound 12ND610 speakers? Thank you for answer.

I am not Peter Morris, but after a cursory glance at that driver's spectifications, I would say that that driver could be used, but is beginning to stray from the ideal drivers available for this design. The MMS is notably lower and vas is higher. I threw all three drivers into hornresp to show how they compare to the original RCF MB12N351. First, the B&C 12NDL76:
1572376337359.png
Next, the 18Sound 12ND610:
1572376498956.png

So, could you use it? Yes, sound will come out. However, it's not what the cabinet was designed around, it provides no benefit in long term power dissipation over the 12NDL76 at a higher cost, and it'll have to be more aggressively high passed/limited due to the smaller Xmax.

I'd suggest just going with the B&C. Here in the states it's 40% cheaper than the 18sound, and provides similar if not better performance. I still think the RCF is the strongest performer due to the larger VC area, but the cost for that 2dB midrange improvement is about 30% and comes with a slight LF penalty due to reduced maximum excursion compared to the B&C.

As a quick follow up, the output relationship between the B&C and the RCF is shown below. The B&C is the gray trace:
1572377544169.png
As you can see, the limiting factor quickly becomes excursion below 300Hz. There's not much difference, really, however.
 
Last edited:

Jaroslav Mikyska

New member
Mar 7, 2018
17
6
3
Czech Republic
www.mikesound.com
I am not Peter Morris, but after a cursory glance at that driver's spectifications, I would say that that driver could be used, but is beginning to stray from the ideal drivers available for this design. The MMS is notably lower and vas is higher. I threw all three drivers into hornresp to show how they compare to the original RCF MB12N351. First, the B&C 12NDL76:
View attachment 209010
Next, the 18Sound 12ND610:
View attachment 209011

So, could you use it? Yes, sound will come out. However, it's not what the cabinet was designed around, it provides no benefit in long term power dissipation over the 12NDL76 at a higher cost, and it'll have to be more aggressively high passed/limited due to the smaller Xmax.

I'd suggest just going with the B&C. Here in the states it's 40% cheaper than the 18sound, and provides similar if not better performance. I still think the RCF is the strongest performer due to the larger VC area, but the cost for that 2dB midrange improvement is about 30% and comes with a slight LF penalty due to reduced maximum excursion compared to the B&C.
OK, thanks Max
Jaroslav Mikyska
 

Peter Morris

Senior
May 8, 2011
1,056
140
63
Australia
The 12N610 should perform almost the same as the RCF 12N351 except as Max points out it will be limited by Xmax (3.5mm Vs 6mm for the RCF)

@Max - I get a slightly different curve to you ... but if I make the input voltage about 102 volts and look at the cone excursion I don’t see the problem at 200Hz you are showing ... BUT ... if I try the "Tools - Max SPL" it shows up like yours - Hmmm something strange going on.

Its also worth mentioning that with a 100Hz 24dB LR crossover the input power will typically be much less (-6dB) around 100Hz; where the Xmax limit occurs. Although if you still put maximum power in you will exceed Xmax (linear) but not Xmax (mechanical) with the RCF or B&C.

FWIW your plots looks like the box is tunned a little high - the sim will not give you the correct result because the ports are next to the walls, this makes the port perform as if it were longer.
 

Attachments

  • Xmax RCF 1300W.jpg
    Xmax RCF 1300W.jpg
    46 KB · Views: 24
  • Like
Reactions: Jaroslav Mikyska

Max Warasila

Graduate
Feb 20, 2013
1,194
62
48
Richmond, VA
The 12N610 should perform almost the same as the RCF 12N351 except as Max points out it will be limited by Xmax (3.5mm Vs 6mm for the RCF)

@Max - I get a slightly different curve to you ... but if I make the input voltage about 102 volts and look at the cone excursion I don’t see the problem at 200Hz you are showing ... BUT ... if I try the "Tools - Max SPL" it shows up like yours - Hmmm something strange going on.

Its also worth mentioning that with a 100Hz 24dB LR crossover the input power will typically be much less (-6dB) around 100Hz; where the Xmax limit occurs. Although if you still put maximum power in you will exceed Xmax (linear) but not Xmax (mechanical) with the RCF or B&C.

FWIW your plots looks like the box is tunned a little high - the sim will not give you the correct result because the ports are next to the walls, this makes the port perform as if it were longer.

Peter,

My point on the 18 sound driver actually that it is both excursion and power limited before the RCF, and more excursion limited than the RCF, all while being one of the most expensive options for 12" drivers in the cabinet.

I'll admit I was unsure how to model the design in hornresp - I don't really have much experience using hornresp for, well, horns. But I figured it was close enough to compare the drivers. Not sure what weirdness is going on with the drivers. Maybe because the black line is power limit, which shows a disproportionately high peak for impedance maxima?

I concur with the point about reaching the limits of the linear range but not the real world operating limits of the driver higher in the passband, though I didn't state it quite so eloquently...
As you can see, the limiting factor quickly becomes excursion below 300Hz. There's not much difference, really, however.

In an interesting turn of events, my hornresp install and exports from this simulation have vanished and I'm now back to an install of version 48 somehow, so I can't verify this, but I remember tuning my ports to 80Hz or so by hand in the simulation, or at least what I thought of as, "low enough," to be representative. If it was too high, that would mean that the excursion I show above is actually not high enough for the lowest octave of the passband, but still makes for a valid comparison.
 

Peter Morris

Senior
May 8, 2011
1,056
140
63
Australia
Hi Max,

FWIW the 18 sound driver has slightly more AES power handling than the B&C, but the program rating is slightly less - B&C always quote the program power as the nominal power not the AES.

12NDL76 - 400W AES, 800W Continuous program, Xmax (Lin) 7mm, (Var), 6.5mm (Mech) ?mm
12ND610 - 450W AES, 700W Continuous program, Xmax (Lin) 3.5mm, (Var)?mm, (Mech) ?mm
MB12N351- 650W AES, 1300W Continuous program, Xmax (Lin) 6mm, (Var)?mm, (Mech) 39mm - total I think

B&C's Xvar - "Beyond this excursion limit, the magnetic field seen by the voice coil, or the total suspension compliance, or both, drops to less than 50% of their small signal value, producing high distortion levels, strong variations from small signal behaviour and power compression."
 

Max Warasila

Graduate
Feb 20, 2013
1,194
62
48
Richmond, VA
Hi Max,

FWIW the 18 sound driver has slightly more AES power handling than the B&C, but the program rating is slightly less - B&C always quote the program power as the nominal power not the AES.

12NDL76 - 400W AES, 800W Continuous program, Xmax (Lin) 7mm, (Var), 6.5mm (Mech) ?mm
12ND610 - 450W AES, 700W Continuous program, Xmax (Lin) 3.5mm, (Var)?mm, (Mech) ?mm
MB12N351- 650W AES, 1300W Continuous program, Xmax (Lin) 6mm, (Var)?mm, (Mech) 39mm - total I think

B&C's Xvar - "Beyond this excursion limit, the magnetic field seen by the voice coil, or the total suspension compliance, or both, drops to less than 50% of their small signal value, producing high distortion levels, strong variations from small signal behaviour and power compression."

Peter,

This may be true, but without a significant increase in voice coil area, there will be very little appreciable difference in the power handling of the two drivers. This is why I consider the RCF to be the real upgrade one can make.
 

Salmirabile

Freshman
May 8, 2013
21
12
3
hey guys, I'm finding it hard to choose between the PM60 and PM90....I'm assuming the only different is the sound coverage? if so, I'm leaning towards the PM90 as i don't have the funds nor the van space for 4 PM60s. By the way, bloody amazing design, man!

I've already got 2 18sound 12ND930s so would be good if they worked well in the design as then i'd only need to buy another pair. Just wondering if anybody could sim it as I've never used Hornresp before. No worries if not.

Cheers,
Sal
 

Don J Davis

Sophomore
Jan 17, 2011
195
14
18
Orange County, CA
hey guys, I'm finding it hard to choose between the PM60 and PM90....I'm assuming the only different is the sound coverage? if so, I'm leaning towards the PM90 as i don't have the funds nor the van space for 4 PM60s. By the way, bloody amazing design, man!

I've already got 2 18sound 12ND930s so would be good if they worked well in the design as then i'd only need to buy another pair. Just wondering if anybody could sim it as I've never used Hornresp before. No worries if not.

Cheers,
Sal
Hi Sal,
I have 2 of the 90s and 4 of the 60s. I find I'm using the 90s way more often. Just depends on the coverage you need to provide.