Tweeter frequency response

[flatisgood]

Can someone please explain this frequency response drop? Tweeter is Hi-Vi SS1II measured at 1 inch centered with Arta. Green curve is free air sitting on table. Yellow is inbox. No crossover.

 

[eriksquires]

You measured both with the same input level, on tweeter axis?

 

[eriksquires]

Also, in the box, did you have the woofer in place, or is there a hole there?

 

[giralfino]

Can someone please explain this frequency response drop? Tweeter is Hi-Vi SS1II measured at 1 inch centered with Arta. Green curve is free air sitting on table. Yellow is inbox. No crossover.

Neither curve provides a valid starting point for performance assessment and crossover design. You need one of the following:
- an infinite baffle response
- a response in the intended box/baffle

For an infinite baffle, you need a very big baffle, in that way you can get closer to the driver but not at only 1 inch.
For a response in the intended box, you need to have a curve that shows also the effect of the baffle. In order to do this you have to measure gated and far field. Anything between 40 and 80 cm will be fine. Put the woofer in place, if you don't have it recreate a flat baffle with a piece of paper covering the hole. Assuming that the published graph is on an infinite baffle, you can simulate the effect of the baffle with excel based tools, and your measure, if done properly and the published graph accurate, should resemble the simulation.

Ralf

 

[flatisgood]

You measured both with the same input level, on tweeter axis?

Yes. Same setup.

Also, in the box, did you have the woofer in place, or is there a hole there?

The woofer is in place, but not connected.

 

[flatisgood]

Neither curve provides a valid starting point for performance assessment and crossover design.

Can you explain the yellow curve? I don't remember having this problem before...

 

[TNT]

Narrow baffle makes tweeter see less support at lower frequencies?

//

 

[giralfino]

Can you explain the yellow curve? I don't remember having this problem before...

I cannot explain a curve obtained with a wrong methodology. 1" or better less is only useful for a woofer near field measure. Attached you can see what is the effect on the driver response from your baffle.

Ralf

 

[flatisgood]

Narrow baffle makes tweeter see less support at lower frequencies?

Tweeter's own faceplate extends flat to 1.5k... Baffle step occurs further down...

 

[DonVK]

Can someone please explain this frequency response drop? Tweeter is Hi-Vi SS1II measured at 1 inch centered with Arta. Green curve is free air sitting on table. Yellow is inbox. No crossover.

As long as the measurement technique is the same in both cases, its fine to make the "relative" comparison.

Both curves fall off at 15Hz which is fine for this tweeter. It's a little lower than spec but then its probably not broken in either. This will improve (extend) with break in time.

Normally, the baffle effects higher freq, and that not what you're seeing. Besides, it would cause a "lobing" ripple and that's not what you're seeing either. It looks like the woofer is not electrically connected and is acting like a "passive absorber" (opposite of passive radiator). That's why you see it only in the cabinet and the effect slowly rises with freq as the mass of the woofer reduces its response impact.

Here is a trick if you're comfortable with it and want to test further. Only do this if no amplifier is connected. Enough warnings. Short the woofer's terminals together. When the woofer tries to move the back-EMF from the coil+magnet will "stiffen" the response and you'll see less of that "yellow curve sag". Remember to remove your test short.

 

[flatisgood]

OK. Sorry everyone. My mistake. It seems I used a defective tweeter for the inbox response. Here's the more appropriate inbox and free air 1 inch response with THE SAME tweeter this time! I have been scratching my head for hours on this. No more wasting your time. Thanks for your answers.