K3 Audio Low Pass Filter

[Alexander Ponomarenko-5]

I think 100uF caps is enough value. Look these plots (with 100uF in
AF-amp),
we see identifical plots for Line-out and for Phone-out.

logarithmic view: http://forum.cqham.ru/download.php?id=43583
linear view: http://forum.cqham.ru/download.php?id=43582

73! Alex

Paul Christensen :
>> I've been informed Elecraft is already replacing those caps with 100uF
>> caps in the current production units.
>>    
>
> My guess is that scaling the caps beyond 100 uF is only of interest to a
> small minority of ops.  The majority will probably be happy with the current
> choice of 100 uF.  Going from 10 uF in the prior design to 100 uF is already
> a significant step in the right direction.
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[P.B. Christensen]

>I think 100uF caps is enough value. Look these plots (with 100uF in
>AF-amp),
> we see identifical plots for Line-out and for Phone-out.
>
> logarithmic view: http://forum.cqham.ru/download.php?id=43583
> linear view: http://forum.cqham.ru/download.php?id=43582
>
> 73! Alex

Alex,  The S/T FFT program you have allows the user to display in semi-log
mode.  That's the preferred way of looking at audio response. Resolution of
low-end response will show much more detail that way, especially the area of
interest below 200 Hz.

Also, keep in mind that the choice in coupling cap value is wholly dependent
on the load Z.  When making response tests of the headphone output with the
FFT program and sound card, try loading the headphone output with various
values of resistance in increments starting at 8-ohms, working your way up
to say ...50 ohms.  You'll end up with an interesting family of curves that
can then be saved in the S/T program.

Using a sound card without external input termination often results in
termination Z onto the circuit of only about 2K-ohm or so, depending on the
model of the sound card.  Pro sound cards (typically with XLR pig tails)
have the ability to select various degrees of input Z.

Paul, W9AC

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[KK7P]

> If you want better low-frequency response, just stick a resistor in
> series with the speaker.

There is an 8.2 ohm resistor in series with each headphone output.

73,

Lyle K7P

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[P.B. Christensen]

> Actually, 100 uF into 32 ohms gives 50 Hz bandwidth at -3 dB.  Into an
> 8-ohm load you get 200 Hz at -3 dB, still fine for a voice signal
> although marginal for music.

Right, nearly 200 Hz on the money.  I think you're trivializing the
difference between 200 Hz and 100 Hz with voice, not just music.  If a
transmitter is capable of transmitting down to 100 Hz, the ear can hear
substantial difference between those two extremes.  Arguably, we can say
that all we really need is 300 Hz or even a bit higher for maximum
articulation.  But, I'm discussing headphone circuit improvements primarily
for AM as well as ESSB.

> If you want better low-frequency response, just stick a resistor in
> series with the speaker.

No doubt, as long as the output is not already volume-starved.  Also,
another benefit of a low-Z source is the higher damping factor involved into
low Z.  Adding pure series R to the output kills that attribute.

Paul, W9AC

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Joe Subich, W4TV-4]

Paul,

You neglect the effect of the 8.2 Ohm resistor in series with
the headphone outputs.  With that resistor considered, the
-3 dB point with 100 uF capacitors as a function of headphone
impedance is:

   Z       Frequency
----------------------
   4         130 Hz
   8          98 Hz
  16          66 Hz
  32          40 Hz
  64          22 Hz
 200           8 Hz
 600           3 Hz

Although 8 Ohm phones might be "marginal" for someone wanting
maximum low end response, 98 Hz is probably suitable for most
communications uses.  One of the reasons I have not felt the
need to change the coupling capacitors in my K3 (although I
will be installing 100 uF caps when I make some other changes)
is that the -3dB point is 80 Hz with the 200 Ohm transducers
in my Heil ProSet.  There is a noticeable difference compared
to the 980 Hz -3 dB point when I use an old set of Yaesu
headphones with their 8 Ohm transducers.

The only reason I plan to change the capacitors is to avoid
problems when I grab a cheap set of headphones or the rig is
used by a visitor with their own headphones.

73,

   ... Joe, W4TV
 

> -----Original Message-----
> From: [hidden email]
> [mailto:[hidden email]] On Behalf Of Paul
> Christensen
> Sent: Monday, July 06, 2009 11:23 AM
> To: Brett Howard
> Cc: [hidden email]
> Subject: Re: [Elecraft] K3 Audio Low Pass Filter
>
>
> > I've been informed Elecraft is already replacing those caps
> with 100uF
> > caps in the current production units.
>
> 100 uF is the maximum size compatible with the existing PCB
> pads for C9/C13
> on the main DSP Board.  Moving up beyond 100 uF presents a
> problem.  Even at
> 100 uF, I believe the SMT caps are now rated at close to 10
> WVDC.  There's
> just no more scaling back on the SMT cap's voltage rating in order to
> accommodate a fit with a higher C value above 100 uF.
>
> Why would anyone want a higher value than 100 uF at C9/C13?
> Consider a set
> of modern, medium-Z headphones with a Z of 32 ohms at 100 Hz.
>  The -3dB
> turnover point calculates to roughly 150 Hz.  Anyone wanting
> to take full
> advantage of the lowest vocal fundamental (e.g., AM or ESSB
> modes) wants
> that point down to ~ 70 Hz.   So, a cap value of 220 uF would
> be preferred.
> I went to 330 uF in a radial style to allow extra margin for
> a wide range of
> headphone vintages and types.  To make most use of the 100 uF
> coupling caps,
> I would ensure that my headphones are of the newer, medium Z
> type and not
> the older style with 4-8 ohm drivers.
>
> My guess is that scaling the caps beyond 100 uF is only of
> interest to a
> small minority of ops.  The majority will probably be happy
> with the current
> choice of 100 uF.  Going from 10 uF in the prior design to
> 100 uF is already
> a significant step in the right direction.
>
> Paul, W9AC
>  
>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Brett Howard]

Isn't the equation 1/2*pi*r*c?  If I do the math on my headphones at 27 ohms
(a little worse than your 32 ohm example) I get:

1/(2*3.14159*(8.2+27)*100e-6) = 45.2145Hz

The 8.2 ohms is actually in the K3.  If I do it again for your 32 ohm
example I get 39Hz.  With your 220uF cap that moves it down to ~18Hz for a
pair of 32ohm cans.

Personally I think the more robust connection is a good point to keep and
I'm ok with the 45Hz output.  I tried running things through my headphone
amp which has an 18KOhm input impedance (which should give a -3dB point of
0.883Hz even with the 10uF caps) and I wasn't hearing stuff down too low.
I'd have to dump it into a spectrum analyzer to see how low it really goes
and I'm not so sure I have too many sound cards that are all that clean down
there.  Guess I'd have to go looking...

~Brett

-----Original Message-----
From: Paul Christensen [mailto:[hidden email]]
Sent: Monday, July 06, 2009 8:23 AM
To: Brett Howard
Cc: [hidden email]
Subject: Re: [Elecraft] K3 Audio Low Pass Filter

> I've been informed Elecraft is already replacing those caps with 100uF
> caps in the current production units.

100 uF is the maximum size compatible with the existing PCB pads for C9/C13
on the main DSP Board.  Moving up beyond 100 uF presents a problem.  Even at

100 uF, I believe the SMT caps are now rated at close to 10 WVDC.  There's
just no more scaling back on the SMT cap's voltage rating in order to
accommodate a fit with a higher C value above 100 uF.

Why would anyone want a higher value than 100 uF at C9/C13?  Consider a set
of modern, medium-Z headphones with a Z of 32 ohms at 100 Hz.  The -3dB
turnover point calculates to roughly 150 Hz.  Anyone wanting to take full
advantage of the lowest vocal fundamental (e.g., AM or ESSB modes) wants
that point down to ~ 70 Hz.   So, a cap value of 220 uF would be preferred.
I went to 330 uF in a radial style to allow extra margin for a wide range of

headphone vintages and types.  To make most use of the 100 uF coupling caps,

I would ensure that my headphones are of the newer, medium Z type and not
the older style with 4-8 ohm drivers.

My guess is that scaling the caps beyond 100 uF is only of interest to a
small minority of ops.  The majority will probably be happy with the current

choice of 100 uF.  Going from 10 uF in the prior design to 100 uF is already

a significant step in the right direction.

Paul, W9AC
 


______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Brett Howard]

Also don't forget there is 8.2ohms in series inside the K3 so its actually
going to half the frequencies you have here.

-----Original Message-----
From: Alan Bloom [mailto:[hidden email]]
Sent: Monday, July 06, 2009 8:37 AM
To: Paul Christensen
Cc: Brett Howard; [hidden email]
Subject: Re: [Elecraft] K3 Audio Low Pass Filter

On Mon, 2009-07-06 at 11:23 -0400, Paul Christensen wrote:
> > I've been informed Elecraft is already replacing those caps with 100uF
> > caps in the current production units.
...
> Why would anyone want a higher value than 100 uF at C9/C13?  Consider a
set
> of modern, medium-Z headphones with a Z of 32 ohms at 100 Hz.  The -3dB
> turnover point calculates to roughly 150 Hz.  

Actually, 100 uF into 32 ohms gives 50 Hz bandwidth at -3 dB.  Into an
8-ohm load you get 200 Hz at -3 dB, still fine for a voice signal
although marginal for music.

If you want better low-frequency response, just stick a resistor in
series with the speaker.  That does mean you have to turn up the volume,
but you do get plenty of bass (assuming your speaker or phones can go
that low).  A 33-ohm resistor would give as much bass as your headphones
are likely to be able to reproduce.

Al N1AL



______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[P.B. Christensen]

> You neglect the effect of the 8.2 Ohm resistor in series with
> the headphone outputs.

Joe, you're right.  I did not take into consideration the 8-ohm build-out
resistance that Lyle just discussed.  Those build-out resistors do not
appear on the headphone amp schematics (page 28/52) so until today, I doubt
anyone else knew they existed <g>.  So, they must appear off-page along with
the speaker coupling cap that also does not show up on the same schematic.

Older headphones with low Z drivers would still benefit from a value above
100 uF.  But with most modern headphones in the 32-ohm area, and I'll
repeat...the majority of users would not benefit from higher C values above
100 uF at C9/C13.  My choice of going to 330 uF was primarily driven by the
fact that I didn't want to be concerned with the operating Z of a new set of
headphones if tried, and I had an older DSP Board wiith 10 uF caps.  Since
re-work was required anyway, I settled on the 330 uF value and would still
choose that value if given the chance today.

Although not as critical for communications audio, the use of increasing the
build-out resistance does in fact lower the damping factor of the amplifier.
Not as important with newer headphones, but it can be a factor with older
low-Z headphones with traditional moving coil drivers where more mass is
moved.

Paul, W9AC






With that resistor considered, the

> -3 dB point with 100 uF capacitors as a function of headphone
> impedance is:
>
>   Z       Frequency
> ----------------------
>   4         130 Hz
>   8          98 Hz
>  16          66 Hz
>  32          40 Hz
>  64          22 Hz
> 200           8 Hz
> 600           3 Hz
>
> Although 8 Ohm phones might be "marginal" for someone wanting
> maximum low end response, 98 Hz is probably suitable for most
> communications uses.  One of the reasons I have not felt the
> need to change the coupling capacitors in my K3 (although I
> will be installing 100 uF caps when I make some other changes)
> is that the -3dB point is 80 Hz with the 200 Ohm transducers
> in my Heil ProSet.  There is a noticeable difference compared
> to the 980 Hz -3 dB point when I use an old set of Yaesu
> headphones with their 8 Ohm transducers.
>
> The only reason I plan to change the capacitors is to avoid
> problems when I grab a cheap set of headphones or the rig is
> used by a visitor with their own headphones.
>
> 73,
>
>   ... Joe, W4TV
>
>
>
>
>> -----Original Message-----
>> From: [hidden email]
>> [mailto:[hidden email]] On Behalf Of Paul
>> Christensen
>> Sent: Monday, July 06, 2009 11:23 AM
>> To: Brett Howard
>> Cc: [hidden email]
>> Subject: Re: [Elecraft] K3 Audio Low Pass Filter
>>
>>
>> > I've been informed Elecraft is already replacing those caps
>> with 100uF
>> > caps in the current production units.
>>
>> 100 uF is the maximum size compatible with the existing PCB
>> pads for C9/C13
>> on the main DSP Board.  Moving up beyond 100 uF presents a
>> problem.  Even at
>> 100 uF, I believe the SMT caps are now rated at close to 10
>> WVDC.  There's
>> just no more scaling back on the SMT cap's voltage rating in order to
>> accommodate a fit with a higher C value above 100 uF.
>>
>> Why would anyone want a higher value than 100 uF at C9/C13?
>> Consider a set
>> of modern, medium-Z headphones with a Z of 32 ohms at 100 Hz.
>>  The -3dB
>> turnover point calculates to roughly 150 Hz.  Anyone wanting
>> to take full
>> advantage of the lowest vocal fundamental (e.g., AM or ESSB
>> modes) wants
>> that point down to ~ 70 Hz.   So, a cap value of 220 uF would
>> be preferred.
>> I went to 330 uF in a radial style to allow extra margin for
>> a wide range of
>> headphone vintages and types.  To make most use of the 100 uF
>> coupling caps,
>> I would ensure that my headphones are of the newer, medium Z
>> type and not
>> the older style with 4-8 ohm drivers.
>>
>> My guess is that scaling the caps beyond 100 uF is only of
>> interest to a
>> small minority of ops.  The majority will probably be happy
>> with the current
>> choice of 100 uF.  Going from 10 uF in the prior design to
>> 100 uF is already
>> a significant step in the right direction.
>>
>> Paul, W9AC
>>
>>
>> ______________________________________________________________
>> Elecraft mailing list
>> Home: http://mailman.qth.net/mailman/listinfo/elecraft
>> Help: http://mailman.qth.net/mmfaq.htm
>> Post: mailto:[hidden email]
>>
>> This list hosted by: http://www.qsl.net
>> Please help support this email list: http://www.qsl.net/donate.html
>
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Brett Howard]

What would be slick is if you had a pair of headphones and a headphone
splitter Y cable.  Plug the headphones into one port and the sound card
input into the other and then take measurements.  That would be a bit more
true to life.

~Brett

-----Original Message-----
From: Alexander Ponomarenko [mailto:[hidden email]]
Sent: Monday, July 06, 2009 8:40 AM
To: Paul Christensen
Cc: Brett Howard; [hidden email]
Subject: Re: [Elecraft] K3 Audio Low Pass Filter

I think 100uF caps is enough value. Look these plots (with 100uF in
AF-amp),
we see identifical plots for Line-out and for Phone-out.

logarithmic view: http://forum.cqham.ru/download.php?id=43583
linear view: http://forum.cqham.ru/download.php?id=43582

73! Alex

Paul Christensen :
>> I've been informed Elecraft is already replacing those caps with 100uF
>> caps in the current production units.
>>    
>
> My guess is that scaling the caps beyond 100 uF is only of interest to a
> small minority of ops.  The majority will probably be happy with the
current
> choice of 100 uF.  Going from 10 uF in the prior design to 100 uF is
already
> a significant step in the right direction.

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[P.B. Christensen]

> The 8.2 ohms is actually in the K3.  If I do it again for your 32 ohm
> example I get 39Hz.  With your 220uF cap that moves it down to ~18Hz for a
> pair of 32ohm cans.

And that jumps to ~ 200 Hz if you use 8-ohm cans.  If you stay cognizant of
the headphone Z, you're in good shape.  If you want to pull a set of Koss
Pro4 AA phones off the shelf, you may want to revisit that.  It's all about
the series C and knowing the headphone Z as a function of frequency.  It's
doubtful your 27 ohms headphones remain 27 ohms across their frequency
range.

> Personally I think the more robust connection is a good point to keep and
> I'm ok with the 45Hz output.  I tried running things through my headphone

Speaking of robust, when I started discussing the matter with Lyle many
months ago, I was concerned that increasing the C value above 10 uF at
C9/C13 would begin to introduce switching thumps that would become apparent
especially in CW.  Stated another way, I was hopeful that the small C size
was not masking keying artifacts that were there all along, only to be
discovered by increasing the cap size!  But at 330 uF, I hear no artifacts
as a result of the change.

Pul, W9AC

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[KK7P]

> ...I did not take into consideration the 8-ohm build-out
> resistance that Lyle just discussed.  Those build-out resistors do not
> appear on the headphone amp schematics (page 28/52) so until today, I doubt
> anyone else knew they existed <g>.

They are in the originally released October 2007 schematic set (page 32
of 43) and marked as K3 DSP IF Rev A1 sheet 7 of 9.

Likewise, they appear on the current (October 2008) schematic set (page
32 of 52) and marked as K3 DSP IF Rev A1 sheet 7 of 9.

I expect they will still be in the next released set of schematics, too :-)

73,

Lyle KK7P

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[P.B. Christensen]

> And that jumps to ~ 200 Hz if you use 8-ohm cans.

Before getting pounced on, that's 100 Hz with the internal 8-ohm build out
resistor.

Paul,W9AC

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[P.B. Christensen]

> They are in the originally released October 2007 schematic set (page 32
> of 43) and marked as K3 DSP IF Rev A1 sheet 7 of 9.

Sigh...

But they don't appear on Page 28/52 where the headphone amp circuit is shown
on the schematic.  They are "off page" components with respect to the
headphone amp schematic.

Paul, W9AC

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Brett Howard]

-----Original Message-----
From: [hidden email]
[mailto:[hidden email]] On Behalf Of Paul Christensen
Sent: Monday, July 06, 2009 12:13 PM
To: [hidden email]
Subject: Re: [Elecraft] K3 Audio Low Pass Filter

>And that jumps to ~ 200 Hz if you use 8-ohm cans.  If you stay cognizant of

>the headphone Z, you're in good shape.  If you want to pull a set of Koss
>Pro4 AA phones off the shelf, you may want to revisit that.  It's all about

>the series C and knowing the headphone Z as a function of frequency.  It's
>doubtful your 27 ohms headphones remain 27 ohms across their frequency
>range.

Actually its 100Hz for 8 ohm cans because of the 8.2ohms inside the K3.

>Speaking of robust, when I started discussing the matter with Lyle many
>months ago, I was concerned that increasing the C value above 10 uF at
>C9/C13 would begin to introduce switching thumps that would become apparent

>especially in CW.  Stated another way, I was hopeful that the small C size
>was not masking keying artifacts that were there all along, only to be
>discovered by increasing the cap size!  But at 330 uF, I hear no artifacts
>as a result of the change.

Good to know!!!

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Alexander Ponomarenko-5]

Paul Christensen:

>> I think 100uF caps is enough value. Look these plots (with 100uF in
>> AF-amp),
>> we see identifical plots for Line-out and for Phone-out.
>>
>> logarithmic view: http://forum.cqham.ru/download.php?id=43583
>> linear view: http://forum.cqham.ru/download.php?id=43582
>>
>> 73! Alex
>
> Alex,  The S/T FFT program you have allows the user to display in semi-log
> mode.  That's the preferred way of looking at audio response. Resolution of
> low-end response will show much more detail that way, especially the area of
> interest below 200 Hz.
>
> Also, keep in mind that the choice in coupling cap value is wholly dependent
> on the load Z.  When making response tests of the headphone output with the
> FFT program and sound card, try loading the headphone output with various
> values of resistance in increments starting at 8-ohms, working your way up
> to say ...50 ohms.  You'll end up with an interesting family of curves that
> can then be saved in the S/T program.
>
> Using a sound card without external input termination often results in
> termination Z onto the circuit of only about 2K-ohm or so, depending on the
> model of the sound card.  Pro sound cards (typically with XLR pig tails)
> have the ability to select various degrees of input Z.
>
> Paul, W9AC

Paul, of course you are right if we talk about Phone-Out,
also on these plots you can see the graph from Line-Out.
This out have to matching ideally with a sound cards.
If we don't have needed LF from Line-Out, where it LF take for Phone-Out?

73! Alex
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Joe Subich, W4TV-4]

Paul,

> Joe, you're right.  I did not take into consideration the
> 8-ohm build-out resistance that Lyle just discussed.  Those
> build-out resistors do not appear on the headphone amp
> schematics (page 28/52) so until today, I doubt anyone
> else knew they existed <g>

They appear at D3 (front) and B3 (rear) of the "K3 DSP IF:
Interconnects and Miscellaneous" (pg. 32/52).  I found them
a while ago after Lyle mentioned their existence in another
e-mail.

> So, they must appear off-page along with the speaker
> coupling cap that also does not show up on the same
> schematic.

The speaker coupling capacitors are 470 uF and appear on
the K3 RF Board (1 of 4) - page 3/52 connected to pins
1/2 and 3/4 of J76 KIO3 (upper right corner).  

> My choice of going to 330 uF was primarily driven by the
> fact that I didn't want to be concerned with the operating Z
> of a new set of headphones if tried, and I had an older DSP
> Board wiith 10 uF caps.  

That's why I will change the caps when I next get into each
of my radios.  I don't have a problem because I primarily
use the Heil.  I will probably use 100 uF, but with leaded
caps I will use what I can make fit comfortably.

73,

   ... Joe, W4TV
 

> -----Original Message-----
> From: Paul Christensen [mailto:[hidden email]]
> Sent: Monday, July 06, 2009 3:02 PM
> To: [hidden email]; 'Brett Howard'
> Cc: [hidden email]
> Subject: Re: [Elecraft] K3 Audio Low Pass Filter
>
>
> > You neglect the effect of the 8.2 Ohm resistor in series with the
> > headphone outputs.
>
> Joe, you're right.  I did not take into consideration the
> 8-ohm build-out
> resistance that Lyle just discussed.  Those build-out
> resistors do not
> appear on the headphone amp schematics (page 28/52) so until
> today, I doubt
> anyone else knew they existed <g>.  So, they must appear
> off-page along with
> the speaker coupling cap that also does not show up on the
> same schematic.
>
> Older headphones with low Z drivers would still benefit from
> a value above
> 100 uF.  But with most modern headphones in the 32-ohm area, and I'll
> repeat...the majority of users would not benefit from higher
> C values above
> 100 uF at C9/C13.  My choice of going to 330 uF was primarily
> driven by the
> fact that I didn't want to be concerned with the operating Z
> of a new set of
> headphones if tried, and I had an older DSP Board wiith 10 uF
> caps.  Since
> re-work was required anyway, I settled on the 330 uF value
> and would still
> choose that value if given the chance today.
>
> Although not as critical for communications audio, the use of
> increasing the
> build-out resistance does in fact lower the damping factor of
> the amplifier.
> Not as important with newer headphones, but it can be a
> factor with older
> low-Z headphones with traditional moving coil drivers where
> more mass is
> moved.
>
> Paul, W9AC
>
>
>
>
>
>
> With that resistor considered, the
> > -3 dB point with 100 uF capacitors as a function of headphone
> > impedance is:
> >
> >   Z       Frequency
> > ----------------------
> >   4         130 Hz
> >   8          98 Hz
> >  16          66 Hz
> >  32          40 Hz
> >  64          22 Hz
> > 200           8 Hz
> > 600           3 Hz
> >
> > Although 8 Ohm phones might be "marginal" for someone
> wanting maximum
> > low end response, 98 Hz is probably suitable for most
> communications
> > uses.  One of the reasons I have not felt the need to change the
> > coupling capacitors in my K3 (although I will be installing 100 uF
> > caps when I make some other changes) is that the -3dB point
> is 80 Hz
> > with the 200 Ohm transducers in my Heil ProSet.  There is a
> noticeable
> > difference compared to the 980 Hz -3 dB point when I use an
> old set of
> > Yaesu headphones with their 8 Ohm transducers.
> >
> > The only reason I plan to change the capacitors is to avoid
> problems
> > when I grab a cheap set of headphones or the rig is used by
> a visitor
> > with their own headphones.
> >
> > 73,
> >
> >   ... Joe, W4TV
> >
> >
> >
> >
> >> -----Original Message-----
> >> From: [hidden email]
> >> [mailto:[hidden email]] On Behalf Of Paul
> >> Christensen
> >> Sent: Monday, July 06, 2009 11:23 AM
> >> To: Brett Howard
> >> Cc: [hidden email]
> >> Subject: Re: [Elecraft] K3 Audio Low Pass Filter
> >>
> >>
> >> > I've been informed Elecraft is already replacing those caps
> >> with 100uF
> >> > caps in the current production units.
> >>
> >> 100 uF is the maximum size compatible with the existing
> PCB pads for
> >> C9/C13 on the main DSP Board.  Moving up beyond 100 uF presents a
> >> problem.  Even at
> >> 100 uF, I believe the SMT caps are now rated at close to 10
> >> WVDC.  There's
> >> just no more scaling back on the SMT cap's voltage rating
> in order to
> >> accommodate a fit with a higher C value above 100 uF.
> >>
> >> Why would anyone want a higher value than 100 uF at
> C9/C13? Consider
> >> a set of modern, medium-Z headphones with a Z of 32 ohms at 100 Hz.
> >>  The -3dB
> >> turnover point calculates to roughly 150 Hz.  Anyone wanting
> >> to take full
> >> advantage of the lowest vocal fundamental (e.g., AM or ESSB
> >> modes) wants
> >> that point down to ~ 70 Hz.   So, a cap value of 220 uF would
> >> be preferred.
> >> I went to 330 uF in a radial style to allow extra margin for
> >> a wide range of
> >> headphone vintages and types.  To make most use of the 100 uF
> >> coupling caps,
> >> I would ensure that my headphones are of the newer, medium Z
> >> type and not
> >> the older style with 4-8 ohm drivers.
> >>
> >> My guess is that scaling the caps beyond 100 uF is only of
> interest
> >> to a small minority of ops.  The majority will probably be happy
> >> with the current
> >> choice of 100 uF.  Going from 10 uF in the prior design to
> >> 100 uF is already
> >> a significant step in the right direction.
> >>
> >> Paul, W9AC
> >>
> >>
> >> ______________________________________________________________
> >> Elecraft mailing list
> >> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> >> Help: http://mailman.qth.net/mmfaq.htm
> >> Post: mailto:[hidden email]
> >>
> >> This list hosted by: http://www.qsl.net
> >> Please help support this email list: http://www.qsl.net/donate.html
> >
> >
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Alexander Ponomarenko-5]

Just finished this "true to life" experiment . :-)
200 Ohm ProSet phones was loaded to phone-out parallel with sound card.
I don't see any difference with 200Ohm-loaded and not loaded pictures:
http://forum.cqham.ru/download.php?id=44172
RX-EQ = Flat always.

Alex

Brett Howard:

> What would be slick is if you had a pair of headphones and a headphone
> splitter Y cable.  Plug the headphones into one port and the sound card
> input into the other and then take measurements.  That would be a bit more
> true to life.
>
> ~Brett
>
> -----Original Message-----
> From: Alexander Ponomarenko [mailto:[hidden email]]
> Sent: Monday, July 06, 2009 8:40 AM
> To: Paul Christensen
> Cc: Brett Howard; [hidden email]
> Subject: Re: [Elecraft] K3 Audio Low Pass Filter
>
> I think 100uF caps is enough value. Look these plots (with 100uF in
> AF-amp),
> we see identifical plots for Line-out and for Phone-out.
>
> logarithmic view: http://forum.cqham.ru/download.php?id=43583
> linear view: http://forum.cqham.ru/download.php?id=43582
>
> 73! Alex
>
> Paul Christensen :
>>> I've been informed Elecraft is already replacing those caps with 100uF
>>> caps in the current production units.
>>>    
>> My guess is that scaling the caps beyond 100 uF is only of interest to a
>> small minority of ops.  The majority will probably be happy with the
> current
>> choice of 100 uF.  Going from 10 uF in the prior design to 100 uF is
> already
>> a significant step in the right direction.
>
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[W6ODJ]

On 3 Jul 2009, at 13:04, Mike Scott wrote:

> I just completed an installation of the K3 audio low pass filter  
> board, the
> one Lyle put out to individuals who wanted to try it. I was busy  
> until now
> as the boards came out some time ago.
>

Can someone point me to the reflector article in which Lyle mentioned  
this mod?  I put up with a horrific noise level here.  Something like  
this might make it more bearable.  Are these boards available?

Thx and 73,

Oliver Johns
W6ODJ
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Brett Howard]

They are not yet available.  They are still in testing...


On Tue, 2009-07-07 at 19:18 -0700, Oliver Johns wrote:

> On 3 Jul 2009, at 13:04, Mike Scott wrote:
>
> > I just completed an installation of the K3 audio low pass filter  
> > board, the
> > one Lyle put out to individuals who wanted to try it. I was busy  
> > until now
> > as the boards came out some time ago.
> >
>
> Can someone point me to the reflector article in which Lyle mentioned  
> this mod?  I put up with a horrific noise level here.  Something like  
> this might make it more bearable.  Are these boards available?
>
> Thx and 73,
>
> Oliver Johns
> W6ODJ
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html