NorCal's Herring Aid 5
With that in mind, I packed up the board, the schematic, the layout, the parts placement drawing and the original article and drove the 125 miles one way to Dave's house. (One advantage to having lived in western Kansas, distances don't mean a lot. Ask anyone who has lived in west Texas, the Oklahoma Panhandle or anywhere out west. We are used to driving.) When I arrived Dave greeted me at the door and we sat out to debug the receiver (Dave doing the work, I was watching and listening as he very patiently explained what he was doing and why).
One problem we had was this was a new layout, done by me, and it was not tested. I watched as the old master himself went to work. He told me that he was going to use the "Paul Harden, NA5N" method. He would start at one end of the schematic and check each section. I sat down and watched as Dave did the following.
First of all he checked for 12 volts at the supply and then for 9.1 volts on Q5. (The schematic calls for a 10V zener in series with a 270 ohm resistor, but Dave had suggested using a 9.1V zener in series with a 390 ohm resistor over the phone when I couldn't find a 10V zener.) Had the 12 volts at the supply, but oops, there was only 1 volt or so at the drain of Q5, which is the FET in the VFO. Dave turned to me and said, this is not good, we must have 9 volts or so. Something is loading down the circuit. He then checked the layout against the schematic, R16, D3, C22, R17 and D4 were all soldered in correctly and in the proper position. Dave had a bewildered look and asked where I had gotten the transistor. I replied that it came from my stash, but it was not the MPF102 called for but rather a J310. (I thought they were interchangeable, and as I was soon to find out, they are not always.) Dave replaced the J310 with an MFP102 and bingo, 9.1 volts, just as it should be.
Next he said that we would see if the VFO was VFOing. He hooked up the scope, and nothing, not a sine wave in sight. He looked at the circuit, checked all the parts and determined that the coil was wound out of phase, so he reversed the leads for L6, and again, like magic, bingo we had an oscillator. He measured the frequency, and it was at 6.8 MHz, which was a little low. It was easier to lower the capacitance in the circuit by changing a cap than it was to change the turns on the toroid, so Dave replaced a 39 pF with 5 pF and we measured the high end. Oops, now it was at 11.2 MHz for the high end. He looked at me, and I know that he knew, but he asked what value we should try next. I suggested 22pF as it was inbetween the two values that we tried. He soldered it in and voila, the top end was about 7.8 MHz and using trimcap C19A he was able to set the bottom of the tuning range at 6.998 MHz, which was close enough for us. One thing, the 5 pF that we tried first was a regular disc ceramic, it drifted several hertz per minute, but when he replaced it with a 22 pF Silver Mica, the vfo was much more stable. The point is use NPO's, Silver Micas or Polystyrene's in VFO's not 20% disc ceramics.
Ahhh, now we had a vfo, and Dave hooked up the signal generator at 7.040 MHz and he listened for it. It was there!! But weak, very weak. The mds of the receiver was only at -80dB, not very good. Lousy in fact. Dave said not to worry, that we had some tweaking to do. The circuit has two tuned circuits that need to be peaked and are set by peaking a trimcap in each circuit (C2A and C4A). The first one came up very nice with two peaks, which is what you are looking for when you use a trimcap. Bruce Florip wrote a nice article on this in a past issue of QRPp. But the other one only had one peak. Dave explained that this meant that the circuit was not reasonat at 7MHz. So he tried removing capacitance (again much easier than adding or removing turns) and it worked. Soon he had two peaks in the adjustment. He adjusted both of them back and forth, as they were interactive. Finally he was satisfied. Now, lets see what it does. Only -100 dB of mds. Terrible. Dave read the original article and saw that Jay Rusgrove got an mds of -130 dB!! Something was wrong.
Dave looked at the circuit, and then he asked about the transformers. Were they the same? No one was a 10K to 2K and the other was a 1K to 8 ohm. He checked them out for proper placement of the primary and secondary and they were correct. Then he asked if I was positive that I put the right transformer in the right place on the board. I wasn't so we checked the part numbers with a Mouser catalog, and sure enough, I had reversed them on the board.
Dave had me repeat after him, "Doug, I am a dummy. Doug, I am a dummy."
That is supposed to exorcise the dummy demons or the badgers or something.
Dave unsoldered the parts, exchanged them and tested the rig. Maybe -110
dB. But still something was wrong. Then Dave looked at the bias resistors
on the 3 2N2222 transistors and he said, "Ah Hah! I see the problem." One
of the bias resistors was 560 ohms instead of 5.6K, it does make a
difference. He switched it out, and tested the receiver. He still wasn't
satisfied. Then he sat down and studied the schematic. He decided that we
weren't getting enough coupling at L3 and L4. This was one of the coils
that was changed from the original Radio Shack modified part to a toroid.
He rewound it and changed it from L3 at 45 Turns and L4 at 5 turns to L3 at
45 turns and L4 at 20 turns. He resoldered the toroid, and what a
difference. Dave tested the mds, and it was at -123 or so. Very good for
such a simple receiver. The rig works!! Thanks to Dave for doing such a
neat job, and helping me to learn some more about trouble shooting.
The Herring Aid 5 Parts Layout
Back to the Future with NorCal's Tuna Tin 2
The Herring Aid 5 Schematic
The NorCal Page