Changing the Q of FR125S with Bucking Magnets

[xfmrengr]

I recently purchased three CSS FR125S drivers. I was less than thrilled with the numbers I tested. Fs and Qts was much higher than published. I have read on diyaudio.com that the manufacturer tests their drivers at 2.83V (1W/8 ohms). I increased the drive voltage and Fs came down, way down.

Speaker #2:
Drive voltage: 0.5118, Fs=70.41, Qts=0.768
Drive voltage: 1.0067, Fs=64.82, Qts=0.723
Drive voltage: 1.5197, Fs=62.34, Qts=0.714
Drive voltage: 2.0408, Fs=60.79, Qts=0.722
Drive voltage: 2.8777, Fs=60.79, Qts=0.770

With only millivolts on the speaker the Fs is about 82.0!

As you can see drive level accounts for a wide range of Fs and to some extent Qts.

Qts bottom out at 0.714 with 1.5197 volts of drive. I decided to use 1 volt of drive for all my testing.

I use the constant voltage method with a current meter and another voltage meter at the speaker terminals. All meters are 4.5+ digits.

The specification of the speaker are Fs of 67.4 and a Qts of 0.56. I was now pleased with Fs, but was unhappy with Qts as I wanted a bass reflex design. With a Qts as high as I tested, the power handling capabilites was very limited and the group delay was well over 30ms.

I had another speaker that tested lower than the other two, so I started to do some experimenting. I calculated BL for this speaker, which I labeled #3. Here are the base measured specs done at 1.0041 volts:
Fs=52.97, Qms=3.179, Qes=0.766, Qts=0.617, Re=7.30, Mms=4.66g, Bl=3.85.

Bl is lower than the specification of 4.47. Bl effects Qes by the square, and since I wanted to lower Qes and hence Qts, I needed to increase Bl. Since the magnet strength effects Bl, I decided to try to add magnets to the back of the driver.

This speaker is shielded with a bucking magnet and a steel cup. I tried to add magnets to the outside of the steel cup, but it did not do much. I carefully taped the entire speaker and cut the steel cup off with a dremel tool. I then removed the bucking magnet, well at least as well as I could. I did not test the speaker with only the bucking magnet, but I should have. Oh well!

I had some bucking magnets I bought for another speaker project and I started adding them. One way they "stick" really good. This increased the Q of the speaker. Woops, wrong way. I reversed the magnets so it repels when you try to stick it to the back of the driver. Once it gets really close, it "sticks". Not great, but well enough for testing. Here are my results of speaker #3 with 60mm OD, 32mm ID, 10mm thick magnets:

No Magnets: Drive voltage: 1.0041, Fs=52.97, Qts=0.617
One Magnet: Drive voltage: 1.0071, Fs=53.08, Qts=0.578
Two Magnets: Drive voltage: 1.0034, Fs=53.24, Qts=0.532
Three Magnets: Drive voltage: 1.0166, Fs=52.75, Qts=0.492
Four Magnets: Drive voltage: 1.0147, Fs=52.85, Qts=0.479

I measured a Vas of 8.84l. As you can see, Qts fell 22%. Also Bl went up 15% and SPL went up 1.2 dB.

Next is the box design, but at least I have a usable driver. Of course I have the other two to modify. I want to get some bigger bucking magnets and try them out as 4 magnets seems excessive.

Has anyone else tried something like this?

Cya L8tr

[freddi]

I've noted it on coaxial - heres a post with figures for Silver Iris

http://db.audioasylum.com/cgi/m.mpl?forum=hug&n=116146

[gmilitano]

Originally posted by xfmrengr
I recently purchased three CSS FR125S drivers. I was less than thrilled with the numbers I tested. Fs and Qts was much higher than published. I have read on diyaudio.com that the manufacturer tests their drivers at 2.83V (1W/8 ohms). I increased the drive voltage and Fs came down, way down.

Speaker #2:
Drive voltage: 0.5118, Fs=70.41, Qts=0.768
Drive voltage: 1.0067, Fs=64.82, Qts=0.723
Drive voltage: 1.5197, Fs=62.34, Qts=0.714
Drive voltage: 2.0408, Fs=60.79, Qts=0.722
Drive voltage: 2.8777, Fs=60.79, Qts=0.770

With only millivolts on the speaker the Fs is about 82.0!

As you can see drive level accounts for a wide range of Fs and to some extent Qts.

Very interesting. Your measurements of Fs are consistent with Dave's comments:
http://www.diyaudio.com/forums/showt...29#post1134629

Looking at your data, there is a clear trend that Fs decreases with increasing drive, leveling off at ~2V.

However, what I find strange is that this is not the case with Qts vs drive V?

[consort_ee_um]

Interesting observations. I found the problem with the FR125 was the air leakage past the phase plug, so when you measure the twin peak impedance in a ported cabinet you do not get the expected results. I wonder if your measured results are down to this problem as well.
Its much easier to use a negative resistance amplifier to get the Qes and hence Qts down to an acceptable level. I am using FR125 in a 4.5 litre cabinet with said amplifier.