Elecraft › [K3]

2.8kHz roofing filter

Classic

List

Threaded

8 messages Options

ab2tc

2.8kHz roofing filter

Hello all,

I recently replaced the 5-pole 2.7kHz roofing filter in my K3 with the 8-pole 2.8kHz hoping to get rid of the difference in tonality between USB and LSB. But the difference is still there. These two links show the noise spectrum of the receiver in USB and LSB.

http://ab2tc.getmyip.com:8000/pictures/usb.png
http://ab2tc.getmyip.com:8000/pictures/lsb.png

In USB there is a definite upward tilt of the passband and the opposite for LSB. This is quite audible. Is this normal and is there anything I can do about it (other than asking Elecraft to add by mode equalizer)?

AB2TC - Knut

N5GE

Re: 2.8kHz roofing filter

On Sat, 23 Oct 2010 10:52:40 -0700 (PDT), ab2tc <[hidden email]>
wrote:

Hi Knut,

That adjustment should be made by adjusting for the offset of the
center frequency of the filter. Even the 8 pole filters can be off by
a little and it's the most noticeable when switching sidebands.

Adjust the center frequency of the filter until you hear the same
pitch with both filters.

BT 73 ES GUD LUK
DE N5GE,
QCWA LIFE MEMBER 35102,
LICENSED SINCE 1976 AR SK

[hidden email]
http://www.n5ge.com

>
>Hello all,
>
>I recently replaced the 5-pole 2.7kHz roofing filter in my K3 with the
>8-pole 2.8kHz hoping to get rid of the difference in tonality between USB
>and LSB. But the difference is still there. These two links show the noise
>spectrum of the receiver in USB and LSB.
>
>http://ab2tc.getmyip.com:8000/pictures/usb.png
>http://ab2tc.getmyip.com:8000/pictures/lsb.png
>
>In USB there is a definite upward tilt of the passband and the opposite for
>LSB. This is quite audible. Is this normal and is there anything I can do
>about it (other than asking Elecraft to add by mode equalizer)?
>
>AB2TC - Knut

______________________________________________________________
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

Amateur Radio Operator N5GE

Bill W4ZV

Re: 2.8kHz roofing filter

In reply to this post by ab2tc

ab2tc wrote

In USB there is a definite upward tilt of the passband and the opposite for LSB. This is quite audible. Is this normal and is there anything I can do about it (other than asking Elecraft to add by mode equalizer)?

Probably nothing you can do other than tweaking EQ per mode. It's caused by the filter having an actual amplitude tilt in its passband. In USB the upward tilt is on the upper side and in LSB it's on the lower side. I don't think you would notice the small 8-pole offset (typically 50 Hz) in a 2.8k filter, but they are more noticeable in CW filters.

73, Bill

Bill VanAlstyne W5WVO

Re: 2.8kHz roofing filter

Given the fact that the filter passband shapes aren't absolutely flat, as
Bill W4ZV pointed out, you can't expect PERFECT balance between the
sidebands. But you CAN tweak a lot of this audible difference out by
adjusting the filter offset using the K3 Utility, as has been stated
earlier. Even the 8-pole filters can show some required offset in any
specific K3. And the specified offset for a 5-pole filter can benefit from a
little tweaking, depending on which K3 you put it in. I have a 400 Hz
8-pole, a 1.8 kHz 8-pole, and a 2.7 kHz 5-pole. Of the three, only the 1.8
kHz came out pretty close to where it was supposed to be (only -0.01 kHz
offset applied made the sidebands sound pretty much the same). The 400 Hz CW
filter required an offset of -0.11 kHz to make CW and CW-R sound the same,
and the 2.7 kHz 5-pole filter, which was marked with an offset of -0.90 kHz,
actually showed a better audible center with an offset of -1.09 kHz in my
particular radio.

An earlier poster on this (or a related) thread explained how to do this
without test equipment, using just the K3 Utility, the antenna noise floor,
and your ears. You can search for that post in the archives.

Now, in tweaking the offsets using this by-ear method, it's entirely
possible that you're actually shifting the filter a little past center to
compensate for the filter's unflat shape. I can see an argument that this is
not a good thing to do, as well as an argument that it is a fine thing to
do. :-) I'm of the latter persuasion, at least at the moment,
notwithstanding any future arguments to the contrary.

Bill W5WVO

-----Original Message-----
From: Bill W4ZV
Sent: Saturday, October 23, 2010 20:51
To: [hidden email]
Subject: Re: [Elecraft] 2.8kHz roofing filter

ab2tc wrote:
>
> In USB there is a definite upward tilt of the passband and the opposite
> for LSB. This is quite audible. Is this normal and is there anything I can
> do about it (other than asking Elecraft to add by mode equalizer)?
>

Probably nothing you can do other than tweaking EQ per mode. It's caused by
the filter having an actual amplitude tilt in its passband. In USB the
upward tilt is on the upper side and in LSB it's on the lower side. I don't
think you would notice the small 8-pole offset (typically 50 Hz) in a 2.8k
filter, but they are more noticeable in CW filters.

73, Bill

--
View this message in context:
http://elecraft.365791.n2.nabble.com/2-8kHz-roofing-filter-tp5665995p5666368.html
Sent from the [K3] mailing list archive at Nabble.com.
______________________________________________________________
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

Joe Subich, W4TV-4

Re: 2.8kHz roofing filter

On 10/24/2010 9:58 AM, Bill VanAlstyne W5WVO wrote:

> Now, in tweaking the offsets using this by-ear method, it's entirely
> possible that you're actually shifting the filter a little past
> center to compensate for the filter's unflat shape. I can see an
> argument that this is not a good thing to do, as well as an argument
> that it is a fine thing to do. :-) I'm of the latter persuasion, at
> least at the moment, notwithstanding any future arguments to the
> contrary.

Bill,

I'll take the contrary position - at least with respect to narrow
filters.

1) when I last checked the K3 firmware placed the edge of the CW
filter to cut off low frequency audio response around 100 Hz
and eliminate the audio image the other side of the BFO. That
means any tweaking by ear is likely to be inaccurate because
the user is not directly controlling the offset.

2) When tweaking the offset by ear on CW there can be substantial
errors in the true center frequency which will substantially
effect performance on RTTY.

I measure the center frequencies of CW filters in the radio by
disabling the other filters, clearing any RX EQ, turn off AGC, set
the CW pitch to 800 Hz and DSP bandwidth to 2800 Hz. Then I look
for the -6dB points using an XG-2 and the dBV capability of the K3.
With wider filters one would need to measure them in DATA A mode
with FC=2.00 and BW=4.00 to minimize DSP effects on the measurement.

I am rather disappointed with the consistency of the Elecraft/INRAD
400 Hz, 8-pole filters. I've measured four of them in my rigs and
the center frequency is all over the map. The measured deviations
are: -15 Hz, +24 Hz, +45 Hz and +55 Hz.

+45 and +55 Hz is getting close to the point that it can negatively
impact RTTY. I plan to move the filters to put the -15/+24 filters
in one rig (with 0.00 KHz offset) and the +45/+55 filters in the
other with a -0.05 KHz offset.

73,

... Joe, W4TV

On 10/24/2010 9:58 AM, Bill VanAlstyne W5WVO wrote:

> Given the fact that the filter passband shapes aren't absolutely flat, as
> Bill W4ZV pointed out, you can't expect PERFECT balance between the
> sidebands. But you CAN tweak a lot of this audible difference out by
> adjusting the filter offset using the K3 Utility, as has been stated
> earlier. Even the 8-pole filters can show some required offset in any
> specific K3. And the specified offset for a 5-pole filter can benefit from a
> little tweaking, depending on which K3 you put it in. I have a 400 Hz
> 8-pole, a 1.8 kHz 8-pole, and a 2.7 kHz 5-pole. Of the three, only the 1.8
> kHz came out pretty close to where it was supposed to be (only -0.01 kHz
> offset applied made the sidebands sound pretty much the same). The 400 Hz CW
> filter required an offset of -0.11 kHz to make CW and CW-R sound the same,
> and the 2.7 kHz 5-pole filter, which was marked with an offset of -0.90 kHz,
> actually showed a better audible center with an offset of -1.09 kHz in my
> particular radio.
>
> An earlier poster on this (or a related) thread explained how to do this
> without test equipment, using just the K3 Utility, the antenna noise floor,
> and your ears. You can search for that post in the archives.
>
> Now, in tweaking the offsets using this by-ear method, it's entirely
> possible that you're actually shifting the filter a little past center to
> compensate for the filter's unflat shape. I can see an argument that this is
> not a good thing to do, as well as an argument that it is a fine thing to
> do. :-) I'm of the latter persuasion, at least at the moment,
> notwithstanding any future arguments to the contrary.
>
> Bill W5WVO
>
>
> -----Original Message-----
> From: Bill W4ZV
> Sent: Saturday, October 23, 2010 20:51
> To: [hidden email]
> Subject: Re: [Elecraft] 2.8kHz roofing filter
>
>
>
> ab2tc wrote:
>>
>> In USB there is a definite upward tilt of the passband and the opposite
>> for LSB. This is quite audible. Is this normal and is there anything I can
>> do about it (other than asking Elecraft to add by mode equalizer)?
>>
>
> Probably nothing you can do other than tweaking EQ per mode. It's caused by
> the filter having an actual amplitude tilt in its passband. In USB the
> upward tilt is on the upper side and in LSB it's on the lower side. I don't
> think you would notice the small 8-pole offset (typically 50 Hz) in a 2.8k
> filter, but they are more noticeable in CW filters.
>
> 73, Bill
>
>

ab2tc

Re: 2.8kHz roofing filter

In reply to this post by Bill W4ZV

I want to make two macros, one that goes to USB mode and sets the RX EQ accordingly, another that goes to LSB and sets the RX EQ for that mode. This would be very simple to operate. But there is no RX equalizer set command that I can see (there is one for TX EQ). Another poster had this same questions a few weeks back and I do not think he got an answer. Any ideas anyone?

AB2TC - Knut

Bill W4ZV wrote

Probably nothing you can do other than tweaking EQ per mode. It's caused by the filter having an actual amplitude tilt in its passband. In USB the upward tilt is on the upper side and in LSB it's on the lower side. I don't think you would notice the small 8-pole offset (typically 50 Hz) in a 2.8k filter, but they are more noticeable in CW filters.

73, Bill

ab2tc

Re: 2.8kHz roofing filter tilt

Hi again,

I am still looking for an answer to this. Or let me rephrase the question: Is there a plan to provide a command to set the RX equalizer in the foreseeable future? Or alternatively to provide separate equalizer settings for USB/LSB?

AB2TC - Knut

ab2tc wrote

I want to make two macros, one that goes to USB mode and sets the RX EQ accordingly, another that goes to LSB and sets the RX EQ for that mode. This would be very simple to operate. But there is no RX equalizer set command that I can see (there is one for TX EQ). Another poster had this same questions a few weeks back and I do not think he got an answer. Any ideas anyone?

AB2TC - Knut

Bill W4ZV wrote

Probably nothing you can do other than tweaking EQ per mode.

Brian J. Linn

Re: 2.8kHz roofing filter tilt

I second a desire for an RX set command similar to the TX set command.

Brian KD0HII

Hi again,

I am still looking for an answer to this. Or let me rephrase the question:
Is there a plan to provide a command to set the RX equalizer in the
foreseeable future? Or alternatively to provide separate equalizer settings
for USB/LSB?

AB2TC - Knut

ab2tc wrote:

>
> I want to make two macros, one that goes to USB mode and sets the RX EQ
> accordingly, another that goes to LSB and sets the RX EQ for that mode.
> This would be very simple to operate. But there is no RX equalizer set
> command that I can see (there is one for TX EQ). Another poster had this
> same questions a few weeks back and I do not think he got an answer. Any
> ideas anyone?
>
> AB2TC - Knut
>
>
> Bill W4ZV wrote:
>>
>> Probably nothing you can do other than tweaking EQ per mode.
>
>