STEMlab RED PITAYA SDR Transceiver Build Part 2

Now that the Stemlab Red Pitaya 125-14 module has been modified for use as an SDR (software defined radio), it is time to design and build a front end module with filtering, RF amplification and digital attenuation.

The Red Pitaya board has very wideband, high-speed ADC's that will digitize everything from DC up to tens of MHz and while this is desirable for a software-defined radio receiver, it is not without some downsides.

So what are the downsides:

• The ADC sees all RF energy within its analogue frequency bandwidth.
• Strong out-of-band signals separate from those we want to receive:
• Overload the ADC.
• Create unwanted intermodulation products. 
• Reduce dynamic range.
• Mask weak signals you want to be able to receive and hear.   

What is the solution:

Add band-limiting filters in front of the Red Pitaya's inputs to shape and limit the RF bandwidth presented to the ADC. It is crucial to prevent frequencies above half the ADC sampling rate (125 Megasamples per second) from entering the ADC.

As my Red Pitaya has a sample rate of 125 MSPS, the half-frequency value is therefore 62.5 MHz. This value is technically referred to as the Nyquist frequency and it is essential to prevent frequencies above it from entering the ADC.

Practical Implementations:

• Add a Chebyshev or Elliptic Low-Pass filter to the Red Pitaya inputs with a cutoff frequency of 50 to 60MHz.
•  Fit switchable band-pass selection filters to the Red Pitaya Inputs.
•  Add some form of RF limiting to the Red Pitaya inputs.

The Red Pitaya does not contain any input filtering or protection to the ADC inputs so it should be considered mandatory that both filtering and protection are added before connecting an antenna in order to protect your investment!

Design Goals:

Having considered the practical implementations to filter and protect my Red Pitaya board, my thoughts turned to designing a suitable module that would meet the following criteria: 

• Provide strong out-of-band rejection of unwanted signals.

• Limit the level of signals presented to the ADC.

• Provide good dynamic range.

For the rejection of unwanted signals, I decided to utilize the band-pass filter that I designed for my Irwell HF Transceiver.

To limit the signal levels present at the ADC, I fitted a BAV99 high-speed back-to-back  diode pack on the input mod board that I featured in my previous blog post.

I added a 0.5 to 31.5dB  digital step attenuator to control the signal level presented to the ADC and therefore avoid it being crushed. The digital attenuator helps prevent overloading of the ADC,  reduces intermodulation and improves linearity, all of which helps preserve the dynamic range. 

My experiments and prototype work accumulated into a design that I present in the schematic diagram shown below:
 

Click image to enlarge to full size! 

The final design includes the following features:

• ESD protection at the RF input by way of a static "Gas Discharge Tube" and a "TVS diode pack".
• Placement for up to 10 dedicated bandpass filters.
• Option to fit a PE4302 or PE4312 digital step attenuator.
• Switchable low noise RF Amplifier to boost signals aton the higher frequencies.
  End PCB Design Supports the population of various MMIC device types.
• Solid-State RF signal path switching.

• Filter selection by either BCD logic or Serial Data over the I2C bus.
   

PCB design: 

Using my PCB design software, I laid out two PCBs for the project, first a main board  measuring 175mm by 80mm and next a small sub PCB measuring 80mm by 20mm. The sub PCB is used for building the filters, and each filter built requires a sub board.

The image below shows the main board populated with nine filters, all of the toroid cores are Amidon T50 types and there is a lot to wind!

Completed Red Pitaya Filter Module With RF Pre-Amplifier & Digital Attenuator.

The hardest part of the build was soldering the PE4302 digital attenuator IC, as it is a legless device but not in the intoxicated sense :). The device measures just 4mm by 4mm and has 20 exposed solder pads in its tiny QFN package.

 

The PE4302 is now listed as an obsolete part and has been replaced by the PE4312, which at the time of writing this blog post is still in production. It should be noted that the PE4312 has improved specifications over the PE4302 and is a direct replacement.

Here's another image of the module viewed from each end of the board:

End views of the completed Red Pitaya Front End Module.

And finaly an image showing the underside of the mainboard:

The board has been populated with a CD4028 CMOS BCD-to-decimal decoder IC to provide the logic switching for the filters. The CD4028 is the IC that is visible on the right-hand side of the image. Just above it you can see the component location where a PCF8575 16-bit I2C-bus I/O expander can be fitted for serial data control of the filters.

As I mentioned previously, the actual bandpass filters are the same as the ones I used in my Irwell HF Transceiver with the only difference being the switching. The Irwell module uses relays for filter switching, whereas the Red Pitaya module uses solid-state switches.

Relays vs. Solid-State RF Switches:

Each of the switching methods has its strengths and weaknesses, and the right choice depends on various factors that are mainly attributed to; power handling, switching speed, isolation and insertion loss.

A closer look at the four factors outlined helps paint a clearer picture.

1. Power Handling

• Relays
  
  • Can handle tens to hundreds of watts depending on the type used.
  • Excellent for HF and lower VHF frequencies.
  • Zero distortion at RF levels used for amatuer radio.
     

• Solid-State Switches
  
  • Typically limited to handling just a few watts.
  • Excellent for frequencies into the GHz.
  • Can compress or distort when operated near their limits.

When it comes to power handling relays are the clear winner.

2. Switching Speed

• Relays
  
  • Slow with speeds typically around 3-10 milliseconds.
  • Not suitable for fast T/R switching like QSK operation. 

• Solid-state Switches
  
  • Extremely fast operating in nanoseconds to microseconds.
  • Excellent for fast T/R switching.

When it comes to speed Solid-state switching wins by a huge margin.

3. Isolation

• Relays
  
  • Outstanding isolation and in the region of 60-80dB.
  • Ideal for preventing filter leakage or cross talk.
  • Isolation is broadband and conditionaly stable.
     

• Solid-State Switches
  
  • Good isolation and typically in the region of 30-50dB.
  • Isolation degrades at higher frequencies.
  • Leakage can be detected in sensitive applications.

When it comes to isolation relays win again.

4. Insertion Loss

• Relays
  
  • Extremely low insertion loss <0.1dB.
  • Loss is stable across a broad frequency range.
  • Relays are expensive to buy for frequencies into the GHz.
     

• Solid-State Switches
  
  • Higher insertion loss - typically 0.3-0.6 dB depending on frequency.
  • Loss increases with frequency but not massively.
  • Power consumption is typically very low.
  • Operate from DC into the GHz region.
  • Lot cheaper to buy than RF relays in the GHz region. 

When it comes to isolation it is difficult to define a clear winner as it depends on the application and frequency of operation.

5. Final comments 

So which should you use?

I think when it comes to signal routing and the choice of relays vs solid-state switching, you have to weigh up all of the factors and make a decision based on your own application, but as a general rule of thumb, consider the following:

Use relays when:

• You’re switching high‑power HF/VHF filters.
• You need excellent isolation between bands.
• You want ultra‑low insertion loss.
• You care about linearity and IMD performance.
• Switching speed is not critical.
• Longterm reliability is not an issue, relays wear out solidstate switches do not suffer mechanical wear. 

Use solid‑state switches when:

• You need fast switching (SDR, T/R, agile systems).
• Power levels are low (receive paths, low‑power TX).
• You want compact, low‑power, long‑life solutions.
• You’re building a multi‑band SDR front‑end with rapid band changes.
  

If you want to delve deaper into the design process of my bandpass filters or take a closer look at the technicalities, then please check out my other blog posts where I built a bandpass filter for my "Irwell HF Transceiver".  Just click the button below.

In my next blog post, I will build a controller module for the Red Pitaya SDR transceiver using a low cost STM32 development board. But for now...

Project files will be made available via the Groups.io platform by joining my G6LBQ community group, where you can discuss my projects, ask questions and help others.

Joining my group is free. Just click on the button below.

 

 

Until next time...

73's From Andy G6LBQ

It's all about the Radio Ga Ga...