active 2 way speaker using FaitalPro HF144 and Lorantz C390X B1 continued...

[zog]

Following on from this post in the what are you building thread I thought I'd make a new thread as its gone on for a few posts.

 

I did some "directivity" type plots of both the LTH142 horn and PH916 waveguide outdoors, using the lazy susan. Although my garden is pretty small it still is more open than my loungeroom.

 

These were done at 2M distance, with no crossover except the 66uf protection capacitor, and smoothed to 12/octave for readability.

Each reading is done rotated 7.5 degrees off axis, so I've got from 0 degrees (direct on axis) to 60 degrees available:

 

 

 

 

Now I'm using a miniDSP EQ correction that makes the PH916 fairly flat on axis.. and I forgot to change this for the LTH142 when I switched over to this, which measures louder in the high frequencies.

 

But it appears the PH916 has mostly better directivity as the FR falls off slower (narrower spread of lines), except it has more chaotic behaviour (notches appearing) above 12khz, while for the LTH142 although it has a notch on axis.. it actually gets smoother off axis at these high frequencies.

 

I'm now toeing in my speakers @ 22.5 degrees to the listening position (and have adjusted the miniDSP EQ to suit), but as Paul mentioned getting the measurements across different angles has opened my eyes a bit.

 

I'm still preferring the PH916, though it looks to me like for a more beamy response the LTH142 will actually give me more even high end frequency response.

 

I'm currently set up crossing over at 700hz with a somewhat complicated MiniDSP EQ curve flattening out the response from the 800-3000 region. Still learning as I go here, but it's sounding pretty good.

 

[davewantsmoore]

Good going!   :-D

 

 

But it appears the PH916 has mostly better directivity as the FR falls off slower (narrower spread of lines),

 

Right.   What you are looking for (well maybe you are) is the distance between each line to be the same at each frequency ... That means that when each driver is EQed perfectly flat on axis .... that the off axis response will be closer to flat  (just lower in level).

 

One thing that might help the analysis is to mount the horn in a very large "baffle" to see past any diffraction effects.

 

 

 

 

I'm now toeing in my speakers @ 22.5 degrees to the listening position

 

I often get confused about what people mean with things like this.   Does this mean you point the speaker straight forwards... and then turn it 22.5deg towards the centre ...... or point the speaker directly at the listener, and then turn it inwards a further 22.5 degree?

 

I'm currently set up crossing over at 700hz with a somewhat complicated MiniDSP EQ curve flattening out the response from the 800-3000 region. Still learning as I go here, but it's sounding pretty good.

 

Ah, this is where I'm talking about the diffraction.   The peaks and dips you are seeing in that region are (mostly!?) diffraction, and you should be very very careful about using EQ on them.       Measuring the driver on some very large make-shift pseudo-giant baffle, even if done roughly will give you an insight into what is going on.

 

 

My gut instinct is that that is too low a crossover point ....   but I'm probably just shootin' the breeze now   ;-)

[davewantsmoore]

Lorantz C390X B1

 

 

Woah.   Nice t/s specs for this applicaton, but rather expensive.

[Paul Spencer]

Looks like I was expecting, with a narrowing pattern to HF. With the horn you also see a dip on axis that disappears after about 10 degrees - something to consider in your crossover.

[zog]

Good going!   :-D

 

 

 

Right.   What you are looking for (well maybe you are) is the distance between each line to be the same at each frequency ... That means that when each driver is EQed perfectly flat on axis .... that the off axis response will be closer to flat  (just lower in level).

yes, thats my thinking with the speakers almost pointing forward but not quite.. the 15-30 degree range, where the FR seems reasonably consistent except for the levels, is where I'll EQ for people to be listening at.

 

One thing that might help the analysis is to mount the horn in a very large "baffle" to see past any diffraction effects.

 

 

 

 

 

I often get confused about what people mean with things like this.   Does this mean you point the speaker straight forwards... and then turn it 22.5deg towards the centre ...... or point the speaker directly at the listener, and then turn it inwards a further 22.5 degree?

 

See carefully constructed technical drawing I have attached to what my idea of 22.5 degree toe in means.. effectively in my loungeroom the speakers are almost pointing straight ahead.

 

 

Ah, this is where I'm talking about the diffraction.   The peaks and dips you are seeing in that region are (mostly!?) diffraction, and you should be very very careful about using EQ on them.       Measuring the driver on some very large make-shift pseudo-giant baffle, even if done roughly will give you an insight into what is going on.

 

 

My gut instinct is that that is too low a crossover point ....   but I'm probably just shootin' the breeze now   ;-)

yes the Pyle waveguide is specced @ from 1000hz, I'm just playing around to see if I can go lower. I will try 1000hz as well.

This is where the miniDSP is a major win.. with my Econowaves I was soldering up new crossovers and going through piles of capacitors, resistors  etc just to do minor crossover adjustments. Now I can dial in stuff on a whim just to check it out.

 

re: big baffle  - this could be a use of some big cardboard boxes I have before I throw them out, good idea.

 

re: Lorantz C390X B1

Woah.   Nice t/s specs for this applicaton, but rather expensive.

 

I had a number drivers from B&C, Beyma, RCF on the short list.. I'd also looked at Lorantz but too pricy for me.. then got lucky with an Ebay auction on some second hand ones.

I'm quite chuffed after checking out the exotic foreign  speakers, I've ended up with quality units made a few km from where I live.

 

 

Looks like I was expecting, with a narrowing pattern to HF. With the horn you also see a dip on axis that disappears after about 10 degrees - something to consider in your crossover.

 

Do you mean the dip at about 15khz on the LTH142 graph?

 

next step: experiment with time alignment of the drivers, and look more closely at the phase, which I've been putting to one side for now, including some reading up.

then: get some woofertest3 measurements of the actual T-S parameters of the Lorantz woofers so I can start working on bass reflex enclosure ideas.

Edited April 11, 2013 by zog

[zog]

Found this while googling about - except for the passive crossover and wooden waveguide sort of a rough idea of the kind of enclosure layout I'm aiming for, with the upper waveguide enclosure separate to I can try different waveguides/comp drivers:

 

 

http://www.sq-products.com/sq_H15_Speaker.htm

 

I'm also amused comparing

http://www.enjoythemusic.com/superioraudio/equipment/0410/sq_products_sq_h10.htm with my older speakers:  http://pix.minirig.org.au/main.php?g2_itemId=4303

Edited April 13, 2013 by zog

[Newman]

I'm also amused comparing

http://www.enjoythemusic.com/superioraudio/equipment/0410/sq_products_sq_h10.htm with my older speakers:  http://pix.minirig.org.au/main.php?g2_itemId=4303

 

That is amusing, heh heh. You should claim prior art!

[Newman]

If you are going to use EQ, I would suggest a CD horn.

 

The exponential and other flat-response horns, with their inherent poor directivity, are compromised when you consider the clear research showing a home audio listener preference for good directivity. I don't care if they have exotic French or Japanese surnames attached; the sole justification for these horns is if you use full-range very low power amplifiers that can't cope with delivering an EQ curve. Then one has no option but to settle for second best in one's horn profile (but an exotic name seems to offer some compensation).

 

[zog]

If you are going to use EQ, I would suggest a CD horn.

 

I am using a Pyle CD horn at the moment.. so far I think I like the sound better. And I'm coming from using more "classic" design econowaves which are also CD.

 

The exponential and other flat-response horns, with their inherent poor directivity, are compromised when you consider the clear research showing a home audio listener preference for good directivity. I don't care if they have exotic French or Japanese surnames attached; the sole justification for these horns is if you use full-range very low power amplifiers that can't cope with delivering an EQ curve. Then one has no option but to settle for second best in one's horn profile (but an exotic name seems to offer some compensation).

 

I will run the numbers to see if it's worth doing a passive crossover once my configuration settles down, but with my current layout (approx 130W tripath amp on LF, 50W tripath amp on HF, via miniDSP) I have lots of headroom and EQ flexibility.

[zog]

Apart from playing with miniDSP curves and gradually tuning in a sound I like,, I've measured the T/S parameters of the Lorantz woofers so I can spec out a "keep it simple stupid" bass reflex cabinet get the best bass from these speakers, with something a non-woodworking person like myself can do by getting the bits of wood cut to spec for a simple glue and clamp.

 

I found my measured parameters differ from those on the Lorantz spec sheet slightly - but the transfer functions on the boxes I have been looking at only vary by about 1dB between my measured numbers and the Lorantz numbers so no matter which ones are more correct I should be OK.

 

I'm thinking now of a 60cm wide, 75cm high, 40cm deep cabinet (138.9 litres) tuned to 33hz. The transfer function is fairly non linear (high f3?), and I'm the box doesnt allow the full 400W power of the driver to be used, but I only have a 130W or so amp. I'm already doing a fair bit of miniDSP correction so I think I can live with the 6dB rolloff. I'm still a bit of a newb with WinISD and Unibox however.

 

I'm hoping visually the boxes will look a bit like those older wide-not-deep cabinets ala Altec, etc... but with the driver up high to be close to the waveguide, and 2 10cm plastic bass reflex ports. This time I'll upgrade to using plywood, as people seem to consider it sonically better and as these might be in the loungeroom for the long haul I can work on getting a nice finish happening down the track.

 

 

excessive graph posting and crappy sketches here:

http://minirig.org.au/2013/06/02/cabinet-ideas-for-the-lorantz-c390x-b1-8-woofer/

Edited June 2, 2013 by zog

[svenr]

With all respect, the two horns you've tested are 30+ year old designs. Their directivity behaviour is ok for their original design purpose (clustered sound reinforcement systems), but ansolutely outdated by modern engineering standards.  Both comprise very distinct resonances below 3 kHz which will cause significant colouration and advert the use of a suitably low crossover frequency.  I recommend using an oblate spheroidal horn design which gives you better directivity and significantly reduced resonance at the lower cut off frequency.  One candidate I have had good experience with is the 18Sound XT1464.

 

Have you asked Michail B for the directivity plot of the 15" driver?  Looking at his cone design, their could be a reasonable chance that the membrane's directivity is somewhat controlled up to 800Hz .. 1kHz - I strongly suggest checking on this to see whether there is a realistic chance of pairing this 15" with a single 1.4" horn without risking a massive glitch in either the directivity or the power response.

 

And also - as you seem to take this serious - use Floyd Toole's approach of spatially averaged response curves (see "Sound Reproduction" for full references) when designing the crossover.

[zog]

With all respect, the two horns you've tested are 30+ year old designs. Their directivity behaviour is ok for their original design purpose (clustered sound reinforcement systems), but ansolutely outdated by modern engineering standards.  Both comprise very distinct resonances below 3 kHz which will cause significant colouration and advert the use of a suitably low crossover frequency.  I recommend using an oblate spheroidal horn design which gives you better directivity and significantly reduced resonance at the lower cut off frequency.  One candidate I have had good experience with is the 18Sound XT1464.

 

Thanks for the info - Paul @ redspade also recently pointed me at the 18sound waveguides, so I will take a look! Budgetwise I have some flex as I landed the 15" drivers considerably under budget thanks to ebay.

 

One reason I'm leaving this "modular" (WG sitting on top) is so I can change directions if something better comes along I can try without too much "rework".

I'd chosen the Faital WG as it is featured in the "Cornscala" design,  which is avowadly "retro", and then just purchased the Pyle WG as it was not expensive and claims to be constant directivity.

 

Have you asked Michail B for the directivity plot of the 15" driver?  Looking at his cone design, their could be a reasonable chance that the membrane's directivity is somewhat controlled up to 800Hz .. 1kHz - I strongly suggest checking on this to see whether there is a realistic chance of pairing this 15" with a single 1.4" horn without risking a massive glitch in either the directivity or the power response.

 

And also - as you seem to take this serious - use Floyd Toole's approach of spatially averaged response curves (see "Sound Reproduction" for full references) when designing the crossover.

 

I have not contacted the manufacturer for further info on the drivers, mainly as I go them inexpensively second (or even third) hand on ebay and considered I can measure them myself (though maybe not particularly well) - http://pix.minirig.org.au/main.php?g2_itemId=4726&g2_imageViewsIndex=2

 

Funny you should mention the Floyd's "Sound Reproduction" as I was just reading the Amazon page for it while reading up on how to make my approach more "correct".

Kindle edition available as well, excellent.

The speakers are working now as works in progress, and I have considerable scope for non destructive alterations, so I'm happy to take my time learning and revising as I go.

[zog]

Ok... I ordered them on the 6th of this month, but they FINALLY showed up (ordered from Germany,/www.thomann.de ):

 

 

A pair of 18Sound XT1464 waveguides.

 

 

 

I'm starting to build up quite a collection now :-)

 

 

 

Annoyingly,  despite the 18sound web site exhortations of super high tech design this did not extend to having the same tolerance for the spacing of the holes as the other 2 waveguides and the compression driver.. it would not fit on, the measurement between the bolt holes 1mm smaller! after stopping my attempts to force it on as I was worried the plastic would crack, a few minutes work with a file got them fitting fine.

 

Just quickly ran them up with a quick sanity test measurement to get them close in response to the Pyle PH916 slightly off axis, sounding good so far without really trying.

Now back to the books and getting it all done properly.

[Guest atilsley]

Nice project....I enjoyed that FP horn and matching FP comp driver in the Cornscala....Bob Crites gives them a very good wrap.

[svenr]

Good on you. I recommend using the listening window averaging described by Floyd Toole as the basis for your crossover design.

[zog]

Nice project....I enjoyed that FP horn and matching FP comp driver in the Cornscala....Bob Crites gives them a very good wrap.

 

I like the Faital comp driver but I think I preferred even the low cost Pyle horn over the FaitalPro tractrix horn - but yes the two way Cornscala was one of my reference points for starting this. Though I'm going for an active 2 way that mixes some new stuff and old stuff while the Cornscala seems more a 2 way tribute to the classic Klipsch style speakers.

 

Good on you. I recommend using the listening window averaging described by Floyd Toole as the basis for your crossover design.

 

Found a few descriptions of that and it seems, like any clever idea, so obvious in hindsight! - I've been working through a rented Kindle copy of Floyd Toole and I'm finding the early chapters heavy going, though partly although a kindle 4 is great for sequentially read, easily consumed fiction its rubbish for reading a reference work where I'm trying to hop around and look at diagrams and search for stuff. So I've ordered a "real" copy. First paper reading material apart from bills etc I've got in a long time.

 

Did a "from scratch" quick and dirty MiniDSP config, this time crossing over @ 600hz and sounding pretty nice, will definitely do until I sort things out properly.

 

Bunnings paver changed to housebrick for sitting the compression driver on to suit the different size of the waveguide.

Edited June 30, 2013 by zog

[davewantsmoore]

Floyd Toole and I'm finding the early chapters heavy going, though partly although a kindle 4 is great for sequentially read, easily consumed fiction its rubbish for reading a reference work where I'm trying to hop around and look at diagrams and search for stuff. So I've ordered a "real" copy. First paper reading material apart from bills etc I've got in a long time.

 

$ well spent.   Persevere.   Very very good reading.

[svenr]

Floyd's book is like any good textbook - overwhelming in the first place, but much clearer once one took the time to think things through.

 

600 Hz between the mid and the high?  What does the resonance frequency of the driver say to that?  I do not recommend operating the driver below its resonance as its only a matter of time until you slam the diaphragm into the phase plug.

[zog]

600 Hz between the mid and the high?  What does the resonance frequency of the driver say to that?  I do not recommend operating the driver below its resonance as its only a matter of time until you slam the diaphragm into the phase plug.

 

Hopefully I should be OK. when I had a mates WT3 rig on loan I did get that info while I was grabbing the impedance curves:

 

In FaitalPro LTH142 waveguide:

 

* f(s)= 407.10 Hz

* R(e)= 5.67 Ohms

* Z(max)= 40.15 Ohms

* Q(ms)= 8.908

* Q(es)= 1.465

* Q(ts)= 1.258

 

In Pyle PH916 Waveguide:

 

* f(s)= 432.70 Hz

* R(e)= 5.64 Ohms

* Z(max)= 41.25 Ohms

* Q(ms)= 8.049

* Q(es)= 1.274

* Q(ts)= 1.100

 

No waveguide:

 

* f(s)= 463.60 Hz

* R(e)= 5.67 Ohms

* Z(max)= 28.16 Ohms

* Q(ms)= 2.701

* Q(es)= 0.682

* Q(ts)= 0.544

 

I would expect the 18sound waveguide to be closer to the Pyle. I've got the 67uF of protection caps (plus 9dB of LPAD to reduce amp hiss), which works out to about 400hz HPF in case I mess up with miniDSP config.

[Newman]

 I've got the 67uF of protection caps (plus 9dB of LPAD to reduce amp hiss), which works out to about 400hz HPF in case I mess up with miniDSP config.

 

Yes, good idea re protection. Always keep a cap on a compression driver, even after DSP is sorted and you have finalized the system. If you want to be oh-so-clever, size the cap for measured HP at your intended crossover point, then add the rest of the HP slope with DSP. Or, if you want maximum flexibility for playing with DSP, use an oversized cap and do all the crossover HP with DSP.

[zog]

Yes, good idea re protection. Always keep a cap on a compression driver, even after DSP is sorted and you have finalized the system. If you want to be oh-so-clever, size the cap for measured HP at your intended crossover point, then add the rest of the HP slope with DSP. Or, if you want maximum flexibility for playing with DSP, use an oversized cap and do all the crossover HP with DSP.

 

I've gone for oversized cap so I've got full flexibility with the miniDSP.  Except for getting better resistors and capacitors for the HF protection, and gathering all the parts (bass reflex ports from Germany, binding posts from LSK, etc) I haven't done much over winter, hopefully the warmer weather will be more amenable to building the enclosures outside.

As I've already got adequate but ugly enclosures for the LF drivers the system is functional, so I've been taking my time. The loungeroom does look like it's got a pub PA in it now though.

 

As I documented with maybe too much info in http://www.stereo.net.au/forums/index.php?/topic/54175-capacitor-recommendtions/#entry929112 the capacitor thread I'm now using "82uF" protection caps that actually measure about 76uF. I was messing around with piggyback capacitors to identically match the values, but after working out the miniscule difference in the transfer function didn't bother.

 

 

[Newman]

Nice cap, well done.

 

I can see from your plots that the horns roll off well before the cap -- so I fully agree with your approach. I'm guessing that F3 (cap, electrical) will be about 260 Hz.

 

I do like projects like this! Keep us posted, cheers.

[zog]

OK, it's been a long time since an update, but I've finally got around to ordering some pre-cut pine plywood for making better cabinets for the Lorantz woofers.

 

133L (75cm high 60cm wide 40cm deep) which according to WinISD I can tune to approx 33Hz with my PVC port extenders, or approx 37Hz just using the 20cm plastic B.R. ports as is.

 

I haven't clamped and glued yet, just a done quick sanity check that all the bits fit together:

 

 

I've also been dabbling with different amplifiers to drive the HF portion of this system for amusement (SET valve amps, JLH 1969 class A, even $6 chipamps), but so far my TA3020 Tripath amp seems best for the Lorantz woofers.

Edited March 13, 2014 by zog

[henry218]

nice cabinet

woofers like power, i would suggest an amp no less than 150w.

[hochopeper]

Looking good zog!

 

 

nice cabinet

woofers like power, i would suggest an amp no less than 150w.

 

isn't this a 15in woofer with reasonably high sensitivity? And active so it's not going to have a bizarre impedance curve either .... 

Edited March 13, 2014 by hochopeper