A Big Error Discovered in the 1976 QST "Herring Aid 5" Article (After 47 Years)

Click on image for a better view.  See arrows for Q1, Q5 and the parts list.

I tried as a teenager to build the Herring Aid 5 direct conversion receiver from the July 1976 issue of QST.  I could not get it to work. Important points: 1) I was attracted to the fact that this receiver used only parts available from local Radio Shack stores (I bought them) and 2) The article provided a PC board pattern (I also bought ferric chloride and etched a board). 

Looking back, I concluded that I had failed to get the VFO to oscillate.  I remember hearing signals when I tuned my HT-37 transmitter (on Cal, on 40) near the receiver.  I was very close, but I never got that Herring Aid 5 VFO to oscillate.

Thirty-eight years later I tried again to build the receiver.  Important points:  This time I used mostly junk box parts and Manhattan-style construction (no etching).  Still I could not get the VFO to work.  ZL2DEX spotted a problem -- I had wound the L6 and L7 coils with the wrong "winding sense."  I corrected this, and BOOM! the VFO sprang to life.  I assumed that I had made a similar "winding sense error" way back in 1976.  This, I thought, explained my failure to get the receiver working.  The QST article had warned that proper phasing of L6 and L7 was necessary.  I figured that I just hadn't fully understood what "proper phasing" meant.  So it was, I thought, all my fault.    

 

Here is the PC board pattern from QST.  Arrows show Q1 (DGS) and Q5 (DSG).  

But it is the same kind of FET!  Click on the image for a better view.

But then on October 17, 2023 a comment appeared on the SolderSmoke YouTube channel.  Rick WD5L had also -- back in the late 70's -- tried to build this receiver.  He recently looked closely at the recommended parts list (that we used!) and at the QST PC board pattern (that we also used). AND HE SPOTTED AN IMPORTANT ERROR IN THE QST PC BOARD PATTERN.   

Take a look at the pattern for Q5 above (see arrow).  That is the VFO FET.  A Radio Shack 2035 FET has a DGS pinout (see below).   The Gate is the center pin.  If you put this transistor into the PC board pattern above you would definitely be grounding the Gate.  There is no way the VFO would work under these circumstances.  Note too that the only other RS 2035 FET in the receiver is Q1 (the RF amplifier).  In the PC board pattern above Q1 is marked correctly as DGS.  This confirms the error in the Q5 PC board pattern.  There is no way an RS 2035 transistor can simultaneously have two different pinouts! 

Wow.  So this failure to get the VFO working may not have been my fault after all.  I may have actually gotten the transformer winding correct, but even if I did, there is no way this VFO would have worked using the part called for by QST and the PC board pattern shown above.  As a teenager I just did not know enough to spot the error or the inconsistency.  I kind put blind trust in QST.  I just couldn't get the thing to work.  

Rick searched the QST archives to see if they ever put out an errata on this.  So far, nothing. Worse yet, the scan of the PC board pattern on the QST site is very unclear and may have the pin designations on Q5 scratched out. This would make it more difficult to spot the problem.  (The image above is not from ARRL.  It is from a high quality scan of the original QST article done this week by a fellow Vienna Wireless Society member.)  Please let us know if you find any kind of errata or any acknowledgment of error.  

This is really pretty bad.  This was a project aimed at novices.  Far from encouraging homebrewing, this type of mistake is the kind of thing that would push people away from the soldering iron.  

Ironically, I may have been doomed by opting to use the QST PC board.  If I had used Manhattan-style construction (as I did in my more recent build) I would not have fallen victim to this PC board pattern error.  Also, if I had built this thing stage-by-stage (as we always now recommend) I would have more clearly realized (back in 1976) that the problem was in the VFO stage.  But I was 17 and didn't know.  I put blind faith in the QST article.  It never occurred to me that something in print could be wrong.  This realization came much later. 

There is more to talk about on this ill-fated project. In future posts I will discuss another error, this one in the AF amplifier.  And possible additional errors...  And I'll write about the 1998 resurrection of this project by NORCAL QRP and the New Jersey QRP clubs.  

Thanks to Rick WD5L for spotting the PC board error.