Verifying port tuning

[Triticum Agricolam]

A comparative study was made with 5 different tuning equations and if I remember right the difference was less than 2 Hz in all cases. Unhappily I don't have a link for it right now.

If you ever find where you saw that study please let me know, I'd love to see it!

I checked JL's equation against the one I was using, it was within 1/3 Hz for a bunch of different sized/tuned enclosures. I'm not surprised all the different equations are so close. I just wish they could more accurately predict what happens in reality.

[Triticum Agricolam]

I found this article on port tuning: http://www.troelsgravesen.dk/vent_tuning.htm

From what it says it looks like having your tuning end up lower than you want in a common occurrence.

[Joe X]

One has to be careful when reading non validated research, from what I've read those coefficients for the end correction were verified from many different sources before becoming accepted. A test has to be done on defined conditions (like an ISO standard) and with specific methodology for the results to be comparable. I wouldn't be too concerned about this for a daily setup though, likely a guy looking to do a 180 dB burp should be.

[Joe X]

That's way off Wicks there may be a calculation error there.

Everybody seemed to say that Torres was the best calculator so that's what I used.

What else do you suggest?

I was concerned that my tuning would end up higher then I'd wanted so I was pleasantly surprised to see that 32Hz number when I measured it in the car and got a real world number.

Torres calc will not be off by 5 Hz despite it's inaccuracies, by comparison sketchup net volume and port length calculation would be more accurate, if the port os choked or the sub becomes loaded from a different source other than the port alone, real world tuning may be not as predicted.

[Triticum Agricolam]

One has to be careful when reading non validated research, from what I've read those coefficients for the end correction were verified from many different sources before becoming accepted. A test has to be done on defined conditions (like an ISO standard) and with specific methodology for the results to be comparable. I wouldn't be too concerned about this for a daily setup though, likely a guy looking to do a 180 dB burp should be.

I'm sure you are right and the difference in being off a few Hz is going to be completely inaudible. Its just the perfectionist in me, I want to design and build the best enclosures as humanly possible.

I also just read somewhere else that measuring port tuning using impedance sweeps isn't the most accurate way to do it and that could be where the deviation is coming from.

They way they suggested was using a near-field microphone and looking for a null in the output from the driver. I've got a calibrated mic so I might give that a try.

Thanks a lot for your input Joe, as usual, you have been more than helpful and I really appreciate it.

[kirill007]

One has to be careful when reading non validated research, from what I've read those coefficients for the end correction were verified from many different sources before becoming accepted. A test has to be done on defined conditions (like an ISO standard) and with specific methodology for the results to be comparable. I wouldn't be too concerned about this for a daily setup though, likely a guy looking to do a 180 dB burp should be.

I'm sure you are right and the difference in being off a few Hz is going to be completely inaudible. Its just the perfectionist in me, I want to design and build the best enclosures as humanly possible.

I also just read somewhere else that measuring port tuning using impedance sweeps isn't the most accurate way to do it and that could be where the deviation is coming from.

They way they suggested was using a near-field microphone and looking for a null in the output from the driver. I've got a calibrated mic so I might give that a try.

Thanks a lot for your input Joe, as usual, you have been more than helpful and I really appreciate it.

Red: True, impedance sweets won't get you the correct port tuning because you can't measure the impedance introduced by the box itself. It's a sum of multiple impedances.

Blue: I doubt your microphone will be good enough, when has it been calibrated and how has it been used? (and which one is it?) This is a better solution then the previous one.

And as has been stated, building the box will introduce errors.

Why do you want to find the tuning of a box? Just experimenting or for a goal?

[kirill007]

When you do the testing with the microphone, follow these steps.

1. Wait at least 10 seconds for enclosure+port to “warm-up” to new frequency.
2. Take at least 3 frequencyreadings about 5 seconds apart. If the values seem to fluctuate a fair amount (± 0.5 dBA), then more readings may be required to get a good average value.
3. Turn off the tone when finished taking readings.

This is also why a sweep won't work, you can always use a SLOW sweep to get a general idea.

[Triticum Agricolam]

Red: True, impedance sweets won't get you the correct port tuning because you can't measure the impedance introduced by the box itself. It's a sum of multiple impedances.

Blue: I doubt your microphone will be good enough, when has it been calibrated and how has it been used? (and which one is it?) This is a better solution then the previous one.

And as has been stated, building the box will introduce errors.

Why do you want to find the tuning of a box? Just experimenting or for a goal?

Thanks for your input. The reason I want to find the tuning is so I can more accurate build enclosures in the future. I'm always trying get closer to perfection.

I got out my mic and did some testing. FYI my mic is Dayton UMM-6 and is basically brand new. Here is what I got:

The lower line is the near-field output from the driver and the upper line is the output from the port. I used a very slow sweep to get this, I think it looks pretty conclusive.

For comparison here is the results of an impedance sweep from 20-50 Hz with my DATS:

It looks like the DATS is pretty close to what I measured with the mic, which is re-assuring.

[kirill007]

How does the calibration file of that microphone look like?

(on this site you can get it, I assume you already have it, because it looks like their program)

It's a decent mic, but most microphone aren't very good below 50Hz. (even microphones 5X the price of yours)

But it's good enough to get a decent reading, +- 5Hz I would guess would be the uncertainity.

By the way if you are doing this for getting the best sound of the box, it's almost not worth it because your ears won't hear a 2-3Hz difference of tuning.

It also matters where you perform this test, is it inside a car or house?

[Joe X]

Both measurements are so close that seem to me pretty conclusive to me (I wouldn't think the test was done inside a car) but not how the test was done and how the box was done. It would be relevant to know the box design sketch, the place where the test took place and the temperature at the time, not necessarily asking for these things just saying. One other thing that seem puzzling is how you managed to isolate the output of the port from the output of the sub, this has been an issue in other tests.