Building a Direct Conversion Receiver: Part 2

In the first installment of this series, we discussed why we’re building a Direct Conversion receiver and talked about some basic ideas. In this installment, we explore what it takes to make the leap from a printed schematic to something physical that works. Follow along!

Schematic Semantics

The DC40 was originally designed by Ashhar Farhan as a receiver that’s easy to construct using parts that are easily and inexpensively available in India. For that he is to be applauded, and his BITX20 design was designed similarly. The parts list calls for things like TV balun cores, nylon tap washers, and 50uF capacitors. Unfortunately those things are rare as hens teeth in the US or even online. Specifications are much too specific for “tv balun core” to mean anything unless one knows a good substitution. 50uF capacitors aren’t a standard value either. What to do? 

I was discussing the ins and outs of the receiver with Robin, AC7LX and she had some great suggestions that I’ll be sharing along the way. The first suggestion she had: Don’t use the published schematic! There are errors in the double balanced diode ring mixer section that will cause some headaches until corrected. She built the circuit in LTSpice and tuned it for best performance (see below).

Robin also adjusted the DC40  so that it can be built with more easily accessible parts. For example, the diode ring mixer transformers are now FT37-43’s instead of TV balun cores. The 50uF capacitors are now 47uF. I am in her debt for her great contributions to this project.

The front end filter (originally built with a nylon tap washer) can be built with a type 6 toroid such as a T50-6 shown on the QRP Parts Catalog. You can calculate the turns for your particular toroid at toroids.info. For a T50-6, 35 turns will work fine for this filter. I will be using a different filter. More on that later! 

What isn’t shown is on the schematic below is the 3dB pad going to the si5351a based QRP Labs VFO. The schematic does include the 10k volume potentiometer, which I already built into the audio amplifier. I’ll leave your audio amplifier up to you. The LM386 circuit shown on the original schematic, or something similar, will work. It’s not critical. In fact, you might even enjoy building one out of discrete components!

CLICK HERE to download the schematic in PDF format

Robin suggested a different band pass filter, a design straight from W7ZOI that is just beautiful on the 40m band. I’m not going to be covering its construction in this series, but it is here for your reference should you choose to build it:

Gathering the Components

I started out by going through the schematic and writing down each resistor and capacitor and their associated quantity. I have a transistor kit that I bought that has many different values, so I just grabbed some 2n2222’s from that, and then went hunting around for the rest of the parts.

As I’ve mentioned in other posts, I’m a bit of an electronics scavenger. I don’t buy anything unless I absolutely have to. There are far too many surplus electronics floating around, and I collect them whenever I can. My favorite find so far was an old Sony TV with a Trinitron tube. It was a high quality TV made in 2000 or 2001, and the components in it were very good. I normally wouldn’t salvage electrolytic capacitors, but these looked pristine and were from an era before the counterfeit electrolytics hit the market. 

A spare FT37-43 was left over from a QRP Labs kit, but the second one… Lets just say I did a bad thing. My U3S will be out of commission until I replace that transformer! 

The resistor kit contained all the needed resistors for the project with the exception of the 50ohm resistor- I just happened to have one left over from another kit. I also could have put two 100ohm resistors in parallel. 

The last thing I needed was the 1n4148 diodes. Don’t be tempted to save cash and order four diodes. You’ll need a bulk amount- I bought 40. We’ll be selecting the four most electrically similar diodes later on.

The diodes aren’t in yet, so I reached out to the parts TV and found 4 unmarked glass diodes that looked the part, and I’ll replace them when the proper part arrives. In the picture below you can see my parts list, along with the copper clad board that I am using for construction. 

The board doesn’t need to be very big, but smaller than 70x100mm would not give you a lot of room to build the receiver. You can also choose to use a through-hole prototyping board such as one of these (click link)

Preparing to Build

For this project I am embarking on several new skills. I’m learning to translate a more complex (although still somewhat simple) schematic into a physical layout, and also learning how to use Manhattan style construction. 

Manhattan style describes using a bare copper clad PCB as a ground plane and gluing small squares of PCB to it as ‘islands’ for components to connect to. It’s a favorite method of radio homebrewers, and so it’s time to give it a shot. 

To prepare for building Manhattan style, I went into my workshop and grabbed a couple of 2×4’s. One was put in my vise, and then laid down a 70x100mm PCB flush with the edge. Next I clamped the other board on top, and left about 3/16″ (5mm) of the PCB showing. I also did a couple of narrower strips about 1/8″ (3mm) wide. 

I used the top board as a fence/guide, and sawed a strip off of the PCB using a regular saw that I got at a garage sale for $1. The copper was facing upward, but next time I’ll face it down so that any scuff marks are not on the copper. Cut several strips to get started. 

You can see the saw marks from previous cuts.

When you’ve go the strips cut, cut one of them into squares. I used my Hakko cutters (see the QRP Catalog for a link) to do the cutting. Be careful, the squares will go flying! 

Strips and squares.

As I mentioned in Part 1, I couldn’t find an LM386 so I used the TDA2822M I had and built a circuit ugly style on the PCB. Before I did that, I decided to dedicate one side of the PCB to a power buss. This was done by the same method as cutting the strip, but the board wasn’t cut all the way through. Here’s how mine ended up (see below).

You can see that I’ve got the red and white wires connected, the red to the power strip and the white to the rest of the PCB ground plane. The power supply is a standard 13.8v (12v nominal) bench supply. If you don’t have a clean power supply, then you might want to get some AA’s together.

You can also see that I decided to put the audio section right in the corner, directly next to the power buss. There’s a reason for that, so read on.

Digesting the Schematic

I must admit that translating the schematic into a physical thing which was to be built by a method I’d never used was extremely intimidating! But I know that many others have done it, and so I felt confident that I’d figure it out. 

Since digestion always starts with chewing, I decided to chew on it a bit. I just read over the schematic, making notes in my mind about which components were physically connected to each other. One thing that really helped me was to draw the pinout of a 2n2222 and label Collector, Base and Emitter (C B E). Then on the schematic, I also labeled each transistor’s CBE. 

From there I needed to wrap my head around the transformers in the diode ring mixer. We’ll get to winding the transformers later, but to simplify it, each transformer has 3 sets of windings. There are three leads on the left side, and three on the right. Each lead has continuity to a lead on the opposite side. Once it’s wound, the leads are checked for continuity, and they’re put in the same order so the top, middle and bottom each have continuity from left to right. If that sounds confusing, don’t worry- we’ll dive more into that in the next installment.

I drew a physical transformer on paper, and labeled the left side A, the right side B, and the windings 1, 2, and 3. Then, I labeled the top connection of each inductor in the transformer on the schematic the same way: 1, 2, 3, with the top connection being A, the bottom connection being B. Then I drew it out a few times until I understood how the two correlated. And lastly, I drew the entire schematic as physical devices, drawing on the paper as if it were a ground plane. You can see my actual drawings and markings in the pictures below. It’s messy!

The proof is in the Putting

Now it’s time to put parts on the PCB and start making an Actual Real Circuit. As was mentioned, the audio circuit is in the corner on purpose. If you look at the original DC40 schematic, the audio amplifier is on the far right. The physical circuit mimics the printed schematic somewhat closely, although that’s not strictly necessary- only the electrical connections matter. But, long lead lengths are usually a sign of poor layout, so we’ll try to keep it reasonably compact.

To place the pads, I cut the legs of the component being used to a reasonable length and placed it on the PCB. Then I marked the spot with a fine tipped marker and glued each pad down using Gorilla Super Glue. A very very small amount of glue is needed! Hold down the pad for a few seconds to make sure the glue sets in the right spot. I used tweezers to place the pad, and a pen or pencil to hold it down. From there, add the first few components:

One thing I noticed right away is that it is vital to do a tug test on each component leg. It’s easy to get a cold solder, and it’s small enough that depth perception was an issue for me, so tugging at the lead with my tweezers really helped. For some resistors, it helps to bend a very small L into the end of the lead to give the solder more surface area to bond to. Be sure to get the PCB nice and warm with your iron before adding solder, and avoid using more solder than necessary in order to keep it looking nice. 

Once I got that bit done, it became simpler to translate the schematic into physical form. I didn’t use my scratch sheet at all. I kept going and before you know it, I’d completed all of the amplifier section!

You can clearly see the layout. Now, you might also notice some funkiness with a couple of resistors right there in the center. I decided to “save” adding a pad by soldering a resistor and transistor leg in series. This was a bad idea, because two more resistors connected to it! I got it figured out, but my quest for saving a few seconds putting a pad down caused some unnecessary ugliness. But hey, it adds character, right? 

You can also clearly see my power run to the power buss, and a solder spot with nothing on it- a misplaced component and a corrected build. After I got this far, I stopped and checked all the capacitors. Sure enough, I had a couple of them mixed up. Also, the schematic calls for a 56K resistor. I didn’t have one, so I substituted a 51K resistor in its place. 

What’s Next?

Wow, that’s a lot to take in! Read it over and get a feel for the build. Draw things out until you understand them, and then take the plunge! You’ll have a working receiver in no time.

In the next installment, we’ll go over building the diode ring mixer circuit, the band pass filter, and VFO.  

Are you building along?

If you’re building along with us, please comment below and share your pictures! You can also follow us on Instagram @miscdotgeek.com for some previews of the next parts of the build. 

Click Here to continue to Part 3

73!
W7RLF

2 comments

2 pings

  1. You may also read my webs pages😇 vu3inj.blogspot.in for many simple DC receiver experiments doing reinventing the wheel. De vu3inj

  2. Its a lot easier to get a drill grind square points on it and cut islands on the pcb board.You then solder components to the island. Cutting up pieces of PCB , filing and gluing your way is too much work. It can also add unwanted added capacitance with the copper pads. Try it on the next project.73

  1. […] a Direct Conversion receiver design to build and getting started with the most basic parts. Part 2 dove into Manhattan construction and building the amplification stages. For this third installment, […]

  2. […] month I built the DC40 receiver (and wrote a giant series on it: Part 1, Part 2, Part 3 and Part 4. And somewhat unofficially Here!) and at the same time, Hackaday.com posted […]

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.