EMRFD Message Archive 13946

Message Date From Subject
13946 2017-05-31 17:00:07 xb6zmeyanbsxhnbt2... KK7B kits

Kanga is gone... any more sources of kits for the KK7B designs?


-Tom N1OOQ

13947 2017-05-31 19:29:08 wb8yyy_curt Re: KK7B kits
Check out XFD design, but I know nothing of them.  plan b is to forgo the pc board and source your own parts for a ugly construction or other style build. 

Curt
13953 2017-06-02 16:50:44 cwfingertalker Re: KK7B kits
Kanga had a long good ride.  Maybe it is time for another kit builder/provider to pick up the baton.  Losing the KK7B kits would be a loss.  The uR2 reciever is head and shoulders above any regenerative or direct conversion receiver kit available out there.  I have built two and they really are excellent receivers.  I hope to see another kit builder/provider pick up the KK7B kits and make a good run with it.

Bill N7EU

13954 2017-06-03 00:10:51 lmarion Re: KK7B kits
After the super high performance  of the R2 Pro, R2 and miniR2,  I found the uR2  to be a flop. I was never able to get even close to the claimed sensitivity, side band rejection was pathetic, poor stabilty, limited coverage, etc. So I think that was the end of the line for good reason.
 
Leroy AB7CE
 
 
 
13955 2017-06-03 09:29:29 adwsail@gmail.com Re: KK7B kits
I tried on at least 2 occasions to get an R2pro over the last year and never could get a response. So sad. I have a Kicad layout for the ASP for the R2mini. I guess it's time to order some boards and build my own. 

Dale Ward
WB4LIP
13956 2017-06-03 19:24:28 kb1gmx Re: KK7B kits
I've built two uR2, they both exceeded the stated performance.  However, I've helped more than a few other that managed to so something wrong with similar complaints.

Both of mine check out at 37-38DB sideband rejection. Stability is like any other VFO built right they are decent build loose they drift.

Limited coverage is not uncommon if the hybrid is messed up but also the RF amp has to be tuned and there is no provision for wide tuning.    Can't fix a limitation designed in.   

The fix is a computer, SI570 and a simple front end... but that will never draw less than 80ma at 12V.

Then again I modded one for 36mhz as a high if (crystal controlled rather than VFO) and use tuneable upconverters and the end result was quite good and very frugal on power.

Its unfortunate that kanga is going away but time has a way to changing.


Allison 
13957 2017-06-03 20:44:53 AD7ZU Re: KK7B kits
The fix is a computer, SI570 and a simple front end...

another idea:  use  SI5351 which has 3 outputs.  setting the phase control register on one channel to provide a quadrature output without the use of a divider or the need to run at 4x to derive the quadrature LO.   I realize the 5351 part has a wider jitter spec and the duty cycle will be looser than the SI570 since the 4x quadrature  edges are aligned on SI570 rising edges but by eliminating  pcb runs at 4x  frequency and 2 dflops likely the results are  equivalent?   the remaining channel can be selectively enabled and tuned to an image frequency for alignment.  once phase/amplitude are trimmed disable the 3rd output.  fewer parts,  lower power consumption,  built in alignment signal source, and lower cost.

Randy
AD7ZU



13958 2017-06-03 22:32:30 Bill Carver Re: KK7B kits
Doesn't software fine tune the phase already?
I'll take the lower phase noise of the Si570 when it's a receiver LO.
W7AAZ


13959 2017-06-03 22:54:26 Ashhar Farhan Re: KK7B kits
Bill, Randy, Allison,

Hans G0UPL has already written the code to generate the phase output from the Si5351. The code isn't public but his kit supports it. It is done exactly how randy described it.

Now, about the Si570. The jitter and hence the phase noise is about 20 db better than Si5351. This is out of the box. It can probably be improved. Given that the Si5351 is mostly the very Si570 with an external reference oscillator, i have tried it with low noise external oscillator instead of the built-in one. the phase noise improved substantially. i am not sure how well. I need to build a second notch filter to measure it.

The trouble with Si570 is not with the Si570, it is the flip-flops that you will use to generate the phase. The jitter noise of the FFs worsens the phase noise. A simpler option would be to use the hybrid coupler. The hybrid needs a clean, low harmonic drive - it doesnt do quadrature outputs at the harmonics. There are some better but more expensive parts like the very high speed SY10EP52V flip-flops that can do 2 ghz. 

A half done project in the home lab is to do an upconverting phasing receiver. This is a VHF phasing receiver running at a fixed frequency of 70 MHz, the front-end is a passive upconverter. The phasing is easily trimmed to be excellent at one RF frequency. The images, spurs are all taken care of by the high IF. An Si570 provides the clean oscillator for the the first conversion. I will probably take it up once kids' summer holidays end.

- f

13960 2017-06-04 07:15:54 kb1gmx Re: KK7B kits
I've not had issues using counters (D-FFs) for creating quadrature and the noise for 
that should be lower.  The problems with that is the high LO needed (4x) and board 
layout is critical as long lines will ring at those rise and fall times creating havoc.
Same for line lengths (must be equal for both I and Q paths).  Multi-layer boards 
help with this as then controlled impedance is easily done.  

I've been running a high IF phasing RX for a decade.  I call it my 6M crunch proof TRX. 

The idea works well and using crystal for the IF is easy and saves running a bunch of chips.

For the upconverter I used a Si570 which worked well as the IF in the current iteration is
36mhz so lo+36mhz works well.

Not magic only controlling all of the various signals and waveforms.

Allison


13961 2017-06-04 10:41:14 lmarion Re: KK7B kits
Yeah,  I heard this from you when I was building two of them and where crappy and you said all of this back  then.
 
I am also a NIST electronics lab tech(retired) building for 50 of my 62 years, and I know crappy performance when I use it and measure it in the lab.
 
I gave a unbuilt kit to another lab engineer ham. He also got less than 20db sideband rejection, NIST lab measured.
 
  I built all of the converters, and they met or exceeded expected performance. I still use the R2 pro in my astronomy  receiver and the mini R2 in one of my
 
QRP rigs. 
                    The uR2s are in the junk box under the bench.
 
 
 
Leroy AB7CE
 
 
13962 2017-06-04 10:43:33 Dale Ward Re: KK7B kits
I'm still learning and building so what is the audible end result of relatively poor phase noise performance in a direct conversion receiver? Is there enough difference between the Si570 and the Si5351 to even be audible? I have used both in various circuits with , imho, very good results so I have picked one over the other for reasons other than phase noise. 

Currently I'm using a '51 as a nano controlled vfo throttled back to a 2ma ad coupled output feeding an SBL1 directly, no filters, pads or squaring circuits, in a 40M CW DC  xcvr with what seems to be very good results. I'm also using a '51 in an SNAJr II that I've modified to use a color oled. Would a 570 give 'better' results in the SNAJr? The audiophile community has all manner of techniques for super quiet power supplies and exotic frozen cables which they claim to have a large effect on the sound quality from a system. Having built 2 high end class A amps I found caps and transformers do matter, the rest is hogwash. So my initial question re the actual audible difference comes from that background. I have no doubt that phase noise matters mathematically but what is the sonic result?
13963 2017-06-04 10:44:25 Leroy Marion Re: KK7B kits
What modes are you running on 6 meters?





I've been running a high IF phasing RX for a decade. I call it my 6M crunch proof TRX.

Allison









[Non-text portions of this message have been removed]
13964 2017-06-04 11:52:49 AD7ZU Re: KK7B kits

Hans' results a few weeks back were very encouraging... enough so i going to try a 5351 in my current project.

i am just now attempting to measure the SI5351 quadrature output. all i have is my new Rigol scope (happy bday!) and an arduino to control the SI5351.. so measurements may be limited .. going to start at 14Mhz

the issue i am looking at is to determine how the duty cycle varies over frequency and between channels.   It could be the duty cycle varies by the same amount between channels at any given frequency.   if that is the case then its a don't care.

The same may be true for jitter?? 
if both channels of the 5351 exhibit the same jitter offset  then the quadrature signal would remain accurate... i think??
there is the possibility that offsetting the phase in one channel either exacerbates the difference or cancels the difference? or has no impact?

The final question i have with the 5351 is if its driven with a good quality XO  in lieu of the internal oscillator and an external xtal would performance improve significantly? 

it would be nice to eliminate the Dflops.
I suspect the Dflop switching times are not symmetrical between states.. that in combination with  layout issues as Allison mentions may be at the root of the improvement reported by Hans

grrrr...all this stuff always just raises more questions ..
 
Randy
AD7ZU



13965 2017-06-04 12:12:32 Bill Carver Re: KK7B kits
Hi Dale,

Think of an LO with phase noise as a big strong central sine wave, surrounded on each side by other, weaker sine waves. Getting weaker as their distance from the central sine wave gets larger. Each one of those will also be mixing with all the signals coming from the antenna.

When you are listening to a desired CW signal by positioning your LO 600 Hz away from it, you get a 600 Hz tone. If there is another signal 1.6 KHz away you'll get a 1600 Hz beat note which your audio filter will attenuate it, maybe even eliminate. But there's also one of these weaker phase noise sine waves, 1 KHz away from the central one, and it will also mix with all the signals coming from the antenna including that one 1.6 KHz away: except it's only 600 Hz away, so it will make a (weaker) 600 Hz tone.

There will also be weak phase-noise components 601 Hz away, 610 Hz (etc, etc) away from the central sine wave. Each one of them mixing with that strong off-frequency signal. Converting to your 600 Hz filter. If that off-frequency signal is strong you will probably hear  "fuzzy" CW, colored by your audio filter at it's center frequency.  In some cases you can copy that "fuzzy" CW.  I don't know if that rings a bell with you. When I was a novice 60 years ago my Hallicrafters S-85 I could hear "noise CQs".....from  guys who were many Khz away. Happily, things have gotten better since then.

Anyway, when your phase noise is 20 dB worse then those artifacts, the "fuzzy-CQ", will be 20 dB worse. Depending on band activity it could just raise the noise floor of the receiver a bit when you're not tuned to a signal. It's not the end of the world. But there is a performance difference between Si570 and Si5351. We get to choose.

Bill - W7AAZ



13966 2017-06-04 15:37:07 aa0zz Re: KK7B kits
>I'll take the lower phase noise of the Si570 when it's a receiver LO.
>W7AAZ

I'm with you! The Si570's 20dB better phase noise alone makes it an easy decision for me.
Regarding software using the Si5351 phase field to generate quadrature on one of the channels - there are only 7 bits of resolution in the field so I don't think this is enough accuracy to do the job properly.  
The field is CLKx_PHOFF[6:0]
I haven't measure it yet. 
Adding 20dB of phase noise plus inaccurate quadrature sounds like too much of a compromise to save a couple of dollars and a few milliamps of power.
73,
-Craig, AA0ZZ

13967 2017-06-04 15:57:09 Dave Re: KK7B kits
  Most generate I and Q in a divide by 4 circuit.  But, it is very easy to use a divide by 2 circuit if you have the LO and its complement /LO.  All available interfaces for the Si570 EXCEPT CMOS do that.  It is also possible to use a single output and an XOR to generate the signal and its complement.  There could be a skew between the 2 outputs depending on the difference in propagation between the 2 XOR gates.  Since they are on the same chip, this difference is very small.  I used this scheme on my AD995X board. 
   In my UHFSDR, I use a pair of the SY10EP52  ECL dividers and the LVPECL version of the Si570.  They have a max clock freq spec of 4 GHZ minimum!  On 430 MHz(860 MHz clock) I saw around 40 dB opposite suppression BEFORE software correction.

Dave - WB6DHW


13968 2017-06-04 21:04:17 AD7ZU Re: KK7B kits
Regarding software using the Si5351 phase field to generate quadrature on one of the channels - there are only 7 bits of resolution in the field so I don't think this is enough accuracy to do the job properly.  

i didn't think that was sufficient either until reading the application note. 
then reading AN619 .. (more obscure SI documentation) i see that the phase offset resolution is:  vco/4  .. so a value of 1 in that reg should offset the input to the dividers 90 degrees and should track for all outputs on the offset channel.

im ordering  parts this week to measure jitter and duty cycle in the quadrature configuration.  i  really want to try the 8 channel version with a good tcxo driving it in lieu of using the internal osc with an external xtal.  Even the high grade SI570 is not  even close to a 50% duty cycle.  Using  4f and dividers  skirts that problem by only switching on rising edges.  All 2f arrangements will have to switch on both edges to derive a quadrature output and in all cases will be asymmetric.  curious to see what the impact of the phase offset will be on the duty cycle.   The duty cycle  issue  may be the limitation to the 5351 stand alone.


Randy
AD7ZU





13971 2017-06-05 07:28:12 aa0zz Re: KK7B kits
>>Regarding software using the Si5351 phase field to generate quadrature on one of the channels - there are >>only 7 bits of resolution in the field so I don't think this is enough accuracy to do the job properly.  

>i didn't think that was sufficient either until reading the application note. 
>then reading AN619 .. (more obscure SI documentation) i see that the phase offset resolution is:  vco/4  .. so a >value of 1 in that reg should offset the input to the dividers 90 degrees and should track for all outputs on the >offset channel.

Wow - that's very interesting.  I didn't see that in the spec and didn't try the programming.  I'll give it a whirl, just for curiosity.
Thanks,
-Craig, AA0ZZ
13976 2017-06-05 15:23:09 Fernando Krouwel Re: KK7B kits
Hi folks, good evening:
I don't know if this is something new, but there is a Si5351 kit been sold, which they say can generate quadrature outputs. See in the features, 6th line, link is:

http://qrp-labs.com/vfo.html

73's from
Fernando - PY2FZU


13996 2017-06-12 09:46:04 Leroy Marion Re: KK7B kits
Hi Dale, All of KK7B kit circuitry is now available from several kit providers. KK7B was a while back now, so vast improvements and advancements Have been made. It was a great educational experience. Modern radios, such as several of My Elecraft rigs us this type of circuitry to great advantage. My Elecraft KX2 Is 100% SDR radio, for instance. So if you are looking for an educational or fully functioning radio experience, There are several higher level options available. 73 Leroy AB7CE
13998 2017-06-12 11:03:20 Ashhar Farhan Re: KK7B kits
leroy,

it would make rick ill to call the sdr radios as 'vast improvements and advancements' over the analog, hardware defined radios that he has designed and build. 
the sdr radios and rick's phasing rigs have just this much common : both of them directly convert to baseband. nothing else is common. i could do an entire table of the differences, but the essence of difference between the two approaches is that rick's designs are artisanal, analog. probably the very epitome of what the analog radios can be when you work from the first principles.
sdr are a brute force method of achieving similar design gains. it is a little like saying the CD is the improved vinyl.

- f

13999 2017-06-12 11:46:03 Dana Myers Re: KK7B kits
14000 2017-06-12 18:16:27 AD7ZU Re: KK7B kits
sdr are a brute force method of achieving similar design gains. it is a little like saying the CD is the improved vinyl.

one only has to count the 10turn pots in a design to realize that sdr is far from a brute force approach
sdr can be a 10th row center seat at the Musikverein compared to worn scratchy vinyl


Randy
AD7ZU

14001 2017-06-12 21:58:29 Ashhar Farhan Re: KK7B kits
my analogy didnt make it, a vinyl can have a perfect recording of a signal, you dynamic range is literally limited to the size of vinyl molecules and the dust times in your room. 
a cd on the other hand, need compression to sound the way it does. there is an early limit to how good it will sound because the limit is designed in. the bits in the adc is your sky.

an HDR, as opposed to an SDR, has an unquantized signal that the ear can make out easily. one must experience this to believe it. i am not claiming that the signals noise floor is any lower etc. just that it sounds better.

a month ago, i stood in front of dali's persistance of memory at the MoMA. as i kept looking into it, i also realized that it was subtley different from the thousands of pictures of it. the colour play, even the size, how the illumination in it play, everything ws different, some parts if the canvas has thicker paint that the others.. a week later, i was tuning around an kx2, and i couldnt help reminding myself of the other painting in the same room of MoMA - rene magritte's painting of a pipe that said - this is not a pipe. it is a representation of a pipe, not a real pipe.

the same goes for an analog phasing receiver. it has a natural, un-digitalized quality that comes a very well engineered, flat and the shortest path to your ear. this is also why the drake TR4 is still the best contest receiver according to sherwood.

- f

14002 2017-06-12 22:49:22 Dana Myers Re: KK7B kits
14003 2017-06-13 06:18:00 John Levreault Re: KK7B kits
14004 2017-06-13 07:49:23 Dana Myers RIAA equalization (was Re: [emrfd] Re: KK7B kits)
14005 2017-06-13 08:15:46 Alberto I2PHD Re: KK7B kits

the same goes for an analog phasing receiver. it has a natural, un-digitalized quality that comes a very well engineered, flat and the shortest path to your ear.
This reminds me of the speaker cables the self-appointed golden ears of the HiFi world buy for prices that very often exceed 100 USD per meter...
Cables made with copper that has oriented molecules, oxygen free, whose two ends are not swappable...  :-)

--
73 Alberto I2PHD
Credo Ut Intelligam


14006 2017-06-13 09:00:05 Ashhar Farhan Re: KK7B kits
here is the link on sherwood about the drake :

quantification of 'goodness' goes only so far. it is also about what you want to quantify. in a contest, or a dx pile up, the important criterion is the in-band distortion. if the guy down the block is calling the same station with his 100 watts. can you still hear the dx over and above that din?
now, there are two ways to do it:

1. use extensive digital processing to clean up the signal and isolate the dx from the rest.
2. allow all the signals to linearly process with high dynamic range, analog circuitry and allow your ears to figure things out.

what is unnoticed in this entire discussion is the higher oscillator phase noise of the digital system as opposed to the very low phases noise of a well designed free running VFO.

although completely subjective, allowing for natural sound without the digital clutter and ringing is a far more pleasing and easy-on-the-ear experience. if you have been following linrad's evolution, you will realize that Leif's main thrust was a better hard receiver for his EME work. 

i am not dissing DSP, i am merely suggesting that as amateurs, who can craft their radios, analog route offers a far more pleasing reception that going digital does. surely, analog radios are more difficult to build. noise blankers, filters, notchers, can all be done in analog circuitry, offering quite a rewarding receiver experience. you can program a receiver either way. digital processing cannot retrieve a bad receiver front-end. nor can a poorly designed VFO always beat an Si570.


- f


 

14008 2017-06-13 10:30:19 John Levreault Re: RIAA equalization (was Re: [emrfd] Re: KK7B kits)
Sorry, Dana, but to suggest that there is "vanishingly little bass" is
misleading. You're right that when a tape goes to disc, the bass is
pre-equalized to a relatively low level. But this happens whether the source is
tape or a live feed. This is to prevent over-excursion of the cutter head. At
the same time, the treble is boosted greatly to improve the SNR. Implementing
the RIAA EQ curve on playback, which boosts the bass and attenuates the treble,
flattens everything out. Or it should.

Am I missing something in what you're saying?

John NB1I


14009 2017-06-13 10:57:51 Dana Myers Re: RIAA equalization (was Re: [emrfd] Re: KK7B kits)
14010 2017-06-13 11:02:23 Dana Myers Re: KK7B kits
14013 2017-06-13 15:22:35 AD7ZU Re: KK7B kits
Interesting thread..
Having built 2 receivers/transceivers and 1 kit (BITx, thanks Farhan!) since getting re-licensed a few years back though the homebrew designs were minimal compared to commercial offerings.
Both hb designs are phasing architecture, one analog, one DSP.
Each has a different character: good points and faults (that are always going to be addressed in the next version). Can’t say I prefer one over the other.   Now there is a new DSP transceiver project in work now using an Analog Devices DSP enabled codec for all mod / demod,  sb selection, filtering, agc, audio, a switch based sampling detector / exciter, and an 8 channel SI5351 VFO directly clocking everything (RX, TX, codec).
 
IC technology advancements over even the last 10-15 years is astounding and rapidly enhancing the performance with ever fewer components.  The downside is that this makes for (lazy?) engineers replacing entire analog systems that previously required skilled design with single off the shelf ICs costing only a few $..
 

Randy
AD7ZU



14014 2017-06-13 15:48:22 Dana Myers Re: KK7B kits
14015 2017-06-13 19:38:10 Andy Re: RIAA equalization (was Re: [emrfd] Re: KK7B kits)
Saying that vinyl has vanishingly little bass is akin to saying that a CD has horrendous harmonic distortion because it can only record clipped (square) waves, not even sine waves.

Andy


14016 2017-06-13 19:57:15 Dana Myers Re: RIAA equalization (was Re: [emrfd] Re: KK7B kits)
14018 2017-06-14 08:21:05 Bob Re: KK7B kits
I believe the two streams (I and Q) can be output from an SDR dongle using gnuradio. 

Is this what is special about the radio under discussion?

73
Bob
WA2I
14019 2017-06-14 14:03:15 Andy Re: RIAA equalization (was Re: [emrfd] Re: KK7B kits)
Dana wrote:

"I'm really just trying to say: the signal recorded onto the actual vinyl is not a "pure"
or "natural" representation of the audio. It's processed to work around the limitations
of vinyl recording."

True ... but irrelevant.

Andy


14020 2017-06-14 17:15:43 Jim Strohm Re: KK7B kits
Actually, what's really special is that Silicon Labs designs their chips only a few dozens of metres away from the famous Congress Avenue Bridge in Austin TX, from whence hundreds of thousands of Mexican freetail bats exit every evening during the summer.

It's better than a sunspot maximum -- but only if you like watching bats.

I could care less than zero about the bats -- but if Austin commuter traffic weren't so abhorrently bad, I'd offer up my considerable tech writer skills to them for a penance.  Working for them is a job I'd nearly do for free, except for traffic.

73
Jim N6OTQ

PS -- I need those bats to come my way to eat my millions of mosquitoes, except that their commute is just as bad as my commute.


14308 2017-09-15 20:57:36 Bill Duns ore Re: KK7B kits
I am in the market for a MicroR2 kit, but Bill Kelsey has closed down Kanga. I want to use it for a 9 MHz CW tail end to a superhet with a 4 pole crystal filter. Any kits available out there?

Sent from my iPad