New DIY Mid High (90deg) - AKA PM90

Karl how did you get the wood cut for yours? if I might asked.
I have not started the build yet - still collecting drivers! I shall be doing all cuts on a bench saw and radial-arm saw, with a hint of copy- and template-guided routing thrown in. To keep the weight down and reduce resonances, I am toying with the idea of sandwich panels, using a PVC honeycomb core (Nidaplast) with vacuum-bagged skins of aluminium or marine ply.
Currently my time is taken up with a new batch of 18" sealed subs to keep up:208922208923
 
  • Wow
Reactions: Bennett Prescott
I am feeling the need to brace the horn mouths horizontally in additional to vertically - my engineering brain just can't tolerate the unsupported sides! I am assuming these should be angled at half the included angle of the horn exit and kept relatively narrow.
 
That sounds really, really nice. I'm building a pair at present, and am apprehensive about moving away from the Beyma TPLs I'm currently using, but your sound check has really boosted my confidence in the build. Thanks for sharing.
 
Subs are similar to the xo1c design, but I made changes to the internal design and layout.
Hey, that’s funny. I have the exact same combination: TH18/118 + PM90. The TH118 seems to have just enough punch to keep up with the PM90 (for general band/weekend warrior stuff). I made the subs active with Powersoft plateamps which also power the tops (4 channels: 3kW sub bridged, 1.5kW mid, 1kW hf/vhf). I think this is currently one of the most compact, high output, high quality, general purpose sub-top sets. I don’t know of a commercial set that comes close with the same compromises (each design is a compromise...).
 
Hey, that’s funny. I have the exact same combination: TH18/118 + PM90. The TH118 seems to have just enough punch to keep up with the PM90 (for general band/weekend warrior stuff). I made the subs active with Powersoft plateamps which also power the tops (4 channels: 3kW sub bridged, 1.5kW mid, 1kW hf/vhf). I think this is currently one of the most compact, high output, high quality, general purpose sub-top sets. I don’t know of a commercial set that comes close with the same compromises (each design is a compromise...).

Mine are 4 way active on cheap Chinese plate amps, but I’m hoping to run them with Linear Research amps in the near future. My ver of the TH118, is tuned slightly lower than the xoc1 version. I get pretty strong output to 30hz, low enough for live, not really planning to get into EDM, but I’m sure it would manage that too. ?
 
How did the event go?

Very well. I do a number of events for my local council each year, but much of my work is for bands. I’m lucky in that all my clients just trust me and I can use what I want. Never had to make a rider for my direct clients. I have another client who is all named brand speakers, d&b, l’acoustic, etc, but I told him I was going to use these for his job, he was concerned, but I told them it would be better and he trusts me. ?
 
Quick question, I have finished cabinets and have a pair of 8&8 ohm 4594 and will be using switchable passive crossover for when I need the extra amplifier channels, I am about to pick up some 12NDL76 all 8ohm, am I right in doing this or would there be any sense to getting 16 ohm drivers.... TIA
 
Quick question, I have finished cabinets and have a pair of 8&8 ohm 4594 and will be using switchable passive crossover for when I need the extra amplifier channels, I am about to pick up some 12NDL76 all 8ohm, am I right in doing this or would there be any sense to getting 16 ohm drivers.... TIA
Passive from mids to hf/vhf? I don’t recommend doing that. The mids need quite some processing to play nicely with the bms. You can get an amp for the money of the passive x-over these days.

I recommend 8 Ohm mid drivers (4 Ohm per cabinet): 8 Ohm recones are much easier to get and it further lowers the requirements on the amps (lower price).
 
There is also the time alignment difference that might be quite hard to get right without processing & bi-amping between the 12’s and cd but a thought i read in the post somewhere that the cd can be crossed passively but the box bi-amped.
 
Thanks for the info Peter, I am aware the PM90 is to be run 3 way active and this is how it will be run with full dsp but I have provisioned the box with passive xover and switch for running 2 way should I need the extra amplifier channels, has been mentioned many times that this is ok to do should you need to..... will order 8ohm drivers to finish off now, thanks!

BMS passive xover for the 4594 is 25 quid , where are you getting amplifiers for this kind of money?
 
Thanks for the info Peter, I am aware the PM90 is to be run 3 way active and this is how it will be run with full dsp but I have provisioned the box with passive xover and switch for running 2 way should I need the extra amplifier channels, has been mentioned many times that this is ok to do should you need to..... will order 8ohm drivers to finish off now, thanks!

BMS passive xover for the 4594 is 25 quid , where are you getting amplifiers for this kind of money?
Aha, I see. Because you asked about the impedance of the middrivers, I assumed that you wanted to use a passive x-over from mid to hf/vhf. But now I understand you intend to use a passive x-over for hf to vhf only. That should work fine: I’m using the passive x-over as well on the BMS. You can get a flat phase response with some allpass filters easily. ;-)
 
I have read through every single post (really) and can find only vague hints at phase response. Clearly Peter Morris has got it well sorted as evidenced by the square wave response. Can anyone please elaborate upon the measures required to obtain a flat phase response, and clarify if this is speaker- or room-dependent, or both. Thanks in advance, Carl.
 
I have read through every single post (really) and can find only vague hints at phase response. Clearly Peter Morris has got it well sorted as evidenced by the square wave response. Can anyone please elaborate upon the measures required to obtain a flat phase response, and clarify if this is speaker- or room-dependent, or both. Thanks in advance, Carl.

I'm sure Peter will be along soon and elaborate, but initially during his measuring and processing of the system, he used all pass filters to flatten out the phase response of the pass bands, and delay to then line the three pass bands of the tops together properly. That was most of the work in getting the box to sound the way it does, though I'm sure, if I remember right, he also looked at adding FIR processing to further tweek and dial things in.

K
 
Hi Carl,
Assuming a reasonable design, it is really not that hard (just a little patience) to achieve a very flat phase response with more modern DSP offerings that allow you to use allpass filters, linear phase FIR crossovers, etc. You will find flat phase even making its way into the MI market - for example Mackie and Yamaha have been doing this for quite a few years now in some of their products. As DSP costs decline FIR has been steadily appearing everywhere.

Here's a listing of some of the better featured DSP platforms currently:
 
This is what I do with the Lake:-

1) I flatten the amplitude response of drivers with in-band PEQs as much as possible. I try to do this so that it is flat at least half an octave past the crossover point. With the 4594 and the VHF crossover this is not possible as the low section will struggle to go above 6K3 and the high section will struggle go below 6K3 - you just need a sensible compromise and if you are using brick wall crossovers its not that important what happens out of band.

2) Then I use all-pass filters to flatten the phase of each pass band ... this can be a difficult to do. There is a limit to how low you can flatten the phase without introducing too much latency when used to live music – I try get it reasonably flat from about 200Hz up.

3) Then I adjust the delay times of each band so that they are all time aligned, and set the levels for each pass band.

4) Next I adjust the crossover frequencies to where I want them and use either a brick wall FIR or 24dB linear phase crossover with LR slopes. The 24 dB curve is use on the PM90/60 mids as it results in better pattern control through the crossover region.

5) With the Lake I can then use the system PEQ which uses raised cosine filter to tweak the finial result.
 
Last edited: