2018 ~ The Year of SSB Transceivers

5/6/2018 ~ Update


I Love the Smell Of Napalm in the Morning!

That specific line came from the actor Robert Duvall in the movie "Apocalypse Now" but is reminiscent of May 7, 1965 when as a member of MCB 10 [Mobile Construction Battalion Ten ~ SeaBees] and part of Regimental Landing Team #4, we made the very first amphibious landing since the Korean War. 

The landing zone was at Chu Lai, [at then South Vietnam]  some 90 "clicks" south of DaNang. Several months later (July 1965) I remember sitting on the veranda of the "homebrew" officer's club where at about 6 to 7 PM we would, with drink in hand, watch the nightly napalm run south of us on  Dung Quat Bay. Our camp (originally called Camp Shields) is now the site of a KIA plant as a part of the Chu Lai Open Economic Zone. My how times change? The 10K foot long  airfield that was built by RMK is now known as Chu Lai International Airport
Here are some photos that have survived ....


Lt. (jg) Peter G. Juliano CEC, USN Chu Lai 1965

Yes guys that is me about 53 years ago --also note I was  a "regular" and seriously contemplated a career in the US Navy. Two trips to Vietnam convinced me that was probably not a good idea. Note the "squared away" US Marine style hat. I even spit polished my combat boots. The gig line was straight and the blouse was "dressed". Noteworthy I was picked up 'early for full promotion to Lt. CEC USN. 


The above photo is the "Officers" shower facility and the man near the right hand center of the photo is taking a "whiz" into an outdoor urinal. This device invented by the Seabees is comprised a 12 inch diameter piece of PVC pip sunk into the sand and that is it. I must admit that after a few drinks the 12" target was hard to hit and the same problem late at night.

The next photo taken in 1965 should conjure up thoughts of the movie and TV series "MASH" -- I think the movie stole our lifestyle. That is my rack in the corner of the hut. Please note that my 782 Gear is packed and at the ready. Oh there are those asking what is 782 gear. Well the Seabees worked closely with the US Marines and adopted many of their processes and procedures. The term 782 refers to the form used to delineate the field gear that a US Marine (or Seabee) is issued. Typically this is a pack, helmet, entrenching tool etc --you know 782 gear.


Pete's rack and living space in 1965. I used to sleep with my 45 under my pillow. Funny story about that as in December 1965 I was selected to be the OIC of the advanced party for rotation back to the states. Because our battalion had been dramatically augmented by about 20% the day before I left I was asked to turn in my weapon as they were short of guns!

That very night our camp was attacked by the Viet Cong and we actually had VC inside the camp --and here I was without my issued weapon. At that time the sort of loose rules were you could have a personal side arm and I did have a .357 magnum which luckily was handy. Here I thought I was about to go home it might have been in an aluminum box. It was scary. I sort of wished I had a grease gun as a few of the guys did have those.

I do have to mention that our most significant piece of communication gear was a Collins AN/PRC-47. Which only went to 12 MHz and USB and CW. But I did manage to get some listening time in the radio shack and thought how cool it would be to operate from south Vietnam. We did have a Collins ham station KWM-2 30L1 but because it was so early in the war --prior to our getting there the total US forces were only about 10,000. That May it jumped to 20K so MARS and other stations were just getting approved. 

Too bad some our government leaders never spent time on the front lines. But we all know that story.

73's
Pete N6QW


Improving the Prototype #1 4/29/2018!


Starting by doing some analysis I looked at two possible factors that would improve the low frequency transmitted audio response and we have two significant players among many.

1. First is the Filter Bandwidth. I am now using a Ladder Filter where all the Capacitors are the same. Initially that was not the configuration. The values of the five capacitors play an important role in the bandwidth as lower values make the BW wider and higher values make it more narrow. For the reworked filter the values of all five caps in 68 PF. I did find that this change opened up the bandwidth and that is obvious in tuning across signals. Further evaluation may find that some sort of happy medium may be in the 75 to 82 PF range.

The filter changes make obvious better/worse conditions on the receive end --- but there are other factors that affect the transmit side including the filter.After making several contacts it was found that the signal reports were quite good with excellent characteristics in the mid to high range (a DX machine) but still a little low in the lows. This lead us to the other player in this game

2. The second factor is the response of the audio amplifier stage, which in this case is a single 2N3904 -- a bare bones stage. This is where having LT Spice at your fingertips is a critical tool. BTW just got it re-installed on my new computer after struggling to get the new libraries for the JFET's to work which now includes the J310. 

So here is some data for you. I have reason to believe the first prototype has this stage installed.



In looking at the plot data we can see there is about a 6 dB variance in the signal level at 300 Hz below what you see at 1.2 KHz. So that is why perhaps we are seeing the lack of lows.

Now if we diddle with several of the capacitors we get the following plot. Now the signal level is flat from 300 to 3kHz. Much better!


The revised schematic involves changing 3 Capacitors.

I will physically make those capacitors changes and report the results. LT Spice is an amazing tool and a few minutes with the computer let's you fix things.

Get cracking -- time to build this rig.

73's

Pete N6QW


Another Transceiver on the Boards 4/25/18

Yes it is on the boards -- breadboards.





Often you hear me talk about building several versions of transceivers with the 2nd version having the value of the "learning and improvements" from the first version that is incorporated in the later version . You see that later on in this post with the Shirt Pocket Transceiver.

One of the most successful projects I have undertaken was the Simpleceiver Plus which was completed in late 2017. That was version two and the version one prototype was simply put on the shelf. One of the main differences between versions was the crystal filter and the filter frequency. Version One was a four pole homebrew filter at 12.096 MHz and Version Two was a commercial 9 MHz six pole filter. Most of the other circuitry was the same with the exception of an SBL-1 used in V1 for the Rx/Tx mixer and in V2 it was an ADE-1. 

Critical to the success of this project was the use of J310's configured as a Dual Gate MOSFET. I also used extensive relay switching of the signal path proven with V1 so that in the final form there are essentially three Dual Gate MOSFETs that are the heart of the project. One J310 DGM serves as the receiver RF Amplifier while also serving as the Transmit Pre-Driver. A Pair of J310 DGM's act as amplifiers ahead and following the crystal filter. Relay switching essentially uses a single pass system so that on receive the signals coming out of the Mixer stage pass through the amplifier block onto the Product Detector. On transmit the Product Detector is now the Balanced Modulator and its output is relay switched to the input of the first IF amp. The output of the second IF amp is directed to the Mixer stage and then on to the former Rx RF amp which is now the Transmit Pre-driver. The signal is now switched to transmit RF amp stages.  LOTS more relays but lots less circuits. 




Two days ago, while looking for something entirely different, I spotted V1 sitting on a shelf, looking somewhat forlorn as I had robbed a few parts from the transmit linear amp stage; but otherwise unmolested. Given this was such a critical development tool for the later project, it was time to make it whole again.

About 10 minutes worth of work involved with liberating some boards from yet another project and the Version One was back on the air. A few on the air contacts with some of the Flex6700 ops indicated that the signal sounded OK but seemed a bit narrow lacking the lows they like to hear. There were no comments about distortion or burbles or spurs just that it sounded narrow. The further comment --it would be a great signal for busting pileups but for the "40 Meter Illuminati " a bit narrow. 

I looked at the coupling caps that were originally used in the homebrew filter and concluded that the filter would indeed be some what more narrow than the 2.8 KHz the Flex guys like to hear. My calculations indicated about 2 KHz bandwidth. So about 10 minutes of additional work and some new caps resulted in a  a newly refined homebrew filter. I can tell the received bandwidth is much wider .

Last night about 8 PM local left coast time --the 40M band was wide open and I made about a half dozen quickie contacts --some sort of contest. Half of these contacts were with east coast stations including one right in the heart of PA District 8-- Connor Lamb country. BUT there were a  few unsolicited reports -- like "you have a great sounding signal." So those are some data points; but we need more rigorous data collection before declaring victory. That I hope to do today. BTW --no QRP during the operation last night --was running 600 Watts. For those who wonder -- no beam just a droopy dipole with an apex at 25 feet. 

Using the SEARCH function on this blog type in simpleceiver for the many details of the build.

Some Photos of the "Al Fresco" build: [Yes that is a breadboard C Clamped to the work bench.]

The Rig 

The Display


EMRFD Driver and IRF510 Final

The Mess ~ But it works so well!
There is a total of 8 boards and with a bit of effort can be made a lot more compact. That is the plan. Get crackin -- you can build this rig.

73's
Pete N6QW

N6QW Innovative Transceiver Projects!

One of my goals for 2018 is to share information about my past projects in hopes that other homebrew enthusiasts will take up the iron and "roll their own" Many of the projects go back over ten years; but are relevant today.

You can make great contacts using homebrew gear! But I should caution you that you may get comments from the other end such as "well I am looking at your signal with my FlexRadio 6700 on a 72 inch screen and it appears you are 10 Hertz low in frequency". You gotta have thick skin--just think your rig cost $100 and his maybe $10,000. But even being 10 Hertz low --you are still having a QSO. 

There are three specific projects I would like to cover today and these include the following:

The first is a 20 Meter SSB transceiver that used MMIC's (Monolithic Microwave Integrated Circuits) arranged in a bilateral configuration. There were six total MMIC's employed in three sets of two for the bilateral amplifiers. This was kind of leading edge in 2011 for a homebrew ham transceiver. Initially I used a VTO (varactor tuned oscillator) but later this was changed to a Si570 VFO via a K5BCQ kit. [This was another rig given away when I had to move in 2013.] This by my definition an innovation in a ham rig.]

The second rig is a 17 Meter SSB Transceiver originally built in 2007. The innovation here was to use a crystal switched Super VXO to give a wide frequency coverage. The innovation here was the selection of the IF 4.9152 MHz and using 11.52 MHz crystals in a diode frequency doubling Super VXO circuit produced a signal; a 23.04 MHz and in doing the math 23.04 - 4.9152 = 18.124 MHz. Using the doubling circuit gave twice the swing in a Super VXO circuit. That worked so well that I custom ordered additional 11 MHz crystals so that I can cover almost the whole 17 Meter phone band.

The third project was a Shirt Pocket Sized 20M SSB Transceiver which was built in two versions with the first requiring a rather large pocket. The second version is 2X2X4 inches --still needing a big pocket. A third version planned for 2018 is to be Altoids tin sized. This rig also employed a 4.9152 MHz IF and this time a crystal switched heterodyne VXO. The first version in 2017 was retrofitted with a Si5351 and OLED Display.


Some Project Details:

The first is the MMIC Transceiver which was the very first "big time" article I ever wrote about one of my projects --this even made the cover of QRP Quarterly. The embedded you tube video was made at the time the rig first had the VTO and the drift is pretty apparent. It was real crap. The final version with the K5BCQ Si570 VFO kit --was like night and day. I regret having to give this rig away when I moved. In 2017 I used the bilateral MMIC stage in a compact 20 Meter SSB transceiver. The AG-303-86G is a superb device. As I mentioned in the article I got help with the bilateral circuit from a ham at Tri-Quint who was the manufacturer of the MMIC.

In passing at times you bump into things that amaze you. A friend sent me an eBay listing for the AG-303-86G MMIC device from a seller in Israel. In the ad, my schematic for the bilateral stage was featured to show how they could be built into the circuit. The MMIC's are termination insensitive good from DC to 6 GHZ and feature 20 dB gain. Should mention that the data sheet includes the required resistors for various sources voltages which enable you to run the devices on anywhere from 5 to (I think ) 9 VDC. I picked  5VDC. Another neat feature of my circuit is diode steering so that when you "bias on" the MMIC that also steers the signal right left or left right. So OK another bit of innovation --for 2010.

To the best of my knowledge I am unaware of any other homebrew rig which used MMIC's in a bilateral configuration in 2010. I am sure there are commercial designs but my application may have been a first. I am sure blog readers will let me know otherwise.

In the last photo is shown the MMIC board (underneath the green perforated board) and a shot of the completed rig. There are two stories here (actually three) with first being the suitability of the MMIC's for our beloved rigs. The second is that the MMIC's can be effectively used in a bilateral configuration and the third -- great for making compact rigs.













The 17 Meter SSB Transceiver Project:



The second project is a 17 Meter SSB Transceiver first started in 2007 where some fundamental circuit blocks were initially developed that found their way into future projects. The heart of this rig was a bilateral amplifier circuit developed by G4GXO [SPRAT # 128] which used either a Dual Gate MOSFET or two J310's configured as a DGM.

Based on some information gathering in preparation for the build, I discovered that in the Elecraft K2 the IF Frequency was 4.9152 MHz. Hmmm good enough for Elecraft so good enough for me. This was a wise choice as I soon discovered that using a stock computer crystal of 11.52 MHz in a diode frequency doubling VXO circuit produced an output at 23.04 MHz. The frequency scheme in a down mix resulted in tuning smack in the middle of the 17 Meter Phone band. Subsequently I purchased special frequency commercial crystals that provided the upper phone band coverage of 17 Meters. A trick I used was to switch the VXO crystal banks was a relay that was controlled by a panel mounted toggle switch. Thus I had a high range of VXO frequencies and also a low range. The yellow cube in the photo below is a DPDT 12 VDC DIP Relay. Because of the frequency doubling two crystals per range in the Super VXO Circuit provided a large frequency excursion.  

Also note the use of single sided copper vector board where a solid ground plane is the norm. After using this (somewhat expensive) medium you never think of Manhattan construction for projects like this. All wiring is point to point underneath the board. With some circuit noodling of the layout it is possible to achieve a layout that minimizes the circuit wiring and reduces the possibility of unwanted coupling. Some nice innovations here.




Complete details of this transceiver can be found on my 2nd Website Believe it or not this project started out as an attempt to build a shirt pocket SSB transceiver which resulted in a working rig; but definitely not something to fit in your shirt pocket. The boards above were finally fitted into an aluminum chassis measuring 7X7X2. There is room in that chassis to fit an Arduino/Si5351 and an OLED Display. That is in the queue.






The 20 Meter Shirt Pocket Transceiver:

As I stated earlier my initial quest to build a shirt pocket SSB transceiver started in 2007; but it was not until 2011 that I arrived at a shirt pocket sized transceiver. A rig 16 cubic inches is small BUT you still needed a large shirt pocket --Paul Bunyan sized more like it. In 2018 I am looking forward to building this same rig into an Altoids sized tin. 

In 2011 I was overtaken by the desire to really get serious about a shirt pocket sized transceiver. W3TLN, Ed Vester built an all solid state SSB rig running 100 Milliwatts on 20 Meters that ran off of 6 VDC. The choice of 6 VDC was because it was a mobile rig fitted into a Volkswagen Beetle -- an early version with a 6 volt battery. It was a small rig and used a homebrew crystal filter and a VXO. The Vester rig was published in a QST article and subsequently published in the ARRL Sideband Handbook. 

I built the Vester rig in the 1970's from the ARRL Sideband Handbook. But I could never get it to work. In the course of preparing some articles for QRP Quarterly, in 2010 I mentioned to John W5DIA about my non-success with the earlier project and he sent me a photo copy of one of the pages in the original article. BOOM --the article and the Handbook version schematics were different. That was my problem! I had a difficult time reconciling how there could be two different versions of the same schematic --in a QST publication...


Thus having found the problem with the Vester rig I decided that I would just build my own version -- the Vester rig was 7x7X2 --not a shirt pocket rig.

I built two versions of the shirt pocket rig with the idea that the first version would be the proof of concept and the second would capitalize on that experience to make it really small.

The is information on my  website  http://www.jessystems.com/20MSHIRTPOCKET.html about this project including some schematic diagrams. This project also was a subject of a QRP Quarterly article. 

Noteworthy the original rig used the same crystal switched VXO approach as in the 17 Meter Rig  with a twist in that the Super VXO operated at 12.906 MHz and was mixed with crystals in the 6 MHz range to give about a 120 kHz spread on 20 Meters. The reason for this choice/approach was the availability of standard computer crystals.







Fast forward to 2017 and these rigs have been further modified with new case colors and Version 1 and been retrofitted with a an OLED/Si5351 and Nano digital LO.



This is a lot to absorb ... but is provided to open up a view of the homebrew possibilities. It goes without saying -- most of these are not first time projects but they do offer an insight into the amazing (and cheap) technology that is available to us today.

73's
Pete N6QW

Comments

  1. Perfect Pete, i m back after few mounths e andiamo avanti, 9a3xz

    ReplyDelete
  2. Hey Pete got your note the other day and thanks for the kind comments.
    I am working on fixing the blog so comments will appear public.
    Anyway, the 70MHz rig is almost done just need to add the PA and antenna relay, however I am already chasing revision 2 of the radio design and I wanted to ask you about your experience with TFT displays and noise - I was caught out by the OLED and don't want to end up in the same boat so wanted to find out first hand if the TFT modules are RF silent?

    Cheers and 73

    Rob
    2E1IIP

    ReplyDelete
    Replies
    1. Hi Rob,

      Send me an email to n6qwham@gmail.com and we can have a lengthy discussion about TFT's. No Mystery --I find they work great.

      If you don't use lots of connectors --it won't work!

      73's
      Pete

      Delete
  3. Pete

    I ran the LTSpice sim on your amp and I have a couple comments:

    The reason R1 works better at 47K is that the bias point with 10K was horribly high with about 11VDC on the collector It shouldn't be run this way under any circumstances. Optimum collector voltage would be around 6VDC. It looks like you never ran the time-domain analysis or you would have spotted that driving the circuit with 0.5VAC horrible saturates it with R1 = 10K, exacerbated by the horribly wrong bias point. The emitter bypass cap needs to be about 500uf for best low end performance, and the coupling caps should be increased to about 1uf for the same reason. With these changes, the input impedance is around 1600 ohms and the gain is about 43 dB. Driving the circuit with lower or higher impedance previous stages can change the gain by as much as +/- 10dB.

    ReplyDelete
  4. Nice story about Vietnam, Pete...i didnt know you were there,it was a long time ago,but i wanted to tell you something else,yesterday i listened soldersmoke podcast and i heared that you talk about Bitx a lot, i like your way of building transciviers, and i would like you to stay on that way because i like your rigs....look at this:
    https://youtu.be/rEbp4GAhX-Y

    73 de Mikele, 9a3xz

    ReplyDelete

Post a Comment

Popular posts from this blog

The uBitx is inhaling RF at N6QW

21st Century Homebrew SDR SSB Transceiver Project

Simpleceiver Plus Version 2 SSB Transceiver