Horn Drivers

[Phil Starr]

I'm after upgrading our stage monitors by replacing the compression units in our horns. They don't need to create more than 120db so the drivers I usually use are overkill, too big and heavy and a waste to have something capable of 130dB padded right back.

The thing is to find something that sounds as good as possible and without any nasties which would make the monitors feedback prone. Crossover frequency isn't a problem as I'll design and build my own.

1" screw on preferred as then I can use the existing horns

Any recommendations?

Edited November 5, 2013 by Phil Starr

[LawrenceH]

Hi Phil,

I don't know about screw-on drivers, but I recently salvaged a pair of decent Celestion cast woofers from PA cabs we built in the late 90s and recycled them into sealed box stage monitors - for tweeters I used the 18sound XD125 (as used in fEarfuls) because they are dead cheap for comp drivers+horn, and was very pleasantly surprised with the results. Tried them out doing sound for quite a big-name vocalist who has discerning ears and they were enthusing about their monitor mix afterwards. They are MUCH nicer sounding than both my dog-eared Yamahas (Eminence ASD1001s into Yamaha custom CD horn - mid-biased so they cut through well as vocal monitors but it is a coloured sound) and my JBL Eon G2s (JBL 2412H-1" I think? A bit trashy up top), and a definite step up from the ubiquitous Em APT80.

I went with an acoustic crossover of ~2.4k, 4th-order LR (electrically 2nd order asymmetric) - a bit higher than the ideal for 12" woofers so the response pinches in a little around 2k horizontally, and gives an alarming notch in the vertical polars. In practice thanks to the steep slopes the subjective sweet spot is quite wide (ideal dispersion characteristics of stage monitors is a whole other topic, I notice that many high-end designs use relatively narrow-directivity horns, something I'm not a fan of for smaller gigs). I did use measurement and modelling software to design the crossovers and they measure pretty flat throughout the crossover region; if I redesigned the crossovers for the Yamahas it might be a fairer comparison, but I can say overall they are very smooth-sounding speakers, relatively free of coloration in the vocal range. If I hadn't already had the woofers I'd have gone with 10s or even 8s, I think this tweeter would make a superb little vocal PA cab coupled to a mini woofer.

One final thing - my understanding is that mating drivers, especially screw-ons, with random horns can cause issues due to resulting abrupt discontinuities in the flare profile. Apart from price this is the other reason I chose a driver+horn combined solution.

Edited November 6, 2013 by LawrenceH

[Phil Starr]

Thanks for this Lawrence, one I hadn't looked at. It has quite a droop in the upper frequencies but that could help in a monitor. You are happy with the sound and clean mids are what a decent monitor needs more than extreme highs.


The only downside for me is that I was looking for something to drop into my existing cabs without needing a new baffle, the upside is that you seem to be suggesting a pretty decent result with something that in my case started as a cheap and cheerful solution to a problem using bits left from other projects. I've used Yamaha S112V and S115's as monitors, the 12's are far too mid biased due to the horrible peak in the EM delta's and lack of intelligent crossover design. The S115's are much better but just too big for the small stages we play.

I've had some experiences with matching drivers to horns and experienced exactly what you describe. I bought a bunch of cheap plastic moulded horns and tried them with the horn drivers I was using with really quite dramatic differences in the subjective results. Most of them have flat sides with sharp angles in the flare which I understand you might need to do if you were forming them from sheet materials but which seems crazy in a moulded component.

Did you see my post about the yamahas? I've reversed the polarity of my horn drivers in my S112's which has knocked the nasty mid peak back a bit, really cleaned up the vocals and tamed some of the feedback problems.

[Phil Starr]

PS don't you think it is shocking that you and I (and any number of others on here) could design a better crossover than Yamaha in a design that must have sold in tens of thousands?

[LawrenceH]

[quote name='Phil Starr' timestamp='1383724992' post='2268072']
Thanks for this Lawrence, one I hadn't looked at. It has quite a droop in the upper frequencies but that could help in a monitor. You are happy with the sound and clean mids are what a decent monitor needs more than extreme highs.
[/quote]
The extreme top end of one of mine (10k and above) is quite ragged, doesn't give a smooth roll off at all, I did wonder about sending it back but it really doesn't seem that important live. I often just filter out everything in the monitors above 10kHz anyway and no-one ever notices! I used series-parallel resistors in the crossover to balance it up as much as possible, I like that way of doing it because resistors are cheap

[quote name='Phil Starr' timestamp='1383724992' post='2268072']
Did you see my post about the yamahas? I've reversed the polarity of my horn drivers in my S112's which has knocked the nasty mid peak back a bit, really cleaned up the vocals and tamed some of the feedback problems.
[/quote]

I didn't see this no, I'll have a look. I have the IV series and I don't think the drivers are deltas, maybe betas? Been a while since I had them apart though - classic chipboard construction IIRC but with a lot of wadding for a commercial cab. The mid-bias is colourfully present but they have always been good performers in terms of volume before feedback, and the region where the boost is seems to be both quite broad and quite useful for a lot of vocalists to hear themselves. I built the new monitors with a similar form-factor to the Yams as I find it a useful one, but using braced 9mm birch. At some point I might scrap the Yamaha cabs and recycle the drivers, if this happens I'll measure them and tweak the crossover accordingly. The current one looks pretty cheap and cheerful; even with a 2nd order and no Zobel or resonators, and being a bit stingy on component ratings, my design was not that cheap unfortunately. But whatever compression driver you choose you face this same problem in my experience. Time alignment is a bloody pain as well with deeper PA horns in passive designs. Active DSP all the way (if only)

Edited November 6, 2013 by LawrenceH

[Phil Starr]

Mine are IV's too but they have the same drivers as the S112V's I used to borrow before I bought them, They sound the same too.

For a long time I liked the mid-range peak as it punched the vocals right forward, and they by no means sounded bad in practice, a bit harsh when the room was empty but vocal articulation came over well when the room was packed. The trouble came when I added subs and started putting the whole mix through, the guitars always sounded harsh. We always had limited gain before feedback too because of the peaky response.

It is standard practice in hi-fi cabs with 2nd order crossovers to reverse the tweeter because of the phase shift in the crossover, so I tried it and yes it sounds better both in an A/B comparison and in a live performance.

What did you do about inductors? I found them hard to source and expensive and in the end had them wound by IPLacoustics.

[LawrenceH]

[quote name='Phil Starr' timestamp='1383780972' post='2269185']
We always had limited gain before feedback too because of the peaky response.

It is standard practice in hi-fi cabs with 2nd order crossovers to reverse the tweeter because of the phase shift in the crossover, so I tried it and yes it sounds better both in an A/B comparison and in a live performance.

What did you do about inductors? I found them hard to source and expensive and in the end had them wound by IPLacoustics.
[/quote]

I will have to check the wiring in mine - perhaps they are already wired that way! I've quite often used them alongside other people's monitors and they have almost always given fewer feedback issues, and by ear the response isn't so much peaky as one broad midrange hump. You are convincing me that I should measure them properly when I get the chance.

Inductors were Jantzen from audio-components.co.uk, there were other sites that listed cheaper but the postage for a small order was prohibitive. I used air core on the high-pass which was around 0.5mH, and 'p-core' for the woofer at 1.5mH. £20 for 2 of each, and another £30 for the other bits so £25 per crossover. I wish I had a better source of these bits!

[Phil Starr]

Our singer wasn't very loud, so I guess we used more gain than most.

When you can pay as much for an inductor as you do for a much bigger mains transformer you begin to wonder about the cost, but they are difficult to track down at all, I wondered about winding my own but I found it difficult to track down cores for the higher values.

My S112's both had the Delta12A driver [url="http://www.eminence.com/pdf/Delta_12A.pdf"]http://www.eminence.com/pdf/Delta_12A.pdf[/url] I'd rejected these as PA speakers in my home brews because of the peak, The cone break up is pretty uncontrolled it seems and i decided that designing a crossover to tame it was going to be too difficult.

Back on track though, any other suggestions for a small but sweet sounding compression driver capable of 120dB?

[LawrenceH]

This Celestion might be worth a look for a screw-on solution, it's pretty new to their range which might be a good thing in relation to previous generation units:
http://www.thomann.de/gb/celestion_cdx1_1010.htm

It's another with a polyester diaphragm, from what I've read these reportedly sound a bit less harsh than a lot of titanium designs when driven hard, certainly that correlates with my subjective impression of the 18sound - n of 1 though!