PA Top with AMT

Moritz Bachmann

New member
Nov 7, 2019
24
8
3
28
Austria
Hey guys,

although we never designed speakers by our own (we mainly built speakers with building kits), I´m trying to set up a PA System with Hi-Fi qualities together with 2 friends.
"Unfortunately" we decided to go with Pro AMT´s from Mundorf, precisely the 197PP27R-7- H (see PDF below) and the Beyma MC700ND series, for testing we bought both, the 10" and the 12" Version. Unfortunately because we found too late Peter Morris´ DIY TOP, which might not have the HF as perfectly clear and precise as the AMT (- they got an absolutely insane impulse response and very low distortion compared to Compression drivers) but seems to be the best solution for a high quality Speaker we found on our long search through the internet.

Now, there are many possibilities for us, and maybe some of you could help us going into the right direction.

We want to try out our own designs for sure, even if it is just for learning - double 10"/ 12" in MMT (MTM seems not to be an option as it´s not possible to arrange the two Mids even near half the Wavelenght for a crossover to the the AMT)

Therefore we would also build the double 10" design from Jaroslav Mikyska from the "Double 10 and Horn" Thread to see which fits better for our application (Sound Quality is still a little more important than extreme SPL), and also there, a very low crossover frequency is required

Maybe someone can tell me from the Datasheet of the AMT if it could be possible to go with the Crossover low enough to match the DIY Top, still we will also ask Mundorf what they think (We got the Version with fan, so they should be capable of taking some Watts also in the lower end without burning instantly- but if they do, it would be a lot of money burning)
If there is a chance to use this AMT in the PM Top, that would be absolutely perfect of course, even if the AMT might be the limiting factor in SPL then...

Let me hear what you think about our plans, and what design you would suggest for our application / Speakers

Thank you all in advance for your help!
Cheers, Moritz

(I´m not sure if the attached pdf works- it would be page 19 of the following link http://www.mundorf.com/PDF/MUNDORF_proAMT_Catalog.pdf)
 

Attachments

  • Mundorf Pro-AMT.pdf
    433.1 KB · Views: 66
Last edited:
I think the AMT still is not suitable for the PM cab. Even though power handling is up, efficiency is still not that high.
However I don't see why not use a MTM arrangement (if going the dual 10/12 range). Such crossover problems do also occur if you use MMT.
 
  • Like
Reactions: Carl Klinkenborg
The Mundorf AMT bears more than a passing resemblance to the Beyma TPL150/H as far as specs go. A friend and I have been using the Beyma units on our separate systems for a few years with single 10" ported mids, and I would say this is about as much as they will keep up with, even crossed-over at 1.8kHz, where the subjective sound quality was much better, bizarrely, than at 1Khz or lower, indicating just how good the mids are.
Now I am building a pair of PM90s as I plan to cover a larger audience, since the AMT route (unless a line array) is a dead-end output-wise and are quite unsuitable drivers for the PM90/60 unless output is reduced to a ridiculously low level (around 110dB max. if crossing-over at 600Hz) making the project wholly pointless.
The single ported Beyma 10LW30/N drivers I use (they're very nice!) produce an honest 120-odd db/m from 100Hz at maximum power which matches well with the AMTs flat-out. This combination I feel fulfills the 'HiFi PA' requirement but it's as far as one can go, and multiple upper bass/mid drive units, or those with higher output would be overkill and wasted expense in my opinion.
I am quite apprehensive regarding the sound quality of the BMS coaxials compared with Beyma AMTs (which are the best-sounding HF drivers I have heard with quality that does not degrade at high levels) since the BMS distortion levels look horrendous in comparison, but one has to ask oneself whether or not they are audible. I truly hope not...
I believe that Peter Morris built line arrays using Beyma AMTs which I am sure were absolutely wonderful, but you would need deep, deep pockets to reproduce this setup, and perhaps a larger van...
I'm sorry if the above is not what you want to hear, but there's no messing with the laws of physics unfortunately! Good luck with your project whichever route it takes you, and keep us posted.
 
Hi Carl,

My line array uses Beyma TPL200, they are 2 dB more efficient and take 2 dB more power than the TPL150 making them more or less a match for a typical 1” compression driver. The low frequency driver is a 300/600 watt 8” Eighteen Sound with a “dog bone” shaped phase plug.

Yes the TPL’s sound better than a BMS4594HE but have no were near the output. The distortion level of the HE version is very good, noticeably better than the standard version.

@ Moritz -To get the PM90/60 to work you need a horn and driver that can go down to around 600 – 700Hz at full power like the BMS4594HE or the BMS464.

FWIW EM Acoustic Halo – C uses a Pro AMT from Mundorf
 
Thank you all for the response, I didn´t really expect the AMT to work in the PM90/60 well- still i wanted to know your opinion.

We already built a test box for the 12" (20L, tuning frequncy around 65Hz) and crossed it at 100Hz to our sub and to the AMT between 700 and 1200Hz- everythink fine here, although the AMT is alwyas a little too intense, from most points of view, also we are missing a lot of output in the lower mids- therefore we hope the double arrangement, probably with the 10" will have more output between 100 and 300Hz.

At the moment we try to figure out which size and tuning frequency fits the 10" and 12" best, but we are struggling a lot to get them flat to even 100Hz, the start to roll off at ~130Hz.... (Box size varied from 35 to 25L (we will test down to 15L) and always 60to 80Hz tuning frequency in 5Hz steps.) We measure with REW and the recommended MiniDSP UMIK-2 Mic, at 1,5m height, free field. In the distance we moved from 15cm to 1m, the further we move, the earlier the rolloff starts. (?)
We will post some measurements tomorrow, but maybe someone can already find a problem why we dont get it played under 100Hz?
 
Sorry that it took so long.. we got a little ill over here
Here are some measurements of the 12" and the 10". We did a lot of tests of different setting but it didn't change the frequency responds that much.
Especially the differences of the measurements of the 12" are not that big (and a worse low end).. We tried 35L, 30L and 25L volume boxes with tuning frequencies from 60Hz–80Hz each (this time in a distance of 30cm, centered between the speaker and the port).
The measurements of the 10" shows the changes of the tuning frequency much clearer, but it still doesn't play that low like we expected.
What can we do to achieve a better low end?

Also there is a cutoff around 900Hz. But I think we didn't have these problems with other cabs, still it keeps us wondering.. (the cab is stuffed btw)

If we finally find the tuning frequency, we would adjust the port diameter that the 1st port resonance fq is not within the frequency range (which is 100hz-1000hz)? So something like 1200hz?

Thank you all :)


Bildschirmfoto 2020-01-22 um 20.31.24 Kopie.png
 

Attachments

  • Bildschirmfoto 2020-01-22 um 20.53.13 Kopie.png
    Bildschirmfoto 2020-01-22 um 20.53.13 Kopie.png
    839.5 KB · Views: 16
  • Bildschirmfoto 2020-01-22 um 20.26.37 Kopie.png
    Bildschirmfoto 2020-01-22 um 20.26.37 Kopie.png
    926.2 KB · Views: 17
Last edited:
The tiny port will be hurting many aspects of your enclosure's performance. I used 100mm ports in the rear of my 20l enclosures with 10" drivers. It is a very common mistake to use ports which are too small, causing low port output due to port compression (which can easily exceed 15db) along with high port distortion.
The starting point of every port should be one having an area equal to the cone effective area then working backwards towards a practicable enclosure design. Large single flared circular ports perform best and multiple small triangular ports are the worst by a considerable margin. 'Professional' manufacturers of subs take note - you know who you are!
 
@Peter Morris
we will try this. I guess the measurement should stay quite the same...

@Carl Klinkenborg
So do you think one port with 100cm2 or 150cm2 (rectangular) is too small?
You said yours were about 100mm (78cm2)? Did you have multiple ports?
The far field measurement will differ from the near field, the job is to learn how to use both.

Re ports:
Larger port area will contribute more SPL, one can see this in Hornresponse Sim by using the backloaded horn to simulate the vented design.
Larger ports will become longer for the same tuning, port resonances go lower.
As always there is no free lunch:
If the vented enclosure is used for mid range also, long ports are not a good idea IMO.
The chuffing of the small port is not as important, if the mid range device will be high passed above tuning when using a dedicated subwoofer.

For the smallish port I found, that there is less gain in reality than some SIM programs suggest, so it´s easy to overdo the tuning in the quest for more effect.
Tuning too high will make the box sound boomy.

Re the drop in SPL around 900 Hz:
Measure under a lot of different angles ( always a good idea ) and see what happens.
 
A couple of comments about ports – yes triangular port have more losses basically because the surface area of the port is large compared to the volume of the port, but they also provide bracing to the box walls …. So it can be advantageous to use triangular ports to save weight and minimise losses and resonances associated with the box walls, you just need to make them bigger so the port losses are acceptable.

The next thing to decide with the double 10” is how its going to be used … as a mid / high box you want to avoided a port resonances (as Uwe mentioned) in the operational range of the 10”. In this case the port air velocity is much reduced and the port efficiency is not such an issue so you can use smaller shorter ports than you would if you were optimising it for bass out and high SPLs.

...... and when you measure the box in the far field as i suggested above I suspect the results will be different, its just by how much (?)
 
Thank you all for your reply, we had a lot of other stuff to do, and managed it just now to continue with our project.

In the meantime, we met Bennet at the ISE in Amsterdam, which was very interesting for us. We talked a little about Josh´s Skram and the Ipals, but also a little about our problem with the AMT Top. The outcome for us was to go with a closed box and to EQ the lower end to where we need it. Amp power shouldn't be the problem, as we got two FP10k for our 4 Mids and the 2 AMT´s (and they will later get their own amps, so we will then have for each driver its own bridged Channel ... so that should be a lot of headroom).
Yesterday we tested different CB sizes for both, the 10" and 12" until we got problems with our neighbours- now we need to find another place for testing.
Again we changed sizes from 20-40L, but couldn´t find any big differences... the small ones with a little better IR and GD, the bigger ones with a aslightly better Frequency response. Also the difference between 10" and 12" is again marginal - here and there the 12" has 2-3dB more, but the 10" has again the better IR and GD.
So all the differences between the enclosure and driver size seems to point to one preference decision: a smalll enclosure/driver (with the better IR/ GD?!) but lower SPL and more EQ needed, or a larger one with natural worse IR/ GD, but less need of EQ and more output?
As we will have 2 drivers per AMT, and sound quality is more important than SPL and also the coupling to the AMT´s will be easier with the 10", we will built 2x 25 or 30L enclosures for the 10" and test wich alignment works better- MMT or MTM.

Now there poped up new questions for the cllosed box designs:

1. is it common/recommended to use the natural "HP filter" of the closed box? So to take the 2nd Order HP of the box and add just another 2nd Order LR/BW filter on the DSP to get a 4th order, instead of EQ it straight down to 70/80Hz, just to have the "real" 4th Order HP filter at 100Hz? of course we would use the 4th Order filter on the Skram then and adapt the Phase to fit the HP filter to the phase of the LP filter.

2. How much of EQ in the lower section is "allowed"? The DBX Venu 360 is cappable of doing a gain of 20dB, with one +12dB Bell we get the 12" down to 95Hz, the 10" needs 15dB Bell to come down to 90Hz in a 25L enclosure....

The measurents show the different driver sizes with and w/o eq, in 25L and the second with 40L... There is no filter used, and still the low end falls down rapidely in the lower end?
(all measurements are made without knwoing the voltage/ wattage, but the SPL meter showed on white noise ~89dB... we also made few measuremts with +10dB, wich look better, but they are also the reason why we can´t measure in our garden anymore :/)


Peter: the 3rd Image shows the 25L on ground, mic 2m away
and the last picture the 12" in our room with EQ

PS: "with EQ" means for al measurements a single 16dB Bell at 87Hz, Q=3.5... not optimised of course, but the best we found in 5min testing
 

Attachments

  • 25L.jpg
    25L.jpg
    111.8 KB · Views: 15
  • 40L.jpg
    40L.jpg
    107 KB · Views: 19
  • 25L ground.jpg
    25L ground.jpg
    105.5 KB · Views: 21
  • 12in room.jpg
    12in room.jpg
    121.8 KB · Views: 20
time to update this thread, since we did a lot of development during this quarantine time.
We finished our final Protoype. MTM alignment, made of 15mm poplar, braced with an Carbon- Epoxy layer outside. We desided to go with the 12in but lower the Volume to ~18L and to build the enclosure without any parallel Walls.

Unfortunately we had to move into a room for testing, so we did only neafield measurements, 1-2cm from the cone to the mic.

The first pic shows the frequenzy response with different EQ Points.: 1. without EQ, 2. with a 8dB high shelf at780Hz and a 10dB Bell at 87.5Hz. the 3.-5 show always one more EQ point to minimize those breakins in the higher frequencies.

Of course we would like to prevent them without EQ:
The second pic shows some of our stuffing tests with sheep wool and hemp, but we also tested classics like Sonofil or Bondum stuffing with only the 2 main EQ pointsactive.
we couldn´t measure big differences between them, they dont even look that much better than the unstuffed box... maybe because the enclosure design has no parallel walls and reflections are already minimized? Of course except those big break downs at 750Hz, 1160Hz and so on... what are your explainations/ preventing tips for them?
-We tried to solve them with an internal diffusor out of MDF on the back, but it seems that it doesn´t function the way we would like it, so tomorrow we will try it with another diffusor on the paces in the enclosure.

What are your thought about our design and testing results, do you know better stuffings or how to place them?

hope you all stay healthy!
 

Attachments

  • EQ testing.jpg
    EQ testing.jpg
    282 KB · Views: 47
  • Stuffing tests.jpg
    Stuffing tests.jpg
    211.7 KB · Views: 53
  • photo_2020-04-13_20-53-33.jpg
    photo_2020-04-13_20-53-33.jpg
    170.8 KB · Views: 59
  • photo_2020-04-13_20-53-24.jpg
    photo_2020-04-13_20-53-24.jpg
    218.3 KB · Views: 56
  • photo_2020-04-13_20-53-30.jpg
    photo_2020-04-13_20-53-30.jpg
    276.1 KB · Views: 51
  • photo_2020-04-13_20-53-35.jpg
    photo_2020-04-13_20-53-35.jpg
    125.5 KB · Views: 48
  • photo_2020-04-13_20-53-17.jpg
    photo_2020-04-13_20-53-17.jpg
    78.3 KB · Views: 46
Last edited:
The diffuser contours will only have an effect at wavelengths of about two times the depth of these features. Have you tried dense synthetic absorbents in the free space of the enclosure?

I think that's the problem ... here is my SIM including the reflection from the back of the box. On trick is to put hard barriers inside the box to break up those modes plus the dampening.
 

Attachments

  • SIM 12.png
    SIM 12.png
    10.5 KB · Views: 19
Last edited:
I was thinking this very morning about the possible usefulness of dividing a sealed enclosure to both break up and raise the frequencies of standing waves so they might be more readily damped. The bracing I built into my subs (used up to 120Hz) was purely for structural purposes, but something along these lines might have advantageous properties in an enclosure used for a broader range. Acoustic foam was added in each of the six enlosurelets (!) aiming for a Q 0.5.

Apologies for the flip phone image quality...
 

Attachments

  • Subs.jpg
    Subs.jpg
    423.1 KB · Views: 21
Hmm, I alredy thought about the correlation between max. disturbed wavelenght and the feature depth of such an diffusor... after we built and glued it into the box.

our next idea is close to yours Carl, we would place dense wood plates, shaped like an open V when looked on the top view -> \ / behind the driver, with an cutout for the magnet, as the box is not that deep...

We will also try softboard seperators in different positions, but at the moment we are very busy with the construction of our carbon fibre subwoofer
 
  • Like
Reactions: Carl Klinkenborg
bad news: we tried many different seperators, softboard as well as birch ply, in different positions and also clamped the side walls to prevent resonances there, but still we were able to reduce the first buckling at 750Hz only...
We then tried to measure the speaker open baffle- we got the same bucklings and more or less the same frequenzy response. Maybe the last time we used the speakers for a little party in some test cabs, we drove them little too hard and they took some damage
As we need to order 2 more anyway, we will see in a week or so, if the drivers are damaged