[vw_michael]

(I will update the pic links once I reach 5 posts

)


In case anyone interested, this is how I ended up convering a STF-3 sub to support a subsonic filter with a cutoff of 20Hz instead of 25Hz, as shipped from the factory. More on the HSU Research forum.


NOTE: Do this at your own risk, I do not take responsibility for property damage or injury resulting from this mod. Lethal voltages are available inside the sub (especially on the BASH amp), so take good care and unplug the AC source before venturing inside.


As it turns out, the STF-3 pre-amp board is already set up for a switch, as seem in the following pic (area of interest is circled red):

(also available as an attachment)


Also inside the circled area is the resistor responsible for the subsonic frequency cutoff, Ry.


This is a crude schematic I drew while tracing the circuit around the missing EXTENSION switch Sy. It is a classic example of 12 dB/oct Linkwitz-Riley highpass:

(also available as an attachment)


Through experimentation with a signal generator and a scope, I have found that by substituting the existing resistor Ry of 0 ohm with 100 kohm, the frequency at which the bass boost begins falling off is at 20Hz instead of 25 Hz. Also, I found out that for a given input signal the max voltage gain at 20Hz is 1.35V compared to 1.15V at the drain of the FET in the above circuit. This means that the max low frequency boost is raised by 1.35/1.15=1.174 or 17.4% from 25Hz to 20Hz.


Some math to back it up:


Original Fs with Ry=0 kohm:

Fs=1/(2*pi*sqrt(239000*20000)*1e-7)=23.02Hz


Modified Fs with Ry=100 kohm:

Fs=1/(2*pi*sqrt(339000*20000)*1e-7)=19.34Hz


ANOTHER Fs with Ry=80 kohm (harder to find than 100k, but it gives a Fs very close to 20Hz):

Fs=1/(2*pi*sqrt(319000*20000)*1e-7)=19.93Hz


I did a couple (crude) sonic tests, and the output at 20Hz was around 4dB higher, while the output at 16 Hz was around 6 dB higher than with the original circuit. The measurements were taken with one of the ports plugged up. The driver movement at 16Hz is much, much more extensive with the 100 kohm resistance, so take care while testing this mod.

 

[vw_michael]

I would be really interested in seeing results from an outside test after performing this mod using something like TrueRTA (which I don't have). Any1?


I have a plan for doing another mod, which would involve a 3-position switch (1-2, off, 2-3), which would allow 20Hz, 22Hz and 25Hz subsonic cutoff using the existing PCB layout. PM me if interested.

 

[UNICRON-WMD]

Very cool man! Keep updateing

 

[frottage]

I wonder if this mod would also work for the STF-2? I haven't looked inside the woofer yet so I can't comment on how the electronics compare, but looking at the HSU site, it appears that the STF-3 "just" has a bigger woofer and more power.

 

[Ed Mullen]

When you plug a port, the enclosure tuning frequency is lowered, but the alignment typically also becomes more overdamped, and the FR tends to slope downward a bit as it nears Fb.


On a true variable tune subwoofer (like the VTF-3), there is a switch to accomplish exactly what you are working on. The switch simultaneously lowers the high pass filter set point, and also adds some EQ to help restore a more critically damped alignment. Sometimes the slope of the HPF is altered too.


The key for success is to match the new tuning point, EQ, and HPF (set point and slope) so the final FR is smooth and flat down to Fb, and then the high pass kicks in to damp cone motion below Fb. Cone motion will become excessive and uncontrolled very quickly below Fb because the woofer has decoupled from the resonator and the subwoofer starts to act like there is a big hole in the box and there is no restoring force.


So if you are seeing "much, much more excursion" at 16 Hz with your modifications, and a port plugged, your set point for the HPF might need to be raised a bit, and the slope might need to be steeper. I think the Fb with a port plugged is about 20 Hz on the STF-3, so you should be hitting it pretty hard with the HPF at 18-19 Hz to help control cone motion at 16 Hz.

 

[vw_michael]

Quote:

Originally Posted by frottage I wonder if this mod would also work for the STF-2? I haven't looked inside the woofer yet so I can't comment on how the electronics compare, but looking at the HSU site, it appears that the STF-3 "just" has a bigger woofer and more power.

Unfortunatelly, the STF-2 cannot be tuned any lower - it has only one port, so you would certainly be asking for trouble if you were to plug it up I think it is tuned exactly like the VTF-2 in the extension mode.

 

[vw_michael]

Quote:

Originally Posted by Ed Mullen When you plug a port, the enclosure tuning frequency is lowered, but the alignment typically also becomes more overdamped, and the FR tends to slope downward a bit as it nears Fb. On a true variable tune subwoofer (like the VTF-3), there is a switch to accomplish exactly what you are working on. The switch simultaneously lowers the high pass filter set point, and also adds some EQ to help restore a more critically damped alignment. Sometimes the slope of the HPF is altered too. The key for success is to match the new tuning point, EQ, and HPF (set point and slope) so the final FR is smooth and flat down to Fb, and then the high pass kicks in to damp cone motion below Fb. Cone motion will become excessive and uncontrolled very quickly below Fb because the woofer has decoupled from the resonator and the subwoofer starts to act like there is a big hole in the box and there is no restoring force. So if you are seeing "much, much more excursion" at 16 Hz with your modifications, and a port plugged, your set point for the HPF might need to be raised a bit, and the slope might need to be steeper. I think the Fb with a port plugged is about 20 Hz on the STF-3, so you should be hitting it pretty hard with the HPF at 18-19 Hz to help control cone motion at 16 Hz.

Good point ... But I think the excessive excursion is inevitable with a 12dB/oct filter. I will try to set the filter to roll off at 20.0 Hz instead of 19.3 Hz as it is now, and see if it makes a difference. It would be interesting to find out whether the VTF-3 uses a 24 dB/oct filter - I was under the impression that the STF-3 uses the same PCB as the VTF-3, less the switch and a few passive components.


Perhaps I should mention that while the cone motion at 16 Hz is high, it does not distort (i.e. reach past Xmax) at very loud levels. It's pretty much what I would expect from lowering the subsonic cutoff.

 

[Ed Mullen]

If you measure the actual Fb you might be able to fine tune your tweaks around that additional information. A very close (like 1/2") mic FR sweep of the cone will work.


Right, the clean stroke of the woofer will dicate how far you can push it below Fb before you need to ramp it down. With a really long throw unit, you might be able to eek out another 1/5 octave of extension below Fb without bottoming it, but that's about all you can reasonably expect. If the woofer complains, move up the high pass to right below the tuning frequency.


A few of STF-3 amps were originally accidentally made with the max extension settings, so they came shipped with a port plug - sort of collector's item now I guess. Maybe you can ask Hsu Research what those amp settings are.

 

[vw_michael]

Quote:

Originally Posted by Ed Mullen If you measure the actual Fb you might be able to fine tune your tweaks around that additional information. A very close (like 1/2") mic FR sweep of the cone will work. Right, the clean stroke of the woofer will dicate how far you can push it below Fb before you need to ramp it down. With a really long throw unit, you might be able to eek out another 1/5 octave of extension below Fb without bottoming it, but that's about all you can reasonably expect. If the woofer complains, move up the high pass to right below the tuning frequency. A few of STF-3 amps were originally accidentally made with the max extension settings, so they came shipped with a port plug - sort of collector's item now I guess. Maybe you can ask Hsu Research what those amp settings are.

Thank you for the info. I broke down and got TrueRTA. I will do some tests once the neighbours are up

 

[vw_michael]

Ok, so I finally got the microphone and the switch required for the new mod.

I have modified it to allow the selection of one of three subsonic filter frequencies.

The new schematic is as follows. The switch allows the selection of 23 Hz (original), 20 Hz and 18 Hz subsonic filter cutoff frequencies.

http://209.195.118.104/~vw_michael/new_schematic.jpg


I'll post pics of the mod soon.

 

[vw_michael]

I have finally done the sound tests with a Panasonic-like electret mic at around 1 inch away from the speaker cone using pink noise maximums in TrueRTA. These are the results. This seems to prove that the tuning frequency FS is around 23 Hz with both ports open (pink) and 20 Hz with one port open (green). Any comments?

I have a feeling my microphone has a slight slope at low frequencies.

It would be nice to compare them to 'clinical' STF-3 responses.


Has anyone here experience with TrueRTA? Should I rather be doing a 'Quick Sweep'?

I wished there was a 'slow sweep', as the 'quick sweep' didn't seem to work too well (the results varies from run to run).


BTW. The graphs should not be interpreted as the frequency spectrum of the sub - it is only that of the driver (speaker). At tuning frequency, most output is from the port, so the actuall freq response of the whole subwoofer is much lower than on the graph. I will do an in-room frequency chart one of those days.

 

[vw_michael]

Pictures of the three-frequency mod: