Home Theater Forum and Systems banner

141 - 160 of 181 Posts

·
Registered
Joined
·
284 Posts
Hi again Rodny,

I just had another look at your photos and see that the inside end of the ports is unflared. This would invalidate any test results because any chuffing you hear is probably coming from inside the port.

Thanks anyway
Collo
 

·
Senior Shackster , Platinum Supporter
Joined
·
828 Posts
Discussion Starter #142
Hi again Rodny,

I just had another look at your photos and see that the inside end of the ports is unflared. This would invalidate any test results because any chuffing you hear is probably coming from inside the port.

Thanks anyway
Collo
I did build the flare on the inside, so inside and out use a 1" 1/4 flare!!
 

·
Registered
Joined
·
7,514 Posts
I did build the flare on the inside, so inside and out use a 1" 1/4 flare!!
Aren't you adding length to the port with those flares and so lowering your tune?

Or did you calculate that into the original design.......

brucek
 

·
Registered
Joined
·
284 Posts
I did build the flare on the inside, so inside and out use a 1" 1/4 flare!!
OK, that's great.

The voltage measurement is AC RMS, which is what you get with a multimeter on the AC volts setting.

It's OK to measure this at the amp.

When calculating the power, use the DC resistance seen by the amp. If the drivers are in series, this would be Re * 2, if they're in parallel it would be Re /2, (where Re is the value for one driver)

You would need to feed the sub a sine wave and slowly increase power until port noise is first detected, sitting close to the port.

I used the signal generator built into WinISD. You just have make sure you're getting a clean sine wave out of your PC

The frequencies I used in my testing were 20, 25, and 30 hz.
Since you mentioned that you heard some noise at 10hz, I would start there.

If you get a result, then move up in frequency by 5hz steps - ie 15, 20, 25, 30hz

Because of your low tune, as you move up in frequency, the port velocity will fall away. More power will be needed to get the velocity back up.

Also, port performance increases with frequency, requiring higher velocities for "chuffing", needing even more power.

Because of these two factors, you'll quickly get to the point where you can't add any more power. If you make 30hz, that would be fantastic, but I suspect that 20hz is about as far as you'll get!

regards
Collo
 

·
Senior Shackster , Platinum Supporter
Joined
·
828 Posts
Discussion Starter #146
When calculating the power, use the DC resistance seen by the amp. If the drivers are in series, this would be Re * 2, if they're in parallel it would be Re /2, (where Re is the value for one driver)
The subs are dual 4 ohm running in series to 8 ohms and then parallel to 4 ohm load.:scratch:

I used the signal generator built into WinISD. You just have make sure you're
getting a clean sine wave out of your PC
Can I use the REW or I need WinSD??
 

·
Senior Shackster , Platinum Supporter
Joined
·
828 Posts
Discussion Starter #147
Aren't you adding length to the port with those flares and so lowering your tune?

Or did you calculate that into the original design.......

brucek
I cut the port to 33", the original design was 35", I measure the port after installing the flares and they are 34.5", its pretty close, I dont think 1/2" will make a difference!:T
 

·
Registered
Joined
·
284 Posts
The subs are dual 4 ohm running in series to 8 ohms and then parallel to 4 ohm load.:scratch:
A little detective work.....
This shot from the Sound Splinter page for the RLP-15



For the Dual 4 ohm driver, it shows the Re as 6.1 ohms. This would be the total dc resistance with the voice coils in series. Since you are running a pair of these in parallel, the value we need for our calculations is 6.1 / 2 = 3.05 ohms

Can I use the REW or I need WinSD??
You can use the generator in REW
 

·
Registered
Joined
·
7,514 Posts
Just thinking out loud...:daydream:

I know for the purpose of INPUT POWER, WINISD is using RMS E^2/Re, where they're assuming that at some point the complex impedance of the drivers will be their DCR, (and that would be the near maximum power, so it's a decent assumption), but I suspect Rodnys impedance will have a wild ride at those various frequencies if you examine the impedance curve of his LLT shown below.

Qty 2 SS D4 drivers
10" ID vent, 35" long, two flares.
38 cu. ft box
11.29Hz tune.....

I added an electronics 1st order rolloff at 10Hz for these - a standard assumption I suppose.

IMPEDANCE CURVE.
Minimum impedance at ~12Hz. That's where the highest AC current will be.

Impedance.jpg


AIR VELOCITY
Shows maximum velocity at ~10.5 Hz. The highest chuffing possibility I suppose.
I used 2500 watts for Input Power for this graph.
I had to go to Input Power of ~630 watts to get the velocity down to 17m/sec (=5% - 340m/sec)

air velocity.jpg

EXCURSION
Again I used 2500 watts input power for this diagram. Somewhat over spec for excursion, but I guess he'd never be applying this much power.

Excursion.jpg


Since the EP2500 is likely in bridge mode, one channel will be inverted across its positive terminal, so you'll be measuring a double rail voltage between the speaker terminals.

I suspect that amplifier would realistically produce ~500 watts per channel into 8 ohms, so it would be using about +100 DC volt rails. RMS of that, less the drop across the output stage of the amp, would yield a maximum of about 64 AC volt RMS capability. So 64^2/8 = 515 watts. Seems about right.

So, you have the theoretical AC voltage of 64 + 64 across the two bridged positive terminals. But since it's in bridge mode they likely have reduced the rails a bit so not to fry the thing, but anyway, be aware and careful when you are measuring not the short anything out.......

I guess you'll only be turning it up to the point where it chuffs........

brucek

edit: changed voltage to impedance..... shouldn't write stuff at night......
 

·
Registered
Joined
·
284 Posts
Hi Bruce,

I ran the idea of using the voltage measurement to determine input power past the WinISD people when I first started the experiments.

Janne replied... "Power means actually that sinusoidal RMS voltage of U=sqrt(Power*Re) is applied to the driver. This does not change with the impedance"

I seem to remember from my Electrical engineering days, that whilst impedence moves around with frequency, the power used remains resistive.


Looking at your estimate of velocity vs frequency, I would say we're probably only going to get a result for 10hz. I do have a range of figures in mind, but don't want to predudice any results by mentioning it just yet.
 

·
Registered
Joined
·
7,514 Posts
Yeah, I do mean (and changed) that the impedance will be riding up and down as shown in the first diagram. Its lowest point is around 12Hz, so the current will be the greatest and so the applied power. The phase curve shows where the load becomes near purely resistive when it reaches 0 degrees. I think the power used will be close to the inverse of the impedance curve, will it not.

Looking at your estimate of velocity vs frequency, I would say we're probably only going to get a result for 10hz.
The velocity sure spikes at that frequency. Do the flares allow a speed over the 5% to not create the chuffing noise?

brucek
 

·
Registered
Joined
·
284 Posts
The science of inductive loads is a bit beyond me, so I defer to the experts on this one.

I do know that when a load contains a resistive and an inductive component, the voltage and current aren't exactly in phase.

If the power delivered to the driver followed the impedence graph, you would expect the SPL graph to do the same, yet this is not the case.

Since I'm using WinISD to do the analysis, and the voltage measurement method was given the OK by one of the WinISD authors, I'll use it.


----

Adding flares definitely allows you to use a higher velocity before chuffing occurs.

This is not as simple as saying a flare of x mm allows a velocity of 20 m/sec

You need to take into account port diameter, flare radius and frquency.

You can also allow a small amount of chuffing which is not heard due to the distance to the seating position and masking by content.

If you follow the link in my earlier posting, you'll see how this was discovered.
 

·
Registered
Joined
·
7,514 Posts
Since I'm using WinISD to do the analysis, and the voltage measurement method was given the OK by one of the WinISD authors, I'll use it.
Yep, and I completely agree. I was just musing about this stuff - I don't want to get the thread off track.. :)

If the power delivered to the driver followed the impedance graph, you would expect the SPL graph to do the same, yet this is not the case.
Well, I think the fact that the SPL units being logarithmic (dB) have a lot to do with that. Doubling the applied power only nets a +3dB increase in SPL. Kinda smooths the SPL graph out in relation to large swings in power. You could increase the power from 500 watts to 1000 watts and the SPL graph would only show an increase (for example) from +113dB to +116dB.
But certainly the reactive components will store and release energy that can be reflected back to the load and result in heat being dissipated by the output stage of the amplifier itself. The power graph that WINISD provides is apparent power ( expressed in volt amps). This would be a combination of the RMS or real power dissipated in the resistive load (Re) plus the reactive power as a result of the inductance and capacitance of the driver. It would simply be the product of the current squared times the impedance over the frequency range.

brucek
 

·
Registered
Joined
·
109 Posts
Yeah, I do mean (and changed) that the impedance will be riding up and down as shown in the first diagram. Its lowest point is around 12Hz, so the current will be the greatest and so the applied power. The phase curve shows where the load becomes near purely resistive when it reaches 0 degrees. I think the power used will be close to the inverse of the impedance curve, will it not.

brucek
Hi Bruce,

I wanted to jump in as this matter is one often confused. Despite common terminology and specifications, we drive speakers with Voltage, not power. Power ratings on amplifiers are defined by driving specific, resistive loads (nothing earth shattering here). Real electromechanical devices have impedances that vary with frequency. Nominal impedances are about as precise as a "2x4" is 2" x 4" :rolleyes:.

You are correct that actual power dissipated in the voice coil is inversely proportional to the impedance curve for a constant Voltage input. The feedback from the creaters of WinISD might have been from confusion of terms. "Power" input could be, and by their description might be, defined by Voltage into the Re or Dcr of the driver. Different programs handle input signal a little differently in how you specify it, but the results should be the same when using the same modeled Voltage.

In short, if you want to correlate the driving signal to a theoretical vent velocity, you want to measure RMS Voltage. Many meters are not True RMS meters, and you will have less confusion in calibrating the Voltage at 60Hz with the meter and then using the frequency response measurements of the electrical signal to extrapolate from there. In taking such measurements, do be sure to include some casual observation to correlate what you are modeling and observing. For example, from the above model you would expect to see ~1/2 the excursion at the tuning frequency as at ~20Hz.
 

·
Senior Shackster , Platinum Supporter
Joined
·
828 Posts
Discussion Starter #158
Collo and I conducted some testing via PM's to save cluttering up this thread. It was concluded that the airspeed was too low to account for the noises heard at 10hz. The most likely culprit was driver noise.


:nerd:
 

·
Registered
Joined
·
1,018 Posts
Collo and I conducted some testing via PM's to save cluttering up this thread. It was concluded that the airspeed was too low to account for the noises heard at 10hz. The most likely culprit was driver noise.
Hi Rodny

What sort of levels dB(C) were you running at to hear noise at 10hz?

Was this on sinewaves? I'm wondering if the signal you were feeding it had heavy distortion which produced an audible harmonic.

I ran my IB array at 10Hz up to nearly 100dB(C) (uncorrected RS meter) using REW in the "Frequency follows cursor" mode.

The IB was completely silent though the room was shaking violently. I stopped at about 1/4" cone movement because I feared the windows would break if I pushed it any harder.

Are you sure your box or something else isn't vibrating in sympathy somewhere?
 

·
Senior Shackster , Platinum Supporter
Joined
·
828 Posts
Discussion Starter #160
Hi Rodny

What sort of levels dB(C) were you running at to hear noise at 10hz?

Was this on sinewaves? I'm wondering if the signal you were feeding it had heavy distortion which produced an audible harmonic.

I ran my IB array at 10Hz up to nearly 100dB(C) (uncorrected RS meter) using REW in the "Frequency follows cursor" mode.

The IB was completely silent though the room was shaking violently. I stopped at about 1/4" cone movement because I feared the windows would break if I pushed it any harder.

Are you sure your box or something else isn't vibrating in sympathy somewhere?
It was about 90dbs with only one side, the signal was from the REW sine waves, I was using the RS meter, and yes the house made some real loud noises:holycow:

I went back and listen to the subs at 10Hz and one of the subs its making a noise, its like the aluminum cone rubbing on something:dontknow:, if I have time this weekend I will check all the subs, it could be a speaker wire behind the sub.
:T
 
141 - 160 of 181 Posts
Top