Tuesday, December 8, 2015

Simpleceiver ~ Part 19

Time for a bit of a Review ~ Simpleceiver
 
It is always good to take a step back to review where you have been to see where you need to go. The Simpleceiver project so far has taken us from a Direct Conversion Receiver to a fully functional Superhetrodyne Receiver with a crystal filter. Our next step will take us to building an accompanying transmitter so that the project will morph into a fully functional 40 Meter SSB transceiver. The project was broken into modules so that each module could be built and tested before moving on to the next stage.

A backbone to this project has been the extensive use of LT Spice and my attempt to provide a detailed documentation of virtually all aspects which hopefully has been useful. I must confess to receiving a bit of criticism to such an approach. An email read in part "stop the blabbing and just provide me a schematic and parts list. I know how to solder two wires together." So to that end I must duly apologize to others who have felt this way but perhaps have not vocalized such an input.

A word here about parts lists. I do not provide parts lists based on some bad experience I had with a publication of an article in QRP Quarterly. I was asked to provide a very detailed parts list for a project article using only two or three suppliers. That effort took longer than to design the radio, build it and then write the article. After publication I received an email asking where to buy two 10K 1/4 watt resistors. My response was to find a local Radio Shack and make that inquiry of them. The return email "Hey you didn't give me the Radio Shack Part Number". I did not respond to that email nor will I ever make a parts list again --sorry folks but You Do Have To Do Some Of The Work!!!! 
 
In my approach to radio receiver construction, the first module to be built would be the audio amplifier stage. Earlier in Part 4, I presented three designs any of which will work. For the hardcore, dyed in the wool*, "I must homebrew everything or it is not a true homebrew radio", then the discrete component version would be your obvious choice. For those who want plenty of reserve audio and have no concerns about using black boxes (IC's) then the NE5534 driving the LM380 is the amplifier of choice. [* Hopefully I used the correct form as inputted to me by the pedantic observer.]

The current configuration of the Simpleceiver has the IC approach. By building this circuit first, those new to homebrewing can develop the skills necessary to proceed with the follow on modules. I am convinced that the discrete audio amplifier should be built once and then forgotten. With that one time build you will learn about every component and its function. But then get over it --you do not need to repeat that experience every time. Based on our experience with the Lets Build Something project --that circuit has too many parts, too many opportunities for screw ups and too little output for the parts invested.

Direct Conversion Receiver

The Direct Conversion Receiver use five modules including the audio amplifier (build first), the product detector employing the J310s configured as a Dual Gate MOSFET, the Local Oscillator (LO), the Band Pass Filter and finally, if needed, an RF amplifier. In the case of the LO my preference is to use the Arduino driving the AD9850 as this has many advantages including the LCD Display and the fact that this combination can be used as a signal generator for building other portions of the radio. The second and third choices in that order are the LC VFO and then a VXO. Below is our Block Diagram for the Direct Conversion Receiver.

 
 
For the hard core discrete component builders who abhor the use of such modern technologies like  the AD9850 or Si5351 and will only use a conventional LC Oscillator or a Crystal VXO feel free to use your favorite deisgns. Many designs abound for LC Oscillators and quite frankly I will leave that to the builder. However I will provide a design for a crystal switched VXO that can be used with the Simpleceiver.
 
Mind you the simple VXO will not give full band coverage and in short order those who opted for this approach will most likely gravitate  to the LC Oscillator or the Arduino Driving the AD9850. Before I receive a rash of emails about VXO's there is a problem with using 5.0 MHz crystals in a VXO --the amount of swing is proportional to the base frequency. So a few possibly up to 10 KHz is pretty normal. Using 15 MHz crystals in a VXO enables a far larger swing especially in a Super VXO configuration.

One prior trick I used for a wide frequency excursion was a crystal switched heterodyne VXO. This circuit involved an 2N3904 simple crystal oscillator operating with several different 6 MHz crystals that could be panel selected and an NE602 that used 12 MHz Crystals in a Super VXO on pins 6 & 7. The 2N3904 signal was fed into the NE602. The resultant mixed signals enabled about 120 KHz at 19.2 MHz for injection into a 20 Meter transceiver with a 4.9152 MHz IF.  Caution you need a Band Pass filter following the NE602 so you only pick off the 19.2 MHz component! A simple panel mounted switched enable two 60 kHz slices with 6 MHz crystals. When using cost as a determinant in a decision as to which approach --you might find this hard to believe but they all cost about the same and that is around $20.

Another consideration for an LO is a heterodyne VTO (varactor tuned oscillator). The circuit below was employed in a 30 Meter CW transceiver I designed and was published in 2013 in a QRP Quarterly article. I was encouraged to write this article based on the fact that readers would love a 30M CW transceiver --well I had zero emails about how much this was loved BUT one thing did come from this article and that was how I piggy backed on another person's work on how to install RIT in an LO.

The circuit below has an RIT functionality that only works on Receive! Basically the VTO operates around 2 MHz and that is mixed in an NE602 with a fixed frequency crystal operating at 12.96 MHz. The IF was at 5.0 MHz. So the output of this VTO was actually at 15 MHz (12.96 + 2.14) and was above the incoming frequency at 10. MHz (30M) and the subtractive mix was 5 MHz --the IF. The diagram shows some constants for keeping the VTO range and by using a 10 MHz heterodyne crystal mixed with the 2.14 MHz VTO produces 12.14 MHz. The subtractive down mix is 5.0 MHz on 40M. There is a tuned network following the NE602 to assure output in the correct range. Somewhat with pride I think this was a bit of innovation on my part --BUT nobody gave a crap given the lack of interest in the project. Purposefully the VTO was chosen for a low frequency (2 to 3 MHz) as it is much easier to treat drift in the VTO. Regulated voltages are a must and you will note that there is not a singular capacitor in the VTO tank network --there are multiple caps so that this minimizes circuit heating in the capacitors and using high quality NPO caps minimizes an capacitance change with temperature. Boiling the inductor in water also puts some "magic mojo" to stress relieve some of the inductor properties.

[I have not modified this VTO to use with the Simpleceiver and these are merely thoughts you should have if you attempt to use this circuit!]

For the DCR we have to supply an LO directly in the 40 Meter band so that filter constants for 40 Meters are already there for you. But how do we get the output to be in the 7.0 MHz range? We have several choices and the first is to use a 9.216 MHz crystal (a standard value) since a subtractive mix is 9.216 - 2.14 and that gets you there picture perfect. Another option is to use the 10 MHz crystal and then move the VTO up to 3 MHz where the subtractive mix is 10 - 3 = 7. In either case you must use the 7 MHz tuned network on the output of the NE602.

If a builder would want to employ this approach for the Superhet then two things have to happen: 1) and crystal/VTO combination must result in a tunable 5 MHz output and the output network must be changed to 5.0 MHz. Just thinking ahead here --there is a sound reason to move the VTO to  range 3.4 to 3.7 MHz and use an 8.5 MHz heterodyne crystal (a standard value). Here is why --the possibility of unwanted mixing products. with this range and because we are up mixing to the 12.096 MHz IF (keeping in mind this is tuning backwards -- the higher the LO frequency the lower the received frequency). With the VTO at 3.4 MHz the subtractive mix is 5.1 MHz and the up convert to 12.096 means a received frequency of 6.996 MHz and with the VTO at 3.7 MHz the upper received frequency is 7.296 MHz. The concern is that the VTO frequency would drop to 2.5 MHz as the second harmonic of that frequency is 5 MHz which could pass through the tuned network. At 3.4 the 2nd harmonic is 6.8 MHz and at 3.7 the 2nd harmonic is 7.4 Mhz which is also outside of the ham band. You must always use care when making frequency selections. Again this is presented solely as an idea piece and was not modified for use with the Simpleceiver



Now for you dye in the wool only use discrete component builders you should be salivating at this point! But that is a lot of hardware that could easily be done with an Arduino and a AD9850!

The following are the remaining schematics for the Direct Conversion Receiver:







In Part 19 we will recap all of the schematics for the Direct Conversion Receiver and Part 20 will be dedicated to the Heterodyne version. Mind you most of what is done with the DCR will find its way into Part 20.

That's it for the DCR.

73's
Pete N6QW