2018 ~ The Year of SSB Transceivers

Why Build Another Transceiver?

10/16/2018 ~ See the latest Video on the Heathkit SSB Transceiver


10/10/2018 ~ Breaking the Guinness Book of Records.

1014/2018 ~ The Heathkit Rig is on the Air

Running 100 MW with just the Driver Stage the Heathkit Rig is doing WSPR. Just installed a Refined Driver Stage with the 2N2219 -- a healthy 100 MW. The Transistor with the heatsink in the lower left corner is the 2N2219






This is the WSPR Data at 100 MW



Get off the couch and start soldering!

73's
Pete N6QW



The SMD board above is the Plessey Bilateral Amplifier in surface mount. This is the work of Nick, G8INE who recently acquired a CNC machine. You can see the small size as compared to the ADE-1 DBM. Nick has done a superb job --and he reports it has been proofed and is a working board!

If you go to my website www.n6qw.com under the Sudden Transceiver link there is a link to G8INE and he has offered to provide the Gerber files for this really small board. Thanks Nick!

Pete N6QW

What are you thinking -- I am not trying to break any world record? My XYL asked me that question today -- why are you building another rig? Followed up by a snide comment that I had so many rigs now why do I need another one. Well the answer plain and simple because I can! 




For the longest time in the late 60's early 70's my success rate with homebrew SSB transceivers was miserable. At that time I lacked the more sophisticated test gear and let's face it some of the technology wasn't that great. Crappy Analog VFO's were high on the list of impediments! I also had to work and to give a fair share of my time to the family -- it is that balance thing.

But today that is all changed --better test gear, better technology like Digital VFO's and a bit more time. The latest project is to demonstrate that some of the components out of boat anchors can indeed be reworked to provide a very modern, very capable rig. I looked through eBay and have seen many crystal filters from heathkit, yaesu, icom, kenwood etc that can be had at very reasonable prices. A good friend just picked up a heathkit filter like the one in my latest rig for less than $15 including shipping.

Previously a big problem was having only the filter did nothing for you without also having the BFO crystals. With the Si5351 -- that no longer is the problem as you can have any BFO frequency between 8 kHz and 220 MHz-- that should cover a lot of filters.

The other factor is that with the currently popular bilateral or as I have demonstrated in the Sudden Transceiver, the single pass with relay switching the total amount of components needed can be had for very little money. Imagine a whole bilateral amplifier strip complete with filter for something in the $20 to $25 range. A digital VFO with BFO and color TFT is another $25 -- a homebrew complete rig for about $100 is a reality. Even less if you have a big junk box. You are only limited by your imagination.

BTW the Digital VFO includes two independent VFO's, USB/LSB select and a tune function. 

Stay tuned for more details on the heathkit SSB rig from the N6QW Laboratories.



73's

Pete N6QW


10/09/2018

The Heathkit 40M SSB Transceiver (N6QW Version) Hears Well!


The N6QW SSB Transceiver Complement:

Receive Section: (Currently working)

  • 2 X 2N3904 (Two SMD)
  • 2 X 2N3906 (Two SMD)
  • 2N3904 Rx RF Amp (Future AG303-86G)
  • 2 X ADE-1 (DBM)
  • 3.395 MHz Heathkit Surplus Filter
  • NE5534 Audio Pre-Amplifier
  • LM 380 Final Audio Amp
  • Arduino Nano
  • Si5351
  • 160X128 Color TFT
  • Band Pass Filter

Transmit Section: (Under Construction)



  • Transmit Pre-Driver, AG303-86G (MMIC)
  • Transmit Driver, 2N2219
  • Transmit Final, Mitsubishi RDHF RF FET
  • Low Pass Filter



This transceiver is on par with the Sudden Transceiver and yes Virginia you can even find some of the circuits in EMRFD. The 2N3904/2N3906 bilateral amplifier circuit (from Plessey) as used in this rig came from there. So it does have street creds for those who think that they will only look at circuits coming from that publication. 



73's

Pete N6QW

10/08/2019 ~ Columbus Day. Chris discovered America (or one of the ones that did) and now you can discover how to use old boat anchor filters in your homebrew SSB Rig!

You can even do WSPR with the new Heathkit Rig. This is receive only using the N6QW Digital Adapter. 



New transceiver  Heathkit 3.395 MHz crystal filter with two Plessey 2N3904/2N3906 bilateral amps. NE5534/LM380 Audio Amp, 2N3904 Mic Amp, 2XJ310 Relay Switched Rx RF/Amp & Tx RF ( a 2N3904 Rx RF for this video). 2N2219 Tx Driver, Mitsubishi RDHF FET for the Final and 2 X ADE-1's for Rx/Tx Mixer and PD/BM/ Arduino Nano, Si5351 and Color TFT.

Some thoughts on switching the BPF's and LPF for 4 Band Operation. On several of my rigs that were multiband and used a common buss for the filters I found bleed through problems and difficulty with drive through the BPF at higher frequencies. With six relays this give 4 bands with either the BPF or LPF and provides better signal isolation over that with a common buss. It will take 12 relays ( @ 60 cents a piece) for four bands and switching both the BPF and LPF. Bottom Line: only one set of BPF's and LPF's connected in line at any one time!



73's
Pete N6QW

Comments

  1. Pete, I'm not wrong. just not as right as I want to be...
    But I did signed up to GQRP and am now member #15909 so I can have a front row seat as you unveil your latest "Knack-fest".
    I'm on "French leave" from the bench as I deal with some Real Life Issues. I did get a new AD9850 module for the LBS part 1 nd I need to take some more measurements. Hopefully my Summer project doesn't turn into the Winter Project! HIHI
    73
    Ralph AB1OP

    ReplyDelete
  2. Hang in there Ralph. The next issue of SPRAT will be in the mail on September 18th. My website will contain additional information about this project and that too will be lit off on the 19th.

    http://www.n6qw.com


    One you see the full project you will see that it is not complex and easily replicated. Just a few more days. Thanks for joining the GQRP Club.

    73's
    Pete N6QW

    ReplyDelete
    Replies
    1. Thank you for the reply and encouragement. I received the 2018 Spring and Summer Issues so I saw your article on using the GQRP sales parts to mod or upgrade one of your previous projects. Quite inspiring.73 Ralph AB1OP

      Delete
  3. Pete, my friends Simon, Ben and I were only discussing just last night the idea of doing a CW transceiver based around the Sudden on a 10x10cm PCB for use as a SOTA radio. And here you have already done it and made it digital. Awesome work.

    ReplyDelete
  4. 'Hi Rob, (and friends)

    I am told that the latest SPRAT should be in the "Royal Mail" as of yesterday so depending where you are in the world you could have your copy today or as in my case on the Left Coast of the USA (currently enjoying the title Trumpnation) it won't be until a week or so from now. The Digital interface has been tried with a half dozen or so homebrew transceivers and wsjt-x software. Flawless is a really good term.

    The nut I had to crack was the switch over from the Serial Port to USB as many modern computers lack a Serial Port. Good thing I grew up in an era of transistor switches and relays as that is what I understand and what parts I had in the junk box. The more modern ham would have done it with PFET's and a bit of magic. Concurrently with SPRAT release I will be providing additional resource information via my website at http://www.n6qw.com.

    Oh I have used several different computers including an old Windows 7 Netbook (current configuration). I have loaded wsjt-x onto a Raspberry Pi2 but sadly have only been able to make it play with an earlier version of wsjtx which skips WSPR and FT8. I have seen some videos and tutorials on how to get the wsjt-x 1.9.1 to work with the Pi but defiantly my Pi2 says "no way Jose". So maybe your group can crack that nut as that would make for a really portable rig.


    73's
    Pete N6QW

    ReplyDelete
  5. Hey Pete, great work mate it looks like you're a bit of a celebrity over here in the UK :)
    I am a GQRP club member and have read your article, keep up the great work we need more of this.

    Cheers

    Rob.
    2E1IIP

    ReplyDelete
    Replies
    1. Hi Rob,

      Glad I didn't jump the gun by posting the info on my website. I wasn't quite sure the SPRAT was actually in distribution. Good to know you have the article. Be sure to visit the website as there are many more photos and details of the build. Two NE602's sure pack a lot of punch.

      73's
      Pete

      Delete
  6. half of the transciever I already have :) :)

    ReplyDelete
  7. but i must put att. on si5351 because ne602 need only 0dbm and less :)

    ReplyDelete
    Replies
    1. Hi Mikele,

      What you see is the max output of the Si5351 straight in to the NE602. Now if you look in detail at the code for the Si5351 you can set the output level in software by a simple line entry

      si5351.drive_strength(SI5351_CLK0,SI5351_DRIVE_8MA);
      si5351.drive_strength(SI5351_CLK2,SI5351_DRIVE_8MA);
      The 8 MA is full bore and if you change that to 2 MA then that is minimum output level. Typically you have four choices 2, 4,6 or 8. Thus you will not need an attenuator. Just some code changing in the sketch. I have had it set to 8 MA.

      Pete

      Delete
    2. Isn t 8ma too much for ne602 Pete ?

      Delete
    3. Hi Shane,

      Now that is the kind of posting I like to hear. Usually it is something like "your sketch doesn't work"! You might want to change the Splash Screens -- but I just recently learned how to do that and so couldn't resist.


      Let me know if you fully build the Sudden and get it working.

      73's
      Pete N6QW

      Delete
  8. Hi Pete, Can you tell me what version of Arduino IDE you are using.
    I'm using 1.8.3
    I'm getting compile errors concerning the Si5351:

    si5351.init(SI5351_CRYSTAL_LOAD_8PF);
    si5351.set_correction(100,);

    This looks like another great radio.
    Regards Shane G0JNR

    ReplyDelete
  9. Hi Shane,

    I am using IDE 1.8.5 I have seen those error messages and that has to do with what version of the Si5351 you are using. There are older and newer versions and you will see first line error message line that gets resolved with
    si5351.init(SI5351_CRYSTAL_LOAD_8PF, 0);
    or
    si5351.init(SI5351_CRYSTAL_LOAD_8PF, 0, 0);


    On the webpage I included the si5351.h and si5351.cpp files that work with 1.8.5 and these are included files with the sketch and must be in the same folder as the sketch. Send me an email to n6qwham@gmail.com and I will send you the sketch and the files as were used for the project.


    Thanks for your post and I know you must be anxious to get this project started.


    73's
    Pete N6QW

    ReplyDelete
    Replies
    1. Hi again Pete. Thanks for the files - I installed IDE 1.8.5 and installed your libraries and sketch - it works perfectly!
      73's
      Shane G0JNR

      Delete
    2. Hi Shane,

      Good news! Let me know if you fully build the Sudden and get it working.

      Pete

      Delete
  10. i m going to work serious on this project like i worked Belthorn ...they are all english boys ;) :)

    ReplyDelete
  11. Hi Pete, after looking at all the modules I would like to start with the Arduino nano /Si5351 / Color TFT. I would like to try and adapt it to a very early Bitx 20 and a MMR40 by Steve Weber, plus get my feet wet with Arduino code. I have been looking for a schematic on that module but have not found one, I might be able draw one from your detail description of the circuit but I would probably create some errors. can you point me to a location of the schematic? Thanks

    John kg9dk

    ReplyDelete
    Replies
    1. Hi John,

      Thanks for your interest in the project and your post. The information on the website and the wiring instructions for the Color TFT embedded in the sketch should be a good start but I can well understand if this is the first introduction then it can be confusing. I purposefully made it a wire listing as you can use several of the different Arduinos such as the Uno, Nano, Pro-mini or even the Mega 2560. They are all different save for the Digital Pins are the Digital Pins and the Analog Pins are the Analog Pins. The difference is the physical location and in the case of the Pro-mini it lacks the 3.3 V regulator so you have to have an external regulator and it also does not have an on-board USB to Serial Interface so a bit more tricky to program.

      Might I suggest you start with the Nano as it is a small footprint. Download the Pinout connections and study the digital pins and Analog Pins. Then look at my wire listing and the embedded wiring of the Display and you can connect the dots. This will help you gain an understanding of how the circuit elements work. You can email me at n6qwham@gmail.com and we can continue this discussion off of the blog.


      I have used the Arduino / Si5351 with a Bitx and that is where studying the code is requisite. You can take the code for the Sudden and you will have to change several inputs like the USB and LSB frequencies and the band you have chosen. Farhan's original Bitx20 vintage 12 years ago used an IF of 10 MHz so you will need to know the USB and LSB frequencies. Let us say it is a true Center Frequency of 10.0 MHz then the BFO Frequencies would be about 1.5 kHz above and below that. Since I typically place the LO above the incoming signal there is a sideband inversion so that the Lower BFO frequency is really LSB and the higher one is USB. for a start up frequency I chose the one that mates with the band. Say you are using 40 M then operation is typically LSB and so if I took 10.0 MHz - 1.5 kHz then the LSB is 9998500. If the start up is 7.2 MHz then the entry is 9998500 + 7200000 = 17198500. Typically you add an "L" after that such as 17198500L to indicate it is a Large Number. The code is set up so that if you change sidebands then automatically the display will change to read then true transmit frequency. When you change sidebands the display will jump by 3 kHz to account for the sideband change. For instance if your mode switch was in USB on 40M then the display will read 7.197 and if you switch sidebands to LSB the display will now read 7.200 Mhz.

      But there is a caution with using Bitx or other rigs especially ones using homebrew crystal filters-- Farhan ships the Bitx40 kit with a 12 MHz filter. But if you read his websites there may be as many as three different Center Frequencies depending on the crystal sorting process. At his factory he mates the proper BFO crystal for that specific filter. BUT most likely it is not 12 MHz and typically less than 12 MHz. So you may need an external signal source for the BFO to find the exact and correct BFO frequencies and then use that data for the final sketch. I will use a known signal like 7.2 MHz and then with an external oscillator for the BFO dial in the exact BFO frequency to give the correct USB and LSB signal at 7.200 MHz. That data is then input to the sketch. In the case of commercial filters they are a bit more accurate as to Center frequency and you can just enter the data provided by the manufacturer.


      Like I said email me with your additional questions. By taking the time to learn how to use the wire listing and pin out info will give you a greater understanding of how this all plays. It is like giving a man a fish or teaching him to fish.

      73's


      Pete N6QW

      Pete N6QW

      Delete
  12. Your work is inspirational. Thank you Pete for taking the time
    and effort to share with everyone.

    73

    ReplyDelete
  13. Hi John,

    Thanks for your post and kind words. We live in an amazing time as homebrewers --lots of really cheap technology that add orders of magnitude to the capability of our rigs,


    73's
    Pete N6QW

    ReplyDelete
  14. Hi Pete,

    Very nice work and I know how much effort it takes to document and share this with the community. Keep it up!

    Your design made me a bit curious. The NE602/NE612/SA602/SA612 family of Gilbert Cell mixer based designs has many attractive features (low noise figure, etc.) but they always warn you of low dynamic range - IIP3 = -15 dBm, meaning you can't really get beyond -25 dBm input before things fall apart due to compression.

    So I continued: in your design, I don't see any RF preselectors and your page on the RF amp stage shows a simulation that is wide open above about 5 MHz with 20 dB of gain. For a 100 Hz wide bandwidth (e.g. CW signal), typical noise at 40 meters is about -100 dBm according to ITU-R. That gets increased to -80 dBm by your 20 dB front end. So if we keep the signals to -25 dBm max, we have 55 dB of headroom.

    This would tell me that a S9 signal or higher _anywhere_ in your input would start to cause the poor 602 to have headaches and behave badly.

    So if this is right, how do you protect the mixer on receive from strong signals causing birdies and other unfortunates? Or (more likely) what did I mess up? Did you see this behavior when testing the rig?

    Note that I just renewed my GQRP subscription so haven't seen your article writeup; apologies if it's already discussed there.

    Again, fantastic job.

    ReplyDelete
  15. Thanks Phil,

    I spent about 20 Minutes replying only to find out my comments didn't get uploaded. The real burden is that I am a two finger typist. So here I am again.

    Can only say listen to the two you tube videos that have both strong and weak signals --you might be able to hear overloading but I don't and neither do I hear any birdies or other junk.

    There is RF Preselection ahead of the IF Module in the form of a band pass filter. Thus the RF Amp is not "straight in". The RF Amp also has a pot on the output to provide a balance between just the right amount of RF gain while not terribly increasing the noise.

    The plots of the 40M and 20M BPF's are provided and as you can see -- nice curves and strong out of band attenuation.

    I did not specifically prepare any analysis such as you suggested. It works and works to my satisfaction. It is not a piece of laboratory instrumentation nor is it a slightly discounted ICOM 7300.


    But now that you have come this far why not build the IF Module and RF Amp and then judge for yourself of how good or not so good of a design.

    Thanks again for your posting.

    73's
    Pete N6QW

    ReplyDelete
    Replies
    1. Hi Pete,

      I missed the BP filter ahead of the IF module - that would do it. See; I knew I had messed up something! So as long as you don't have anybody strong within your BPF, things should be A-OK. I checked out the Youtube videos and sure enough, I don't hear anything weird which agrees with your analysis.

      Definitely worth a try - will do. Fun!

      73
      Phil W1PJE

      Delete
    2. Hi Phil,

      I tell my four kids I don't need Google because I know everything --that shuts them up. But the truth is that I do not know everything so much appreciated your input and the issues you raised and what you found should be helpful to other prospective builders.


      You will find that I frequently use and reuse various circuits/modules. One of the reasons for the BPF location in many of my designs which are often bilateral is that a single BPF can be used both on transmit and receive. Frequent comments to me is why not have the BPF and then the RF Amp. The reason is it won't work too well for the bilateral approach. You will note in this design there are actually two BPF's --- one on the Rx side and another on the Tx side.

      Please let me know if you spot anything else.


      73's

      Pete N6QW

      Delete
    3. Hi Pete,

      I've gone down the Google rabbit hole and learned a lot in reading a bunch of things today triggered by this thread. So the Sudden design is teaching in those ways too!

      No matter where you have your BPF, the sticking point though here might still be when you have strong signals within the passband of your BPF (say, some other splattering SSB guy 3 kHz off your frequency and just down the road from you, or some of the AM fanatics with a Johnson Desk KW). The close-in stuff exposes the first NE602 to them since they punch right through, and since the '602 doesn't have a great IP3 of -15 dBm, you might run into trouble with close-in images/birdies in say a Field Day or contest situation. I wonder if you could spot this by trying the rig in a crowded band and A/B compare it to whatever you use for a reference rig.

      (You already have a "roofing filter" in there protecting the second NE602 - the 9 MHz GQRP 6 pole 2.2 kHz narrow one - so it's really the first NE602 that is the vulnerable part.)

      Ultimately, one could use a much better mixer, typically diode based, which has more headroom. For example, a diode based double balanced MiniCircuits RMS-1+ has an IP3 of 20 dBm - a full 35 dB better! It is also a Level 7 mixer so needs +7 dBm. Don't know if the Si5351 output would be up to the job - or would it? But the RMS-1+ has a conversion loss of 6 dB since it's passive, so you then might have to make it up downstream in the IF chain somewhere. No free lunch. It also is $12 in single quantities!

      Of course, now I'm being very unfair and applying resources well beyond your goal, which is to wring a lot of performance out of a purely NE602 based design. If I ever get time (!!), it would be very interesting to build the RF and IF stages of your receiver and simply play with a two-tone test where the in-band strong adjacent signal is increased. What does the NE602 do when it falls apart? How bad does it get? I would find out! A good exercise.

      I realized as well that in real life, maybe the background noise level is better than ITU-R claims (or turn the RF gain down), and so you wouldn't run into this problem until those nasty other people got stronger than my numbers.

      Great stuff, and I haven't even touched the soldering iron yet! Stimulates the thinking, for sure. Hats off to you once again.

      73
      Phil W1PJE

      PS: I really enjoy you and Bill chewing the fat on the Soldersmoke podcast. Do another one soon!


      Delete
  16. Phil,

    There is a simple answer. Build one, run your tests and then you can see with certainty if the NE602 folds. The Sudden Transceiver is a fun homebrew rig that was not designed as a lab instrument or contest grade transceiver. The project started out to see if I could take a two device Sudden PSK receiver and turn it into a transceiver. I did that successfully and in the webpage I mentioned the "bonus" is that the Sudden Transceiver is an experimental platform. Here is your chance to experiment.


    73's
    Pete N6QW

    ReplyDelete
    Replies
    1. You read my mind. I'll be back once I've done that.

      73s
      Phil W1PJE

      Delete
  17. Hello Mr. Juliano,

    Thank you for your reply on the other blog. And thanks to Si5351, just as you said, we don't need to get the crystals for the USB/LSB as long as we got the crystal filter. Excellent! Not only have your works given me inspirations, but your passions give encouragement. Thanks again,

    Дэн Рэндом (Dan)

    ReplyDelete

Post a Comment

Popular posts from this blog

The uBitx is inhaling RF at N6QW

21st Century Homebrew SDR SSB Transceiver Project