The Black ADF4351 Boards

 Well,

Some time ago, I did some expermimentation with black ADF4351 evaluation boards that are available at all good sites selling wares from the land where copyright means "lets copy it right".

I've bought some more recently and found that I couldn't get them to work.

Thanks to this post, I was able to deduce that the newer boards seem to be missing some bits - most importantly the CS line isn't tied high by a SMD resistor. Once I had pulled that pin high via 10K to the 3V3 line, the board worked as expected.

I've been using mine with a Pi Pico2:

You can see the additional 10K on the fritzing diagram above.

I have written some very simple code here to upload the register values to the device, and also have a spreadsheet here that will do the necessary maths from a target output frequency, channel spacing and clock frequency and generate the 6 register values needed to program the device.

Good stuff!

Spectrum Analysis

Well,

I've been messing about recently with some inexpensive alternatives to a Spectrum Analyser. I find my Rigol DS-815 one of the best investments I've ever made, it represents excellent value for money. However, its still an expensive purchase and I wondered if the inexpensive alternatives were any good.

First off, I tried one of these that you can source on all good auction sites, here is a link to the device on Amazon:

Without a box

and also:

With a box (if you are posh like me)

They look like the image above and represent exceptional value for money - you couldn't purchase the bits for the price.

Here's my evaluation of the device:

Second I tried an SDRPlay and associated software as a Spectrum Analyser:

Next will be a Red Pitaya... so watch this space.

STM STM32F103RC8T6 - really?

Well,

On popular auction sites you can find these boards for less than 2 GBP delivered:

They contain (on the underside) a STM32F103RC8T6 processor from ST Microelectronics. Looking at my musings from here, these boards will support the necessary double precision maths and also have the SPI interface needed for frequency synthesyser interfacing - amazing!

These will work with the Arduino IDE with a very small amount of effort.

Assuming you already have the Arduino IDE installed (if not go get it), the steps are:

1. Go to the Tools -> Boards -> Board Manager and install the Arduino SAM boards (Cortex-M3)
2. Download the STM32 support package from here.
3. Unzip the download to create the Arduino_STM32 folder
4. Copy the Arduino_STM32 folder to My Documents/Arduino/hardware (Note: if the hardware folder doesn't exist you will need to create it).
5. Navigate to My Documents/Arduino/hardware/drivers/win and run the install_drivers.bat file - right click and run as Administrator
6. Restart (or start) the Arduino IDE and select "Maple Mini" as your board, "Original" as your bootloader
7. Attach the board to the PC with a USB cable, you should see a "Maple DFU" device under "libusb-win32 devices"

You could then try this sketch and upload it to the device; please note that at this point you don't have a COM port for the board - it is using the bootloader via the DFU device instead.

Once the sketch is loaded you will then see a COM port "Mapel Serial" which you can now also select in the Arduino IDE tools -> Port.

I've connected my board to my SV1AFN ADF4351 board as follows:

STM32F103RC8T6 Pin 7 -> SPILEA (pin 3)

STM32F103RC8T6  Pin 6 -> SPICLK (pin 1)

STM32F103RC8T6  Pin 4 -> MOSI (pin 4)

STM32F103RC8T6 VCC -> 3V3 (pin 7)

STM32F103RC8T6  GND -> GND (pin 5)

and have run this sketch - seems to work just fine.

I then ran this sketch - and followed the instructions in the Serial Monitor to upgrade the bootloader - once I'd done that I then needed to select Tools -> Bootloader Version -> 2.0 for any future comms with the board.

More ADF4351 - really?

Well,

You may recal my dissapointment back here with the Practical Electronics GPS-synced Frequency Reference.

It turned out to be a load of Dingo's Kidneys.

You may also recal the most excellent work of Brian, GM8BJF back here using the ADF4351.

Well, I have been playing with the Micromite that was used in the Practical Electronics project, and have created a touch screen signal generator using the ADF4351.

The ADF4351 modules I have are here and this one by SV1AFN.

It's very much stolen from a project here.

I've modified the code a bit to have an additional digital output to show the PLL locked state, and also changed the code to use a 10MHz clock rather than the default 25MHz clock.

It's quite neat!

Albeit upside down:

I'm 3D printing a case for the front, then we can decide what to do with it:

The output is as I expected:

I've uploaded my bodged version of the code here.

After I finished the software mods and fiddled some more, I stuck this all in a box:

You use the touch screen to set the frequency:

and there's an LED for power and another to show the PLL is locked.

I've learnt loads building this and I'm rather pleased with it!

I've done a video on the detail here:

Here is our very beautiful and rather special Miss Florrie Cat:

13GHz Sig Gen - you can't be serious?

Well,

You may recall back here when I became very inspired by a project by GM8BJF using some eBay sourced Chinese modules to create an ADF4351 based signal generator.

There's another variation on that theme that has been published in the May 2019 Scatterpoint. Scatterpoint is the journal of the UK Microwave group, membership is about 3 bob and a conker so you have no excuses for not joining.

This time it uses an ADF5355 to generate signals from 52MHz all the way up to 13.6GHz.

The 4.4GHz generator has proved extremely useful, this one is a must.

I've modified the design and the software a little so that the 100MHz clock is generated by am ADF4351, rather than an OCXO as per Brian's design.

Brian has made the software available here, and my modified version is here. The AD4351 is a version of the evaluation board from SV1AFN - and it gets its reference from my 10MHz shack frequency reference.

The software runs on a SAMD21 ARM Cortex M0 which runs at a 48 MHz clock rate. This is supported by the Arduino IDE but is not something I had used before.

Its still WIP, but the project is up and running OK.

How cool is this?

Here's Miss Maggie Cat and Miss Pepper Cat doing what they do best:

A tracking what?

Well,

As part of the fiddling I've been doing on 13cm, I've been using the new to me (read old) spectrum analyser I have. It covers from about 9KHz up to 22GHz.

You may also recall not so long ago, that I made a signal generator that covered up to 4.4 GHz.

As this spectrum analyser has a 1st IF output socket, it struck me that I could probably make some kind of tracking generator to go with it. Actually the IF output will be doing the tracking, all I need is a signal and a mixer.

Some experimentation allowed me to discover that on the low range, the Spectrum Analyser has an IF output of 3910.7 MHz plus the tuned frequency.

I've made myself a simple Arduino Nano and AD4351 combination:

The source code for the above is here. I've not done anything clever at all, just used the Analogue Devices software I showed here to calculate the required registry values and then hard coded them into the Nano.

That gives me the 3910.7 MHz signal required. We then subtract that from the IF output from the Spectrum Analyser using a simple and small Mini Circuits mixer:

Then I've added a low cost return loss bridge from ebay:

Whilst it's not lab grade, in this example you can clearly see the resonant frequency of the antenna that's connected as the Device Under Test:

The difference between the trace with the DUT socket open (the thicker line) and the other trace is the return loss at the specific frequency.

You can see that the open circuit sweep is nowhere near flat - so there are all sorts of issues with this setup, but as a basic antenna analyser up to about 3GHz this works just fine.

All the while, Florrie the ham cat has been sitting on my rotator manual which I am consulting as the display bulb has died:

Local conditions.

Boxing Time

Well,

You may remember the 35MHz to 4.4GHz signal generator made from cheap eBay components I was fiddling with last time.

I've stuck it in a box now:

I've removed L1 from the board:

so that there is no power to the on-board 25MHz XO. I'm running mine using the shack 10MHz frequency standard I made back here. This external oscillator input simply connects to the SMA socket on the back of the board.

I've recently made a distribution amplifier for the frequency reference so  I have 8 50R outputs at 10MHz to lock a recently acquired Microwave frequency counter and also my HPO signal generator.

Good egh?

4.4GHz Signal Generator - Really?

Well,

I attended the GMRT event this year, one of the highlights was seeing this in the flesh:

I also spotted a great project by GM8BJF which was a great signal generator covering 35MHz up to 4.4GHz using very cheap Chinese and eBay sourced components.

Here are the bits, firstly a ADF4351 evaluation board:

 

Secondly, a CY7C68013A development board:

And finally a broadband amp module:

Total cost of these three modules delivered was less than 45 GBP.

Now, I've tried two ways to drive the ADF4351, firstly using some excellent software from F1CJN - this worked very well and would make an excellent stand alone sig gen.

However, I have decided to replicate the project by GM8BJF exactly and use the three boards above together.

I followed the instructions on his website and have programmed the EEPROM on the CY7C68013A board and am now using the software from Analogue Devices from my PC to set the output frequency:

This software controls the ADF4351 from a PC using a USB connection and allows you to see all the register values and fiddle with all the device config - most excellent!

Here's a quick shot of the spectrum analyser connected to the output with the device set to 1.2GHz:

This is an appalling picture of the boards under test:

And here, Miss Florrie Cat assisting:

Neat, egh?