Options for 3 inch driver for near field monitor

[deandob]

All,

I'm looking to upgrade my PC speakers (monitors) and wondering what is the current SoA / popular / recommended drivers, with an emphasis on mid range & treble performance, as I will be crossing these over with a subwoofer around 100 - 150Hz.

Currently I use Jordan's JXR6HD drivers, and before that Tangband W3-817, and like the high end clarity & detail of the Jordan drivers, but after 5 years use of the Jordan I want to improve the setup as I spend 90% of my time listening to music on my PC.

I see Jordan is coming out with a new range however I'd like to try a different make driver so looking to see what is currently recommended.

Thanks!

[planet10]

Top tier (i've not heard the Audience A3 or the small Jordan).

Fostex FF85wk, Mark Audio Alpair6.2 (paper or metal cones). If you don't need bass that gives the Fostex a slight edge. 3" driver audition Fostex/Mark Audio

Next tier, Fountek FR88ex (the cheaper FR85 with ceramic magnet might make it more cost competitive to tier 1). Possibly a TB in here too, i haven't heard many of the current ones.

3rd tier, TB 871 (or equivalent), 881 others,

ScanSpeak & Vifa have some reputedly very good 3", i have the ScanSpeak, haven't done anything with it yet, the price was a bit of a turn-off.

Listening to A6.2PeN (they normally have a blue cone) at the moment (with a pair of stealthWoofers)



dave

[deandob]

Thanks Dave for the feedback, the linked post is informative.

Would be nice to know how the JXR6 Jordan compares with other drivers listed here, especially in resolving details which was the main improvement over the Tangband 817. But the price of the Fostex FF85WK is relatively cheap to do experiments on.

I ran up a quick WinISD plot and for a 2L cabinet I can get -3db at 150Hz which is sufficient.

[big russ]

I've just ordered a pair of the Vifa TC9FD18-08 for the same purpose following some favourable comments in another thread. Was thinking of running them sealed crossed over to a sub.

Zaph has just measured them on his site: Zaph|Audio and the price seems very reasonable.

[Bigun]

I've had very satisfactory results with the FR88ex

MARTELLO enclosure for FR88EX

[deandob]

I had a closer look at Zaph's excellent site and compared the graphs (frequency response, cumulative spectrum & harmonic distortion) for the 4 drivers tested on his site that are possible full range contenders:Vifa TC9FD18, FR88EX, ScanSpeak Discovery 10F/4424 and Dayton RS100-4.

Interesting comparison, these drivers are fairly close to each other, with the vifa having a nice extension to about 100Hz, scanspeak with the best cumulative plot (decay vs frequency) and not much comparison in the harmonic distortion, the Dayton has good F3 response but poor F2, with Vifa the best all-rounder distortion wise.

I ran these through WinISD and saw that for my 5L box, the scanspeak and fostex don't have sufficient Xmax or extension below 200Hz to be suitable for my application (subwoofer rolloff at 120 or 150Hz), with the vifa having a nice 1.5db peak at 200Hz then quickly rolling off to -3db at 100Hz.

Overall on paper the Vifa looks to be the best all round performing full range driver, with good low frequency extension, fairly flat frequency response to 20Khz and good distortion figures.

I'm fairly new to full range design, so I'm not sure what are the most important parameters to look at when comparing driver tested response graphs. I'm assuming the following:
- Frequency response smoothness and low/high freq extension (this one is pretty obvious)
- Harmonic distortion, a lower F2 is important as it will be the most audible, especially in the mid frequency range where the ear is most sensitive, but also look for an overall lower response across all the harmonics, again in the mid frequencies.
- Cumulative decay plot, higher values immediately after the initial signal won't be resolvable by the ear (<1msec), so look for the lowest values again in the mid frequencies especially after 2msec.

Am I making the right assumptions here for a full range driver? Do I need to look at other parameters like impedence or phase plots?

Unfortunately I can't find any similar tests for the Jordan JXR6 driver to make a paper comparison.

[Bigun]

Originally Posted by deandob

I'm fairly new to full range design, so I'm not sure what are the most important parameters to look at

If you build a bass reflex type of enclosure I found some advance from Dave (planet10) to be most helpful - don't set the port dimensions for maximum bass response (lowest F3) but pull it back slightly so that a) you don't get such a steep roll off from f3 to f10, b) more tolerance to variation in driver parameters from manufacturing variations and with temperature etc.

[deandob]

Thanks BigGun. I'm hoping to use these sealed to get better transient response (tighter and lower distortion) as I have a subwoofer for frequencies lower than 100Hz. I was wondering more about the best way to interpret Zaph's performance graphs for full range design.

I have found the specs for the JXR6 (from the EAD site) however its hard to compare with Zaph's measurements due to the differences in measurement environment. However from what I can see from the specs the JXR6 is a low distortion unit and it has good treble but it suffers from resonances.

[tuxedocivic]

You're gonna get some varied advice on reading graphs in this forum.

1. H2 is not the most audible. Many say H5 is, but is usually measures lower. Look for a driver that has all H distortion types to be low, BUT don't put much stock in it. Look for large trends, like massive amounts below 200hz where a small driver starts to wheez.

2. Those measurements are on an infinite baffle. So the in box response will be different. I use EDGE, a baffle correction type software. I size my baffle to best conform to the frequency response. Example. If there is a dip at 1500hz I'll make the baffle very small to allow the inevitable diffraction fill that dip. Or a rising low frequency response will offset baffle step.

3. Don't put much stock in the <200hz response. Again, IB measurement. It'll change once in a box. Use your box model for this. CSD is also not a high priority, like H distortion. Use it lightly.

4. Not all the info is there. Zaph doesn't provide polars, power level FR compression, or Klippel tests. And there are other tests as well I'm sure. On top of all that, many here will tell you our ears are the ultimate test, and, well, who would argue that

Edit - oh and look for little blips on the impedance plots. Those are usually little resonances and can be different things. Smooth imp is good. Pro drivers usually have lots of squiggles in their imp plots.

[deandob]

tuxedocivic,

Great tips, especially the comments about not putting too much weight on the distortion/CSD data. Thanks!

I assume its better to have the impedence "hump" (resonance) lower than the driver -3db rolloff point? By the charts the Vifa driver looks the best but its impedence peak is right at the box rolloff point, while the other drivers have their peak as the driver/box is rolling off.