Ken G4APB reports that on 14-Apr-2017 his all-QRP Labs transceiver made the first two-way 4m QSO using the QRP Labs receiver, with G4FKK. To use the receiver kit on 4m it is necessary to drive it with a quadrature LO. VFO kit firmware version s1.04 has a quadrature output mode in which the Si5351A Clk0 and Clk1 outputs may be set with 90-degree phase offset. It is necessary to drive the FST3253 quad switch in the receiver kit directly with this quadrature signal, rather than use the onboard 74AC74 divide-by-4 counter, since 4x 70MHz is 280MHz which is way above the maximum frequency of the 74AC74 (and possibly too high for the Si5351A synth too). 

04/14 18:39        First U3 RX 4m spot? 2017-04-14 18:32 G4FKK 70.092507 -16 1 IO91wi 5 G4APB JO01cl 27 59
04/14 18:45        And the other way on U3 Tx! 2017-04-14 18:34 G4APB 70.092515 -5 1 JO01cl 0.5 G4FKK IO91wi 27 239

Ken provides the following information:

U3 Rx setup


  • Relay switched BPF board and RF routing and link options changed for all positions in parallel (posn 1 NOT in series). You don’t have to do this for 4mteres, I just wanted 6 band coverage.
  • Quadrature direct to Rx module using PHASE 90 setting and CLK1 as well as CLK0 achieved by:
  • tracks cut on Rx module to IC1 pins 14 and 2 and coax feeds from CLK1 and CLK0 to these pins, however, as this stops the RX being used below 3.2MHz, I added
  • Quadrature ‘new’ to ‘original’ changeover relay using AUX feature (pd4 connection) (AUX 0 = original wiring, AUX 8 = new wiring). This was achieved by inserting a small pcb with c/o relay and transistor driver underneath the RX module BPF position. Here an extended length pin breakout board feeds coaxes from the Relay switched BPF board.
  • Four metre (70MHz) BPF passive only gave a sensitivity of around 20uV so I replaced it with a dual gate mosfet , single tuned, amplifier giving a sensitivity of around 2uV, not great but that’s all I could fit into the standard BPF footprint.
  • On 144MHz, sensitivity is around 30uV with no BPF or amplifier. Makes a good ‘home tx’ monitor RX for now until I build a preamp.

Programmed settings

  • Multiplier x1 (Change this to x4 below 3.3MHz)
  • Quad mode ‘Phase 90’, (change to ‘Fixed’ below 3.2MHz)
  • All 15 Preset positions (0-F) used (136KHz to 144MHz), WSPR mode.
  • LINE1 and LINE2 active displaying lots of data.
  • GPS Mode 2 9600 (but see later note)

Setting Problems

  • If an attempt is made to set the Presets while QLG1 is connected, it corrupts the edit screen, allowing odd characters to enter the editing field and be programmed.
  • For instance, the relay selection should only be 0-5, but I have one field (posn E) showing an ‘=’ sign that I can’t edit as it is just skipped over to the AUX edit field (even with QLG1 disconnected).
  • Also, I tried to set in the posn F field a relay ‘5’ but it changed to ‘7’ and although I can change it on screen, the memory does not get up dated

[QRP Labs: this will be investigated and fixed in the next firmware version]

Additional notes

I am using AUX (0/8) to switch a relay to changeover between the original single CLK0 LO mode and the new CLK0 and CLK1 direct modes wiring to the relevant ICs, so I can cover below 3.2MHz (original mode) and up to 144MHz (new mode). Although I can choose which mode of quadrature I want in the Presets by setting the AUX, I can’t set the Presets to change the MULT from x1 to x4 as required nor the Quad mode Phase 90 to Fixed, so I have to do these manually for now.

Also, I suffer severe 1 second pulsing interference from the serial data burst from the QLG1 GPS above 30MHz, so I have to set GPS MODE to 0, 9600 once the REF frequency has been set.

For TX on 4m, I have an early U3 TX module with an OCXO driving a single BS170 PA as a driver, the 'AUX' feature to control a hex decoder chip CD4028 feeding transistor drivers to a set of resistive dividers to an attenuator chip PAS-1 then to a small PA (2N4427) to boost the 10's of mW on the highest freqs. This way I can preset the outputs on 28,50,70 and 144Mhz to similar levels with fixed bias levels on the BS170 and my 2N4427 PA, fed back to a homebrew 70MHz LPF (in a Relay switched LPF board), I get 500mW on 70MHz.

Antenna is a multiband 28/50/70/144MHz dipole in the attic.

Hope this inspires some of you to try things not ‘out of the box’. Don’t give up when it all goes wrong! The Tx above has lost most of its tracks due to extensive re-modifications and has been on fire at least twice but is still giving useful service.