Sept 30, 2024. Spice up your rigs with a FAT Color display.

A hallmark of many of the current homebrew rigs is compact size aimed at portable operation and the ever-popular POTA / SOTA events. Half Size OLEDs used as an SDR Spectrum Display are a reality. 

But going the other direction for those of us whose eyesight can no longer spot a 29A from a 44DD at a half a mile, then we need bigger displays. The next couple of postings will cover the use of larger displays such as those that employ the ILI9341 scheme. Note you can see the pixels on the display. 

Feast your eyes on this display using the ILI9341. I should add that I have several of these displays ranging in size from 1.8 inches to 3.2 inches. The pinouts and the code are the same for either. This is the 3.2 Inch display.

ILI 9341 on a Rig

Resist simply taking your current display using the ST7735 wiring and "plug er in". You risk smoking something as the ILI 9341 Display inputs like to see no more that 3.3 VDC. Thus, you need a CD4050 level shifter or a simple voltage divider, to limit the input.

A resistive divider, a simpler approach is shown as connected to the Arduino Uno. But now crisis mode as you only have an Arduino Nano and so how do you wire that.

There is no connection to Pin 9 on the Display. 

My P3ST rig which I think can still be purchased from K7TFC as a kit was never moved from the wooden breadboard and it looks like hell! So, moving to the ILI9341 will not only give me a display that is easier on the eyes. It will also be a chance to make the 7-transistor rig look uptown.

But this is just a start as the initial Hybrid Wireless code was for a single frequency display. The next steps will add in two frequencies with memory and the tune up tone. The original use of the 3.2 Display was on the Hybrid Wireless that had a 12AV6 tone oscillator. 

The 1st task is to build a new digital LO/BFO with the ILI9341 and then add the changes mentioned. Something that caught my eye early on was that even though the display was much larger the frequency changes seemed seamless and smooth. There did not seem to be a latency of you move the tuning knob and you await the display to change. Look closely at the video.

TYGNYBNT. Don't eat your pets. Speaking of that, in 1965 while temporarily stationed on Okinawa (at that time call sign prefix KR6) we were cautioned not to order burgers at the local A&W Root Beer stands as they might contain dog or cat meat. Yuk! Maybe Vance while in the USMC was stationed on Okinawa. Ohio, Okinawa -- they both start with an O, so who knows maybe a distortion of location.

73's

Pete N6QW

Sept 29, 2024. Where in the Hell is Coober Pedy?

AI is at work on my Pixel 7 when stuff shows up on the phone like the town Coober Pedy.

Well, firstly Coober Pedy is the Opal Capital of the world and so that would put it in the middle of nowhere in downtown Australia.

At first, I thought it might be a CIA site like Pine Gap near Alice Springs, but it is not. Other notable info is that it is so hot there that many folks live underground in what is known as dugouts. They even boast dugout motels!

Dugout Motel ~ Comfort Inn Coober Pedy

[Flash News: A friend in the CIA advises that CP is the home of Fred and Ethel of the AI duo. You heard it here 1st.]

There is a real issue in Coober Pedy because so many of the opal mines are shallow and located not too far underground. Many of these mine shafts pass right under the city. Thus, a high risk of sink holes

I grew up in Arnold, Pennsylvania and the same issue with the coal mines. One year a whole residential city block started to collapse into an old mine shaft. 

But you tuned into this blog for radio stuff and so we look at another of my past projects which is a rebuilt Version 1, of the Shirt Pocket SSB Transceiver. The rebuilding was a result of an unfortunate accident were most of the original transceiver got smoked!

 

The heart of this rig is a 4 Pole Homebrew Crystal Filter at 4.9152MHz with the 2N3904/2N3906 Plessey Amp Module either side of the filter. The Balanced Modulator /Product Detector and the Rx/Tx Mixer stages are TUF-1's. The Mic Amp is a 2N3904 and the Audio Amp is a 2N3904 plus LM386. Another Plessey circuit is used as the Rx RF amp and Tx Pre-Driver. A 2N5109 is the Driver Stage and the IRF510 is the Final Amp.

Literally not much hardware but the "fun factor" is off the scale. The Shirt Pocket Transceiver was the subject of a QRP Quarterly article.

Original V.1 Shirt Pocket SSB XCVR

We are only limited by what we think we can't do! Get off that couch and build something.

TYGNYBNT. Don't eat your pet.

Disclaimer as no AI was involved in producing this posting.

73's

Pete N6QW

Sept 28,2024. Artificial Intelligence Impacts Ham Radio

Leave it to Google and their AI Application called Notebook LM to stir up the brain neurons. Several ham friends (N2CQR and KK4DAS) have been dabbling with this AI application -- so yes, I had to try it!

But first a sewer backup. The Google App is billed as an assistant that can take your notes or documents and create a summary for presentation. It can also take a You Tube Video and essentially create a conversation between a male presenter and a female presenter that deep dives into the video. They are nameless so let's just call them Fred and Ethel* and that is what they did, a deep dive into one of my recent videos.

I chose the video on the resurrection of the Swan SW140 radio that was bought as parts only, needing repair. It was a careful choice as it had technical issues aside from a dirty radio. Thus, I wanted to see how Fred and Ethel did on that part. I had to convert the soundtrack to an MP3 and then make it a video so it could be shown here.

The process involved loading the You Tube link and then waiting until Notebook LM does its magic. I was impressed that for the most part what came out was fairly accurate but that is a caution, a big caution, as human intervention is needed to fact check the final product.

You can also create a whole podcast using Fred and Ethel and so that has some interesting aspects. Or in my case of time limitations have the AI write the blog each day. Now we are talking something of interest to me.

Now I can see the concern of the Hollywood crowd -- suddenly our favorite actress or vocalist are replaced by Ethel. I do hope Ethel is not a 300-pound size of Mary Jo.

Original SW140 Video

Technology is moving forward and at this point we must realize that AI is in its infancy and should be regulated. In sweatshops across the world masses of low paid or no paid software weenies are creating AI products to scam us. Caveat Emptor. 

TYGNYBNT. Don't eat your pet.

* Thanks to K7TFC who suggested names for the nameless presenters.

73's

Pete N6QW

AKA: The Indiana Jones of Vintage Radio Resurrections

Sept 27, 2024. Troubleshooting ~ starts by carefully understanding the problem.

For our small band of loyal blog readers, we start with a mystery, a final exam and a real-world electronics problem. These all have an element of troubleshooting.

The mystery which may have come from a logic class is as follows. A man goes into a bar and asks for a glass of water. The bar keep reaches underneath the bar and pulls out a gun which he points at the man's head. A short interlude and the barkeep smiles, the man says Thank You and walks out of the bar.

So, what just happened? Pause! The man going into the bar had hiccups and with the action of the barkeep -- the man was frightened and the hiccups stopped. Thus, the smile and Thank You. A stretch and something in a thousand years you would never have guessed but plausible. At times troubleshooting a problem is the same -- think subthreshold conduction.

Next, we turn to a final exam question. The course was Short Story Creative Writing. The question was: Create a short story that has three elements including a mystery, sex and religion. A really hot YL spends about three minutes writing and turns in her paper. The professor had a look of dismay and said aren't you even going to try to do the test. She says: I'm done. The professor reads her work and beams a huge smile and marks an A on the work. It was but one sentence which read as follows. Oh my God, I am pregnant, and I don't know who the father is! She had the three elements, it was spare, creative and left the reader engaged and wondering.

When we trouble shoot a problem quite often it is a simple issue that is not so obvious. You can make contacts with a one transistor Michigan Mighty Mite and get signal reports just like from a FLEX 6700. Fixing a M^3 can be less onerous than popping the lid on that FLEX box.

The IRF510 Final!

For the one or two nitpicker blog readers who are always looking for errors, one of the resistors is not marked -- it is in yellow so that will make your task easier to find the nit. Its value is not the problem. So, this circuit was built, using all new parts including some surface mount parts.

When fired up the output was almost nil. I rechecked the wiring, insured the bias was hot and set for about 3.5VDC. There was 12VDC on the Drain. I looked for solder splash or whisker shorts and all solder connections seemed solid. 

I had about 5 volts of RF at the input side yet hardly any output. So, what is wrong? Oh, the thumb on the IRF510 was cool to the touch.

 

Fault tree analysis kicked in to overdrive. While there was RF into the stage, there was hardly any output following the 10nF input cap. This cap was an 805 SMD rated at 25 volts. The issue was I applied too much heat in the soldering process, and it was "open". I replaced it with a ceramic unit and boom full output. Troubleshooting took more than 5 minutes, but the open cap was not so obvious at the outset but revealed as a part of going through the usual suspects list. 

When we troubleshoot a problem, it helps to noodle through a list of the usual suspects such as bad parts, improper wiring or wrong value parts. Is there power to all parts of the circuit and how about shorts or as in this case open circuits. This is a fault tree analysis.

TYGNYBNT. Don't eat your pets.

73's

Pete N6QW

Sept 26, 2024. It pays to look at the data!

 

One purpose of this blog is to share some Tribal Knowledge. So aside from providing me a soap box to vent about the ARRL, I also try to highlight portions of stuff I have done or to present information on complete projects.

Sticking out like a sore thumb, the one posting that garnered about 6X the normal daily views was the Jan 26th posting on the 30M CW transceiver. This was a rig I designed, built and had published in QRP Quarterly.  Recently ZL2CTM's Simple SSB Transceiver got about 3X. Consistently most other postings get the usual small number of daily views. 

It takes effort to prepare a blog and in reality, is a great outlet to add a bit of diversion from the daily shifts at the Board and Care facility. This in itself is likely the only reason I continue. 

But given we always want something more than day old bread; I really have to stretch picking subjects that cover a wide spectrum of readers. 

On balance realizing that hardware homebrewing is dying off, I do try to add in some items that would appeal to those who have software chops. The recent Orange Pi posting got a ho hum reception likely because you have to do more than just open the box and plug it in.

So, as I search to find subjects that perhaps come close to the Jan 26th posting views, I do hope those few who read the blog continue to do so. 

A Military Field Radio from WW1

 This falls into the category of HDR (Hardware Defined Radio) most likely pretty crude by today's standards and over 100 years old! The requisite photo to catch blog readers.

TYGNYBNT. Don't eat your pet!

73's

Pete N6QW

Sept 25, 2024. One of the best LC VFO's

Those several words: N6QW and LC VFO's said in the same sentence are much akin to wearing Brown Shoes with a Black Tuxedo.

But if you are so inclined to use Analog VFO's then hunt and find an Analog VFO assembly from one of the early Yaesu Radios, like the FT-101. The unit is small, but more importantly is linear, stable and clean! Yaesu found the magic decoder ring. I have two Yaesu radios (FTdx100 vintage 1967 and a FT-101EE vintage late 1970's) and both VFOs are just as I described.

One of my more successful radio projects was the JABOM* which originally started life as a 17M radio with a homebrew 4.9152 MHz Crystal Filter and VXO. In the Rx Tx Mixer Module, I had two sets of BPF's so with a pin header and Jumper pins I could add in 20M.

That worked OK except I got a critical report from someone I trust who said my LSB component was only down about 25 DB. He is an RF Engineer and could actually look at my signal with test gear he has in his shack.

That did it! I ripped out the homebrew filter, installed an 8.998500 BFO and added a 9 MHz commercial filter and wired everything for 20M. The real plus for this rig was the HyCas IF and AGC. I ditched the VXO and because I had a Crystal BFO the DDS of the day was the AD9850 and that is what went into the radio. 

The AD9850 was a predecessor to the Si5351 and was only good to 40MHz and did not have the capability to generate two separate frequencies. Its output needed filtering and could be somewhat tepid. 

So, after 15 years I embarked on replacing the AD9850 with the Si5351. The current approach with my Si5351 is to have two frequency ranges and VFO memory. So, the replacement unit boots up on either 14.2 or 14.074. A simple toggle switch enables jumping back and forth. If the SSB part of the band sounds dead, I need only flip the switch to see if there are FT8 signals and usually there are. 

Here is a brief snippet of the 1st rough checks to see if we have it working.

Once again building in modules facilitates major changes. It also works well with my limited time available. I had the AD9850 out of the JABOM in about 3 minutes. Luckily, I have a box full of Arduino/Si5351 assemblies wired up and so about the same amount of time to pop in a unit to run some tests. This looks promising. 

TYGNYBNT. Don't eat your pet.

* JABOM = Just A Bunch Of Modules.

73's

Pete N6QW

Sept 24, 2024. More from the Land of Oz (Amazon)

Lauren Sanchez's boyfriend has done it again! Behold a Single Board Computer (SBC) for about $30. The price point is in the range of an Arduino Mega 2560!

Meet the Orange Pi Zero, 2W 1GB with expansion board. A Quad Core 1.5 GHz, 64 Bit machine with Wi-Fi and Bluetooth is sitting on that small board assembly. 

Having a choice of operating systems, it has some real possibilities for integration with your existing ham gear. I mention this since those who are addicted to FT-8 and POTA/SOTA ops, this SBC can do the job in the field! 

To make this work with your existing rigs you will need to build a digital adapter using a few readily available parts that won't set you back a king's ransom. Its small size makes it perfect for dragging along up that small hill so you can send CQ SOTA on FT-8. You will also need a small display and a keyboard/mouse.

[If you want more info on my homebrew digital adapter, send me an email.]

Our hobby is changing and even though you may be inclined to resist the change, there is inexpensive hardware that is now available to you so that cost is no longer an excuse. From a quick look at the specifications, it appears that you could load QUISK and run your homebrew SDR from this SBC assembly.

A scary thought: A desktop computer for $30. Read the reviews as that is important to understand what other users are saying. It seems like some users did not get all of the hardware pieces. Of import while this product may have some start up issues, other manufacturers will be jumping into the fray and soon we will have more choices. A new page is added to our hobby.

You can read about the Orange Pi OS here. Most likely you would need to load the Unbuntu or Debian OS software to use QUISK.

TYGNYBNT. Don't eat your pet!

73's

Pete N6QW

Sept 23, 2024. AGC from W7ZOI~ Hybrid Cascode Circuit

 W7ZOI did a seminal piece on a Hybrid Cascode AGC Circuit. This same circuit was used in my JABOM rig. (JABOM = Just A Bunch Of Modules and published in QRP Quarterly)

The original article for Hayward's circuit was published in QST.

Originally published in the December 2007 QST, Figure 1 is similar to what was presented in yesterday's Blog Posting. Hayward found that this circuit, however, was sensitive to Vdd dropping below 12VDC. 

Figure 2 with a simple change of J2 to a 2N3904 enabled using as little as 8.5VDC to give peak performance. The issue was portable battery operation where a fully charged battery in short order drops below 12VDC.

The full-blown circuit is shown below.

Homebrewers are indeed fortunate as Todd, K7TFC sells the circuit board for the HyCAS Module which uses the J310 configuration and so that removes 60% of the effort to replicate your very own HyCAS.

Below is an actual build from N6QW as used in the JABOM. This was the subsequent refinement by W7ZOI where the J310 and 2N3904 transistors were used. My board also includes the SBL-1 Product Detector located behind a shield.

 

 

I can attest that switching in the AGC is like night and day. It really works but does require more parts than one FET and two resistors.

So, I will stop here with any further talk about AGC and as I said most of my rigs use the Armstrong method (manual adjust of the volume control).

TYGNYBNT. Don't eat your pet!

73's

Pete N6QW

Sept 22, 2024. Automatic Gain Control (AGC)

One of the comments posted on the video shared on yesterday's blog essentially said: What no AGC?

That observation was based on seeing me adjust the audio gain pot as I tuned across some really loud signals. Typically, I do not include AGC in my rigs although several have that feature using circuitry from W7ZOI.

In keeping with the season, I wished we had built in AGC for use on several of the political candidates as they speak about immigrants eating your pets.

To implement AGC takes a whole raft of electronics to do it properly. So, for today's discussion let's talk about AGC. 

Let us start by theorizing that maybe AGC was invented for use in AM automobile radios. It would be hard to drive down the road and to keep having to adjust the volume which could present a distraction. Just my theory not necessarily the real how or why. Sitting in front of a rig presents less of a safety problem.

In essence, AGC levels out the final audio gain so that the volume coming out of the speaker is within a small range regardless of a strong or not so strong signal. The process starts with detecting the very strong signals and adjusting the overall system gain so that we do not have blasting signals coming out of the speaker.

That is a simplistic view as behind the scenes is the detection and measurement to some preset level and then enacting the gain reduction at the speed of light so our ear detection system cannot tell what is happening.

AGC takes many forms including at the RF level, the Audio level only and a combo of both Audio and RF. If you check Hayward's SSDRA you will see a combo of the RF and AF AGC in one of his receiver projects. 

The "audio only" approach tends to present a signal that sounds like it is "pumping" -- a shortcoming manifest in the Sideband Engineers SBE-33. Some simple Audio AGC circuits have a FET switch in the audio chain where biasing the FET reduces the signal passing through to the final AF amp stage.

The RF approach would typically apply gain reduction in the RF Amp and 1st IF Amp stages. Think of it -- the strong signal hits the antenna jack and in literally zero time you are trying to reduce the gain of the RF amp. Not an easy task. Some approaches simply bypass the RF stage and use the 1st IF amp stage as the crowbar to knock down the gain.

One "untying of the Gordian Knot problem" is the enactment of the AGC process without introducing distortion in the signal. A second problem especially in transceivers using a common IF is how to have AGC on receive and No AGC on transmit, or how to have AGC on Receive but ALC in the same circuit on transmit. This is more than a single FET and two resistors!

One thing I noted when working with "making" Dual Gate MOSFET's using two J310's is that you could add a trim pot to Gate 2 to adjust the stage gain. Extending this idea, you could have a variable voltage instead which depends on the signal level from some preset standard (AGC). A small relay could switch between the AGC voltage and a Fixed Gain wide open DC voltage on Transmit or the application of an ALC signal.

 

 

If you make R3 a 3.3K resist0r in series with a 10K trimpot wired as variable resistor you can adjust the stage gain Now substitute a variable DC voltage for the 10K trimpot and you have an AGC controlled stage. You would have to have a negative going AGC to reduce the gain on strong signals. By negative going means that a weak signal would be a higher voltage whereas a very strong signal would be a much smaller voltage.

 

Again, all of the added stuff is a high wire balancing act and unless it is done right is nothing short of a disaster. Thus, an Armstrong approach of manually adjusting the volume is not a severe handicap while operating. 

As usual there are commercial devices now available that likely will find their way into future homebrew rigs. Microsemi has a voltage-controlled attenuation device good from DC to 50GHz that can provide up to 27dB of attenuation.

TYGNYBNT. Do not eat your pets.

73's

Pete N6QW