EMRFD Message Archive 8818

Message Date From Subject
8818 2013-07-25 21:33:49 Bob L Receiver Project Choice: Superhet or DC?
For my next project I want to build a well performing 20 meter receiver for SSB.

I've never tried phasing DC, but the design appears to have some real advantages compared to a classic superhet.

But the designs make use of big poly capacitors. Those aren't common in my SMT junk box.

The bigger issue is one of age. Its common to use 1% and even .01% components in these designs. That's fine for today, but how will performance suffer as these components age?
8819 2013-07-26 07:31:11 kb1gmx Re: Receiver Project Choice: Superhet or DC?
8820 2013-07-26 07:49:37 William Carver Re: Receiver Project Choice: Superhet or DC?
Direct conversion "sounds good", because it doesn't have crystal filters
with their extreme group delay variations over the passband. It's much
simpler to build and OK for relatively casual "rag chew" kind of
operating. The audio circuitry does require good components and/or
careful adjustment to obtain rejection of the opposite sideband.

Superhets with good filters have better suppression of QRM from the
opposite sideband. So for DXing, contesting a superhet is probably
better. Building your own good xtal filters is possible but it is not a
casual process, and high quality crystals are very expensive.

W7AAZ
8821 2013-07-26 08:48:51 Ashhar Farhan Re: Receiver Project Choice: Superhet or DC?
Bob,
As Wes said, a good phasing rig is at least as complicated as a superhet.
So, here is the deal : if you want to stay safe, just plonk down a
superhet. I would heavily recommend the S7C from emrfd. It epitomizes
what a minimal superhet can be. It is an incredible performer too!
On the other hand, if you want to be a little more adventurous, you
can start with a binaural dc rx, add software of soft rock to get
single signal reception, and finally swap it with an analog processor
as designed by rick campbell.
- farhan VU2ESE

On 7/26/13, Bob L <bobledoux@proaxis.com> wrote:
> For my next project I want to build a well performing 20 meter receiver for
> SSB.
>
> I've never tried phasing DC, but the design appears to have some real
> advantages compared to a classic superhet.
>
> But the designs make use of big poly capacitors. Those aren't common in my
> SMT junk box.
>
> The bigger issue is one of age. Its common to use 1% and even .01%
> components in these designs. That's fine for today, but how will
> performance suffer as these components age?
>
>

--
Sent from my mobile device
8822 2013-07-27 10:20:26 Rick KK7B Re: Receiver Project Choice: Superhet or DC?
Good comments all. I design and build roughly equal numbers of both DC phasing and superhet receivers for HF. I've mostly published the phasing versions, as there has been less written about that architecture. But I don't select the architecture until sometime after I've made a number of initial sketches of any project.

The question was about a 20m receiver for SSB, so I'll limit my comments to that mode.

For a single band receiver for SSB, I'd almost certainly go with a phasing project. It is relatively easy to obtain SSB receiver performance that allows you to separate SSB signals on an amateur band as well as they can be separated. It is a very rare SSB transmitter that has adjacent channel IM products more than 40 dB below the desired output, and that's what you hear when you tune close to a SSB signal 40 dB stronger than the one you are trying to copy. More selectivity at the receiver doesn't really help.

For contest use, or in parts of the world where there are brutally strong SWBC stations near the band, it is somewhat easier to obtain good dynamic range with a superhet. The highest dynamic range receivers are generally single-conversion superhets. But dynamic range is like horsepower in an automobile. There are trade-offs, and more is not necessarily better, once you can cruise at highway speed on a crowded band. I've never needed more dynamic range than the basic diode ring mixer phasing architecture provides at the home QTH or portable. Some of my friends live in locations where more is sometimes desirable.

I personally enjoy receivers that sound better than commercial gear. It is easier to build those with the phasing architecture. Brick-wall SSB crystal filters with good rejection a few hundred Hz out of the passband affect signals in a way that can be unpleasant, but since most SSB transmitters generate signals that way, the damage has already been started at the transmitter. A good phasing receiver allows you to easily hear differences in transmitted signal quality on the air. A very high performance superhet that sounds good can be built by combining a carefully designed crystal filter with phasing techniques for the close-in opposite sideband rejection.

Regarding component aging. Some of my phasing receivers are now more than 20 years old, and two of them have been in continuous operation for that whole period. I've seen no effects of component aging. I expect it's a much bigger problem with modern digital ICs. Quartz crystals also age, and modern quartz crystals can be either quite expensive or of remarkably poor quality. I've had one narrow crystal filter age to the point where the passband was unacceptable for even a casual receiver. For my last serious superhet project, I spent about $60 on crystals for the filter.

Either way, it's a project, and not a simple one. For minimalist gear, it's hard to beat a simple DSB DC receiver, and I strongly encourage anyone to start with that if they've never experienced it. If you only care about opposite sideband rejection, it's much easier to build a simple superhet. But as so
8823 2013-07-27 12:02:12 Ashhar Farhan Re: Receiver Project Choice: Superhet or DC?
Rick,
There are two kinds of people in the world, those who have heard just the I
channel and those who have been on Q as well. Binaural is worth the extra
five dollars.
Remember the first time you heard JJ Cale (god rest his soul) on a stereo
headset? That!

- farhan (being quite opinionated tonight)


8824 2013-07-27 12:28:48 William Re: Receiver Project Choice: Superhet or DC?
Bob,

I have built both the uR2 phasing receiver and the Progressive Receiver (a superhet).

I am very satisfied with the uR2 receiver for 40 meters and actually surprised how well it performs. Both designs have their pluses and minuses. It is instructive to build both and experience the characteristics of each design.

Look in my photo section of this forum and take a look at both my projects.

Currently my uR2 is "done" only labeling the front panel is needed to be finished. This receiver evolved from a simple single board receiver to multiple "daughter boards" for other functions. I chose to build a separate VFO, add digital readout, separate audio stage, add an audio "S meter", electronic switching for QSK, antenna attenuator and a sidetone oscillator.

The PR receiver is on 80 meters and operational with plans to develop it further.

I would suggest picking your favorite band and build a single band receiver. Then once that is done add other features you like and maybe a converter for other bands.

Regards,

Bill N7EU
8825 2013-07-27 12:28:48 dale_k7eky Re: Receiver Project Choice: Superhet or DC?
It would seem that aging could be a problem over a period of time. Assuming that you are planning to use all pass phase shifters; one solution is to make them adjustable. By using a combination fixed and variable R element in each section of the audio phase shifters, you will be able to set each shift secti
8826 2013-07-27 12:28:48 Bob L Re: Receiver Project Choice: Superhet or DC?
You folks make a pretty good case for a superhet for my next move. I recall the EMRFD S7C receiver provided a spark for the BITX design.

Recently, I have have been reviewing the literature for both phasing DC and superhet designs.

Jim, K8IQY, has developed some simple performers starting with the 2N2xx and the more recent SS-40. These designs are only a bit more complex than the S7C.

His early designs included a two crystal roofing filter which is no longer present in the SS-40. I don't understand the need for a roofing filter in a simple receiver.

I'm leaning toward a version of the SS-40. I'll build my own crystal filter for SSB bandwidth and use a PTO for the VFO. I've had good luck with PTO's. I'll also added a frequency meter of my own design.

The SS-40 uses an SA612 for the product detector. This provides both a crystal controlled BFO and mixer function.

Are there reasons not to use the 612 for the product detector?
8828 2013-07-27 23:08:25 Rick KK7B Re: Receiver Project Choice: Superhet or DC?
Dale--it seems that you missed the part where all of us who have actually built large numbers of phasing receivers using all pass networks with modern components have not experienced any issues with aging.

Furthermore, see the appropriate plots in chapter 9 of phase error versus component tolerance. Adjustment does not buy you anything. Period. The beauty of the phasing receiver is that you can design and build a network using either analog or digital methods that is stable over time and provides more opposite sideband suppression than the rest of the receiver can support. You then need only two adjustments: an amplitude trim to balance the signal paths, and a phase trim to set the offset between the two channels to 90.0 degrees.

Sorry for the curt response, but adjustment of phase shift networks is really not a good idea, in my opinion. There are way too many variables, and they all interact. Once you start tweaking, you can never get back to the performance of the original network.

Best Regards,

Rick KK7B

8832 2013-07-28 08:12:08 blumu Re: Receiver Project Choice: Superhet or DC?
----- Original Message ----- From: Bob L

You folks make a pretty good case for a superhet for my next move.
. . .
The SS-40 uses an SA612 for the product detector.
This provides both a crystal controlled BFO and mixer function.

Are there reasons not to use the 612 for the product detector?

----------------------------------------------

Fine for an undemanding superhet. Easy to implement,
but the IM characteristics are not the best currently available.
For better performance, I would consider a better-performing
mixer, such as the TFM-2.

Michael
UK














------------------------------------

Yahoo! Groups Links
8834 2013-07-28 12:22:43 dale_k7eky Re: Receiver Project Choice: Superhet or DC?
Rick,
I have only a little experience in building phasing receivers, but have really enjoyed the sound that I hear.

My comment was off the mark considering the aging of components since I have experienced little to no age related problems with RC circuits in other phase sensitive applications.

My goal in using adjustable R at each stage is to adjust the RC product to the same as that called for in the original design, after measuring the C. That way I don't have to find precisi
8835 2013-07-28 12:49:25 Rick KK7B Re: Receiver Project Choice: Superhet or DC?
Hi Dale,

Your comments were most welcome, as they allowed me to hop up on one of my soap boxes, hi. That's why I included the apology for the curt response. In any discussion forum there will be active differences of opinion--Bob, Wes and I openly discuss ours.

1% resistors and even 0.1% are so inexpensive and available these days that we don't need to tweak those. For capacitors, if you have the ability to measure them, then the usual plan is to buy a small quantity and measure them into lots of 6 that are closely matched--I think 100 5% poly capacitors is less expensive than 6 trim pots, even in the surplus stores.

But your method and your thinking will definitely work, and is a viable option if you'd rather measure time constants than capacitor values, or if your junk box is well stocked with trim pots.

Before I worked on high performance direct conversion receivers, I focused on no-tune microwave gear, and the goal for both is to reduce the number of adjustments to a minimum. That is one of my own design quirks. If I can design something ahead of time that is stable and tolerates a statistical spread in component values or circuit board dielectric, I think that's really cool.

Thanks much for opening this discussion.

Best Regards,

Rick

8837 2013-07-28 14:51:53 Bob L Re: Receiver Project Choice: Superhet or DC?
Let us suppose we use variable components or series resistors and parallel capacitors to achieve the 0.1% some desire. To check
8838 2013-07-28 15:03:26 kb1gmx Re: Receiver Project Choice: Superhet or DC?
I'll be brief,

For an all pass network variable resistors should be held to
minimum and of the best quality for best long term results,
or better yet not used at all. Even if they don't age they
can vibrate.

The aging thing is a red herring.

That and dialing it in is going to be no real added value
unless you enjoy pain.

The misconception is that more than 40db of opposing sideband rejection is required. Sorry, its not, unless you have a
filter SSB rig with an 120db if gain following it then you need a filter with 140db of stopband and a very good AGC to use it.
Simple hand made filters with less than 8 crystals simply are
not up to that task. Good filters take a lot of work, the many
I've done took at least as much work as the entire rest of the
radio. The common 4 crystal ladder filter is barely good enough
for 40db of opposite sideband rejection.

The concept of selectivity is all messed up. We need rejection
of the guy 20khz up but we can't because he is also pushing the
linear and his buckshot is in our passband where its too late to reject it. Doesn't make a difference how good the RX is, you will
hear their trash.

When your doing simple radios, phasing can work better. But to get
to the next level on either case we are no longer talking about
simple radios.

Allison



8839 2013-07-28 15:42:16 kb1gmx Re: Receiver Project Choice: Superhet or DC?
8840 2013-07-28 15:59:33 William Carver Re: Receiver Project Choice: Superhet or DC?
The saying "There's more than one way to skin a cat" applies to audio
quadrature networks. I've build phasing SSB generators with well-matched
capacitors and precision resistors, and I have also tuned up individual
phase shift network sections with pots.

It's elegant to compute values which, within a given statistical
probability, will deliver "X" opposite sideband rejection as Rick
described. When you're designing for production there's a big emphasis
on minimizing adjustments! In its own way it's also elegant and a
learning process to realize how each stage of the multistage phase shift
network performs its "magic", and adjust each RC product at the
frequency and watch the opposite sideband rejection be perfect.

Jim Tonne (W4ENE) of Tonne Software, creator of "Elsie", has a free
Windows program that will design audio phase shift networks that use op
amps and RC networks to develop two outputs in quadrature. You can
download Jim's QuadNet program at:

tonnesoftware.com/quaddownload.html‎

You can choose the range of frequencies you want to be in quadrature and
how many op amps you want to use. A wider range of audio frequencies
results in poorer worst-case quadrature, more stages produces better
quadrature.

QuadNet tells you the audio frequency for each op amp stage which
produces exact quadrature. So you can apply that frequency and adjust
the corresponding pot for for maximum sideband rejection. There is small
interaction, but it doesn't take long to tune up a 4-stage phase shift
network, for example.

I used nice junkbox capacitors close to the necessary values and
computed the resistor that would produce the same R*C product as Jim's
design software specified. I chose a smaller 1% resistor and put it in
series with a junkbox pot so the sum could be adjusted lower and higher
than the computed value of resistance. I applied an audio signal at the
frequency the software said should produce the deepest null in the
opposite sideband amplitude, and adjusted the pot for that null.

W7AAZ
8845 2013-07-28 20:22:52 Jim Re: Receiver Project Choice: Superhet or DC?
If your budget allows, get an LC2 capacitance/impedance meter, and then buy your phasing network devices in bulk.  For example, if you need 10 each, get 20 each and then bin-sort the 10 best-matched parts.

You'll find that if you get devices from the same batch, their new values are all well within their advertised tolerance, and their aging characteristics will all be similar.  So you should be set for your lifetime.

And varying from the design values by 1 or 2% really won't have much effect on the finished project as long as you've got good matches to start with.

Finally, all Jim Tonne's software is wonderful.  Send him money if you can.

73
Jim N6OTQ



>________________________________
>
8847 2013-07-28 20:57:22 Ashhar Farhan Re: Receiver Project Choice: Superhet or DC?
just a thought,

In an active networks like those of campbell configuration, the two crucial
matches are for the 0.01 capacitors (all of the same value) and the
feedback resistors of odd values. I have easily matched the capacitors by
plugging them into a test oscillator and measuring the frequency variance.
i usually have a box of a hundred 0.1 and 0.01 capacitors to choose from on
the bench. so, getting six of them within very tight tolerance is not too
difficult. the 10k resistors too can be chosen from a large pack of 100
using a digital VOM. You can keep the rest for your junkbox.

for the odd valued resistors, you could add a small value preset to a fixed
value resistor, and set it to the desired resistive value by measuring it
on the VOM meter. you must do this with without the op-amps plugged into
the socket. once set, you should leave the preset untouched. this is
different from any 'alignment', it is just an easier way to get to the
precise resistive value in the network.

as donald knuth famously said 'i have only proved it correct. i have not
tried it'. this is my pure conjecture. this needs to be tested on the bench.

- f


8858 2013-07-30 19:40:56 kb1gmx Re: Receiver Project Choice: Superhet or DC?