PA3AJR - RX RF Board for the Yaesu FT-221 R / FT-225‌ RD

After I obtained my amateur radio C-license in 1977, I naturally had to get a transceiver right away. It became the Yaesu FT-221 R, a 2-meter all-mode transceiver that I thought was far superior to everything else available at the time due to its specifications, mechanical construction, and modular design. A beautiful device. Yet, I had to part with it rather quickly. Since my true passion lay in the shortwave bands, I immediately began working on Morse code to obtain my A-license. A Kenwood TS-430 S came along, and the FT-221 R was sacrificed for it.

But as it goes with true love, the regret lingered.

Forty years later – and after many searches on eBay and Marktplaats – it finally happened. For a price of €85, I managed to get my hands on an immaculate FT-221 R, along with a matching style speaker I had never seen before. A black vintage SP-101? According to the seller, they had both spent 40 years under a blanket. And indeed, it looked better than the one I parted with back then.

The design, with its eleven plug-in cards, makes it particularly suitable for revising parts of the device while leaving the rest untouched. Although it still holds up to reasonable standards today, the RF receiver board is the first candidate for an upgrade, as the receiver's sensitivity could be improved.

The Yaesu PB-1456 'RX RF board' and its replicas

The original Yaesu board

In the photo beside this, the original RX RF Unit from my FT-221R. The receiver sensitivity is specified at 0.5 µV @ 10 dB S/N for SSB and CW, and 0.75 µV @ 20 dB QS for FM. The modest sensitivity is partly due to the noise figure of the MOSFET at the input (a 3SK51). Additionally, four varicaps, used to align the narrow band filters with the band selector switch, contribute significantly to the noise. The overall design doesn’t excel in terms of strong-signal handling either. Forty years ago, different standards would have applied to these specifications.

The (G3SEK) MuTek-board

I remembered that there was much praise at the time for the so-called MuTek board. So I first looked on the internet to see what could still be found about it. Already in 1983, Ian White (G3SEK) designed a high-quality replacement RF receiver board. It was marketed by MuTek Ltd from Oxford, UK, under the name "RPCB 144ub FT221/225 front-end board". Translated from English, Ian cited the following reasons for his initiative:

To address these issues, the input stage of his board is equipped with a low-noise dual-gate MOSFET (a 3SK88 with a noise figure of approximately 1dB). In addition, there is a Chebyshev band-pass filter that covers the entire 4 MHz wide range, eliminating the need for narrow-band tuning with Varicaps. Furthermore, there is a Minicircuits diode ring mixer (an SBL-1 with a 1dB compression point of +1dBm), a number of 10.7 MHz amplifiers operating at a slightly higher voltage than the original board, and there is a 6-pole crystal filter. All in all, a significant improvement over the original design. The only problem is that the board was only on the market for a limited number of years. It is now a collector's item that people are still looking for. Not surprisingly, new designs have been created.

I have found two of them.

The IZ4BEH board

In 2008, an article by IZ4BEH appeared in the Italian magazine Notiziario VUSHF (July 2008, p. 2 and following). The main modifications compared to the MuTek front-end were that the 3SK88 at the input was replaced by the ATF54143 GaAs pHEMT from Avago, and the oscillator level was increased to accommodate 17 - 23dBm LO mixers (such as the SBL-1H, SRA-1H, and RAY-1). Judging by the schematic, I wonder if that 23dBm could actually be achieved in practice without compromising the 50 Ω termination across a wide frequency range of the LO mixer.

The IZ1DYE board

Another design came from Michele IZ1DYE in 2010. It shares many similarities with the previous one. For unclear reasons, an extra ERA-5 MMIC was added between the LNA and mixer. Furthermore, the board primarily uses SMD components.

The PA3AJR board discussed here

As far as I can tell, neither of these successors to the Mutek board has ever been commercially available, and no new designs have appeared since. So I decided to put something together myself! Creating a new design from scratch isn’t too much of a challenge with current PCB design and production options. It also offers the opportunity to make further technical improvements, select modern components, and use up some parts that were already on hand. This led to the board shown in the photo next to this text, and I want to mention upfront that I have no intention of bringing it to market as a commercial product. However, I’m always willing to assist anyone who needs help with words and deeds.

Design Principles

Below are some considerations and points of attention that played a role in the construction.

• No irreversible modifications to the transceiver itself

  Precisely because the set has remained in its original state for 40 years and its modular design encourages it, completely swapping out the board is an excellent option to keep the device fully intact. A simple modification I did make, but which is separate from this RX RF Unit, is a modification in the PLL unit as described by SM5BSZ. This change aims to limit the phase noise of the LO signal. It is well worth the effort as it tackles another bottleneck in reception quality and is not a major intervention.
• No surprises in the final outcome

  When a prototype doesn’t go much beyond a pilot series, it is crucial – given the long lead times for obtaining parts – to avoid critical circuits in the design and keep things predictable. Preferably no trial and error in the final product!
• It must remain affordable

  The entire design (including board production) should not cost more than a few tens of euros. The goal is to bring Yaesu's museum piece into the modern era, but I don’t foresee myself returning to frequent use of the 2-meter band. It needs to remain reasonably proportional. The joy is in the design process, not necessarily in the usage.
• Compatible with various crystal filters

  

  The PCB has enough space for a "full-size" crystal filter found in many VHF/UHF transceivers (such as those from KVG, ITT, Dantronic, Salford) which I had on hand. However, there is also a range of decent and very affordable 6-, 8-, or 10-pole filters in a 23 x 12mm SC-4 package (with a 13.4 x 5mm pin spacing), like those from Toyocom (10M5F3 50) and ECS (ECS-10.7-D). Prices on eBay range from €5 to €15. It’s easy to merge the footprints of both types. I also accommodated the minimalist approach seen in the MuTek and IZ4BEH boards: single or double monolithic filters in an HC-49T package (e.g., 10M15A at €2.50 for 5 pieces). Thus, there are four options just in terms of physical layout.
• Compatible with various mixers

  

  We are now on the second version of the board, which includes – in addition to a footprint for traditional through-hole mixers – a footprint for a range of SMD mixers. Both variants are available for different LO levels. For further details, see the Mini-Circuits selection guide and the technical description below.
• Testing and alignment provisions

  Around the mixer – on the top side of the board – test points are included to measure or inject RF, IF, and LO signals. They are cheaper and much easier to use. To align the board outside of the device, a loose connector is used as a test jig (Card Edge Connector RA805 series, 18-pin, 3.96mm pitch). Once inserted in the device, however, the board is not easily accessible. To allow measurements and adjustments, an Extender Board was designed alongside the receiver PCB, featuring a double row of pins to support other PCBs in the transceiver. Since this extender may be useful for many other devices, details and production files for this board can be found in a separate topic.

  

  

• Attention to Good Availability in Small Quantities of the Components

  The circuit board primarily uses (603) SMD components. An overview of the components (BOM) is available in the attachment, with Mouser part numbers suggested. A few exceptions apply. The 10.7MHz coils/transformers are standard 10x10mm TOKO (10K series) models, often identifiable by a pink, purple, or orange tuning core, and can be found on eBay and AliExpress. For the 145MHz RF and LO coil forms, I initially used 'Neosid 10 V 1 CoilAssemblies' but later opted for pre-made "Molded coils" from AliExpress. For details, refer to the description of the "144-148 MHz bandpass filter."

Block Diagram

Technical Description

The circuit will be reviewed below with design details highlighted. A high-resolution PDF of the schematic is available in the attachment of this article. Relevant sections will be featured throughout the technical description.

ESD/Overload Protection, LNA, LNA Bias, and Bandpass Filter

The RF section of the receiver board includes ESD/RF overload protection, an SAV-514+ LNA, a corresponding bias circuit with a 78M05 LDO regulator, and a bandpass filter. These elements will be discussed in order, using the schema fragmentabove as reference.

ESD and Overload Protection

Directly at the input, two BAS70 diode pairs provide ESD protection without significant noise figure degradation. Schottky diodes were chosen for their 'soft-clipping' characteristic, offering a mild effect on intermodulation. An additional three pairs are placed at the LNA output to protect the subsequent mixer and crystal filter from relatively high RF voltages that may occur due to the high linearity of the SAV-541+.

The SAV-541+ LNA

Initially, I considered an LNA based on an Avago (now Broadcom) GaAs HEMT, the ATF-53189, but I soon had reservations. This was mainly due to the critical input network and the (high Q/costly) trimmer commonly used in designs. Biasing the gate is also challenging due to the high local impedance. Furthermore, this FET has a less favorable reputation for ESD, lightning discharges, and overload from TX power. Fortunately, an excellent alternative is now available: the SAV-541+ from Minicircuits. This (Enhanced-) pHEMT features an integrated “noise matched” 50Ω input/output network, approximately 30dB gain, a noise figure below 0.25dB (!), and +33dBm IP3. With superb specifications, it is unconditionally stable and easy to implement, all for under 2 euros. The 30dB gain also makes it unnecessary to include an extra MMIC as in the IZ4BEH design. The concept for this LNA came from a publication by Gyula HA8ET, who at some point also switched from the ATF-53189 to the SAV-541+ for VHF preamplifiers. Ultimately, this led to the following setup.

To achieve stability over a wide frequency range and optimize the noise characteristics of the SAV-541+, inductive source degradation was applied, following Gyula HA8ET's design. As shown in the "magnifying glass" detail of the layout, the source of the FET is not directly grounded at its pin but connected through a rectangular shape and several vias to the bottom layer. This introduces a slight inductance and negative feedback that secure the goals mentioned above. The air coil at the input is not particularly critical. Construction details are available in the schematic, with further information in HA8ET's article.

The BIAS Circuit

The SAV-514+ is powered by its own 5V voltage regulator. Initially, the 78M05 LDO was powered from 13.5V (on pin 3), dissipating about 600mW. To reduce heat in the compartment, a later decision was made to also allow the 8V RX connection (on pin 11) to be used. This reduces dissipation to about 200mW, decreasing temperature rise and thus reducing the risk of drift in the tuned circuits. For now, it has been decided to offer both options via a jumper. The drain voltage is around 4V, while the drain current can be set to (nominally) 60mA via P1. This results in a voltage between the test points TP1 and TP2 of 75mV, including the current through the bias circuit.

The 144-148 MHz Bandpass Filter

Directly after the LNA, there is a 144-148MHz 7th order Chebyshev bandpass filter, consisting of three parallel circuits with top coupling and capacitive voltage dividers at the input and output. The top coupling (nominally 0.7pF) uses two capacitors in series to achieve values under 1pF and allow a bit more fine-tuning. The LC Filter Design Tool by Marki was instrumental in designing the filter. Below is the result of the passband characteristic, measured from the input of the PCB to the ring mixer. In the second version of the PCB, molded coils will be used.

In the left image, the -50dB points are at 130 and 165 MHz, respectively. The right image focuses on the 144-146 MHz frequency range. The attenuation is less than 0.5 dB at the extreme edges. For the European band allocation, the filter is centered at 145MHz, though it can also be adjusted to 146MHz to cover the full 144-148 MHz range. A 3dB attenuator is placed after the Chebyshev filter to provide a broadband termination for the mixer input port.

For the practical realization of the first version of the PCB, I initially chose "Neosid 10 V 1 Coil Assemblies" that I had on hand (both for the bandpass filter and the tunable coil in the LO-Amp). See the photos below.

Although Neosid still offers the coil sets, they are now less common, harder to obtain, and relatively expensive. Additionally, they require winding, which I find too fragile for mechanical variations and temperature changes without an adhesive that maintains the Q factor. These drawbacks are avoided by switching to ready-made 'molded' coils. I chose Chinese look-alikes of the Coilcraft Uni-7 series. See the photos below.

The coils labeled 950A4.5T cost around €6.00 for ten, including shipping. They come with both a ferrite and a brass core, allowing inductance to be adjusted over a wide range (measured: 54-83nH with brass core and 84-240nH with ferrite core). In this design, the brass core is chosen. The molded coils fit neatly into the original Neosid housing or into discarded 10K series TOKO (10x10mm) housings, although this appears unnecessary. The second PCB version will continue to allow for Neosid's use.

Mixer and Post-mixer Amp

The traditional so-called "plug-in" models from Mini Circuits (such as the IE500, SBL-/SRH-/RAY-series) all have the same connections, a Casestyle "A0x" footprint, but differ in their ability to handle strong signals and the associated required LO power. For the SBL-1+, this is 7dBm, while the SRA-1H+ requires 17dBm. After amplification in the LO Amp - which will be covered later - around 23 dBm is available. The excess power is dissipated in the PI attenuator between the LO Amp and mixer. The resistor values for R38, R39, and R40 can easily be determined with an online "PI attenuator calculator". For a 7dBm mixer, this would result in 68, 150, and 68 Ohms, and for a 17dBm unit, 150, 39, and 150 Ohms, respectively. As mentioned in the introduction, in the second version of the PCB - in addition to the footprint for through-hole mounting - a (98-PL-05x) footprint for Casestyle CD54x and CD63x SMD mixers is also included. Unlike the crystal filter, it was not possible to overlay different footprints. Therefore, they are not 'merged' but placed 'side by side'. Both are included in the schematic, but only one configuration is actually populated. Incidentally, the prices of the ring mixers are only a fraction of what they were in the early 1980s.

It's essential that all three ports of the mixer are reasonably terminated over a wide frequency range. This is achieved on the RF port with a 3dB attenuator. The port most sensitive to correct termination is the IF port. The matching network there serves multiple functions. First, the 50Ω of the (10.7MHz) IF signal needs to be transformed with L8 and C19 to a value acceptable for the following MOSFET (approximately 39k from R14 in parallel with the gate capacitance of around 2pF). Additionally, the diodes in the mixer need a ground potential to prevent intermodulation. This DC return is provided by choke L7. Finally, it's important that the VHF and UHF products find a termination in the mixing process, with the main component being the (2f_rf + f_if) image frequency at around 280 MHz. The RC network R12-C18 provides an impedance that gradually approaches 50Ω as frequency increases.

L8 and L9 are standard Toko 10.7 MHz 10x10mm transformers where only the primary winding is used. The capacitors on the bottom are removed and replaced with external SMDs on the PCB, allowing for adjustments later if necessary. The mixer post-amp is a low-noise BF998,235 dual gate MOSFET in a SOT143B package. The voltage gain is around 20dB. The AGC on gate 2 can reduce this to about -3dB at 0 volts. Warning: there is also a BF998R, but the 'R' stands for 'reversed,' meaning it would need to be soldered upside down on the PCB!

Test points are available around the mixer for measuring or injecting signals.

Crystal Filter, AM/FM-SSB IF Amps, and Noise Blanker

As previously mentioned, a range of possible 10.7 MHz crystal filters has been considered. The specifications of the discrete crystal filters I had on hand could not be determined. Based on similar units, frequency, and bandwidth, I assumed an impedance of about 900Ω // 20pF. Monolithic units typically have a higher impedance, up to 4kΩ // 1pF. Since I found it unsatisfactory to guess, I looked for a method to determine it. I came across a method by Kerr Smith on This email address is being protected from spambots. You need JavaScript enabled to view it. with reference to a post by Chuck WD4HXG on the eham forum. Based on this, I created the following test setup and conducted some measurements.

IF Output Amplifiers and Noise Blanker

The output amplifiers for AM/FM and SSB/CW are nearly identical. Both are based on a BF998 MOSFET, share the DC bias for gate 1, and each has a TOKO 10.7 MHz transformer to transform the drain output impedance to 50Ω. Only the SSB/CW amplifier, however, is equipped with AGC on gate 2 and uses a BA591 for the noise blanker function. The primary role of the amplifiers is to transform the approximately 1kΩ impedance at the crystal filter output to 50Ω. For this reason, the voltage gain is a modest 8dB. Under AGC influence, the gain of the SSB/CW amplifier can be reduced to around a 10dB attenuation.

LO Amp

The VCO signal is fed into pin 17 of the board from the set’s PLL unit. The frequency ranges from 133.3 to 137.3 MHz, with a measured level of 330mV_RMS in my set. This is insufficient to drive a diode ring mixer, which typically requires between 7 and 17 dBm, depending on the type. Level-23 mixers are not considered for this application. The MuTek board used a BF274 transistor to amplify the signal, while IZ1DYE and IZ4BEH chose a BFQ18 and BFQ196, respectively, both in SOT89 packages. In the first version of the board, I opted for a BF591G, a robust transistor in SOT223 packaging. However, I later found that it had long been out of production and that the versions sold on AliExpress and eBay were apparently counterfeit. They worked adequately, but I have decided not to source semiconductors from China anymore as they have frequently turned out to be fake. The BFU590Q is a good "drop-in" replacement and is readily available. Due to slight differences in specifications, I conducted SPICE simulations for both transistors. These simulations revealed that a minor impedance adjustment at the input provided a much higher gain and that the output network offered little additional value and could be simplified significantly. The transistor's output impedance is transformed to 50Ω with a tuned circuit in the collector, followed by a capacitive voltage divider and a 3rd-order Chebyshev low-pass filter at 140MHz to keep higher harmonics out of the mixing process.

For the coils L12 and L14, wirewound-on-ferrite 805 SMDs have been chosen, but the layout also allows the use of air coils to provide some more tuning flexibility. (Especially for L12, this might be relevant to account for the difficult-to-predict output impedance of the PLL unit in the set.)

According to the excellent calculator by ON4AA, we should be fairly close with 4½ turns of 0.4mm wire over an inner diameter (drill bit) of 3.5mm and a coil length of 3.2mm (corresponding to the footprint).

Another option that fits in the same footprint is the MD0505-3.5T Molded coil with a brass core. They can be found, among other places, on AliExpress and have an inductance measured by me of 48-70nH.

Installation

Only a minor modification is required to make the board work in the FT221/225. This involves connecting pin 3 to the 13.5V supply in receive mode. The original board operates at 8V, which limits dynamic range. A detailed description of the procedure can be found in the MuTek manual or those by IZ1DYE or IZ4BEH in the appendix of this article. After installation, it’s necessary to recalibrate the S-meter’s zero and maximum readings. As mentioned earlier, it’s also worth carrying out the PLL unit modification as described by SM5BSZ.

Final Remarks

The second version of the board is now ready for production. In addition to the changes discussed above, there are a few smaller adjustments. The miniature coaxial U.FL connectors on the test points have been replaced by through-hole test beads, as they are cheaper and, in fact, easier to use. All pins that do not carry RF signals (13.5V, 8V, AGC, and NB) have been fitted with preventive RFI filters consisting of a ferrite chip bead and a decoupling capacitor. A mix of 603 and 805 footprints had emerged for the SMD resistors and capacitors; this has now been standardized to 605 unless there were good reasons to deviate. The parts list (BOM) has also been completed, with attention to component quality. High-Q capacitors are used in the LNA, standard thin-film resistors, NP0 capacitors in the signal path and tuned circuits, and low ESR MLCCs for decoupling. The layout has been critically reviewed to improve logical organization. All relevant design documents are available in the download area, and a 3D animation of the board can be found in the gallery below. Photos will be added when available.

Downloads

Illustrations

Version v2

Version v1