New DIY Mid High (90deg) - AKA PM90

A bit of a random question, but on the combined cosmetics/performance front, is there an optimum shape for the main horn edge termination? The exterior walls will be 28mm thick so I have a little more thickness than usual to play with, shape-wise.
 
one more question about the damping: Is it supposed to be applied on the sidewalls of the chamber only? Or should the whole chamber be filled? What material did you use? sonofill? or soundabsorbing foam mats?
I am about to finish the cabinet build and totally forgot about the dampening.
 
one more question about the damping: Is it supposed to be applied on the sidewalls of the chamber only? Or should the whole chamber be filled? What material did you use? sonofill? or soundabsorbing foam mats?
I am about to finish the cabinet build and totally forgot about the dampening.
Filling the whole chamber would impact the air flow, making the drivers heat up faster. I would avoid that. Do put some foam on the free surfaces though, as in, the sides, top and bottom of the chamber.
 
Hey guys, haven't posted in a while but have some progress pics to share when I get them off my phone. Proud to say we've finished building our own pair of PM90's - but I have an important question to ask regarding amplification and worrying that I have royally messed up here.

Basically, I have ordered and built the PM90's as:
2x 12NDL76 12" 8 ohms wired in series @ 16 ohms
1x DCX464 with passive crossover - 16 ohm

Don't ask me why, but at the time I didn't realise I could order the 12" as 4 ohms and wire in parallel to make an 8 ohm system

Currently are wired to run on one NL4 1+/1- = 12" series , 2+/2- = DCX via passive crossover

I measured the ohms with meter and they are coming up as ~7.4 ohms. I believe this is considered normal because they are measuring DC resistance

Now back to the issue at hand - I realised I may have messed up in building this as a 16 ohm system with regards to finding suitable amp/power. From specs and information in this thread, I realise I need 400W for the 12" and 160/220w for the DCX (we'll leave processing aspect to a bit later) so,

1) What power amplifier do I actually need? Does it need to be a 16 ohm 400 watt 2 channel amp for mids and 2 channel ~220 watt amp for DCX? Or can I run the whole tops system on a 4 channel 16 ohm amp @ 400 watts per channel
i.e. https://www.labgruppen.com/product.html?modelCode=P0CPE
2) I know I can still safely but not efficiently amp it with an 8 ohm amp still, but if I do - does the wattage need to be above the continuous watt recommendation to make it work ?

i.e. https://proaudio.com/content/Documents/Specsheet_Linea Research_48M.pdf
< - 48m03 or 48m06 being 400w/750w per channel ?

3) If I run it on an 8 ohm amp - will it sound fucked ? / have large loss in max volume achievable - I read somewhere you will only lose out on 3-6dB of max volume achievable by high mismatching amp and speaker impedence
4) As a learning question - Is it possible to run the whole top system + 1250w rated 18TBX100 subs on say.. the linea research 48m10 @ 1250w per channel

Sorry that my understanding of matching amps is complete newb level. Biting nails here, and feel stupid.
 
Don't ask me why, but at the time I didn't realise I could order the 12" as 4 ohms and wire in parallel to make an 8 ohm system

Not quite. Typo perhaps?
Two 4Ω loads in parallel is 2Ω, which a good amp should be able to cope with but is not best practice.
Two 4Ω loads in series would give you 8Ω, which may have allowed an easier match to suitable amps.

What you have though is 2 8Ω drivers, so your choices are: series wiring for 16Ω total as you initially described, or parallel wiring for 4Ω total.
Getting the most out of the drivers will be more expensive with the 16Ω option as that requires a bigger amp, so I'd recommend putting them in parallel for 4Ω nominal.

Assuming you pay attention to how hard you are driving them and don't simply thrash the living daylights out of the system, then traditionally we'd size the amp at 2x the nominal/AES rating of the speakers to maximise headroom. This is commonly referred to as "program power".

That means targeting 1600W per channel from the amp for each pair of 12s. The LabGruppen you mentioned is therefore rather underpowered.
The 1250W offered by the Linea 48M10 is not far off this (only -1dB), so is a reasonable match.
If cost is critical, the 750W of the 48M06 is only 2.2dB less, so while not technically getting the most out of your speakers, will not be likely to be noticed by most of your audience.

It's perfectly OK to run different parts of your system at different impedances across multichannel amps - this actually reduces the total load on the amp (compared with running all your 12's off one amp and all your DCX's off another smaller one for example).

A 48M10 therefore could certainly run both your DCX's on one pair of channels (vastly over powered, but you'll be able to turn those channels down and not have them work too hard), both your 2x12 low mid sections on another pair of channels, and have 4 channels available for your bass section. Depending on how many bass drivers you have and how they're loaded, this might still end up a little light on LF, as the PM90 (particularly the DCX464) are capable of a lot of output.

Re the DC Resistance you measured, the 8Ω version of the 12NDL76 should be about 5.3Ω per driver, so ~10.6Ω for the series pair (+ a tenth or 2 for wiring).
Your 7.4Ω seems a little low for that to be the pair, or a little high if you measured it individually. I'd re-check each driver individually to be sure that they're closer to the factory spec before applying any real power to them, just in case something's wrong.

HTH,
David.
 
Not quite. Typo perhaps?

Yes apologies that was a typo - such a great and informative answer thanks David. Was planning on building 4 subs running 18TBX100 rated at 1250 RMS / 2400 max - already have the 18TBX's. Perhaps I could power with a cheaper sub amp with simple crossover filters and power the tops with a suitable DSP 4-channel amp for tops. Good to know I am allowed to get too much power and just dial it down for the coaxial and mix impedences across multi channel amps.

Also just confirming when you note the nominal power for each speaker is:
12NDL76 - nominal 400w per speaker - I am gathering you quote 1600w as that is 2x 800w of one group of 12"
= 2x 1600w channels needed
DCX - nominal 110w + 80w = 190w per group
= 2x 380w channels needed
Correct?

Stupid fun question time to build understanding:
1) is it possible to wire the DCX coaxial in series somehow from the amp so that it becomes an 8 ohm circuit powering both coaxials ?
2) I read somewhere that bridged amp outputs double their desired impedence i.e. 2 channels bridged that are rated for 2x 400w 4ohm individually would now be 8 ohm 800w when bridged ? Is this true?
3) If I power a 16 ohm circuit of the coaxial which is also quite sensitive with an 8 ohm amp say 400w channel, is it allowable/suitable to think of the wattage requirement as increasing / doubling ? So something along the lines of an 8 ohm 800w channel is necessary under this application.
4) Or under this same silly philosophy , that powering the 16 ohm coaxial with an 8 channel 300w amp would be suitable given it is more sensitive and the volume reduction would be suitable (and run 12" parallel as 4 ohms)

Kicking myself that I've easily made it that much more difficult
EG
 
Last edited:
Perhaps I could power with a cheaper sub amp with simple crossover filters and power the tops with a suitable DSP 4-channel amp for tops.
Yup, that's an option too.

Also just confirming when you note the nominal power for each speaker is:
12NDL76 - nominal 400w per speaker - I am gathering you quote 1600w as that is 2x 800w of one group of 12"
= 2x 1600w channels needed
DCX - nominal 110w + 80w = 190w per group
= 2x 380w channels needed
Correct?

Yes, spot on for the 12NDL76 pairs, 2 channels needed (one per box) of 1600W at whatever impedance you settle on for maximum headroom or marginally less as constrained by budget. Bear in mind that most amp designs are capable of less output as Impedance rises, so an amp rated at 1600W @ 16Ω would traditionally be rated 3200W @ 8Ω. That's why series wiring for your mids is likely to get expensive, fast. Granted some modern amps are sophisticated enough to manipulate their output voltage & current to overcome this, including the Linea's you referenced, so if you're still considering this manufacturer you'd best contact them for actual 16Ω ratings.
Or just wire in parallel for 4Ω anyway, making the issue moot ;).

No for the DCX's though. What a passive crossover does is split the incoming signal from the amp to the 2 drivers, so at any given frequency only one is operating (slight oversimplification, but it's what's relevant here). That means that you'd only need to choose amp channels rated for the larger part of the pair, ie 220W @ 16Ω to match the program rating of the mid part of the DCX. Conventionally, that means you're looking for amps rated 440W @ 8Ω if you want to power them with a separate amp than your low-mids.

Stupid fun question time to build understanding:
1) is it possible to wire the DCX coaxial in series somehow from the amp so that it becomes an 8 ohm circuit powering both coaxials ?
2) I read somewhere that bridged amp outputs double their desired impedence i.e. 2 channels bridged that are rated for 2x 400w 4ohm individually would now be 8 ohm 800w when bridged ? Is this true?
3) If I power a 16 ohm circuit of the coaxial which is also quite sensitive with an 8 ohm amp say 400w channel, is it allowable/suitable to think of the wattage requirement as increasing / doubling ? So something along the lines of an 8 ohm 800w channel is necessary under this application.
4) Or under this same silly philosophy , that powering the 16 ohm coaxial with an 8 channel 300w amp would be suitable given it is more sensitive and the volume reduction would be suitable (and run 12" parallel as 4 ohms)

1) There would be nothing to gain by wiring both DCX's in series IMO. Firstly, that doubles the impedance to 32Ω nominal, meaning you need to start with a bigger amp to power them; you'd lose the ability to discern stereo, and you'd need a custom patch panel/wiring setup to make it happen. If you were really desperate to save one amp channel, again wiring in parallel to make it an 8Ω load would probably be more practical.
2) Yes, that's exactly how bridging works. Bear in mind though that operating any amp at it's minimum rated impedance as standard practice is not recommended.
Firstly, it means the amp is working harder the whole time, so may have a shorter working lifespan than a larger amp worked less hard. Second, it limits your options to get through a gig if something goes wrong. If you buy an amp capable of running the 12's at 4Ω on one channel, then if a channel dies you still have the option of pairing up both your low mid sections on one remaining channel. You do lose stereo, and a bit of total power, but not too much. If you had bought a smaller amp to bridge in the first place and then had to try and run all four 12's off a single channel to limp through the rest of a gig. you would effectively lose a lot more power.
3 & 4) I think I've covered these above.

Cheers,
David.
 
Hey guys, haven't posted in a while but have some progress pics to share when I get them off my phone. Proud to say we've finished building our own pair of PM90's - but I have an important question to ask regarding amplification and worrying that I have royally messed up here.

Basically, I have ordered and built the PM90's as:
2x 12NDL76 12" 8 ohms wired in series @ 16 ohms
1x DCX464 with passive crossover - 16 ohm

Don't ask me why, but at the time I didn't realise I could order the 12" as 4 ohms and wire in parallel to make an 8 ohm system

Currently are wired to run on one NL4 1+/1- = 12" series , 2+/2- = DCX via passive crossover

I measured the ohms with meter and they are coming up as ~7.4 ohms. I believe this is considered normal because they are measuring DC resistance

Now back to the issue at hand - I realised I may have messed up in building this as a 16 ohm system with regards to finding suitable amp/power. From specs and information in this thread, I realise I need 400W for the 12" and 160/220w for the DCX (we'll leave processing aspect to a bit later) so,

1) What power amplifier do I actually need? Does it need to be a 16 ohm 400 watt 2 channel amp for mids and 2 channel ~220 watt amp for DCX? Or can I run the whole tops system on a 4 channel 16 ohm amp @ 400 watts per channel
i.e. https://www.labgruppen.com/product.html?modelCode=P0CPE
2) I know I can still safely but not efficiently amp it with an 8 ohm amp still, but if I do - does the wattage need to be above the continuous watt recommendation to make it work ?

i.e. https://proaudio.com/content/Documents/Specsheet_Linea Research_48M.pdf
< - 48m03 or 48m06 being 400w/750w per channel ?

3) If I run it on an 8 ohm amp - will it sound fucked ? / have large loss in max volume achievable - I read somewhere you will only lose out on 3-6dB of max volume achievable by high mismatching amp and speaker impedence
4) As a learning question - Is it possible to run the whole top system + 1250w rated 18TBX100 subs on say.. the linea research 48m10 @ 1250w per channel

Sorry that my understanding of matching amps is complete newb level. Biting nails here, and feel stupid.
+++ on everything David said. I habe built the same combination, only difference is theb12NDL76 are wired in parallel -> 4 Ohms.
To give you an example, I am powering the pair of DIY90 with a FP10000q. At 16 ohms for the high section, it has enough reserve (660w power). With 2100w@4ohms for the mids, i have enough power for most scenarios, since I dont want to treat them to maximum output to minimize risk of damage. They will still go louder than any standard top i ised before.
The subs are powered with a seperate amp (FP14000).
 
  • Like
Reactions: David Morison
This is going to sound like a very obvious question, but what is the thinking behind the 16 ohm versions of the DCX464 and 4594 coaxials? They have never struck me as the type of driver which would be used in multiples, and even the 8ohm versions need a disproportionately large amplifier to achieve the required output voltage to drive them properly...
 
+++ on everything David said. I habe built the same combination, only difference is theb12NDL76 are wired in parallel -> 4 Ohms.
To give you an example, I am powering the pair of DIY90 with a FP10000q. At 16 ohms for the high section, it has enough reserve (660w power). With 2100w@4ohms for the mids, i have enough power for most scenarios, since I dont want to treat them to maximum output to minimize risk of damage. They will still go louder than any standard top i ised before.
The subs are powered with a seperate amp (FP14000).
Yes, it is easing to know that that is the right way of doing things - this is my first build so I thought all had to be matched. Thanks for the tip and reassurance on amp choices.

It is right that the subs need ~2500w each if there are 1250 watt rated ?

This is going to sound like a very obvious question, but what is the thinking behind the 16 ohm versions of the DCX464 and 4594 coaxials? They have never struck me as the type of driver which would be used in multiples, and even the 8ohm versions need a disproportionately large amplifier to achieve the required output voltage to drive them properly...

This is where things get weird for me as I've spent a lot of time thinking about this in last 48hrs, and I'm utilizing the passive FB464 crossover. The DCX464 is technically 2 in-phase drivers for HF + VHF, both - according also to website 16 + 16 ohm. If I am powering the crossover itself to both speakers - is this in fact an 8 ohm parallel group of 2 drivers ?

And if they are resistance-matched - they each draw equal power evenly, then should the "program power" amp channel requirement be 320w (220w MF + 160w HF continuous power handling) to suit the more delicate HF driver ?

Sorry, brain going on logic explosion

Here is crossover diagram:
1651648239362.png
 
This is going to sound like a very obvious question, but what is the thinking behind the 16 ohm versions of the DCX464 and 4594 coaxials? They have never struck me as the type of driver which would be used in multiples, and even the 8ohm versions need a disproportionately large amplifier to achieve the required output voltage to drive them properly...
As mentioned here and here, the 16 Ohms version will have less noise. The DCX464 is really sensitive. With the 16 Ohm version, I still get quite noticeable hiss. In PA settings, this is no problem, but for hifi, it would be too loud, imho.

Yes, it is easing to know that that is the right way of doing things - this is my first build so I thought all had to be matched. Thanks for the tip and reassurance on amp choices.

It is right that the subs need ~2500w each if there are 1250 watt rated ?
No, you could go with less. 1.5 to 2 times RMS power is a rule of thumb. You want to have enough reserve on the amp to secure headroom. Otherwise, you might too easily drive the amp into clipping. The signal then gets compressed and the speaker likely gets damaged.

This is where things get weird for me as I've spent a lot of time thinking about this in last 48hrs, and I'm utilizing the passive FB464 crossover. The DCX464 is technically 2 in-phase drivers for HF + VHF, both - according also to website 16 + 16 ohm. If I am powering the crossover itself to both speakers - is this in fact an 8 ohm parallel group of 2 drivers ?

And if they are resistance-matched - they each draw equal power evenly, then should the "program power" amp channel requirement be 320w (220w MF + 160w HF continuous power handling) to suit the more delicate HF driver ?
Not sure, what you are referring to. Do you mean powering to cabs with one crossover? Is your question, if the 16 Ohms HF + 16 Ohms VHF of the coax are wired in parallel on the FB464 crossover = 8 Ohms impedance for the amp channel?
 
Thanks for your reply and insight above.
Not sure, what you are referring to. Do you mean powering to cabs with one crossover? Is your question, if the 16 Ohms HF + 16 Ohms VHF of the coax are wired in parallel on the FB464 crossover = 8 Ohms impedance for the amp channel?
In response - for example with my PM90 which I figure is relatively standard in using the FB464 passive crossover designed for the DCX464 :

1 amp channel supplies 2 poles +/- to the NL4 port, and then to the HB464 crossover board.
On the XO board it has 2 clips for IN +/- , the circuit of lamps, inductors and resistors and 4 output clips for MF and HF drivers respectively.
an example:
440w amp channel +/- --> XO IN +/- --> XO MF OUT +/- -->. DCX MF driver (220w)
___________________________________________ --> XO HF OUT +/- --> DCX HF driver (160w)

Looking at the crossover schematic previously posted, it appears that the two drivers in the coaxial DCX464 are in parallel so - are they 8 ohms total ?
 
Re: New DIY Mid High



I used a BMS 4594HE with an RCF HF950 horn and 2 x RCF MB12N351s. You could use the MB12N405 for a bit more power but it makes the box a bit heavier than I wanted. The 12's are on a folded or bent and ported horn about 2ft (600mm) long.

The HF950 horn flare has a 400Hz cut off frequency and a 300 x 300mm mouth. The mid horn(s) exit above and below the HF so that the spacing will provide some directivity control just as it does with your Flex array but in the vertical plan. The dual diaphragm concentric compression driver can easily operate down to this frequency.

Think in terms of the Flex Array in point source mode with a 90 x 50 horn pattern, using a similar HF / VHF combination and crossover frequencies, and two fully horn loaded 12" drivers instead of the 10's.

"How does it sound?" - It sounds like the best studio monitor you have ever heard. The vocal are so clear and it goes loud. Everyone that has heard it comments about the clarity and that they can hear things in their favorite music they have never heard before.

Processing is Lake FIR. I don't think you need to use the lake, but it won't be able to make a square wave without it. I did the square wave just to see if I could ... and it will do a resonable square wave from 300Hz to 3kHz.

When I have a moment I will post some pictures and some hornresp sims which I have compared it to an EAW KF850 - the idea was the output of an EAW KF850 but with a speaker on a stick. It needed to be a useful size and weight and look good.

At this stage the plots are just the prototype, but I thought some people may be interested.