The Dummyloads

Where ham brains come to fry

Where ham brains come to fry

NE602

Final Write Up Direct Conversion Receiver

The NE602 Direct conversion receiver has been a good project.  I’ve learned a lot.  This last week of studying the schematic and trying to understand what I can do to peak the performance has been instructional.

Since I was focused on the band spread capacitor I first tried to understand what the spread of the tuning on that capacitor should be versus what it was in my implementation.  In the process I manually calculated the max and min value of the schematic.  The band spread should have been somewhere between 3.33pF and 8.3pF.  A spread of 5pF. As it turns out this is all calculated out in the book. Then I calculated out my values.  I have a variable capacitor that ranges from 15-95pF.  With a 10pF cap in series that gives me a range of 6pF to 9pF.  Only a spread of 3pF.  I was right. . . I don’t have enough of a spread to in my band spread capacitor.

What value of capacitor would I need to get my band spread up to the 5pF as suggested in the text.  Well I got to bust out my elementary algebra skills and failed dramatically!  With the help of my sister-in-law N5LRF (who teaches math) I was able to figure it out.

But what does this mean in reality?

So I decided to calculate out the min/max frequency of the circuit as designed in the schematic and as implemented in my circuit.  For this I used an online calculator.  But for your reference F=1/(2π√(LC))

As designed the circuit will tune 6.8MHz-7.52MHz.  With my implementation it will tune 6.70MHz-7.35MHz.

Hmmm. . . . My LO seems to be right where it should be.

My observation of this circuit in reality is that no matter how I tune the LO, the frequency is unchanged.  I clearly have an enormous swing in frequency.  I can observe it in the math and I also can observe it on the Oscilloscope. BUT. . . It doesn’t really seem to change the receive station strength.  Why is that?

Another set of observations are that when I am measuring with the Oscilloscope I can observe no signal on the LO unless I am transmitting into a dummy load.  When I switch the Oscilloscope over to the front end and measure I can observe no signal, whether transmitting into the dummy load or not.  Interesting enough when I transmit into the dummy load I also can hear signals on 40M much better.  I can still hear the entire band, but I hear them better.

After discussing with my friend Myron (WV0H) I came to understand that when transmitting into the dummy load I am increasing the current flow in my LO.  This increased current results in better out of frequency rejection and receiving.  The only real solution to this would be to increase the current of LO into the NE602.  So the band spread is irrelevant.  This receiver is just not very selective.  Especially when there are strong adjacent signals.  A potential solution would be to have a crystal determine the frequency.  I would have a far smaller tuning frequency range, but it would improve the current flow.

I’ve decided that I have learned all that I can possibly learn from this particular endeavor.  It’s time to move on to the next project.  I’m really thinking about tackling a regenerative receiver and just starting from scratch.

Stay tuned to the site for more updates on building projects and outdoor operating adventures!

 

Measurements and Learning

I have been disappointed to discover that my receiver is easily overloaded.  Also tuning the variable capacitor seems to have very little impact on the received frequency where as the adjustment on the front end seems to have a dramatic impact on the received frequency.  Why?  What measurements can I take and what can I learn?  Well I’ll tell you my story. . .

****Warning everything below this line is my best guess, I don’t actually know.  If you know I’d love to hear from you in the comments below.***

First of all I don’t have much equipment, but I do have an oscilloscope that Bill Ellis (N5TXN) gave me years ago.  It is a 15 Mhz BK Precision 1472C scope.  Unfortunately I don’t have any probes. . . Amazon 100Mhz scope probes ordered.  While I’m waiting on those, what else can I do?  First I used a dummy load to transmit 7.000 from another radio at 5W and used the variable capacitor on the front end to make sure I was peaking the signal in the CW end of the 40M band.  I then used the tuning capacitor to try to tune off the signal.  No luck.  Then I keyed down into the dummy load and started spinning the VFO.  I made it to 7.150 before the signal diminished in any way.  Hmm. . . that just doesn’t seem right.  Could my LO be on the wrong frequency?

How can I measure the frequency of the LO?

Many phone calls and emails to many friends.  A couple of notable quotes.

  1.  Tuning doesn’t matter if the receiver is overloaded.
  2.  You can hook up your oscilloscope or frequency counter, but know that you are loading the circuit and the frequency is not the exact frequency in the circuit.
  3. Given your LO arrangement, there just isn’t much signal to measure.

I was advised to use a 1″ form and wind 10 loops of wire around the 1″ form for a receive antenna.  I could then use this receive antenna to try to get a frequency count by using my MFJ-259B frequency counter OR use my oscilloscope and the antenna.  I was also advised that this was likely a futile effort as the total energy in the LO circuit is unlikely enough to get a good reading.  Turns out this was correct. . . The MFJ frequency counter gave me results from 400 Hz to 30Mhz and the oscilloscope could see a signal, but it wasn’t enough that I could actually count the waves.

I took a trip to the local Nortex QRP meeting and brought my little creation.  I expressed my frustration that the tuning capacitor seems to have no impact on the tuning.  After visiting with Dave Lear (NE5DL) and Joe Spencer (KK5NA) I became convinced that the issue was in my band spread capacitor.  It’s not a large enough tuning value.  When I arrived home my Oscilloscope probes had arrived!

I hooked up the oscilloscope probe to pin 6 on the NE602.  I got nothing.  I played with the RF gain pot.  Nothing.  I put my desk radio on 7.050 Mhz and keyed into a dummy load at 5W.  Hello. . . I have signal.  I guess that the LO runs at such a low current there is no observable signal until a threshold is met on the RF input.  I then spun my tuning cap and the frequency changed as did the amplitude.  The amplitude and frequency changed dramatically.  I interpret this as confirming my hypothesis.  The band spread capacitor needs to be adjusted to tighten up the tuning capacitor.  As the tuning capacitor exits the tuned portion of the front end the amplitude drops.  In my case, its REALLY fast!!

I think my next update will probably be finishing my final experiments with this receiver and start planning my next project.  We are approaching the time of year where operating outdoors is nice so I may take a break from building until it gets to be so hot that operating isn’t fun and field day isn’t far away either!

Local Oscillator and Connecting the Chips

If you remember from last post we had the power rail all wired up to the sockets.  The next step is to connect the chips with the various capacitors, finish the audio out and start the Local Oscillator.  You will see in the picture that I have fixed the power wire for the LM386 and it is now wired to pin 6 which is the proper V+ pin rather than pin 7.  You will also note that indeed the power jumper created a small issue jumping from pin 6 and 7 of the NE602 over to the start of the local oscillator.  A good foot note for future builds to think about that a little more.  Other noteworthy lessons on this section are:  1.  My canonical tip for my Weller soldering station is not a great choice for melting the Krylon and getting a good ground solder bead.  Nor is my wet pad a satisfactory device for cleaning my tip.  There is a significant delay in having a clean tip and being able to melt solder that is sufficiently annoying.  As a result I went to Fry’s in Arlington, Tx and purchase an ST2 and ST3 soldering iron tip as well as a Hakko tip cleaning bucket.  This definitively made the soldering process better!

The only other notable step from this section was that I purchased a bunch of alleged NP0 caps on eBay.  They were shipped on the slow boat from China.  When they arrived they looked like ordinary ceramic capacitors and in no way looked C0G or NP0.  So I decided to try to test these out.  I took out a 680pF capacitor and put it on my AADE meter.  It was within tolerance of expected C value.  I then pulled out my YL’s (wife for non-amateurs) hair dryer.  I then kicked it on low and started heating up the cap.  It went from a cool room to hot enough I preferred not to touch it.  Total change is capacitance.  1 pF.  I’m still learning, but that seems to be well within tolerance for an NP0 capacitor.

The audio chain only needs the actual 3.5mm jack.  The LO only needs toroids and the final variable capacitors.  Then I will start on the front end.  This project is getting real!!  Hope to have it make noise soon!  When I do get it to make noise I will try to post video!

Touching the Front End

Since I was using a different gauge wire, I checked my result

At the point I left off last time I nearly had the LO complete.  Winding a toroid and connecting all the pads was the last step to close that stage out.  For this step I opted to go ahead and do the toroids first.  I don’t understand why people dislike toroids so much, it really isn’t that hard.  This build was a T-37-6 with 21 turns of #28 wire.  (The schematic calls for 26, but I didn’t have any.  I did the calculation for what the inductance should be and measured my result.  They matched.)  Every time you go through the center of the toroid, count 1. Easy!

If your wondering about my L/C Meter.  They are unfortunately unavailable.  Its a really sad story. . .

This is the front end nearing completion

Since I already had the pads laid out for the local oscillator I finished up that section.  I still hadn’t decided how I wanted to do the tuning capacitors for this receiver.  I really am not in the mood to use 2-80pF variable capacitors.  I opted to go with a 0-50 variable cap that you need a screwdriver to tune for the band spread and a 20-107 pF air variable for the regular tuning knob.  With that decided and done, laying out the pads for the front end began.

The front end is a pretty simple circuit.  A toroid that matches the local oscillator, an on the board capacitor and a couple of trim caps.  The tricky part of this stage was making sure everything would fit.  The particular variable cap I had chosen was pretty big relative to the remainder of the board.

A picture of the board with all of its hardware and a 9V battery for a size comparison.

The particular day I was working on this the family had a variety of activities planned and I had to stop a couple of times and then start back up.  About 5p I had finished everything.  I decided to go through and double-check all the solder joints.  Then I realized that I had forgotten the 3.5mm jack for a speaker or headphones!  OK fixed that.  I double checked all the solder joints and all looked well.  I installed the ICs.  When installing ICs you will need to bend the leads ever so slightly inward so that they fit in the socket.  It is easiest to do when they are IN the socket.  Lightly press one edge of leads into the socket to bend them ever so slightly inward.  Then gently set the IC into the socket from the back and roll it forward.

 

Feeling very good about my progress and nervous about the next steps.  It’s time to apply power.  More to come!

Power!

initial layout and super glue mess

Now that the board is sized its time to melt some solder.  I’m not sure what the real first step is supposed to be.  I’ve heard Eric Guth on the QSO Today podcast talk to a number of builders and answers vary from the audio amp to the Local Oscillator.  I decided to start with getting power to the chips.  Once I had that figured out I’d have a better idea of where the local oscillator would go on the board.  I also decided to add a couple of steps to the power input.  One is a reverse polarity protecting diode and the second was a 5V regulator.  Since I don’t have a real 5V power supply, I am likely to power this with a 9V battery.  The NE602 has a max V+ of 8V so I want to make sure that my V+ stays at 5V and not fry anything.

Power stage complete. . . almost

You will see on the board I’ve laid out and superglued the 8 pin DIP sockets as well as the initial pads for the power rail.  Lessons learned from this experience.  Don’t put superglue on the pad then try to place it on the board.  Put the superglue on the board and then set the pad on the glue.  If you don’t you will smear superglue on your board as you can see in my picture above.  The other lesson I learned was to be a little more thoughtful on where you run the wire from one side of the board to the other.  I think I would run it a little closer to the chips so that the Local oscillator could have a little shorter leads over the power wire to the NE602.

If you look closely at my picture and study the schematic from the initial post, you will note that I’ve wired something incorrectly.  It can be easily seen in this photo (left).  Can you spot it?  Leave a comment below when you find it.

The build has begun. . . More to come!

 

Prepping Copper

8 pin DIP socket soldered to the MePads

Now that I know how the layout is going to run its time to start planning how much copper will be needed for the circuit board.  This will likely be the post where I get the most wrong!  To be clear this is being written as the project is in process, so we will find out together if I set this up wrong. . . but a few weeks from now probably.

The circuit is set up around the two ICs, an NE602 and an LM386.  Planning how these devices will sit on the board and creating enough room for the various components supporting the inputs and outputs will determine the size of copper board needed. The full size 4″x6″ copper board seems like major overkill for this circuit.

4″x6″ copper board is probably too big for our project.

I’ve decided to do the K7QO method of using 8 pin DIP sockets.  While the socket adds ~$.15 to the build, it guarantees that I won’t smoke the IC with my soldering iron being too hot.  I purchased some MePads when building the Regen last year. They went unused as the particular IC I was using was surface mount and didn’t fit.  As you can see in the picture above, these fit the sockets just fine.  I also intend to use the MeSquares for the other shared connection points.  Now that the sockets are on the MePads, its time to lay them on the 4″x6″ sheet and guess how much board I need. 

It seems to me that you would probably only need 1/2 of the board.  Measuring 3 inches across the top and bottom of the board I drew a straight line with a black sharpie.  I carried the line around the back of the board as well.

Copper ready to be drilled for standoff holes

Cutting the board using the AA7EE method resulted in a slightly crooked cut, but ultimately successful separating of the board.

 I also wanted to do the K7QO method of adding stand offs to the board to raise the soldering level up off the table a little.  To keep the holes semi symmetric I measured 3/10ths of an inch from each edge of the board and drilled a hole where the lines intersected.

MePads and MeSquares together

The last step was to scrub the copper with some steel wool and clean it with alcohol.  Then put a couple of coats of matte Krylon on the board.  This should help protect the copper a little and allow me to build on it over a period of weeks without the copper oxidizing.

The project is well underway, the next step should be actually super gluing pads and beginning to solder circuit blocks.

 

 

Failing to Plan is Planning to Fail

Most of my parts have arrived and I am beginning the process of planning my first official scratch build.  For sure this is not a true scratch built radio like the OMs who had to learn 20 WPM CW in the 50s.  BUT my goal here is to complete a project from a schematic and learn a little about what challenges I will face in the planning phase of circuits.  I’m confident there will be LOTS of little lessons along the way….

Read More