Chris661's PA system thread

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Depends on your definitions. That's when the coil just about starts to leave the gap, but you can keep a useful amount of drive level with some of the coil outside.
I can say that the Beyma units stay pretty linear up to 15mm one-way, at which point distortion starts to become obvious.

Chris
 
Bumping again, as I've got the pair of 6" tops up and running for some tests.

The good:
- Really staggeringly loud.
- Sound pretty good, even loud.
- Small and light

The bad:
- Nothing below 80Hz

They definitely need bass support (getting the 2x6" cabs with probably-B&C drivers in this afternoon) if you want a full-range sound, but for acoustic music they're fine as-is.

For now, I need to add things like a pole mount, handles, SpeakOns, and grilles.
Passive crossover parts are on their way, and the crossover sim looks pretty good to me! 3rd order highpass, 2nd order lowpass. The horn has a big bump in the 1-2kHz range, so I've set the first pole of the highpass to counteract that, with the second poles providing additional attenuation lower down. The lowpass is pretty standard. The result will be an acoustic crossover of 1.6kHz, but that might all change when I've finished tweaking the values.
The Das M3 driver is good down to 1kHz, and the horn loads down to 700Hz, so I think it'll be fine to keep up with a single 6.5" driver. If it was going to be asked to keep up with a 12" unit at high power, I'd certainly be looking at a higher crossover point.

Chris
 
Back again. This is turning out to be a lot of work.

The big problem is this:

59f09e_51a2333124594736a9b0186f97d1b689~mv2.png


This is a full-range sweep of the 6.5" driver. Note the absolute mess below 1.5kHz - that's all port contribution, and when you notice it, it doesn't sound great.

Here's what happens if you stuff the port:
59f09e_3cb7d47b64894baebd0b8f598af55129~mv2.png


Much much better. There's still a bit of something at 400Hz, as well as the cone breakup at 5kHz, but it's something to aim for.

This is what happens if you add some wool carpet underlay as lining:
59f09e_ed54bd2ecb8a49a0ae949c31a54c9e98~mv2.png


Not wonderful, but a step in the right direction.

After getting fed up with constantly opening up the speakers and doing more measurements, I fired up Hornresp for a play around. Simulated as an offset-driver transmission line, and got the midrange peaks and dips to roughly line up.

59f09e_d872641e77e246cd939e6b60d146de72~mv2.jpg


Looks like some dense stuffing in the top half is going to be the way to do it. That'll be my next port-of-call, so I'll do some measurements and report back.

Apart from that, the drivers that look like B&C units aren't 6PS38s, but are similar. Their T/S parameters lend themselves to larger boxes, tuned quite low. When the last cabinet arrives, I'll have a total of six of those drivers, so it'll be three per side in some low-profile subs. They'll likely be ported and tuned in the 40s.

Each driver is 16ohm, so each sub will have 3 drivers wired in parallel, and if I'm running the NU4-6000, one channel will drive both sub cabinets at around 2ohm. Two channels for the tops, and the last channel would drive a couple of wedges. Nice little setup.

Chris

PS - it just so happens that the cabinets with the 2x 6" drivers in there also have a B&C DE250 on an 18Sound horn. I'll be comparing the DE250 with the Das M3.
 
The idea for the stuffing layout from Hornresp didn't work as planned. The reason being, from what I can tell is that while it's pretty close in a lot of ways, I couldn't get the model to be exactly the same as the cabinet, which has square or rectangular cross-sections. The problem was simply that 400-1kHz was making it directly out of the port. No standing waves or anything like that, just direct radiation from the back of the cone.

So, how do we eliminate that?

It's simple - you put a whole load of absorbing material in the direct path between the driver and the port. Here's the clever bit, though - you ensure there's still a path that goes from driver to cabinet volume to port.

In this particular cabinet, that path goes from the back of the driver up into the area around the horn, and then back down to the port. Directly between the driver and the port, there's quite a lot of recycled wool in the form of 10mm carpet underlay. Its important that the path around the wool is unobstructed, which will allow the bass response to remain unaffected. When the top half of the cabinet was stuffed with polyfill, around 3dB of output was lost in the low-frequency region.

Results:
59f09e_be790745313d4c739e56827710f99d85~mv2.png


This is comparing port blocked vs open, both with that internal treatment. Compared to the earlier graphs, this is a fantastic result.

I also threw together a very quick sketch of what the finished PA will look like, for some sense of scale:

59f09e_68d3a5e2a6e948d5ae832a0bd5ad49dc~mv2.png


If Hornresp is right about the output levels (and it's usually very close), you're looking at >115dB at 1m from around 40Hz upwards, per side, in something that ought to fit in the boot of a small car. I'm rather looking forward to having this built and tested.

Crossover parts have arrived. Might get a bit of work done with those on Friday. Since I designed the original crossover I've been playing some more with the software, and it looks like adding a notch bottomless notch filter to completely kill the 5kHz cone breakup might be worth doing. Fortunately, the speakers with B&C units have plenty of crossover parts to play with, so I'll pull a few parts, measure them, and go from there.

Chris
 
Got the first passive cabinet up and running.

After a lot of arguing with my laptop, I've got something I'm happy with.

First up, the arguing:
The software I'm using to design the crossovers is XSim, and I'm using REW to measure the results.
I found that, when I changed a component in XSim, and then swapped that component in (after measuring every component), the change that was measured was not the same as what the XSim would suggest.

So, some degree of trial and error was needed. The result, however, is something I'm very pleased with.

Crossover schematic that I finished up with:

59f09e_6584f9011fcc4b4e99ece22569220ca7~mv2.jpg


Pretty standard, though note the following:

The first component of each crossover (C1 & L2) are different to you'd want for a textbook crossover. The reason for this is that the response of each driver actually rises towards the crossover, so throwing a standard Linkwitz-Riley (or whatever) crossover at it would give a response that's exceptionally midrange-heavy. The large inductor builds in some BSC, and the small capacitor starts the rolloff quite high up, which levels out (to some extent) the natural response of a flat driver in a constant-directivity horn. Since you've started to electrically filter out the response quite high up, the HF unit is having a fairly easy time while reaching a fairly low crossover point.
The rest of the filter is there to keep the rolloff beyond the crossover point nice and steep, so you're not getting loads of out-of-band signals hitting the drivers.

The 5kHz breakup on the 6.5" unit is around 20dB down, though I'm going to add a notch filter to see if there's any improvement.

Turned out the supplier didn't send all the parts I ordered, so I'm hoping they'll send the rest soon, so I can have both speakers up and running.


With all the playing around in the cabinets, port tuning has dropped to 60Hz, and the cabinets sound quite solid when used on their own. The downside, of course, is more excursion used in the 80-100Hz region, so I might take the crossover up to 120-140Hz ish once the subs are built.

For now, the single speaker that's working sounds very nice.

Cheers,
Chris
 
They're both up and running. Passive crossovers installed, cabinets finished with the usual rosewood stain. All that's left to do is cut the grilles, which will be done when I'm not waiting in for a boiler engineer. Again.

Didn't bother with a notch filter in the end. The midrange is nice and clean, and there's not a whole lot of space in there to mount the crossover - it's gone to one side of the HF horns, and it's tight. I changed the 7ohm resistor (which was a 4R7 and a 2R2 in series) for a single 8R2 to keep things neat. The result has been a slight decrease in the 2kHz-upwards range. Acoustic crossover comes in around 1.6kHz.

Unprocessed, they sound good. I think the drive units integrate very well - I can't tell it's two drivers working. Midrange is definitely the strong point - they don't do a lot of bass, and treble could use a little more sparkle.
I've found that a parametric boost of 20kHz, Q=1, +5dB, improves things. It goes from a slightly-soft sound to a much flatter overall sound. Easy enough to add that sort of processing in at the digital desk or DSP unit, but they sound good enough for PA work without always needing that HF boost. Most dynamic mics top out around 10-12kHz, anyway.
The low-frequency response isn't bad for a cabinet of this size and efficiency. There's enough there to know what the kick drum and bass guitar is doing, and things like Fleetwood Mac's Dreams aren't lacking much in a smaller room, but if you play something that goes down low, the lack of extension becomes immediately apparent.

Measurements coming when I can get outside. For now, I'm enjoying these as the main speakers in the living room. I suspect the amplifier (NU3000DSP) isn't doing them any favours as a HiFi amp, but things like double-tracked vocals are immediately very obvious.

Pictures on their way.

Chris
 
Some photos now I've attached the grilles:

59f09e_6a624f5a5906407b9afe133610978a86~mv2_d_5312_2988_s_4_2.jpg


No grille:
59f09e_deaabbb382ae433f8f49749cb88278c0~mv2_d_2988_5312_s_4_2.jpg


Back:
59f09e_cf5fcbf6512e4ddaa1b551ad623a00a0~mv2_d_2988_5312_s_4_2.jpg


Bottom, with second SpeakOn connector:
59f09e_efe7dd24ef0041fba5b0aea46c5a9335~mv2_d_5312_2988_s_4_2.jpg


One more:
59f09e_6b7d816bcc274da699fbd1c7ff84fce0~mv2_d_2988_5312_s_4_2.jpg



I'm rather liking these cabinets. Smallish, pretty light, sound good. All being well, their first outing will be on Wednesday (open mic night - acoustic artists), and another on Thursday (delays/fills as part of a larger system).

Chris
 
Nice!
it's Always enetertaining and educational to follow your thought and build process, the measurements with and without stuffing in particular.
Keep it coming, your contribution is the one i look for in the - quite slow - PA subforum!

Cheers, jo. I was beginning to wonder if I was talking to myself.


Currently they're back in the living room, hooked up to an old Akai AM-2250 that's being fed by a GL-70. Sounds rather nice, though I know the amp is more than a little rough around the edges.

Hoping to start construction on the 3x6" bass modules next week. Should end up a nice little PA system.

Chris
 
Midrange is definitely the strong point - they don't do a lot of bass, and treble could use a little more sparkle.
I've found that a parametric boost of 20kHz, Q=1, +5dB, improves things. It goes from a slightly-soft sound to a much flatter overall sound.
Chris,

You could install a "bypass cap" ahead of the HF resistor to the HF driver. A small value around .5 uF should do the trick, and bring the VHF up to the level it would be without the resistor's attenuation.

Probably not worth messing with since you use processing, but gives the "sparkle" with less power wasted in the resistor.

Art
 
First gig with the little speakers last night.

Setup was this:
2x little speakers
2x 8" coaxial monitors
1x Behringer NU4-6000, running one channel per speaker
QSC TM16 desk, EV mics, etc etc. Processing at the desk was 60Hz highpass filter, +3dB at 20kHz.

Acoustic music plus some spoken word stuff - open mic night at a regular venue.
The stage in this particular pub is on the corner of an L-shape, so it was one speaker firing along each length.

They did well. Nudged the red lights once or twice during the gig (and was asked to turn it down when it got that loud), but they sounded good and filled the space nicely. I was rather hoping someone would bring a cajon or something with some low end, but it wasn't to be this time.
No obvious signs of nastiness when they got loud, and no trouble with feedback, either, which was a concern given the small cabinets.


59f09e_860ea5ae0e8c4c129b138687cb13d200~mv2_d_5312_2988_s_4_2.jpg


59f09e_4cbade4d6b214cb088fd1a7d11ee3c7b~mv2_d_2988_5312_s_4_2.jpg


Rather pleased with these. Today, they're going to see use as fills/delays with the bigger PA system (2x10" tops, 15" subs).

Chris
 
Back again. Yesterday was the second gig for the little speakers. This time, I took them for fill usage in a difficult room.

Program material was mostly spoken word, but they also had dancers. Venue was an Indian restaurant in Sheffield city centre. Everyone was very pleased with the sound, so I'm optimistic about getting more work there.


Main PA covered the length of the room quite well, even in a further seating area that held another 50 or so people. I had wanted to put a speaker in the far seating area, but the areas immediately to the side of the stage needed more attention.

59f09e_0987021fff5648d88a7b4823337446f9~mv2_d_5312_2988_s_4_2.jpg


Then the little speakers came in to fill in the areas to the sides/rear.
59f09e_a3ee68e723cc4beda8e9bbc12f0aa0af~mv2_d_5312_2988_s_4_2.jpg


59f09e_528e140bf82e4c97955f28f022c7cfa8~mv2_d_5312_2988_s_4_2.jpg


Again, they did a good job. That time was more about bringing clarity to the other seats, while the main PA carried most of the work.

So yes, I rather like these. They're back in the lounge playing music again.

Chris
 
Didn't take any pictures, but they were used last Saturday at the same pub. I was hoping for a band, but it was a singer/guitar act again.

This time, the amplifier was an Ecler DPA2000, which puts out around 500w/ch into 8ohm. Didn't see clipping this time, but it's one of those class H amps with a big ol' transformer in the middle, so in the interest of keeping my amp racks reasonably light, I won't be using that again. Perhaps the MA12000i...

Chris
 
Another update - 4-piece band - bass, acoustic guitar, violin, cajon/sax. Everything mic'd or DI'd, in the venue shown further up the page.
Just the 6.5" tops again, plus a couple of 8" monitors.

Did the job nicely, on the NU4-6000. Loud enough for a moderate Saturday night, and enough grunt to feel the bass from the cajon if you're near the speakers.

Really pleased with these little boxes. The 3x6" subs are on their way, though.

Chris
 
Congratulations.
Look very good and I must trust you about the sound :)

Thanks!
Still need to get outside for some proper measurements - I much prefer graphs to audiophile-speak.
I tried to get a short video, but the mic on my phone was clipping, showing in excess of 100dB when within a few metres of the speakers.

Love the look of the stained (oiled?) cabinets!


Reminds me of my old Bag End TA-12C's.

Ah.... youth.

Stained - rosewood stain straight on to the ply. Nice and easy to do, touch-up is simple, and it's a good way of standing out in a crowded market.

Cheers, all.

General update:
It's been a busy couple of weeks, including a musical at a theatre for a week followed by a big-name artist, but I'm hoping to have some time this week to take the newly-constructed 3x6" sub and a 1x6" top outside for some measurements. I'll be testing how well they integrate together using just the mixing desk (QSC TM16) output processing (3rd order filters plus parametric EQs). I'd really like it if I could use the desk and a four-channel amplifier to do all the work for a pair of tops, pair of subs, and a monitor mix.

Photo from the big-name gig:
59f09e_9d29626498914e10b20a305e81598fd9~mv2_d_3840_2160_s_2.jpg


Chris
 
The saga continues...

I've had a thread discussing the 3x6" bass module running in the Subwoofers forum for a little while, but here's the tail end of it, after the design choices had been made:
Small PA Subwoofer - 3x 6.5" drivers.

Here's a quick run-down of the 3x6's performance:

- 117dB maximum output before excessive compression and distortion
- Flat to around 55Hz with some EQ.
- Dimensions: 220mm height, 570mm square on top.
- Crossover: 170Hz

59f09e_ffc0e9bcce7e4dc8bdc4d588f56b3ed5~mv2_d_2012_2486_s_2.jpg


There's a few things left to do to get it usable as a complete PA system:
- Finish first 3x6 cabinet
- Build second 3x6 cabinet
- Buy some longer speaker poles
- Patch panel for the amplifier

With regards to amplification for this system, here's the plan:
- NU4-6000:
Ch1 - Main left
Ch2 - Main right
Ch3 - Both subwoofers
Ch4 - monitor(s)

I'll need to make up a patch panel, as the subs take a 4-pole input and pass the mid-high signal out.

Processing will be done in the QSC desk, using an Aux send for the sub feed.

The primary use case for this PA system will be gigs with acoustic instruments, spoken word, and smaller parties.

Chris
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.