I've put up a page on this project here, just in case anyone's interested: The Subwoofer DIY Page v1.1 - Projects : "Proof of Concept #2"
I haven't measured max SPL, but it was enough to get some of the dust off of the ceiling. I wasn't able to get audible distortion with 120W in.
I'm going to be rebuilding the box in a few weeks, to fix a few issues I came across. Will do some more detailed response measurements then.
According to the sim, a bit over 115dB at 21.5 volts into 2PI space. Remember this is a $75 driver in a somewhat compromised (volume-wise) TH! Output before significantly audible distortion is likely going to be a bit higher, seeing that there's a null outside the upper passband that occurs just around where most of the distortion components caused by in-band excursion should show up.
Because the distortion is non linear I don't think it necessarily helps... anyway in between the nulls there's peaks so overall its a wash...
The Subwoofer DIY Page v1.1 - Projects : "Proof of Concept #2"
Nice to see the impedance and spl charts... I subbed your WT3 measured T/S parameters is some of my sims and got a surprisingly similar response to the "original" measurements.
Do you have any more impressions of this driver/horn now? I'm a bit worried that the suspension is very non linear at higher excursions, even below x-max, giving what you said.
Nice to see the impedance and spl charts... I subbed your WT3 measured T/S parameters is some of my sims and got a surprisingly similar response to the "original" measurements.
Do you have any more impressions of this driver/horn now? I'm a bit worried that the suspension is very non linear at higher excursions, even below x-max, giving what you said.
Nearly OT post:
How could PE muff the Le measurement so badly? After going for the shorting cap and all you'd think they'd be interested in showing such a low actual measurement for its effect in midbass/low mid applications.
But then the actual (presumably unbroken-in) reduced Fs alongside the suspension limited excursion seems to turn this custom PE speaker into yet another barely practical weirdo.
Actually, looking at the published specs for this series of speakers made me think, "Hmm copper cap eh? May as well leave it out since it ain't doin nothing."
In fact I recently chose a Beta12A2 over it because of that, but maybe the PA310 really is a decent high power low mid?
How could PE muff the Le measurement so badly? After going for the shorting cap and all you'd think they'd be interested in showing such a low actual measurement for its effect in midbass/low mid applications.
But then the actual (presumably unbroken-in) reduced Fs alongside the suspension limited excursion seems to turn this custom PE speaker into yet another barely practical weirdo.
Actually, looking at the published specs for this series of speakers made me think, "Hmm copper cap eh? May as well leave it out since it ain't doin nothing."
In fact I recently chose a Beta12A2 over it because of that, but maybe the PA310 really is a decent high power low mid?
Unfortunately, I haven't had a good opportunity since that build to give it a good listen. I was planning to do some distortion measurements last week, but work got in the way. I may be able to do something this week, if work doesn't get in the way again. I also have a 6th order BP design for it that I'm planning to build and test as well. It doesn't go as low as the TH, but the box is a lot smaller.
The driver itself seems to be pretty well built, though likely more suitable for midbass/midrange duty. Zilch on the PE forum used it in one of his Econowave designs.
I'm not too concerned about its suspension going non-linear at higher excursions, given the design purpose of this driver.
Missed your reply Brian,
You want it to be linear at least till 5mm naturally, otherwise you won't even get the spl from the sim.
That level of testing isn't going to happen in a hurry - just gave my amp to Dad for Xmas, LOL. His own blew, and my amp was in storage while we retiled the living room, so...
Right now I'm using an old Denon PMA250II - it's not going to provide enough voltage to move the PA310's cone sufficiently for any distortion measurements, except at very low frequencies if I run it free-air. I might hook it up anyway to see what happens.
Hmm... didn't think of that. I've got a sub amp in my car, but it's capable of doing only 125 watts or so into this driver's 8 ohm impedance. I'll give it a try in the coming week.
Well, I tried to do something, I'm not sure how successful it was. Basically hooked POC#2 to the car amp, set the EQ flat, the LP filter to 100 Hz (-18dB/octave) and set the output to 0dB. I then used my TrueRTA on my netbook to generate test tones (sent to the car deck via BlueTooth) and record the response from the sub (using an ECM8000 with a USB interface). The sub was placed on my driveway and the mic was placed on the ground, a meter away from the sub's mouth (the sub's mouth is small enough IMO that measuring from 10M really isn't necessary).
Results varied from 104dB@42Hz up to 111dB@50Hz, then back down to 107dB@70 Hz. I didn't measure at any higher frequencies than that, as I suspect my next door neighbour was beginning to get pissed off, LOL.
A voltage meter placed across the sub suggested that the amp was providing around 17.4V, but I suspect the meter measures peak, not RMS - if I calculate the RMS value and plug it into HornResp, there's pretty close correlation between the HornResp predictions and what I measured.
If I'm reading the graphs correctly, 2nd and 3rd HD seems to vary between 1% and 3% at that drive level, down to Fb.
Oh, before doing the measurements, I decided to run the sub as loud as it could go off the amp before distortion was noticeable, using music as the source signal, as most people listen to music, not test tones
. Clocked 117dB@1M using an RS SPL meter set to "C" weighting . "G6" was playing at the time, if I remember, and the cone was really moving - not surprising, as the bass peaks @50 Hz, right near the peak of the displacement curve in the sub's passband. My daughter made a video of it, but I don't know if that would be of any use, particularly as the poor camera's mic was clearly overloaded @50 Hz ). During this test, I had to adjust the LP filter down to 80 Hz, 18dB/octave during this test - I suspect the amp was starting to run out of steam and the box (crappy plywood) was adding its own non-linear distortion above 90Hz or so. I'm planning to rebuild this thing at some point, and proper bracing to fix that issue is going to be a priority.I also need something that can do a bit more analysis than TrueRTA. I downloaded HolmImpulse tonight and, if the weather's good, I'll see if I can get some more tests done tomorrow (hopefully the neighbours will be out, LOL).
Oh, one more thing - I didn't run the sub with any HP filter set to below Fb, and it did not blow up, though it was audibly pretty disgusted when presented with the extra-low frequencies in my copy of "Clubbed to Death". While HornResp predicts "catastrophic" excursion below Fb in the model, the reality is that the "B" part of BL quickly disappears after the coil starts leaving the gap (not to mention the suspension wants to bring the cone to a halt too once Xmax is exceeded), so any predictions about excursion near or above Xmax should probably be taken with a grain of salt.
Nice job Brian,
AFAIK,
All the common simulation programs are only valid when you dealing with linear systems. Surpassing x-max, overdriving the amp, reaching the nonlinear SPL range of air, etc mean that the sim is no longer going to be a true simulation.
I would, of course, use the headphone jack to a little mixer or direct to the RCA's on the amp.
Record till 200 or 300 hz if you can... so that you can see how well the sim matches up there. If the amp can handle the lows at that drive level it should be no problem in the higher range. Of course it will be much louder, because efficiency is going up as well as percieved loudness. At least you can drive to somewhere?
Yes meters that measure rms are more expensive and state so.
Those are low distortion numbers for a sub.
AFAIK,
All the common simulation programs are only valid when you dealing with linear systems. Surpassing x-max, overdriving the amp, reaching the nonlinear SPL range of air, etc mean that the sim is no longer going to be a true simulation.
I would, of course, use the headphone jack to a little mixer or direct to the RCA's on the amp.
Record till 200 or 300 hz if you can... so that you can see how well the sim matches up there. If the amp can handle the lows at that drive level it should be no problem in the higher range. Of course it will be much louder, because efficiency is going up as well as percieved loudness. At least you can drive to somewhere?
Yes meters that measure rms are more expensive and state so.
Those are low distortion numbers for a sub.
Well, I tried to do something, I'm not sure how successful it was. Basically hooked POC#2 to the car amp, set the EQ flat, the LP filter to 100 Hz (-18dB/octave) and set the output to 0dB. I then used my TrueRTA on my netbook to generate test tones (sent to the car deck via BlueTooth) and record the response from the sub (using an ECM8000 with a USB interface). The sub was placed on my driveway and the mic was placed on the ground, a meter away from the sub's mouth (the sub's mouth is small enough IMO that measuring from 10M really isn't necessary).
Results varied from 104dB@42Hz up to 111dB@50Hz, then back down to 107dB@70 Hz. I didn't measure at any higher frequencies than that, as I suspect my next door neighbour was beginning to get pissed off, LOL.
A voltage meter placed across the sub suggested that the amp was providing around 17.4V, but I suspect the meter measures peak, not RMS - if I calculate the RMS value and plug it into HornResp, there's pretty close correlation between the HornResp predictions and what I measured.
If I'm reading the graphs correctly, 2nd and 3rd HD seems to vary between 1% and 3% at that drive level, down to Fb.
Oh, before doing the measurements, I decided to run the sub as loud as it could go off the amp before distortion was noticeable, using music as the source signal, as most people listen to music, not test tones
I also need something that can do a bit more analysis than TrueRTA. I downloaded HolmImpulse tonight and, if the weather's good, I'll see if I can get some more tests done tomorrow (hopefully the neighbours will be out, LOL).
Oh, one more thing - I didn't run the sub with any HP filter set to below Fb, and it did not blow up, though it was audibly pretty disgusted when presented with the extra-low frequencies in my copy of "Clubbed to Death". While HornResp predicts "catastrophic" excursion below Fb in the model, the reality is that the "B" part of BL quickly disappears after the coil starts leaving the gap (not to mention the suspension wants to bring the cone to a halt too once Xmax is exceeded), so any predictions about excursion near or above Xmax should probably be taken with a grain of salt.
Stuffing doesn't work (for this TH)...
See image of (smoothed) HornImpulse measurements, before (red) and after (blue) stuffing (the first 1/4 of the line). Looks like broad 2dB loss from about 70 Hz on, and no significant change to the notch. Ugh.
Apart from the stuffing, measurement conditions remained identical for the test. I used a LP filter @400 Hz to avoid annoying the neighbours.
See image of (smoothed) HornImpulse measurements, before (red) and after (blue) stuffing (the first 1/4 of the line). Looks like broad 2dB loss from about 70 Hz on, and no significant change to the notch. Ugh.
Apart from the stuffing, measurement conditions remained identical for the test. I used a LP filter @400 Hz to avoid annoying the neighbours.
Attachments
Bracing...
The attached image shows what I did to the top panel to try to reduce panel flex. Two layers of 3/4 ply, a center brace and steel braces screwed to the bottom. Still vibrates a little too much for my liking. I must have the worst ply in the world to work with...
The attached image shows what I did to the top panel to try to reduce panel flex. Two layers of 3/4 ply, a center brace and steel braces screwed to the bottom. Still vibrates a little too much for my liking. I must have the worst ply in the world to work with...
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.