25 May 2013

W1FB Direct Conversion Receiver experiments 1

As a part of my project to build a field radio NVIS system, the scope of which I have described in previous blogs, I have now built a first version of the W1FB Direct Conversion Receiver. This venerable design is fully described in W1FB's QRP Design Notebook so I won't repeat it here. I built my version using Ugly style construction and specifically Manhattan style. This being a pleasure and possible with only a hacksaw to cut the PCB and a pair of side cutters to cut out the 'islands'. See a pic below.

My challenge was that I did not have the same components for the Colpitts/Clapp design Oscillator. So I set about seeing what I could do with the components on hand. Certainly I did not have a T-50-6 material toroid. I used a T-50-2 instead. This core is not designed for temperature stability use but I had no choice. The trimmers I have are also probably not suitable and are SMD type trimmers. In addition the capacitors are all ClassII X7RO types. Oh well I will learn something for sure.

The calculations I did are as per the included paper below. I found a calculator online that helped greatly and eased the tedious maths a bit. I also measured the actual inductor and capacitor values using my L/C bridge. Once I had the oscillator built I then used my MFJ259 to verify that indeed the circuit was resonating in the 40meter band. My half turn trimmer covered from 6.6MHz to 7.2Mhz. This was going to be a challenge to tune!

I started by building the LM386 based audio amp. Verified that it was working by simply touching the input with my finger and hearing hum in the earphones. I then added the Audio preamp and conducted the same test. I certainly could hear added gain and louder hum in the earphones. The components were all as per W1FB's design except the speaker coupling capacitor which is a 100uF unit. I supposed that the low frequency response is less as a result of this.

I also have not yet built the audio filter, figuring that the mission was to just get a platform design working.

My design has a smaller inductor of 1.38uH and is inherently less stable.

When I first turned on the rig I heard nothing. I could peak the noise on the tuned circuit so I figured that it was working. I then connected my 2meter J pole to the input and could hear a few very low level signals. I then transferred to the freezing patio outside where my 40m efhw is located. Man was it cold and now 2am in the morning. But I had to see if I could hear anything.

I connected my efhw and put in the earplugs. On turning on the rig I was greeted with a very loud blast of sound! I could not believe it! On turning the volume down to a minimum and tuning the oscillator with a tuning stick I was able to copy loud CW and SSB signals coming in from Europe and the US. Excellent! Over the next few days I tested the rig at all times of the day. During the mornings there is loud local broadcast interference. During the afternoons the rig works fine except that the bandwidth is very wide (not surprisingly). The dynamic range of the NE602 seems very good. Tuned to the CW section of the band I can hear local SSB up frequency. So there is plenty of room for improvement. Adding and audio filter will greatly improve daytime operation. I am less sure about how to cure the BCI however without adding excessive filtering on the front-end. Microphonics seem under control and also there is no hum. all good. There are 4 areas for improvement in the following order of priority.

  1. Improve the VFO. First arrange the capacitors to allow narrow band coverage with the trimmer. This is going to be tough since only a few picofarads are needed to tune the 40Khz CW band. Second I can do some things to the physical construction to improve stability. Mainly securing the inductor and windings better. Having said this I am very happy with the short stability as is. Once the rig is tuned on frequency there is no drift discernable during the period of  a typical QSO.
  2. Improve the selectivity by adding an active audio filter. This filter should be optimized for CW reception and should eliminate the higher band ssb interference from loud local stations.
  3. Improve the front-end selectivity by perhaps adding a double tuned front-end parallel circuit. This should help eliminate the SWL broadcast interference when the band goes long in the evening.
  4. Put in a BCI filter and shield the receiver to eliminate the very strong BCI that swamps the radio in the mornings. 
W1FB DC Receiver
ZS6RSH version 5/24/2013

No comments:

Post a Comment