Software Defined Radios for 2020

Some new technology to add to the new technology for 2020.

2-3-2020 Photos of the day..

First we have a photo of Virginia Hall (from Baltimore), a famous WWII Spy working for the Brits and later the OSS. Here we see her pounding brass on a Paraset.

That must be a British version of the GN82 hand crank generator. Notably she had a wooden leg, a result of a self inflicted gunshot wound. She passed in 1982.

Our next photo has a deep meaning...

Maybe this was about their voting for the emperor (with the small e)

After connecting power of the wrong polarity to a circuit how many have said what is on this T Shirt. 

I didn't do that but did put a Teensy 3.5 into a socket, one pin over, and did see a ball of flame come out of the top of the chip. Luckily I had two Teensy 3.5's (now only one). Even more lucky it was not the Teensy 4.0!

******************************************************

2-1-2020 TEENSY 3.5 ala ZL2CTM, Working Again; But needs further evaluation. 

[My plan is get a Teensy 3.5 working and then transition to the Teensy 4.0 MCU. Keep in mind the Teensy 3.5 build is modeled almost entirely after the work done by Charlie, ZL2CTM who has done all of the heavy lifting. My contribution is to routinely solder my fingers together --twice today.]

The Codec Board is the key interface tool.

The trick with this board is the interconnections to pins 11, 12, 13 and 14. These are the Line In / Line Out pins. I have terminated these connections into 3.5 mm stereo jacks one for Line In and the other for Line Out. Short jumpers connect from these two sets of jacks over the board to another set of similar jacks and an all important DPDT relay. 

(Notice I used a mailing envelope as a scratch pad to draw out the interconnections.)

This relay (as I found out in the RADIG V1) switches the I and Q channels on transmit. Yes friend this is also done on the Soft Rock V6.3 and the Ensemble. What I have yet to determine is the polarity ( I and Q) on my RADIG Board. So there may yet be some wire switching to be done.

I was listening to 40 Meters early this evening as the band was dying out --so must await more testing tomorrow when signals are bit stronger. I may call this the PlankSDR since it is mounted on a chunk of plank.

When you know stuff and have a junk box full of built modules, then you can do stuff in about a half a day!

Note the microphone (an electret) plugs into Pins 15 & 16. I may jury rig something so I can have a very low level transmitted signal to evaluate opposite sideband suppression and if I have all the wiring in the correct order.

Stay tuned.

Coronavirus may be a solution not a problem. Just saying!

73's

Pete N6QW

****************************************************

1-31-2020 IYDWTKS,TYMBARUSS.

If You Don't Want To Know Stuff, Then You Must Be A Republican U S Senator. But If You Know Stuff, You Can Do Stuff. 

Have started my Teensy 4.0 Trek by resurrecting my ZL2CTM Teensy 3.5 SDR using the RADIG V2 homebrew SDR Board. (http://n6qwradiogenius.us) 

This approach will enable me to get a working system wherein I can later swap in the Teensy 4.0 and Codec Board as I would know that the remainder of the circuitry is working properly. I have a couple of Teensy 3.5's and so if I smoke something, it won't be the newest Teensy 4.0.

I think I discovered a wiring error in my original ZL2CTM Teensy 3.5 / Codec Board. That might account for a couple of strange observations I had when I was putting it on the air.  Nothing like putting something away and then taking a look with a fresh pair of eyes.

73's

Pete N6QW

**************************************************

1-30-2020 --- My Teensy 4.0 and Codec Boards arrived yesterday!

But I am resisting taking them out of the box and powering them "ON" until I better understand the interconnects and any special application considerations. 

I thought initially I might create a sketch just to drive a Si5351 as a single output. A friend has shared code with me to create the quadrature signals directly. I will get it to work first with a single output and fully understand its capabilities and short comings before heading off to quadrature outputs. Am really trying to avoid smoking my 1st Teensy 4.0.

Do know it has a smaller foot print than the 3.5 or 3.6 -- so that makes physical construction a bit easier. One of my first tasks will be to install Male pin headers on the Teensy 4.0 and then matching Female sockets on the Codec Board. I think they are 14 pins --I have 16 pin types so probably will get the proper size. 

There are some other preparations needed if the Teensy 4.0 is operated with a separate power supply other than being plugged into  USB all of the time. I thinks I saw some notes about adding a SD card. Lots to absorb before the first smoke test. Patience grasshopper!

Stay tuned!

Pete N6QW

1-29-2020 --- Some Noodling Thoughts about using the Teensy 4.0.

On the screen above is a prototype noodling exercise of how a 4X20 LCD might appear using the Teensy 4.0 in a Homebrew SDR Rig. The actual hardware  generating this is a Teensy 3.5 that was used with ZL2CTM's SDR project. I had to keep my hands busy with something, as I await a Teensy 4.0 to arrive from Oregon.

Typically if you use just a single output (versus the quadrature output of the Si5351) and then further run that through the 74AC74 -- you have two unused outputs of the Si5351. Assuming you use CLK0 as the LO (at 4X) to drive the dual flip flop, then CLK1 and CLK2 are idle. 

Thence in a flash of brilliance I said why not program fixed frequencies into these clocks. Therefore  by flipping a switch you could shift to the FT-8 frequency (or WSPR) and also make it shift to USB. A DPDT relay could easily switch the Clocks going into the 74AC74. With a bit more code work you could even make VFO B tunable versus fixed. 

So in the display above -- VFO A is tunable and VFO B is set to the FT-8 frequency using CLK2. Those wishing to really extend the envelope could make a VFO C that would boot up on WSPR. You could even have "enable code" so only when you selected the other VFO's would that clock be active. 

Lots of possibilities that come from a bit of noodling. Keep in mind I am a brute force homebrewer and those even mildly more proficient in software development than me would suggest writing code so that no switches were involved and it was simple menu selections using the encoder push button. So noted -- but when your are few cornflakes short of a bowl, like I am, then you resort to "kludge" approaches to achieve a desired end. Works for me!

Stay tuned for more misadventures from N6QW as he steers a Teensy 4.0 into a mushroom cloud of smoke.

73's

Pete N6QW

*********************************************************************************

Firstly, you all are encouraged to visit the N2CQR http://soldersmoke.blogspot.com Up on Bill's blog is a video of an LT SPICE simulation of a DSB signal. I didn't know you could do that with this program -- but I followed the video and was able to do it. 

This is a new capability for me and the bonus --you get to see a FFT display of the resultant output. This is more new technology to use in 2020!

The two spikes about mid-center of the photo below are the USB and LSB signals. If you add a DC Offset in the simulation factors you will see a third spike in the middle which is the Carrier. The stuff on the display toward the right hand side of the photo -- gives good cause for Filtering in our rigs! 

What I have not explored is how you look at the output of say a 1st stage audio amp to see the FFT spectrum. The video covers two voltage sources in a mixer stage. Some link must be made to see an output of say the audio amp as a voltage source. I just could not push enough buttons fast enough to do that so RTFM is in order.

New Technology for 2020

PJRC (up there in Oregon) has recently unveiled a new Teensy known as the Teensy 4.0. This is a smoking MCU board operating at 600 MHz and there is a matching CODEC Board so a homebrew SDR using something other than a Raspberry Pi is here!

Charlie Morris, ZL2CTM has done a lot of pioneering work with the Teensy series in a homebrew SDR transceiver and even the uBitix V6 can now be fitted with a  Teensy 4.0 in lieu of the Nano. That almost seems like shipping coal to Newcastle --true faster and more umphf -- but it still is a hardware defined analog radio. 

Hans Summers' QSX when it hits your mailbox I believe is using a Teensy. I suspect it won't be too long before Hans will be shipping product.

[A correction here. Hans is not using a Teensy but another processor so that was an inaccurate statement on my part (Thanks, Ben!). I also do not have any real knowledge about when the QSX is being released to production. I along with a lot of other hams are wishing and hoping  that it is soon.]

After a download of an ancillary program (from PJRC) you can program the Teensy 4.0 just like a standard Arduino using the Arduino IDE. The ancillary program has internal libraries that facilitate their use in SDR. 

You should also visit Iowa Hills Software to get the free software that enables you to build Digital Low Pass, High Pass and Band Pass Filters. Data from the Iowa Hills Software filter output coefficients is simply embedded in your IDE sketch.

An interesting note here as the Teensy 4.0 can be had for less than a $20 (less than what the Teensy 3.5 and 3.6 cost)  and the matching CODEC Board is about $15 --so with shipping two $20's should get you both. I mention this in light of the abundance of new technology hitting us in 2020 that won't break the bank. Think if the Teensy 4.0 purchase is a light meal for two at McDonald's.

I am getting a strong urge to use the Teensy 4.0 in a SDR transceiver but a lot more noodling ahead before anything concrete would happen. Stay Tuned to this space.

Just so you don't think this project has died...

 73's (From the Left Coast Wonderland!)
Pete, N6QW

You Saw where this was headed ~ A 1930's Style Station Part 3

The PMR6A Receiver itself will now undergo the re-incarnation process.

1-18-2020 ~ when you know stuff, you can do stuff! Not a uBitx v.6; but a trip back 67 years!

73's

Pete N6QW

1-17-2020 ~ It is Alive! 

More Space Filler.

Ham Safety while working with Thermatrons (vacuum tubes). 

Most of our homebrewing taking place today uses 12 VDC and that is not high voltage. Although I do know of a case where a constructor had his hand on a 5 VDC 30 amp power supply output -- seems like his ring got shorted across the supply and the ring nearly melted around his finger, which subsequently I think required amputation. So any voltage can be dangerous.

The voltages in the PMR6A under load show 260 VDC on the lead to Pin #3 on the power connector. [There is a clue here for those who read every word...] The problem with the HV is that it can start your heart (defibrillator) or stop your heart if applied too long.

One of the additional steps I will be undertaking is to provide shield covers over the power supply so one does not experience an unwanted jolt! HV kills!

73's

Pete N6QW

***********************************************

That process will entail a clean up, tube testing, circuit integrity checks and a lot of getting to know the innards of the PMR6A. But that will not happen in one day. So stay tuned.

To have a space filler for this blog since I will be working in the background on the PMR6A Receiver I would like to share a little more about simple rigs that are now nearing over 70 years old!

Enter the WesKit BN-1 Novice Transceiver. You might think of this kit much like the $3.50 Pixie CW Transceivers that are currently on eBay direct from China. Here is a BN-1 in all of its glory!

There are some notable points to observe here and firstly is that it could be operated over the range of 3 to 8 MHz --shades of a Paraset. It was battery operated (a plus for portability) and you could have as high as 180 volts on the plate and as low as 45 volts. Most likely the power output with say 90 VDC on the plate in the less than one watt range. The plate dissipation was one watt and when the two halves are operated in Push Pull the Pout is 2 watts! So definitely QRP. 

Yes a CW transceiver was all done with a single tube, a 3A5 which is a dual triode intended as a power output stage in a battery operated receiver. I did see one variant where a 6SN7 was used in lieu of the 3A5 which opens the door to a bit more power output. But the 6sN7 needs more filament voltage (6.3 Volts)

One slick advertising photo shows the BN-1 with a loaded coil whip antenna on the box but was never produced or sold that way. I can just imagine the number of hams who were getting postal money orders (the plastic card of the 1950's) and sending off to purchase a BN-1. With that whip antenna it was now a complete rig in a box on 80M CW!

With a large tip of the cap to my friend Bill, N2CQR who recently built and operated a homebrew ET-2 that was not unlike the BN-1 in that it had a one transistor regen and an a one transistor 100 milliwatt oscillator. Bill made two dozen contacts in multiple states. So the BN-1 could likely do the same. Most BN-1 field reviews were not so kind attesting to a poor receiver and low power. But others had remarkable records using the BN-1.

Circuit Forensics

I guess one could argue was this a true transceiver as we think of transceivers today? What we have is a tunable regenerative receiver (1/2 the tube) covering the range of 3 to 8 Megahertz and a crystal controlled modified Pierce Oscillator (the other half) covering the same range. So the true transceiver is only when the receiver is near the Crystal operating frequency. But I guess you could operate cross band with the receiver on one band and the transmitter on another. Unlike the Pixie which forces a received frequency near the transmit frequency.

So lets us look a little more closely at the schematics in our tour of the circuit forensics. Although the two schematics are drawn a bit differently they are the same. 

The first thing that catches my eye is the DPDT toggle switch which is the TR switch (on a small metal box). You had to really be deft when you engaged that switch. You guys with 500 pound gorilla hands, if you pounced on that switch you probably would move the received signal location. There are no vernier drives on those tuning capacitors and with a range of 5 Megahertz over the arc of the capacitor --a velvet touch is mandatory. So that could be an issue. 

The only items being switched is filament power between the two separate filaments. However on receive the crystal is grounded but on transmit connected to the plate of the transmit half.; but there is no filament voltage supplied to the transmit half of the 3A5. I have no explanation about why it would be grounded on receive.

However as mentioned on transmit, the crystal is connected to the plate circuit to form a critical function in the kB=1 criteria. The Pierce oscillator uses this topology. BUT normally one sees a blocking capacitor in series with the crystal on the plate side. 

There are many concerns about crystals used for frequency control. Two concerns with the use of quartz crystals are voltage and current. If say you were running 250 VDC on a plate of the 6SN7 variant, then one side of the crystal would be hot with 250  volts without the cap. The other issue is that huge currents can be developed in crystal oscillator circuits and an excessive amount will shatter your one and only rock. 

Perhaps the cap was left out because of cost --a penny here and a penny there. I have no crisp explanation as to why it wasn't installed. 

Now a big caution about Crystal currents. I see there are many FT-243 style crystals being sold on eBay today that are essentially smaller HC-25 or HC-49 crystals stuffed inside a FT-243 holder and re-labeled with a new frequency. Read the fine print... It says these crystals are only good for QRP rigs where the current is low. I suspect you put one of those crystals in a 6AG7 oscillator circuit you will see a mushroom cloud of smoke in your shack

The transmitter tuning involved the old light bulb trick to show when the antenna was taking  juice -- brighter = more fire in the wire! BUT given the link coupling -- this might be an issue with harmonics when operating say on 3706 kHz ( Novice CW frequency in 1950) and the second harmonic would be 7412 kHz. That might get you a post card from an OO. 

So using that rig today might not meet the current FCC spectral purity issues on transmit. A Band Pass filter, should you build this transmitter, is much in order. Today around $7 will get you a tube and socket from Antique Electronic Supply in Phoenix, AZ

Because the regen receiver is in fact an oscillator even in receive it too is radiating a signal. Think about those hams running 10 milliwatts on WSPR and being heard hundreds and even thousands of miles away.

A couple of issues with the receiver! Your Sony Walkman 32 Ohm earbuds will not work as the phones must be high impedance "magnetic" earphones and they are an active element in the receiver plate supply circuit! Also the headphone interconnects are insulated "pin type" connections so another issue. Typically the 3.5 mm stereo jacks have a common ground. Ditto for the Key and Crystal using the same type pin jacks.


Let us now look at the back side of the BN-1 box and again at the schematic.

Since this is the back side the receiver portion is along the left side where clearly we can see Coil L1 and L2. On the right side we see L3 and L4 as well as the connections to the pilot bulb (antenna current indicator) and the two wing nut antenna posts. Pretty clear.

BUT and a BIG BUT ... The schematic does not show any connections from the antenna to the receiver! The antenna is always connected to the transmitter coil L3.

So how does the received get from the antenna to the receiver. Often in a regen receiver you would have a third coil winding that essentially link couples the off the air RF to the tank circuit of the regen. I thought initially if that if coils L3/L4 were close enough to coils L2/L1 then we would have mutual inductance link coupling. But it is clear that they are too far apart for that to happen.

Here is where we can get fooled. I keep looking at that circuit diagram as two tubes. Like Bills ET-2, where there were two separate FET devices, the 3A5 is two triode tubes in a single tube shell. Not two separate tubes!

I am theorizing that in the signal path from the antenna to the regen receiver is the inter-electrode capacitance of the 3A5. Thus within the 3A5 itself the antenna on receive is connected to the receive half through a small capacitance internal to the tube. 

From the NJ7P Tube Data Base we have data for each triode so we can assume there likewise is a small interelectrode capacitance between the two triodes. That value in the 1 to 2 PF range would give "light coupling" between the antenna and the regen. The light coupling is probably good for a smoother regeneration and with moderate band condition some signals would be heard!

Each Triode
Source ........................................ RCA RC-29 - 1975
Input ......................................... 0.9 pf
Output ........................................ 1.0 pf
Grid to Plate ................................. 3.2 pf

[One of our very astute blog reader's emailed me after this was initially posted and made me aware that there was perhaps even a bit more sophistication to this circuit... lurking there in the bushes. If we think of L3 / L4 as the front half of a Band Pass Filter coupled with a very small capacitance (internal to the tube) and in turn that L1 / L2 is the other half of the BPF. This then  would indeed add some front end selectivity. So having the transmitter tank tuned close to the receiver operating frequency makes for a nice way of achieving a BPF and keeping crud out of the receiver circuits. Bravo WesKit!]

Thus today's filler has covered the WesKit BN-1 Novice two band CW Transceiver that likely does work but in itself might not bring thrills but real challenges. 

(The WesKit BN-1 was sold through Western Radio in Kearney, Nebraska and if you mated that rig with a Gotham Vertical --you could work the world!)

73's
Pete N6QW

It is all about our ham radio roots!

A Trip Back to the 1930's.

1/6/2019 ~ A Popular 1930's Transmitter Circuit known as the Jones Oscillator used two tubes in Push Pull. Below is the creation of W9GT, Jack, which uses two 955 acorn tubes and produces in excess of 1 watt to the antenna. The rig  is built into a wooden cigar box, which was a favorite building medium during the 1930's.

Dave, K4TWJ an avid QRP enthusiast and author who I now think is a SK published a schematic for a single tube 955 version as well as the Jones two tube 955 version which is shown below. 

I have been advised to be wary of simply plugging the output into an antenna as there may be issues in complying with current FCC spectral purity standards. My bent would be to install a Band Pass Filter on the output side ahead of the antenna. 

[For those who built the W1FB "Tuna Tin 2", you may recall that the ARRL revised the original schematic to add additional filtering so that there was compliance with the newer regulations.]

Major Caution Here! The Schematic for the single tube version as published in CQ is in error! If you print it out the Plus plate supply shows a DOT as if connected to what would be the GROUND side. The Correct schematic would have the Plate supply connected to the 2.5mH choke and that junction bypassed to ground through the 1NF to ground. What should be shown is that there is negative side keying to ground of the plate supply. This in itself is an issue as you have about 120 to 140 VDC across the key terminals. One solution is to find a high voltage PNP transistor and insert that into the circuit and key the plate circuit using a low voltage on the base. 

Another approach is to open the Cathode connection to ground and key the Cathode -- you might need a small value resistor for a bit of wave shaping. The two tube version uses Cathode Keying! 

Acorn tubes were essentially UHF devices that had  low filament current requirements and were plentiful as war surplus components. The plate voltage was a modest 130 VDC.  I once bought some at a price 2/$1. Now if you can find them more like $10 each.

More 1930's Transmitters...

Some Examples of 1930's Transmitters!

Recently my good friend Bill, N2CQR had on the air a two FET transistor transceiver operating on 40M CW using a single crystal. The further challenge aside from a single rock, was using but 100 milliwatts of RF power to a wire antenna. This marvel which he dubbed the ET2 had one J310 FET operating as a regen receiver and the second J310 was a crystal oscillator stage. 

Netting some two dozen contacts in multiple states over a  6 week period of casual operation, such a feat today must be viewed with awe, astonishment and down right envy! The ET2 having been mounted on a wooden plank is now affixed to the wall of his shack not only as a testament of accomplishment; but as a piece of living art. 

We should also highlight that such simple rig while very uncommon today was very much the main equipment in ham stations during the 1930's -- although quite possibly the transmitter of those days put out at least 5 watts (about 5 orders of magnitude more fire in the wire). 

N2CQR quite openly admits that operating the ET2 was a challenge -- not because of the 100 milliwatts; but because of a single frequency. He categorically stated that in the "old days" operators had a separate receiver and transmitter and often worked stations not on your same frequency. Today most hams operate using transceivers and if you are not in their receiver pass band --you are simply ignored. That indeed is today's challenge.

Bill's adventure's caused me to reflect on the comparison of a ham shack 90  years ago to one of today. But more importantly an evaluation of the cost/enjoyment/excitement value. 

Things just seem to show up on my phone most of it ham related and I marvel at what I recently saw. Here is what showed up this week: a shack that had a FLEX 6700, an Apache Anon and a Elecraft K4. Behind each rig was a very large (37 inch) 4K monitor and of course the usual array of linear amplifiers, one for each rig. Undoubtedly outback was a huge antenna farm. Quite possibly the investment exceeded $100K. There was also a low budget one I saw that had just a FLEX 6400 and a 19 inch Monitor (perhaps only $4K).

My friend Bill maybe had $20 total invested which is more in line with the ham station costs of the early 1930's. The latter part of the 1930's as things got better economically here in the US, we started to see some pretty nifty receivers that were a huge step up from the simple regenerative rigs. But in large hams still built their transmitters where now the power levels once more took a leap to the 50 to 100 watt range.

Certainly the 1930's rigs were better than the original spark gap transmitters of the 1905-1915 period. The year 1912 was pivotal as that is when regulations (200 meters and down) and the establishment of ham licensing took place. Keep in mind that it was in 1888 that Heinrich Hertz discovered radio waves. So a mere 40 to 50 years later, world wide communications took place with simple one tube receivers and one tube transmitters. Now hams spend $100K to do that and we still have no sunspots --but we have the expensive equipment when Cycle 25 finally peaks about 6 to 8 years from now. 

The IARU stated that hams are hams for two reasons: Contests and Operating! I don't agree with that position as I think there is also a group that likes to tinker. True you can only tinker if you have some skills (The Knack) or as I like to say "you have to know stuff to do stuff". N2CQR took up the ET2 challenge as it was not a simple build this circuit. But Bill actually had to develop some circuitry to make the ET2 (tinker phase) and then got to the operating part. 

I closely followed Bill's adventure during that 6 week sojourn and you could tell he was not only excited that this was a viable working transceiver; but that he was in a sense "reliving" the 1930's. Here he was making contacts with a rig built with his own hands and costing about as much as a trip to McDonalds for a quick lunch! Bill added a bit of flair to his creation -- for every contact he "logged" the QSO on the wooden breadboard with a sharpie pen. It doesn't get any better than that!

But the point I earlier made about cost/excitement/enjoyment aspect. You would hope that after spending $100K that you are having fun -- Friend Bill certainly can claim low cost, plenty of enjoyment and very exciting to make contacts with a rig  developed and built with his own hands. I suspect that Bill's buttons are popping with pride that he can get on the air and have QSO's and there is not $100K worth of gear just waiting for a sunspot!

So even though as of 2007, US hams no longer need to demonstrate a proficiency in CW as a condition for getting a license nonetheless there is still an interest in CW as a means of communication. I frequently hear as I work SSB with my homebrew rigs, the op at the other end says "I spend time on CW as well as SSB". 

The 1930's has been called the "Golden Era" in commercial radio broadcasting and that is also extended to amateur radio operations. I do not wish to overlook today's technology and the fact we have radios that can look at 15 MHz of spectrum on a 72 Inch HD screen to see what bands are open. But that technology for most hams must be purchased as the skills and tools required to homebrew such rigs are beyond the general ham population. But there are hams who continue to do this pioneering work -- thankfully. 

What would you do if you could actually build a 1930's style ham rig? Would it be as much fun as operating the FTdx101D or maybe the FLEX 6600 or the K4. You might have to build one to find out.

Below is a typical early 1930's station! Two #30 Tubes in a Regen Receiver and a #10 in a Harley self excited oscillator. One problem with the self-excited Oscillator -- the antenna! If the antenna shifted in the breeze --its loading changed and that was reflected back into the tank circuit -- the signal moved about in frequency as the wind changed. That was 1930's excitement! Notably missing: no 72 inch 4K OLED screens!

As I looked for some examples of the 1930's transmitters and receivers it was evident that "ugly construction" awaited to be invented. It appears the favorite form of construction was a wooden base plate perhaps totally "al fresco" or perhaps even sporting  a metal or wooden front panel. Clearly there was a sense of pride in their creations!

Stay Tuned. Those who know stuff, can do stuff!

73's
Pete N6QW