Gradient - ESL63 subwoofer filter box schematic? [Archive]

Andre Jute[_2_]

March 11th 10, 06:24 PM

On Mar 11, 7:16*am, Patrick Turner > wrote:
> I have a Gradient subwoofer filter box which is about 20 yo and which
> has a number of LF353N oppamps within. It also has what I didn't
> expect to find, two chips numbered MM5666BN which I think are AD
> converter chips.
>
> Does anyone have a schematic? it is a black box 200 wide, 50mm high,
> and 150mm front to back.
>
> There is an unchangeable 90degree phase difference between the LF and
> HF outputs at the X-over point at 115Hz, and I seek to reduce this.
>
> Patrick Turner.

That's how it was designed, my man. Audio designers don't believe
anyone can hear phase. The way this works, a Gradient setup has mixed
second and third order passes so of course they're out of phase. I
don't understand why you even need the schemo. The way to fix it is to
solder a cap from the input of the Gradient crossover amp to input of
the main amp. I did this to a Gradient I had on loan when they first
came out in order to have a reversible mod on a component I had to
return to QUAD. All I remember now is that I started with 100nF and
worked my way down, tuning by ear. If you still have out of phase
results, just swap the leads over externally and bingo. I couldn't
find my schemo but here, way down the page, is a reverse engineered
schemo: http://user.tninet.se/~vhw129w/mt_audio_design/quad_tweaks.htm
and if you go back up the same page a bit there's a guy who says he
did what I did, only he used 28nF which, without knowing his amp's
input impedance, could be okay, I suppose. Gradient themselves didn't
publish the schemo in their handbook
http://www.onethingaudio.net/FOR/QUA/57/PDF/9512-QUA-GEN-MAN-SW57.pdf
but if I were you I would ask QUAD themselves for the Gradient schemo
(if you still need it after seeing the reverse engineered jobbo)
because they were the distributors of the Gradient.

HTH.

QUAD. Definitely God's own amps and speakers.

Andre Jute
Visit Jute on Amps at
http://www.audio-talk.co.uk/fiultra/
"wonderfully well written and reasoned information for the tube audio
constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site containing vital gems of
wisdom"
Stuart Perry Hi-Fi News & Record Review

Patrick Turner

March 12th 10, 11:05 AM

On Mar 12, 5:24*am, Andre Jute > wrote:
> On Mar 11, 7:16*am, Patrick Turner > wrote:
>
> > I have a Gradient subwoofer filter box which is about 20 yo and which
> > has a number of LF353N oppamps within. It also has what I didn't
> > expect to find, two chips numbered MM5666BN which I think are AD
> > converter chips.
>
> > Does anyone have a schematic? it is a black box 200 wide, 50mm high,
> > and 150mm front to back.
>
> > There is an unchangeable 90degree phase difference between the LF and
> > HF outputs at the X-over point at 115Hz, and I seek to reduce this.
>
> > Patrick Turner.
>
> That's how it was designed, my man. Audio designers don't believe
> anyone can hear phase. The way this works, a Gradient setup has mixed
> second and third order passes so of course they're out of phase. I
> don't understand why you even need the schemo. The way to fix it is to
> solder a cap from the input of the Gradient crossover amp to input of
> the main amp. I did this to a Gradient I had on loan when they first
> came out in order to have a reversible mod on a component I had to
> return to QUAD. All I remember now is that I started with 100nF and
> worked my way down, tuning by ear. If you still have out of phase
> results, just swap the leads over externally and bingo. I couldn't
> find my schemo but here, way down the page, is a reverse engineered
> schemo:http://user.tninet.se/~vhw129w/mt_audio_design/quad_tweaks.htm
> and if you go back up the same page a bit there's a guy who says he
> did what I did, only he used 28nF which, without knowing his amp's
> input impedance, could be okay, I suppose. Gradient themselves didn't
> publish the schemo in their handbookhttp://www.onethingaudio.net/FOR/QUA/57/PDF/9512-QUA-GEN-MAN-SW57.pdf
> but if I were you I would ask QUAD themselves for the Gradient schemo
> (if you still need it after seeing the reverse engineered jobbo)
> because they were the distributors of the Gradient.
>
> HTH.

Thanks for your reply.

Since I posted originally, Gradient have just sent me a schematic. The
so called digital chips are 4 way switches so that for example the
stereo/mono function is switched by a chip, and not through a damn
metal switch. So with just one metal normal switch several things on
the circuit can be switched.

I measured the phase difference between LF and HF at 115Hz and found
90 degrees. It is as it was designed, but hey, there's a lot of phase
difference between what the signal at the ESL amp would be if there
were no x-over unit used.

The biggest complaint I hear about bass with systems using sub-woofers
is that the bass is wooly, un-integrated, fake, awkward sounding etc.
I wouldn't use a sub woofer even if forced at gunpoint. So I have
whopper 12" old Electrovoice drivers in 135L reflex boxes tuned to Fs
of 28Hz. Bass sounds very well and isn't improved by putting a rolled
up towel in the port to convert to a sealed box. I guess its because
all the "non musica"l bass resides below 35Hz. I don't interfere with
the phase of bass between about 40Hz to the x-over to midrange at
250Hz where there is a lot of overlap and over damped second order
filtering. So all my bass between 35Hz and 250Hz is from the front of
the cone and not from behind the cone via a port hole. The trouble
with most reflexed bass boxes is that the Fb is too high.

I have all the info you refer to about placing a cap between Gradient
output for ESL and the ESL amp.
In this case the ESL amp is an Audio research and probably Rin is
quite high so the C required to give more phase shift is rather a
small value. To get an extra 90d at 115Hz, there is also a
considerable amount of attenuation introduced of the ESL amp's bass F
below 115Hz. Maybe this results in cleaner mid bass and MF and HF
produced by the ESL. But I fear a dip in bass response may be created
and it may not be compensated for by reverse phase connections of the
subwoofer drivers. The signal to the sub has 270degrees of phase shift
at 115Hz, rather a lot, in reference to the main input signal, or
absolute phase reference.

In subs I have made, I have had 3 selectable F of 27Hz, 45Hz and 70Hz
and phase difference at the -3dB point is about 50d. But beyond the
selected cut off the phase shift rapidly increases because of 24dB/
octave attenuation above 100Hz.
But that does not matter because the bass is then so far below the
level of the main amps.
I have tried to make the subs I have supplied work with the existing
systems rather than to remove some bass from existing and then make
the subs do what has been removed as well as do the missing octave
down to 20Hz.

The one thing that everyone argues about is bass. Its easy to see why
when I have measured system responses on my way around; bass response
in different rooms and different systems varies widely, and seldom
does one find there are no serious peaks and troughs in the bass
response. And then the bass below 100Hz varies in recording levels so
much that its almost impossible to not to have to trim the sub levels
if the best perceived bass levels are to be had. And some ppl like a
lot of bass and others like almost none. And many ppl have tried to
set up their bass without any measurements and have succeeded in
ruining what was an acceptable bass response without a sub because the
sub beavered away to nicely cancel some of what the main amp and main
speakers were making, so owners turn up the sub then grumble about the
rumble that sounds implausible. Just where the sub is placed in the
room is rather a delicate problem....

Patrick Turner.

>
> QUAD. Definitely God's own amps and speakers.
>
> Andre Jute
> Visit Jute on Amps at
> *http://www.audio-talk.co.uk/fiultra/
> "wonderfully well written and reasoned information for the tube audio
> constructor"
> John Broskie TubeCAD & GlassWare
> "an unbelievably comprehensive web site containing vital gems of
> wisdom"
> Stuart Perry Hi-Fi News & Record Review