K3 Filter response
Locked
 
|
Hello,
I've traced the 2.8 kHz filter response with Mike's software and i get a little "anomaly" i can't explain me. I've a *.bmp file (~825 kB) or a *.txt file (~2kB). How can i "show" this unexplained (for me!) anomaly? I don't find a file upload procedure for this group. Thank you for your help. 73 QRO fr Rudolf, HB9ARI _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
I am not sure where it ends up but if you use the web to write you will see an "upload file" button and an "add link" button.
Clipboard01.jpg |
|
|
|
|
In reply to this post by AD6XY
Hello,
I wrote this message directly from my e-mail software as i do habitually. As most of the time i use Yahoo groups, i've a link for files, pictures uploading. I know i've some "memory" problems (66 y/o) but i never used n2.nabble.com. By looking your 2k7 filter plot, i saw that my "anomaly" (the little peak at the right part) is also present on your plot. For me it's no more a "local" anomaly and the question is partially closed. 73, Rudolf AD6XY wrote: > I am not sure where it ends up but if you use the web to write you will see > an "upload file" button and an "add link" button. > http://n2.nabble.com/file/n2191105/Clipboard01.jpg Clipboard01.jpg > _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
The "anomaly" is not really an anomaly, it is just the other side of the down conversion oscillator. For a filter to pass an SSB audio signal the passband needs to go down to 300Hz or so. For this to happen the attenuation for 0Hz and image frequencies can't be that far down. That is all it is though it is probably better with the 8 pole filter. It can also be eliminated by raising the centre frequency, at the expense of bass response. Unfortunately I don't think you can do this on transmit, though perhaps you could adjust your filter offset. That would not work with the USB/LSB inversion below/above 8MHz.
Currently the filter is centred on 1.5kHz by default. The best place for it but if you narrow it down with DSP it is probably better to raise this to 1.8kHz. I am not sure how the filter shift relates to the DSP filter and the crystal filter, i.e. which is adjusted, probably both. However, I find I always need to reduce the filter centre frequency to maintain intelligibility as I narrow the DSP bandwidth. So to make sense of a signal within 1.8kHz I need a 1kHz centre frequency. That is not what I would expect which makes me wonder what the shift control actually does. Mike |
|
|
Mike,
Thank you for the explanation; i understand now that this false "anomaly" is present with the 2k7 and 2k8 filters. May i "abuse" of your knowledge to explain to me the following behavior : During receive, the 2k7 and 2k8 filter seem to give a ~ 300 Hz at -10dB frequency; it's ok for my use; during transmit, the same filters seems to give a ~ 500 Hz at - 10dB relative to the 1 kHz attenuation; i've tried to shift the bandpass, but naturally, the high frequency is shifted too. As i work essentially with digimodes, i can accommodate with that; but what is the cause of this difference ? (coupling capacitor, etc). I've tried to "play" with EQUALIZER, but the ALC "bars" begin to get non desired values... 73 QRO fr Rudolf, HB9ARI AD6XY wrote: > The "anomaly" is not really an anomaly, it is just the other side of the down > conversion oscillator. For a filter to pass an SSB audio signal the passband > needs to go down to 300Hz or so. For this to happen the attenuation for 0Hz > and image frequencies can't be that far down. That is all it is though it is > probably better with the 8 pole filter. It can also be eliminated by raising > the centre frequency, at the expense of bass response. Unfortunately I don't > think you can do this on transmit, though perhaps you could adjust your > filter offset. That would not work with the USB/LSB inversion below/above > 8MHz. > > Currently the filter is centred on 1.5kHz by default. The best place for it > but if you narrow it down with DSP it is probably better to raise this to > 1.8kHz. I am not sure how the filter shift relates to the DSP filter and the > crystal filter, i.e. which is adjusted, probably both. > > However, I find I always need to reduce the filter centre frequency to > maintain intelligibility as I narrow the DSP bandwidth. So to make sense of > a signal within 1.8kHz I need a 1kHz centre frequency. That is not what I > would expect which makes me wonder what the shift control actually does. > > > > Mike > Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by AD6XY
As I understand it, the K3 switches to a different input bandpass filter
at 10.000 MHz which I believe is the source of the artifact. You will get a cleaner response with K3 Filter if the test signal is not at 10 MHz. That's unfortunate, as it's easy to find a clean strong 10 MHz reference from test gear. Perhaps Mike should add a reminder in the help file, when he completes the higher priority items. Jack K8ZOA AD6XY wrote: > And why not use it. > > http://n2.nabble.com/file/n2191129/27k1.jpg 27k1.jpg > http://n2.nabble.com/file/n2191129/2k72.jpg 2k72.jpg > > http://n2.nabble.com/file/n2191129/400H1.jpg 400H1.jpg > http://n2.nabble.com/file/n2191129/400H2.jpg 400H2.jpg > > http://n2.nabble.com/file/n2191129/13kH1.jpg 13kH1.jpg > http://n2.nabble.com/file/n2191129/13kh2.jpg 13kh2.jpg > Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
I don't know about this but I find my K3 switches the filter and band at 9MHz. I tried using 9MHz just now to verify this. The mode changed as I moved from 9MHz to just above. I don't understand why but the mode changed from USB just below 9MHz to TX data just above it. Odd. This does not happen when you spin the dial so perhaps a bug in Mikes program. I also found another bug. The AGC was turned off and I was supposed to set the centre frequency. The audio was so loud that it was distorted badly and the volume would not decrease with the pot, even the radio power off switch failed to work. I had to cut the signal to regain control. I have noticed that the K3 audio into the loudspeaker does distort very badly indeed but stopping the important controls working was a surprise. Power off should do just that, always, no delay no debate. Mike |
Re: Glitches in K3 Filter response
|
|
I went back and looked at the response data I took in August using my
version of the self-test bandwidth program and see the same artifacts. I work very little SSB and did not pay as much attention to the SSB filter plot as I should have - with the narrower filters for data and CW, the artifact seems either invisible or greatly attenuated. It's not related to input filter switching. Sorry about any confusion. Jack K8ZOA AD6XY wrote: > > Jack Smith-6 wrote: > >> As I understand it, the K3 switches to a different input bandpass filter >> at 10.000 MHz which I believe is the source of the artifact. You will >> get a cleaner response with K3 Filter if the test signal is not at 10 MHz. >> >> > > I don't know about this but I find my K3 switches the filter and band at > 9MHz. I tried using 9MHz just now to verify this. The mode changed as I > moved from 9MHz to just above. I don't understand why but the mode changed > from USB just below 9MHz to TX data just above it. Odd. This does not happen > when you spin the dial so perhaps a bug in Mikes program. > > I also found another bug. The AGC was turned off and I was supposed to set > the centre frequency. The audio was so loud that it was distorted badly and > the volume would not decrease with the pot, even the radio power off switch > failed to work. I had to cut the signal to regain control. > > I have noticed that the K3 audio into the loudspeaker does distort very > badly indeed but stopping the important controls working was a surprise. > Power off should do just that, always, no delay no debate. > > Mike > Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
Re: Glitches in K3 Filter response
|
|
In reply to this post by Jack Smith-6
The "glitch" shown in the USB plots here are the response on the other side of zero beat. It will be visible if the DSP bandwidth is set to 0 - 4000 Hz because the IF filter will be placed with the "corner" at the carrier frequency and there will be response on the other side of the carrier on the filter skirt. The "notch" effect is due to audio response in the DSP as the audio frequency gets close to DC. 73, ... Joe, W4TV > -----Original Message----- > From: [hidden email] > [mailto:[hidden email]] On Behalf Of Jack Smith > Sent: Wednesday, January 21, 2009 7:15 AM > To: AD6XY; 'Elecraft' > Subject: [Elecraft] Glitches in K3 Filter response > > > As I understand it, the K3 switches to a different input > bandpass filter > at 10.000 MHz which I believe is the source of the artifact. You will > get a cleaner response with K3 Filter if the test signal is > not at 10 MHz. > > That's unfortunate, as it's easy to find a clean strong 10 > MHz reference > from test gear. > > Perhaps Mike should add a reminder in the help file, when he > completes > the higher priority items. > > Jack K8ZOA > > > AD6XY wrote: > > And why not use it. > > > > http://n2.nabble.com/file/n2191129/27k1.jpg 27k1.jpg > > http://n2.nabble.com/file/n2191129/2k72.jpg 2k72.jpg > > > > http://n2.nabble.com/file/n2191129/400H1.jpg 400H1.jpg > > http://n2.nabble.com/file/n2191129/400H2.jpg 400H2.jpg > > > > http://n2.nabble.com/file/n2191129/13kH1.jpg 13kH1.jpg > > http://n2.nabble.com/file/n2191129/13kh2.jpg 13kh2.jpg > > _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by AD6XY
Mike AD6XY wrote:
> I don't know about this but I find my K3 switches the filter and band at > 9MHz. I tried using 9MHz just now to verify this. The mode changed as I > moved from 9MHz to just above. I don't understand why but the mode changed > from USB just below 9MHz to TX data just above it. Odd. This does not > happen > when you spin the dial so perhaps a bug in Mikes program. 9 MHz is the transition between two bands (40 M & 30 M) on the K3. When you used K3 Filter Tools to cross the 9 MHz boundary the mode will automatically be switched from the last mode used on 40 M to the last one used on 30 M. Thus, your mode automatically changed from SSB to Data. Users should be aware of these transition boundaries between the bands when they are using K3 Filter Tools. > I also found another bug. The AGC was turned off and I was supposed to set > the centre frequency. The audio was so loud that it was distorted badly > and > the volume would not decrease with the pot, even the radio power off > switch > failed to work. I had to cut the signal to regain control. I have not been able to reproduce this problem, but it is not a bug in the program. K3 Filter Tools does not control the AF gain control at all, but it does turn off AGC. I am modifying the program to turn down the AF gain just before it turns off AGC. That will prevent the loud, annoying burst of sound users are experiencing. Thanks for your comments. Mike, KS7D _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
«
Return to Elecraft
|
1 view|%1 views
| Free forum by Nabble | Edit this page |