New Technology for 2020 ~ You almost can't keep up and here is #48 with #49 and #50 not far behind.

#48 is Finished --- #49 in Work.

October 3, 2020 ~  uSDX NOW Working!

This morning I touched up the solder joints on the encoder and the switches. Boom now working on receive!




BUT THIS IS WHERE I WAS LATE YESTERDAY. Still food for thought about the encoder, switches and display...

I received the missing parts installed same including the new firmware and #49 is best described as a crystal set with only 5 feet of wire attached! I have three ATMEGA 328 microcontrollers that have the very latest firmware installed and the condition is exactly the same with any of the controllers in the radio. I do get some of the functionality so I know the code has been loaded on the 328's.

My belief is that there is a hardware issue other than the 328 quite possibly in the encoder and the push button switches. Here is the issue --the radio does not respond to the command inputs. This is a menu operated radio and it is difficult to get the radio to respond to command inputs. There have been some guides published on how to navigate the menus -- but this radio is much like a small child who ignores instructions and inputs (Maybe like our emperor. See wearing a mask is important!). See the guide below the photo. 

Turn the encoder and nothing changes on the display; but going through some other button pushing you realize the frequency has changed as a new value will show up but you hear nothing. It did work once and I heard a 40M station quite well -- but never again did it work. This lends a weight of evidence to some hardware issue.

Others have their uSDX perking right along -- but alas mine is a like a lump of coal that has not been ignited -- no warmth, no light, no glow. I think I will just move away from it for a period of time as I am fresh out of ideas. 

Here are a couple of things I would do different knowing what I know now. I would (like I saw in one set of constructor photos) dump the cheap detent encoder and the flimsy push buttons and hardwire in a better encoder and switches mounted on a panel. I would also remote the display on that same panel. With the display as a plug in the board -- hard to view. I am not so much interested in a compact radio as I am one that works and has decent hardware. 







Yes Virginia, this is the early stages of a uSDX build. I was shorted some parts by Digi-Key and so had to order from Mouser. Thus I am not done! I also need to build the LPF's and have parts coming for those too.

The build is pretty simple but you have to know some stuff to do this stuff. Example, there are twelve 100nF caps to be installed. I looked over the board and found all 12 locations. While I was doing the hunting, I wrote down on a my construction notes the twelve part ID's like C1, C13. Then as I installed the part, double checking the silk screen with the value, I marked it off the listing. Once I got through the list I checked all twelve were marked off and double verified that I indeed installed the right part value at the right location.

There was a problem with Spurs which has been resolved by Guido PE1NNZ and new Firmware is available for download It is V1.021.

The only thing aside from not having all of the parts yet being able to run the ATMEGA 328 is the crazy menu's. On boot up there are 9 sections with sub menu's in those section such as MODE (USB default, LSB,CW) or Band (40M default). The problem I have been having it seems a mystery once in the section --how to make changes. You essentially have four controls. The encoder which has two -- the encoder tuning and the PB control and then there are two Momentary PB switches. At times I have been able to change bands or change modes but no idea how I did it. The four controls interact but not intuitive what they do. This was designed by a software guy! Have pity on us hardware freaks! Must be a fault on my part as I don't own and ICOM 7300 thus not used to menu stuff. 

There is an issue with the ATMEGA 328 purchased from Digi-Key. It is as dumb as a rock as it lacks the Boot Loader. So you have to Jimmy a way to get a Boot Loader installed and then you can load code. You can either do it with two UNO's hay-wired together or go to Amazon and purchase a USB asp with 6 pin ICSP Adapter. This gizmo connects to your computer and then. connects to an UNO where you have installed the bare 328. Under the Tools Tab in the IDE are select a programmer (USBasp) and then hit Burn Boot Loader. Once you get the Boot Loader installed then you can install the new Guido Hot Fix Firmware. 

An alternative is that All Electronics (out here in CA) sells the 328 with the Boot Loader installed --about $5. The USB asp programmer cost $20 with shipping. 

The new Firmware if you are not using the QSX as the starting device must have the 1st Line nulled out like //#define QCX 1 . The display will now read uSDX on the 1st line

If you are using a stock Si5351 breakout board that has the 25 MHz clock then about line 899 there are three clock selections and choose the following. Null out the other two.

#define F_XTAL 25004000 //25MHz SI5351 crystal

I am sharing more than I really know; but the kit went together fairly quickly -- one full day if you take your time and follow an orderly process. Oh look carefully at the board to see "tough" installs. I started with installing all of the capacitors and there are five electrolytic types. One of those a 10 UFd. sits right in front of a special 5 VDC LDO regulator which is surface mount. 

INSTALL the REGULATOR before the cap otherwise you will burn the cap with the soldering iron as you try to solder the three tabs. Another suggestion do not install the sockets and other hardware parts (switches, jacks, power) until all of the passives (resistor, capacitors, diodes, crystals) are installed.

I used some aluminum pillars one the board so that  when you turned the board over to solder the parts the board was elevated off the work surface. My process is to insert the part into the board and then use masking tape to hold it in place and next turn over the board to solder at the underside. An 80 watt Radio Shack soldering iron with a 3/4 inch tip is inappropriate for this Fine work.

Oh Forget that ROHS solder as you will end up with bad connections and cold solder joints --get the 60/40 type solder and wear a mask.

Ergonomics will be an issue for me with my FFS (Fat Finger Syndrome). The encoder looks cheap and it is doubtful it will last more than about 10K rotations. The BNC connector is at the bottom of the board so packaging WILL BE an issue. 

So for me if I were to build a 2nd unit I would create a top panel where I will install a robust (smoother -- and not a detent type) encoder, and locate the BNC elsewhere. Better PB would also be on the list.

It would be easy to build an extension cable for the LCD. For the uBitx I bought some 4 inch long M/F jumper cables and simply did a pin for pin plug in and was able to mount the Raduino LCD on the front panel.

So #49 is in work and I won't have any news until all of the parts arrive.

73's
Pete N6QW


10/1/2020

Yesterday on Unit #48, I added the back chassis plate and fitted the panel with the Power Cord, Linear Amp Switching Jack and the Antenna Connector. 

You might wonder how I can crank out rigs at a lightning fast pace. It is all about the tools you have PLUS thinking creatively. You just can't go out and buy stuff like one used to do. Pre-Covid19, I made a once a week stop at Home Depot and simply toured the aisles looking for stuff that co  1uld be used in projects. 

Today, not being able to simply run through  Home Depot has forced me to look in the bins and boxes at stuff I already have. Boom, I have been surprised at all the stuff I have in the past ordered and simply squirreled away for future use. The future is here now!

Speaking of tools, there are CNC machines on the market in the $300 to $500 range that do an absolutely creditable job in making electronic hardware. Whether it is making circuit boards or cutting rectangular holes in panels, it can be done by you. You don't have to be a machinist or have special shop skills --just read the manual and use that computer between your ears. 

Unit #48 was easily built using the CNC and some good old fashioned creativity. Take for instance the IRF510 which needs a heat sink. Here is what I did for a heat sink, There is a piece of double sided copper board about 1.5 inches by 1 inches that is the circuit board and heat sink. On the back side I removed the copper over an area about 2X the size of the IRF510. But left the rest intact. 

Next using an insulating mica thermal pad and a nylon shoulder washer I formed a mounting kit so that the shoulder washer was on the underside of the board. The shoulder rests in the PC Board itself so that the mounting bolt never comes in electrical contact with the grounded PC Board

A 3/4 inch 6-32 bolt was passed through the washer, through the PC Board, through the Mica insulator, through the IRF510, through solder lug (this connects the Drain) and finally a 6-32 nut. When this is cinched up the IRF510 is in thermal contact (I did use thermal grease too) with the copper board; but electrically isolated. Next I have four 1/2 inch aluminum pillars that are threaded for 2-56 bolts that connect to the PC board and to another L Shaped aluminum sub-chassis which houses the two main stacked transceiver circuit boards. 

Thus there is a 1/2 inch air-gap between the IRF510 board and the L Shaped bracket. BUT through the four aluminum pillars there is a thermal path for the heat to dissipate to the larger aluminum sub chassis. This has worked well! Bottom line --I was fresh out of TO-220 transistor insulating kits, had no more TO-220 heat sinks and was sort of placed in a position to come up with an alternative. 

The original plan was to have the IRF-510 on the top transceiver board. Too close and feedback issues. This solution kept the IRF510 in close proximity but provided electrical isolation and shielding. All done with the CNC Mill and IUWIH.




The fabricating of Unit #48 has provided a sense of deep satisfaction as it was a challenge to use what was at hand and think out of the box. Regretfully, I have run out of Juliyellow paint as I really wanted to paint this rig a nice bright color -- no running over to Home Depot so used the Juliano Blue (Oasis Blue).

So notable points -- this has a homebrew crystal filter at 4.9152 MHz and is proof that I do know how to build crystal filters. It is also small which was an engineering challenge to fit all into a smaller foot print; but not as small as the Second Shirt Pocket Transceiver which you can see below.




73's
Pete N6QW


Updated Photos of the #48 ~ 9-29-2020


I have made a half dozen contacts just this afternoon including two DX contacts into South America. Got a High Quality Audio Presence report from a station in Florida. So once again the Plessey Amps (all three sets) and the homebrew Crystal Filter rule the airwaves.




Work remaining: installing the back plate and finalize the power wiring and antenna connections. This is a winner! 

The next two photos were taken before I installed the TR relays and LPF board.






SSB Transceiver #48 ~ 20 Meters.

[ The repackaging continues...













For those who still might be members of QRPARCI and have QRP Quarterly publications going back to 2011 you would find I penned an article on a 20 Meter Shirt Pocket QRP SSB Transceiver. 

I am no longer a member of QRPARCI nor get QQ so I am woefully unaware if any other innovative projects such as the Shirt Pocket Transceiver that are being published. I do know  that one of my other projects was recently republished. 

Regrettably I was not even notified that they were doing that. It seems like when you submit an Article to QQ they maintain they own it and as an author you have no rights of publication. Needless to say QQ will not be on the receiving end of any of my current or future work.

I actually built two of the transceivers and this project was the first one where you needed a really big shirt pocket. However V.2 was much smaller and likely could fit in a shirt pocket. There are youtube videos of this earlier variant still on the you tube. The one on the left is the larger version and the basis of this repair/restoration/upgrade. The smaller one is only 16 cubic inches -- 2x2X4



Both versions involved the use of the Plessey Bi-Lateral Amplifiers, a homebrew 4.9152 MHz four pole Crystal Filter and a novel Crystal Switched two range VXO. The larger one had a Rx RF AMP stage and the second one did not (that was a mistake). 

In an attempt in early 2020 to fit an Arduino Nano/Si5351 and 1/2 size OLED display inside the metal case of V1, some wires got crossed and I smoked a lot of parts including the transmit amplifying stages and the receiver RF amp. So I just put it away mainly to not look at it; but to rethink the replacement electronics.

The basic receiver, filter, mic amp and audio stages were unscathed so I only needed to deal with the Rx RF Amp and the Transmit Stages. I made a board the same size as the basic board and used aluminum spacer to top mount the board. On that board is another Plessey bi-lateral amp stage (much like the two in the IF Module) only it serves at the Rx RF Amp and Tx Pre-driver. 

Also on that board is the 2N2219 Tx Driver Board and the Bias Network for the IRF510. Initially I had the IRF510 installed there too. BUT too close and RF Feedback! I made a small auxiliary board and side mounted just that stage. Feedback gone!





The 2020 redo involves the Arduino Nano / Si5351and a Color TFT Display. The transmit chain has been simplified by using the bilateral Plessey stage and the receiver is as hot as ever. I get about 4 watts out and that will drive an outboard linear amp to 100 watts.

Smoking the parts was not good; but it was in that the upgrades and improvements are very nice and increases the functionality and overall operation of the rig.

Let's see your build.

73's
Pete N6QW

*******

The Atomic Pi $39.95 @ Amazon.


Regrettably I smoked at Raspberry Pi3. Well actually it works but the  WiFi no longer does. So kind of useless where you need Internet Time like for WSPR and FT-8. So I needed to find a replacement.

That is when I spotted the Atomic Pi which is branded as a super competitor to the Raspberry Pi. It is sort of like Emperor Trump telling his cluster of supporters only he can fix things and lying in the process. 

The Atomic Pi is an Intel Atom based Single Board Computer that started life as an embed for industrial applications. Somehow that got derailed and there were tens of thousands of these lying (good term) around a warehouse and that is where some genius figured a way to sell these boards as a Raspberry Pi Alternative. They price point is less than $40 and it has a 1.4 GHz processor that can run at 1.9 GHz, and  2GB of RAM. It has 16 GB of built in Storage but an SD Card Slot for up to 256 GB. HDMI output and a 1.5 Watt Stereo amplifier. 

The Atomic Pi is being sold with a bottom mounted breakout board and even comes with a separate Camera. The OS installed is Unbuntu. But you can install Windows 10 (limited) via the SD Card.

You will need a 5 VDC at 4 amp Pig Tail supply that is fed via a terminal block on the breakout board. I have not been able to get Netflix to work (as yet) but I have installed Quisk and WSJTX. They sell a supply for about $10 -- a $50 bill will get you the Atomic Pi and the Power Source.

Some limitations --poor documentation, only two USB ports ( a 3.0 and a 2.0) kind of big and uses passwords for everything. There is a limited supply as no more are being built and once the warehouse is empty much like the Emperor: A lot of hype but no more in the store.

But it does get you a replacement for the RPi3 and puts you back on the air. Oh another Big Bonus --a Real Time Clock on board.

It does have WiFi but you need to buy one of those funny APEX antennas (about $6) . Scroll down for photos 










 In Case you haven't guessed by now #48 is now working. See the videos above... Never sleep as you will miss something important.




New Technology for 2020 ~ A Return To Crystal Sets...

 Crystal Sets, Is That What You Said?

In the 1950's living outside of Pittsburgh I was fortunate to have some very high powered AM radio stations close by namely KDKA. This early pioneering radio station had a 50KW transmitter and was "Clear Channel". Bottom line a simple crystal set with a modest antenna could hear KDKA. I had a crystal set and an exposed bed spring antenna provided many hours of enjoyment. No Crystal set was complete without a pair of Brush headphones. I am a side sleeper and to this day one of my ears (the one one the side facing down)  is slightly bent from an earphone constantly pressing on that  ear.

It was an an amazing time for me as I did not have a kit crystal set but one that I actually homebrewed myself. So it is today that there are many new radios, totally driven by technology, that are like those crystal sets of old. 

Sparse on hardware but driven by software. There is a "hot on fire" new SDR radio that is enjoying viral trending status and that is the uSDX. Costing less than $100 in parts it will change the face of ham radio.

Guido PEINNZ, started with the low cost QCX, CW Transceiver from QRP-Labs which by itself is around $60 and repurposed that hardware into a multi-mode, multi-band 5 watt SDR Transceiver. The rest is history with history being made on a daily basis. The Big Guns of ham radio are more than dipping their toes in these waters.

I predict by Christmas we will be seeing full built radios coming from China (maybe Wuhan) and the price point would be in the $100 to $150 range. I am also suggesting this is the new Bitx type radio with a lot more Bells and Whistles.

Yes simple crystal sets, now simple SDR rigs.

73's
Pete N6QW

New Technology for 2020 ~ Moving beyond the MC1496 Direct Conversion Receiver

 The MC1496 SSB Transceiver ~ #47

9/21/2020 Added Schematics and Layouts





I had 40673's! This shows J310's as a substitute





























It all started with SolderSmoke Podcast #224 and the comment from N2CQR about a transceiver project from Paul, VK3HN. It seems Paul found an assembly at a Radio Rally (hamfest) that was a project in SSDRA pp203-205. This assembly was an IF amp stage (2 X 40673 and Crystal Filter) and a Transmit mixer Stage comprised of the Motorola MC1496 DBM. VK3HN took that assembly and added bits and pieces to create a Four Band Transceiver he called The Prowler.

I was intrigued by that project and started small with a Direct Conversion Receiver using the MC1496. My effort called The Cat-Around (like a left coast Prowler) was but a stepping stone to a full transceiver using the MC1496 in four key locations:
  • The Receiver Mixer
  • The Balanced Modulator
  • The Transmit Mixer
  • The Product Detector
The Balanced Modulator and Product Detector circuits were a direct lift form the Motorola Application Notes. I used the schematic for the Transmit Mixer Portion as found in SSDRA for both the Receive Mixer and Transmit Mixer. The IF Module used the 40673's and the GQRP Club 9 MHz Crystal Filter. 

You can literally use any 9 MHz Filter including the INRAD Model #351 which will be back in Stock on October 1. The IF Module Schematic matches the Zin/out with 510 Ohm Resistors. the Zin/out of the INRAD is 200 ohms so you will need to match to that impedance like a 2.5 to 1 match. A 5 Turn to 8 Turn Ferrite transformer would get you close 5^2 = 25 and 8^2 = 64. 64/25 = 2.56:1.

A truncated Block Diagram of just the modules mentioned would look like this...



Our Board would be rounded out with a 2N3904 Microphone Amplifier  and combo 2N3904/LM386 for the Audio Amplifier.

This is the 1st transmitter test of the MC1496 Balanced Modulator and Transmit Mixer.




The MC1496 SSB Transceiver has been set up for two bands including 40 and 20 Meters. A simple DPDT Toggle Switch signals the Arduino to change bands (1/2 the switch)  while powering on (2nd half) banks of relays two switch the BPF and LPF Networks.

I always consider any rig I build as an experimenter's platform and this rig was no exception! I have used the bilateral Plessey Amplifier in many of my projects; but typically in the IF Module. This time I thought why can this be used as the Receiver RF Amp (in one direction) and Transmit Pre-Driver (in the other direction). Essentially the 2N2907 is the Rx RF Amp and the MPS6514 is the Tx Pre-Driver. (This is another experiment and a change from the 2N3906 and 2N3904)




The board photo above is the two BPF's, the Plessey Circuit and the In/Out relays. Initially I used three pin headers with shorting bar to switch bands. This board has now been upgraded to include two more relays (in the open space) so that band switching is all automatic from the front panel. Once again the Plessey circuit comes through!

Some would ask why are you using 1970's technology to build a rig. Well it was an experiment for me and I have shown it can be done. The cost is pretty small and most importantly --Because I can!

73's
Pete N6QW


TLPD!








New Technology for 2020 ~ A FREE TKT ( Tribal Knowledge Tip)

 How to Handle More Than One Band!



[Arduino Code Snippet ADDED to this post]

So you have just homebrewed a new QRP transceiver and it doesn't matter if it is SSB or CW and as a bonus have included two bands. Now the problem how to switch all the needed circuits so that with a simple DPDT Toggle switch -- the magic happens!

But let us back up in the sewer so we get the full aroma. Switching bands entails several actions including switching the LO range (I am presuming a digital LO) as well as the Band Pass and Low Pass Filters. One half of our DPDT can be dedicated to the Arduino Input for band change while the other half of the switch can "Power On" sets of relays to select the one of two bands.



Above is a two band LPF with the yellow core being the 20 Meter Section and the red core the 40 Meter Section. The two black relays are SPDT. and the Yellow relay is a DPDT.

By my wiring convention I made it so the NC contacts on the DPST are connected to the 40M Section and the NO to the 20M Section. The Center Leaf on the Left DPST is connected to the IRF510. Thus the LPF is always connected to the IRF510. In simple terms the LPF is not in the Receive Loop. This avoids hot switching the LPF to the IRF510.

The DPST relay on the right side of the LPF's has its center Leaf connected to the NO Contacts of the DPDT (yellow) Relay. (Again NC = 40M and NO = 20M) 

For the DPDT (TR Relay) I simply bridged (connected together)  both sets of contacts as we are running 5 watts through that relay. The Center Leaf of the Yellow relay goes to the bulkhead mounted antenna connector. The NC contacts of the yellow relay are connected to the Receive side of the transceiver. That follows that the NO side of the (TR Yellow Relay) is connected to the output side of the  LPF Relay.

Wiring is as follows the two DPST relays are powered  from the Toggle switch! So "power on" switches bands. That switch at the same time switches the Arduino to the new band.  The Yellow (TR) Relay has juice applied at all times BUT only grounded through the PTT switch (or MOX) control on transmit. Thus normal action has it so on receive the signal from the antenna is passed onto the receive section But on transmit the antenna is connected to the selected LPF.

Purposefully I did not draw a schematic as that is left to the reader. If you can't draw a schematic from what I have described, don't build the filter and consider stamp collecting!

Critical Path items! The base for this assembly is single sided copper vector board where the top surface is a ground plane and all connections to ground are simply soldered to the top of the board. All network connections are made on the under insulated side. Prior to passing connections that are not grounded through the board, I take a small drill bit and ream out the area around the penetration hole so there are no shorts to ground. Connections made to ground are not reamed. Look at the Right Side DPST in the lower right hand corner and you will see the reamed hole.

Secondly I use the W3NQN filter design as that specifically addresses the 2nd harmonic issue. The QST article which can be found on the internet gives the constants for all of the ham bands. I found it best to simulate the circuit in LT Spice and if you are a Nano VNA aficionado you can test the network after it is built to check your work. 

Next is some code snippets to show how the Arduino Code is arranged to respond to a LOW condition on either Pins A0 or Pin 5. This is not a standalone sketch but shows some of the elements needed to be added to existing sketches that start by identification that two pins will be used to detect a band change and then when a condition is identified for each LOW that becomes the start up frequency sent to the Si5351. I did not write this code but a ham in VK land was kind enough to furnish this to me. 

A ham friend here in the US is already looking at using the Nextion touch screen so that you touch a button on the screen and it is automatic band change. An additional output pin on the Arduino would trigger a transistor switch to cause the relays on the BPF and LPF to toggle between the bands. 

READ THIS IS NOT A COMPLETE SKETCH! THIS IS JUST ADDED ELEMENTS TO BAND CHANGE. ADDING MORE BANDS AND START UP FREQUENCIES IS NOT DIFFICULT. Added bands require changing a few parameters.

/* Sample code added to your sketch to give two band operation. Pins A0 an 5 are the input pins driven LOW to change bands
   The two numbers 16198500 and 23201500 are the start up frequencies for 40 and 20M with a 9 MHz IF
 */

int SWBANDS = 0;

int BSW = 0;

const int MAXBANDS = 2; // the number of bands we are using
const int SWBAND[MAXBANDS]={A0,5};  // 40m on A0,20m on 5
const int_fast32_t rxfreqs[MAXBANDS]={16198500L,23201500L}; // the default starting frequency for each band
int lastband=0;         // used to keep track of the last band used



void setup() {
 int lp;
 int pin;


 for (lp=0;lp<MAXBANDS;lp++) // set up each bandswitch position
        {
             pin=SWBAND[lp];        
             pinMode(pin,INPUT_PULLUP);      
             digitalWrite(pin,HIGH);
         }
        

}

void loop() {
  
      CheckBand();
}


//************************************************************************************  Band Switch Check
        

void CheckBand()
{
    unsigned char lp;       // temporary variable we will use to step through each of the bandswitch pins
    int bandpin;            // and temporary variable to store the reading from the current bandpin

 
    for (lp=0;lp<MAXBANDS;lp++) {
        bandpin=digitalRead(SWBAND[lp]);      // read the bandpin
        if (bandpin==LOW) {                   // if it is low we do a small delay before checking it again - this is a dodgy way of doing a debounce
            delay(5);                         // might need to change this to 50
            Serial.print("Pin low detected: ");Serial.println(lp);
        }           
        bandpin=digitalRead(SWBAND[lp]);      // after the delay, we get the pin state again, so we can see if the state has changed

        if (bandpin==LOW) {          // okay, if we are still low, then it is probably safe to assume that the switch is fully engaged, and the contacts have settled.
            if (lastband!=lp) {       // check to see if we were already on this band
              Serial.print(lp);Serial.print(" low and not lastband (");Serial.print(lastband);Serial.print(")");
              lastband=lp;            // we were not, so we need to set lastband to this band, so the next time around we won't actually change the frequency
              rx=rxfreqs[lp];         // and finally we can change the frequency
              Serial.print(" new band: ");Serial.print(lp);Serial.print(" new freq: ");Serial.println(rx);
              break;                  // the break here drops us out of the loop - after all, we have changed frequencies so we don't need to waste time checking the other switch positions
            }
        }
    }
    delay(100);             // I've added this line as you had a delay at the end of the change band code previously - not sure if its required or not? 90% sure it can be removed
}

 This project has five bands using the code.




73's
Pete N6QW

Wow what revelations about the emperor (with miniscule e) from Bob Woodward. It is said: Trump Lied, People Died as I heard quoted this morning! We should all get White Hats with the letters TLPD. Hey all you MAGA guys still think he is your man?

Neutralization.

The word neutralization has many interesting meanings. For Mafia Dons, like the Tony Soprano types, the word has the definition of simply el...