Re: Elecraft Digest, Vol 37, Issue 19

[I7SWX]

Hi Bill and Geoff,

thanks for your comments.

I knew I was going to agitate "the elecraft users sea"
but I believe we ham should share ideas, experiences
and knowledge.

I will see to reply to your comments.

Message: 40
Date: Wed, 9 May 2007 08:04:39 +0100
From: "Geoffrey Mackenzie-Kennedy"
<[hidden email]>
Subject: Re: [Elecraft] Improvement changes to K2 and
K1 transceivers
To: "Bill Coleman" <[hidden email]>
Cc: Elecraft Discussion List
<[hidden email]>, Giancarlo
        Moda I7SWX <[hidden email]>
Message-ID:
<000e01c79208$8b1ac6d0$0f9e87d9@your1ff2e8a4b1>
Content-Type: text/plain; format=flowed;
charset="Windows-1252";
        reply-type=response

Bill,

Gian might be suggesting the use of H-mode mixers to
replace the TUF-1
mixer
and NE602 product detector, but I have not asked him.
The H-mode mixer
has a
much higher 3rd Order Intercept than either the TUF-1
or NE602, but I
would
have thought that the very good IMD performance of the
H-mode mixer
would be
wasted in the K2. As I see it the IMD performance of
the 2N5109 post
mixer
amplifier, variable bandwidth IF filter and MC1350 IF
amplifier (and
probably the front end bandpass filters) would become
the weak link,
and
determine the overall IMD performance of a modified
K2.

> Geoff you are right. Certainly the H-Mode Mixer has
higher IMD performances then TUF-1 and NE602.
Some friends in Italy had problems with K2 (and also
K1) on 40meters during contests using good antennas.
This is how the idea of trying to improve the K2
started.

> Certainly the worst performance is in the post mixer
amplifier and this is the reason why the idea of the
simple 2 xtal roofing filter was implemented. THe
following bandpass xtal filter is protecting the IF
and demodulator stages.

> Unless we do trials on changes we cannot really
calculate if the improvements are really worthed or
not ... but our is an hobby and if one like the "risk"
he can do what he like... then he can go back to the
original.
If one is happy with his equipment it should not touch
it ... the H-Mode mixer has been implemented and
tested in many commercial transceiver, old and new
ones. It Is possible to guess if improvements are
worth or not ... but the "adventure changes calls" are
always around....hi

> To give you an idea, a friend in W-land tried the
suggested mods for his TR7 ... the improvements were
limited by the high phase noise of the PLL ...he then
replaced with an AD9951 DDS ... and the big difference
did show up right away.
> Another friend in Italy tested the H-Mode Mixer and
checked the improvements possible in his IC706 first
mixer mod. The original dynamic range at 20kHz tones
spacing was 85-88dB, with the H-Mode implemented it
went to 98dB. Replacing the original PLL with a xtal
oscillator it went to 105dB... clearly phase noise
differences between PLL and xtal oscillator.
> Although you may have not "enough improvements"
because the stahge is limited by the phase noise ... I
believe it is better than having both the mixer and
the PLL limitations.

  Also the
H-mode
mixer requires a 50% duty cycle squarewave LO
injection for best IMD
performance which would require some modification of
the K2's LO and
BFO
output circuitry, although this modification might not
be necessary
because
of the poor (by comparison) IMD performance of the
weak link mentioned
above.

> NO. The H-Mode Squarer has a balance adjustment and
you do not need to put your hands in the PLL. Also the
HMM require 0dBm drive or less, so youmay have less
spurs. The BFO 50-50 balance is not necessary as the
LO frequency is low and the squarer gives quite a good
50/5o outputs.

73,
Geoff
GM4ESD


 On Wednesday, May 09, 2007 at 2:46 AM, Bill Coleman
wrote:

Part of the beauty of the K2 is that it is a very
simple radio,
superbly executed. (To paraphrase a comment K9AY said
to me before I bought one)

> The K2 is certainly a simple and good radio. This
does not means it has no limits. Nothing is perfect.
Being only ham radio bands it helps a lot having
narrow input band pass filters. Try to remove these
and you will see something is changing. Listen to
40meter in US and then come to Europe and listen our
40 meters when a contest is on, CW or SSB.

The K2 uses a TUF-1 balanced mixer with close
attention to the
injection levels and impedance matching. You'd be
hard-pressed to
improve the mixer performance for HF.

The H-Mode Mixer is not sensitive to impedances
terminations like a classic diodes d b mixer,
particularly to the LO. Even the IF output when loaded
with a variable load like a xtal filter does not
reduce too much the IP3 (tests give between 5 to 10dBm
changes.. i.e from +40dBm to +35 or +30dBm as second
of type of filter.)

Since the K2 is single-conversion, the primary crystal
filter IS the
roofing filter. The primary crystal filter is pretty
darn good for CW
or RTTY.

> Certainly the filter is a good filter but if IMD is
reduced in the TUF-1 or more in the 2N5109 post mixer
amplifier, the IF filter cannot do much.


> 2)Demodulator replacement.

What would you do to replace the product detector?
Since the dynamic
range of the signals after the AGC of the IF stage is
pretty limited,
the NE602 is a pretty good choice. It wouldn't be easy
to do better,
especially at the price point for the K2.

> As I wrote, the Demodulator changes have not been
applied... they are reported as to experiment with
this stage and see what happen. You have to think that
the NE602 IP3 is around -20dBm while the 1T DBM has an
IP3 of at least +25dBm.
> I know that my ideas can be heritics for the K2 ...
as they seems to try to change the entire K2 receiver
stages... but these are ideas, right or wrong, and one
should not take them as melted gold... hi

> There are friends that have modified a lot of
equipments and are pleased. Some of these are:
Heathkit SB204, TS940, IC751, FT981 (here we
discovered one RF amplifier self-oscillating at vhf),
FT1000MP (2nd Mixer) anmd some others on trial...

> What I believe is also important to share comments
like your as it is part of the learning process and
share knowledge.

73

Gian
I7SWX





 
____________________________________________________________________________________
Food fight? Enjoy some healthy debate
in the Yahoo! Answers Food & Drink Q&A.
http://answers.yahoo.com/dir/?link=list&sid=396545367
_______________________________________________
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

[Geoffrey Mackenzie-Kennedy-2]

Hi Gian,

Your messages have arrived!!

Giancarlo Moda [hidden email] wrote:
on Wednesday, May 09, 2007 2:48 PM

>> Geoff you are right. Certainly the H-Mode Mixer has
> higher IMD performances then TUF-1 and NE602.
> Some friends in Italy had problems with K2 (and also
> K1) on 40meters during contests using good antennas.
> This is how the idea of trying to improve the K2
> started.

I experience the same problem here with the K2 on 40m all the time when the
BC stations above 7100 kHz are being received at S9 ++, and have to use
another receiver which uses H-Mode mixers etc.

> Certainly the worst performance is in the post mixer
> amplifier and this is the reason why the idea of the
> simple 2 xtal roofing filter was implemented. THe
> following bandpass xtal filter is protecting the IF
> and demodulator stages.

Understood. Without running IMD tests on my K2/100's receiver and its IF
cascade I can only guess that this might not be enough protection when the
BC station carrier levels are up in the -10dbm region, and you are trying to
copy a signal that is in between them. This is the situation that I have
when using SSB between 7150 and 7197 kHz. The situation improves of course
if the beam is pointing away from the BC stations. There is also the
question of roofing filter linearity and Rx phase noise. There is not much
that we can do about BC station Tx phase noise!

>> Although you may have not "enough improvements"
> because the stahge is limited by the phase noise ... I
> believe it is better than having both the mixer and
> the PLL limitations.

Understood!

>> NO. The H-Mode Squarer has a balance adjustment and
> you do not need to put your hands in the PLL. Also the
> HMM require 0dBm drive or less, so youmay have less
> spurs. The BFO 50-50 balance is not necessary as the
> LO frequency is low and the squarer gives quite a good
> 50/5o outputs.

I understand, but I was looking at the Squarer as part of the circuitry to
be added to the K2's LO (VCO) *output* which provides injection to the TUF-1
mixer.

--------------------------------------------
>From your second message:

> The H-Mode Mixer is not sensitive to impedances
> terminations like a classic diodes d b mixer,
> particularly to the LO. Even the IF output when loaded
> with a variable load like a xtal filter does not
> reduce too much the IP3 (tests give between 5 to 10dBm
> changes.. i.e from +40dBm to +35 or +30dBm as second
> of type of filter.)

Agree about terminations, but there is a limit!! The SD5000 FET array (now
obsolete) with three transformers yields an Input IP3 of +50dbm or better at
the worst case input levels below compression, but my test equipment is
becoming stretched. Perhaps the newer arrays will provide the same
performance.

> Certainly the filter is a good filter but if IMD is
> reduced in the TUF-1 or more in the 2N5109 post mixer
> amplifier, the IF filter cannot do much.

Have you measured the filter's input IP3 with the varicaps' back bias at the
minimum value used?

Good luck and many thanks.

73,
Geoff
GM4ESD


_______________________________________________
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

[I7SWX]

Hi Geoff,

--- Geoffrey Mackenzie-Kennedy
<[hidden email]> wrote:

> Hi Gian,
>
> Your messages have arrived!!

I am glad you got the message.

 
>
> I experience the same problem here with the K2 on
> 40m all the time when the
> BC stations above 7100 kHz are being received at S9
> ++, and have to use
> another receiver which uses H-Mode mixers etc.

If we have problems in Italy certainly the same is in
UK on 40m.

>
> Understood. Without running IMD tests on my K2/100's
> receiver and its IF
> cascade I can only guess that this might not be
> enough protection when the
> BC station carrier levels are up in the -10dbm
> region, and you are trying to
> copy a signal that is in between them. This is the
> situation that I have
> when using SSB between 7150 and 7197 kHz. The
> situation improves of course
> if the beam is pointing away from the BC stations.

Yes, ointing the bem to another direction is like
inserting an attenuator

> There is also the
> question of roofing filter linearity and Rx phase
> noise.

I am not sure we can call the first K2 Xtal filter a
roofing filter as this is the main Bandpass filter.
The second filter I believe it is an IF noise limiting
filter to reduce IF noise.
I believe the linearity problem is in the 2N5109 stage
more than in the xtal filter. You need to do
measurement to see which is the stage. In the CDG2000
we have linearity problems with passive components
like front end coils and transformers.

There is not much
> that we can do about BC station Tx phase noise!

Sometime they overmodulate

> >> Although you may have not "enough improvements"
> > because the stahge is limited by the phase noise
> ... I
> > believe it is better than having both the mixer
> and
> > the PLL limitations.
>
> Understood!
>
> >> NO. The H-Mode Squarer has a balance adjustment
> and
> > you do not need to put your hands in the PLL. Also
> the
> > HMM require 0dBm drive or less, so youmay have
> less
> > spurs. The BFO 50-50 balance is not necessary as
> the
> > LO frequency is low and the squarer gives quite a
> good
> > 50/5o outputs.
>
> I understand, but I was looking at the Squarer as
> part of the circuitry to
> be added to the K2's LO (VCO) *output* which
> provides injection to the TUF-1
> mixer.

Yes, the squarer is the one that square the LO signal
to drive  the FST3125 switches in the mixer. The STAR
has the same squarer that drive the 3t or 2t H-Mode
Mixer the user like to implement.
>
> --------------------------------------------
> From your second message:
>

>
> Agree about terminations, but there is a limit!! The
> SD5000 FET array (now
> obsolete) with three transformers yields an Input
> IP3 of +50dbm or better at
> the worst case input levels below compression, but
> my test equipment is
> becoming stretched. Perhaps the newer arrays will
> provide the same
> performance.

Yes... The SD5000 was good at its time... the problem
is that it required several adjustments and this is
not a good solution for people without proper test
equipment. THe FST3125 H-Mode MIxer is very simple, it
does not require adjustments like the original mixer.
Exception the balance if you are looking for the
highest IP3. The other aspect of 3125 vs SD5000 is the
lower insertion losses, > -4.5dB vs -8dB or more.
That's why it gives you a very hot front end not
needing preamp.
If you look at the measurements done by PA3AKE on the
H-Mode Mixer with 3 and 2 transformers using FST3125
and new switches like FSA3157, the average IP3 can be
+40 to +45dBm with 0dBm drive or less.
I had the idea to use FST3125 and the like in mixer in
1998, at this time ... 10 years ago; SD5000 was
implemented in the G3SBI H-Mode Mixer in 1993, 15 yeas
ago.
Now the search for new switches in the H-Mode Mixer
are to reduce DDS spurs as all the other parameters
are already good ... hi  Nothing is perfect...

>
> Have you measured the filter's input IP3 with the
> varicaps' back bias at the
> minimum value used?

Receiver IP3 measurements were done, before and after
mixer and roofing filter mods, with IF xtal filter in
its widest bandwidth.

>
> Good luck and many thanks.

Thanks to you and best 73

Gian
I7SWX

> 73,
> Geoff
> GM4ESD
>
>
>



 
____________________________________________________________________________________
Sucker-punch spam with award-winning protection.
Try the free Yahoo! Mail Beta.
http://advision.webevents.yahoo.com/mailbeta/features_spam.html
_______________________________________________
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