Down or Up conversions designs

[Georges Ringotte F6DFZ]

Out of an extremely interesting private mail, I don't  have got comments on this topic, which was much debated in the past.
I post it once again  because I think the Elecraft team and Elecraft aficionados are more apt to discuss this subject than on other reflectors dealing with down conversion receivers. I added some comments.
Sorry if I am boring you!

George

Here it is :

Hi to the group

Even though I own and appreciate very much my K3, in fact I believe it's the
best sounding receiver I ever had, I have always been an advocate of up
conversion design, namely because it's the only way to get general coverage and
also because some characteristics, for a given cost, are better with an up
conversion design (IF and image rejections for example). The probability is
low, but if in the future a WARC give a new band very near the IF of a down
conversion design, it will not be able to cover it.

Due to the Sherwood chart, and the way receivers are classed, there is a trend
for manufacturers to propose down conversion design, even though some
characteristics may be poor ; I think of the Yaesu FT5000 for example, very
high in the Rob Sherwood chart, but with poor image rejection, and extremely
poor IMD2 figure for the second receiver.
Even Kenwood comes with a high end transceiver with a strictly down conversion first receiver, and a mixed design second receiver.
Only Icom seems to resist. The IC7700, even if costly and not perfect, employs some very interesting features (preselector, relays switching of RF filters, large coils, good gain distribution, etc...)

I always thought the key to success was an up conversion design with very good
gain distribution, like RS XK2100 transceiver for example.
Ten Tec tried this with the Omni 7, but with limited success, even for remote signals IMD3.

Now the 1st transceiver on Rob's chart is an up conversion design, with a VHF
roofing filter 50 kHz wide, proof of the validity of optimized gain
distribution ; it has 105 dB dynamic range with signals separated by only 1
kHz.
In the past, the Signal One CX7 and CX11 were up conversion design, with a first IF from 39 to 40 MHz (the real bandwidth was much wider than 1 MHz of course), and the CX11 managed to get very good IMD3 inside this bandwidth, proof of a very good gain distribution.

Perhaps, after the current trend to develop down conversion receivers
(Elecraft's rigs, Eagle, FT3K, TS-590, TS990) it will revert trend and we will see up
conversion design with large dynamic range for closed spaced signals.

Yes this German made transceiver, built like a measuring instrument, is very
expensive, but surely it's concept can  be used to market simpler rigs (100W,
13,8V, only one receiver, no VHF, not so luxurious, with true general coverage ...) for a more correct price.

Any comments?

Best regards.




Georges F6DFZ
______________________________________________________________
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

[Alan Bloom]

As with any design decision, there are always trade-offs: each option
has its pros and cons.  I think an up-conversion architecture is easier
to design and potentially cheaper.  And as you say, there is no "hole"
in the coverage around the intermediate frequency.

However I believe it is harder to get top performance.  Because the
local oscillator is at a higher frequency it will tend to have more
phase noise, all other things being equal.  Also, with a VHF IF, it is
not possible to make a narrow-bandwidth, high-performance roofing
filter, which is important when working a big pileup or contest where
strong interfering signals are very close to the weak signal you are
trying to copy.

It is possible to mitigate the latter problem with a carefully-chosen
gain distribution using amplifiers and mixers before the second-IF
filter that have very high dynamic range.  However, that generally
requires high-power devices and/or complex feedback circuitry which
works against the advantages of simplicity and low cost.  I believe the
design goals of the K3 included small size and low power consumption,
which is more do-able in a high-performance receiver with a
down-conversion design.

> Perhaps, after the current trend to develop down conversion receivers
> (Elecraft's rigs, Eagle, FT3K, TS-590, TS990) it will revert trend

Down conversion was universal before the late 1970s.  Then up conversion
became the fad.  Now we're back to down conversion.

Plus ca change...

Alan N1AL


On Fri, 2012-10-12 at 08:36 +0200, Georges Ringotte F6DFZ wrote:

> Out of an extremely interesting private mail, I don't  have got comments on this topic, which was much debated in the past.
> I post it once again  because I think the Elecraft team and Elecraft aficionados are more apt to discuss this subject than on other reflectors dealing with down conversion receivers. I added some comments.
> Sorry if I am boring you!
>
> George
>
> Here it is :
>
> Hi to the group
>
> Even though I own and appreciate very much my K3, in fact I believe it's the
> best sounding receiver I ever had, I have always been an advocate of up
> conversion design, namely because it's the only way to get general coverage and
> also because some characteristics, for a given cost, are better with an up
> conversion design (IF and image rejections for example). The probability is
> low, but if in the future a WARC give a new band very near the IF of a down
> conversion design, it will not be able to cover it.
>
> Due to the Sherwood chart, and the way receivers are classed, there is a trend
> for manufacturers to propose down conversion design, even though some
> characteristics may be poor ; I think of the Yaesu FT5000 for example, very
> high in the Rob Sherwood chart, but with poor image rejection, and extremely
> poor IMD2 figure for the second receiver.
> Even Kenwood comes with a high end transceiver with a strictly down conversion
> first receiver, and a mixed design second receiver.
> Only Icom seems to resist. The IC7700, even if costly and not perfect, employs
> some very interesting features (preselector, relays switching of RF filters,
> large coils, good gain distribution, etc...)
>
> I always thought the key to success was an up conversion design with very good
> gain distribution, like RS XK2100 transceiver for example.
> Ten Tec tried this with the Omni 7, but with limited success, even for remote
> signals IMD3.
>
> Now the 1st transceiver on Rob's chart is an up conversion design, with a VHF
> roofing filter 50 kHz wide, proof of the validity of optimized gain
> distribution ; it has 105 dB dynamic range with signals separated by only 1
> kHz.
> In the past, the Signal One CX7 and CX11 were up conversion design, with a first
> IF from 39 to 40 MHz (the real bandwidth was much wider than 1 MHz of course),
> and the CX11 managed to get very good IMD3 inside this bandwidth, proof of a
> very good gain distribution.
>
> Perhaps, after the current trend to develop down conversion receivers
> (Elecraft's rigs, Eagle, FT3K, TS-590, TS990) it will revert trend and we will
> see up conversion design with large dynamic range for closed spaced signals.
>
> Yes this German made transceiver, built like a measuring instrument, is very
> expensive, but surely it's concept can  be used to market simpler rigs (100W,
> 13,8V, only one receiver, no VHF, not so luxurious, with true general coverage ...)
> for a more correct price.
>
> Any comments?
>
> Best regards.
>
>
>
>
> Georges F6DFZ

______________________________________________________________
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

[Bill Frantz]

On 10/12/12 at 12:12 AM, [hidden email] (Alan Bloom) wrote:

>Down conversion was universal before the late 1970s.  Then up conversion
>became the fad.  Now we're back to down conversion.

I look at the K3 as a "middle conversion" design. With the 8.215
MHz IF; 160M, 80M and 40M are up converted and all the others
are down converted. It works very well.

For a real down conversion radio, consider my old Hallicrafters
SX-99. Everything is converted to a 455KHz IF which includes a
crystal filter (I think one poll). It really doesn't do very
well on image rejection. :-)

Cheers - Bill, AE6JV

---------------------------------------------------------------------------
Bill Frantz        |"Web security is like medicine - trying to
do good for
408-356-8506       |an evolved body of kludges" - Mark Miller
www.pwpconsult.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

[Geoffrey Mackenzie-Kennedy-3]

Georges,

Although the Hilberling PT-8000 series of transceivers are very expensive,
this could be because they have been designed for Military and 'commercial'
use and not only for amateurs.  I believe that it should be possible to
design and market a high performance transceiver using an up-conversion
architecture, whose selling price is reasonable for most amateurs.

While I agree with Alan's N1AL comment about the higher density of phase
noise generated by the receiver's first LO in an up-conversion scheme, it
was possible back in 1994 to devise a LO system whose phase noise density
allowed a SFDR3 > 100db at 2 kHz spacing to be obtained from an
up-conversion HF receiver, whose IF was at low VHF.  However the task was
not easy at the time.

With regard to roofing filters there is a problem which is often overlooked,
and is present in both HF and VHF crystal filters.  This is the level of
filter generated IMD vs. filter bandwidth.  The results I have of tests
performed by other people and myself, show that the level of odd order IMD
generated by a crystal ladder filter using a given quality of quartz
increases as the filter's bandwidth is decreased.  This effect can be seen
when the two parent signals used during the tests are either inside or
outside the filters passband, which places a question mark over the use of
very narrow bandwidth roofing filters in a receiver designed for high
performance.

Again I agree with Alan that the use of amplifiers and mixers before the
second IF filter, which have very high dynamic range, should be used to
offset the effects of signals passing through a wider bandwidth roofing
filter.  IMHO these are not very complex circuits, but the amplifiers
usually operate at high values of standing current.  Comprehensive Gain
Distribution calculations are essential, of course, when designing a
receiver.  I would also suggest that the frontend gain between the antenna
input and roofing filter's input should be negative in value to avoid
overdriving the roofing filter, provided that a useful receiver Noise Figure
is obtained.

73,

Geoff
LX2AO



On Friday, October 12, 2012 at 8:36 AM, Georges Ringotte F6DFZ wrote:


> Out of an extremely interesting private mail, I don't  have got comments
> on this topic, which was much debated in the past.
> I post it once again  because I think the Elecraft team and Elecraft
> aficionados are more apt to discuss this subject than on other reflectors
> dealing with down conversion receivers. I added some comments.

<snip>

> Yes this German made transceiver, built like a measuring instrument, is
> very
> expensive, but surely it's concept can  be used to market simpler rigs
> (100W,
> 13,8V, only one receiver, no VHF, not so luxurious, with true general
> coverage ...) for a more correct price.


______________________________________________________________
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

[Georges Ringotte F6DFZ]

Hi to Geoff and to the Group,

Geoff said :
Although the Hilberling PT-8000 series of transceivers are very expensive,
this could be because they have been designed for Military and 'commercial'
use and not only for amateurs.  I believe that it should be possible to
design and market a high performance transceiver using an up-conversion
architecture, whose selling price is reasonable for most amateurs.Yes, it's built like a RS instrument and a lot of money went also with its 200W, 2
receivers, 144MHz coverage, high end power supply, all filters included...  The results I have of tests performed by other people and myself, show that the level of odd order IMD
generated by a crystal ladder filter using a given quality of quartz
increases as the filter's bandwidth is decreased. This is surely one of the reason why the Hilberling uses a 50 kHz wide VHF roofing filter.This remind me
of the 50 years old debate about SSB generation and the choice between filter
or phasing generation.Filter generation was considered like a brute force
method, and phasing a soft one.Down conversion with ultra narrow filters is
somewhat brute force, up conversion with large bandwidth filters is a softer
method.Just my opinion ! Regards Georges F6DFZ
______________________________________________________________
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

[Mike Furrey]

> For a real down conversion radio, consider my old Hallicrafters
> SX-99. Everything is converted to a 455KHz IF which includes a
> crystal filter (I think one poll). It really doesn't do very
> well on image rejection. :-)
>
I had a pristine HQ150 (just finished restoration) and while listening to SW
broadcast in 15 MHz band I heard a ham station. I found him on 20 and told
him I heard his spur up there. He said he would check it out. While
continuing to listen in that SW band, I heard a few other hams in there. I
went Hmmm ... did the math ... I tried to find the other ham whom I told had
the spur ... he may still be looking for it.

73, Mike WA5POK

 

______________________________________________________________
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