In this blog post I test the Fostex FF165WK 6.5" full range driver.
The cabinet and contour network plans can be purchased here.
I purchased these drivers ten years ago for a project which was featured on my blog here.
Testing again after ten years.
I was curious to see how these drivers would measure with my ARTA measurement software considering the vast database I've developed in the past five years. Being able to compare the FF165WK data against, for example, the Markaudio CHR120 was something I was personally interested in, considering how much time and effort I spent ten years ago on this driver. I must admit also that I have a bit of nostalgia for this particular driver, among man other single full range drivers.
Contour Network
Over the years I've improved in my ability to develop excellent passive circuits to help achieve the desired frequency response from various drivers. Examples of this is here and here. So I wanted to see if I could use my latest skill to further refine this driver for even more sound quality. Now, I must confess, that this driver is not necessarily anything stand out in terms of sound quality. It is pretty average if I am honest. But I feel like many enthusiasts have similar drivers where they've been tolerating the faults while enjoying what they get right, which is coherency and soundstage depth. In most cases, these faults can large be mitigated through careful contouring of the frequency response with a passive network. (see more later in this blog)
New Tapered Transmission Line Enclosure
In this blog post I also introduce a new tapered transmission line floorstanding enclosure for the FF165WK as a plan set that can be purchased downloaded on my site. This is also an area where I've improved in my ability over the past ten years. I believe the new TL enclosure maximizes the potential bass performance from this driver. Please contact me if you would like a custom enclosure designed for a specific driver.
Super Tweeter
Over the past six months I've been spending a lot time on super tweeters. The FF165WK lends itself to the addition of a super tweeter by virtue of it's natural roll-off above 10kHz. The driver can be used alone, or with a high frequency driver using a range of crossover points ranging from 2kHz - 10kHz. In the context of the DIY enthusiast, the FF165WK can serve has a good test box to evaluate various high frequency solutions. Considering that the overall cost of the FF165WK driver is relatively affordable, the enthusiast just may have money left over for a super tweeter as well.
Appearance
For me I really like the appearance of the FF165WK with it's various shades of grey. It has a vintage look to it as well which some people really like.
General Features of the FF165WK
The particular model stands out from the rest in that it uses a aluminum dust cap to extend the high frequencies as apposed to a whizzer cone. It's part of a family of FF series driver models which range in size from 3", 4", 6.50", and 8.00".
Features
- Double layer cone
- Pulp base layer with Kenaf top layer
- Kenaf layer provides high speed sound propagation
- Aluminum dust cap
- Proprietary center dust cap ridge to distribute modal breakup
- Low loss urethan foam surround
- Relatively high sensitivity of 92dB/1w for a 6.50" woofer
Measurements
I measured the response at 1m mounted in a 20cm wide transmission line floorstanding speaker. I gated the measurement above 300Hz and then conducted a near field ungated measurement below 300Hz and spliced the two responses.
The on-axis frequency response can be summarized as follows.
- Very high sensitivity through the midrange (95dB 1w)
- +5dB rising response from 300Hz - 8kHz
- -12dB/octave falling response above 8kHz
- Surround resonance at 4.5kHz
- Sharp +15dB resonance at 18.5kHz (outside audible band)
Comparison for Context
Below is the frequency response of the Markaudio CHR120 for comparison.
If I overlay the Fostex FF165WK (Blue) for comparison we can see that the Fostex has significantly higher sensitivity. This is very important if using a low power amplifier since the amplifier is going to produce much less distortion at the lower output voltage. In this particular case the Markaudio will draw two times the power from the amplifier for the same output SPL. When looking at specific low power amps such as the Nelsen Pass Amp Camp Mini, this can be a distortion difference of 39dB in noise floor when comparing distortion between 1v output from the amplifier versus 4.00V (2w).
Off-Axis
Below is the off-axis frequency response overlay for 0, 15, 30 and 45 degrees off-axis. The driver's directivity starts to narrow at around 5kHz. This can be observed by the blue line (30 degree off-axis) at the 5kHz region where things start to fall off as you move up in the frequency spectrum.
Time Domain
Below is the burst decay out to 20kHz. The resonance at 4.5kHz is likely more audible due to the peak in the frequency response more than it's tail out to 12 periods. So if the peak is subjectively offensive then simply notching out this region is likely a good solution. The peak at 18.5kHz is entertaining to see.
Extending the burst decay out to 50kHz highlights this resonance even further. We see full ring out to 24 periods. That's a big resonance! Luckily this is outside the audible band.
The CSD is surprisingly clean with the exception of the 4.5kHz surround resonance.
Correction Network
Based on the above test data it appears the driver performs well with the exception of the following...
- Midrange peak at 1.5kHz
- Resonant peak at 5kHz
The below schematic is what I came up with the correct for the above issues. It is comprised of two series LCR contour circuits. I've blocked out the values and the plans can be purchased here.
Below is an overlay between the raw response (red) and the affect of the filter (green).
I then re-measured the off-axis at 0,15, and 30 degrees off-axis.
Below is the time domain performance with the correction network in place.
Bass Cabinet No.2286 -- Tapered Quarter Wave Transmission Line (TQWL)
The cabinet and contour network plans can be purchased here.
Below is the half space in-room response without the correction network. The speaker was placed about 30cm from the rear wall. No internal damping material was used in the enclosure.
