kipman725

The level seems quite low. I would expect the noise to be constant with level so SNR should improve if a larger signal is used. The Fireface UC is spec: THD+N AD: < -98 dB (< 0.0012 %), THD+N DA: -96 dB (0.0015 %) so you're getting pretty close.* *didn't see your last post I think optimal level for the Fireface will be around 4V from your results.

I did some loopback tests of a presonus firebox, they are ~7dB worse than your results for noise and distortion. However your distortion is much lower. https://www.audiosciencereview.com/forum/index.php?threads/presonus-firebox-24bit-96khz-loopback-measurments.24565/#post-828710 should be more than adequate for PA amplifier characterization.

Hi @Ricci have you tried any horns with exponential throat sections with these mid-range drivers? my understanding is that any horn which doesn't have collapsing HF directivity (conic or OS) will have poor loading at low frequency which (I hope) is what makes them sound bad when run low. I am planning a horn that is used in PA down to 400Hz but it only needs to get to 4kHz so I am basically designing a CE horn (using M200 drivers): http://www.xlrtechs.com/dbkeele.com/PDF/Keele (1975-05 AES Preprint) - Whats So Sacred Exp Horns.pdf There are some nice loading graphs on page two of this: https://www.grc.com/acoustics/An-Introduction-to-Horn-Theory.pdf

Also Klipsch are building a very similar speaker as there new flagship: The lense that they add to the AXI2050 looks interesting!

If you have a lot of money you can purchase the K402 from Klipsch as a cinema replacement part with a driver attached. The K402 isn't actually that fancy a horn though its basically a huge conical horn with a mouth flare. A dual slope conic that is typically used in unity horn designs of the same size will approximate its behavior. This a spreadsheet that can design a horn for you: http://www.libinst.com/SynergyCalc/Synergy Calc V5.pdf This person built a more complex replica incorporating the mouth flare: https://www.stereonet.com/forums/topic/232459-klipsch-k402-replica-build/ I've been designing mid-range horns recently and would caution though that conic horns like the K402 do not provide good loading to throat mounted compression drivers low in frequency compared to there size. This may mean that it might not sound nice crossing at 300Hz, although the AXI2050 has huge surface area compared to a typical comp and may be OK.

Impressive ports. The 15DS should be one of the very best 15" drivers. The best way to find tuning is to measure the impedance of the box W.R.T frequency where it will be at the minimum value between the two peaks. If you only have a microphone you could also try looking for a notch in the near field of the woofer response and peak in the port response but this may be complicated by their close proximity.

successfully used this tool to do some tricky alignments like a 100-400Hz horn with a cluster of 100-400Hz horns of different length. With a decent initial guess (tape measure) you can easily see what delay gives the best result.

cool I was unaware of this tool

Yes all of your results make sense. The basic issue is that your using a driver with relatively low xmax in a low tuned box and as such its excursion limited rather than power limited. There exist drivers with 15-20mm of linear excursion capability that would be able to get much louder in a low tuned box. The Hornresp simulation assumes linear behavior whereas in real life increasing the drive voltage beyond which causes xmax will result in less than the predicted excursion. So when you observe 13mm peak excursion its unlikely the driver would get that far and may be undamaged, it probably won't sound great though. The xmech for this driver is listed as 16.5 mm but it might be almost impossible to reach as the coil will be far beyond the magnetic gap and thus subject to little force before excursion reaches this point. In this thread I measure the excursion capability of a driver, you can see on the scope screen that the driver position measurement starts to 'flat top', increasing drive beyond this point just doesn't result in anymore excursion, just more heating in the voice coil.

this should help with your generator issues as the measurements I have seen of Powersoft amps(x4/x8) show sinusoidal in phase mains current whereas reports on the NU/NX6000 indicate large current spikes.

Series would be 8 ohm and parallel 2 ohm. 2 ohm will cause greater heating of the amplifier but will increase the maximum possible power output. The datasheet specifies that the heatsink temperature should not be allowed to exceed 95C with the note that reduced performance may occur at this temp. I would suggest installing temperature controlled fans, cheap fan controllers with temp sensors that drive PC 12V fans are readily available.

https://www.comsol.com/paper/download/679311/bezzola_paper.pdf might be of interest for a 'fully optimized' design

I think from the datasheet the 18mm is when BL=0.7*MAX(BL). Looks like a great deal for the price, shame that price will not be anything like that in the UK. I wonder how it would do in a FLH like the lab horn?

I'm interested in this popplar plywood carbon-epoxy layer construction, do you have any details of the products used?

Boxes are isobaric bass reflex with all 4 voice coils over two drivers in series making a 16 ohm nominal box. I observe up to 2dB power compression on a 94V sweep but its at the frequencies where excursion is maximized, I don't think I can see any heating effect in the data I have. I don't have data for higher voltage sweeps because that's as much as I could get out of the amp I had at that time. I used to run these drivers in sealed cabs and this resulted in the motor structure glue failing jamming the voice coil, the voice coils suffered mechanical damage (scrapes) but didn't burn. This makes me think that the power handling is way better than a 2.5" voice coil made using conventional pro audio voice coil formers like kapton. Due to the design of my boxes I can touch the magnet of the driver during use of the outer isobaric driver and found that it was barley above ambient when driving all night from a P7000s channel (77Vrms). All suggests I should do more comprehensive testing but I live in a terraced house so this is difficult.

I want to define some long term power limiters for my subs using Pyle plpw15D drivers as I'm going to hook them up to a bigger amp. However as they are car audio drivers with non PA coils I'm not sure what a sensible long term power limit would be: the former is alu which should increase power dissipation capabilities, 2.5" dia, 4 layer with two voice coils. They have a vented pole peice. The manufacturer rates the driver at 1000Wrms which from rules of thumb I have seen would suggest around 500W as a sensible long term limit. I also found this chart from powersoft: from which I would guess the closest is 3" mid-bass giving a long term limit of 150W? (Per driver, my subs use two drivers in series)

High power closed box doesn't work well for PA the air inside the box gets hot so after an hour or so you will get noticeable reduction in output, I have tried it and changed to a design closer to ported. Some reduction in box size should be possible by accepting a non flat un-equalised response.

https://www.powersoft.com/wp-content/uploads/2018/04/Powersoft_X8_DS_EN_WEB.pdf So you have the choice of running the two channels bridged or as non bridged channels. Bridged is rated down to 4 ohms so you can't run all 4 drivers off this. Then looking at single channels maximum rated power is at 2ohms per channel so you should put two 4 ohm coil drivers on each channel. This excludes the IPAL leaving the SW or DS, the DS has a more powerful motor and slightly higher XVAR so should be capable of more output.

The configuration I have used is the right hand side one with a box that has 3 speakon connectors (input and two outputs). I just used a plastic project box and held it all together with a lot of hot glue.

I have run subs in series with an external "series speakon adaptor" box I made without any issues. I think as long as you are using identical subs there shouldn't be any issue running them in series as speakers are fundamentally current mode devices. This is also why I run the drivers in series in my isobaric subs as this means that the force exerted by each motor should be equal.

oof. 2 sub version of that stack would be nice at home.

Thanks for the info! I had already designed a sub with just pipe ports to try and get something built quickly, however I needed to redesign it as my model would break on changing the material thickness parameter and it was overweight for the application (ultra portable human carried ~5km) resulting in a change from 18mm to 12mm (yes I know 12tbx100 is heavy, but I got them cheaply and this is the easiest thing to change). But then I realized how much output at tune I was potentially giving up (JBL HLA port design information indicates around 9dB at tune). In the future I would be interested in a even more port optimization but its a bit outside my area of expertise as I'm an electronic engineer and my only physics is in electromagnetics. From the off the shelf options ( https://www.l-acoustics.com/en/product/ks28/# ) it looks like your on the right track. If you wanted to try lots of port designs the best approach for testing would be to have a box that allowed attaching ports to the outside so that each test only required building the port and not a whole box. Size is very constrained for these subs as they have to go on someones back and also fit under my desk at home.

ah ok I have only used 1/2 of the port here so I guess the height is less than I expected?

I liked this port so much I'm basically copying it for my 12TBX100 sub. The height wasn't listed so I guessed it as 44mm from the pictures (3*18mm boards with 10mm of rebate). Tuning came out at 38Hz. How close did I get? I'm also guessing from how your Hornresp response compares to mine that you used complex inductance parameters?

very tidy!