[ARDF] Re: ARDF Digest, Vol 38, Issue 4

[Dale Hunt, WB6BYU]

    Mike -

       Yes, you remembered correctly.  I was going to point Jen in
    your direction as well, since I knew that you had dealt with the
    problem in ABQ.

       The Montreal controllers built on the FAR circuit boards work
    just fine EXCEPT when they are used with the ON7YD 80m transmitter
    boards.  The problem, as Mike explained, is that in some conditions
    the original version would close the key line without also closing
    the transmitter enable line.  With the ON7YD circuit this would
    overheat the final transistors.  (In my mind, this is a design
    flaw in the 80m transmitters, not the controllers.  But the
    controllers were easier to fix.)  I use the original controllers
    on both my 80m and 2m transmitters with no problems.

      But the changes in the Montreal Controller circuits are not
    that much as far as I can tell.  (I am sitting here with one
    of the FAR Circuits boards in hand, plus the schematic of the
    current version from Jacques' web site.)  I think the only
    change required is to connect R5 to pin 17 rather than pin 18.
    (That would one trace cut and one jumper added.)  There is
    also a change in the capacitors across the crystal: rather than
    a 22pf cap he now calls for 10pf in parallel with a 4 - 20pf
    trimmer to give precise setting of the crystal frequency.  This
    shouldn't be too difficult to wire up - I'd bend the leads of
    the 10pf cap to fit the board, then solder the trimmer across
    the top before sticking the leads through the holes in the board.
    That should allow you to mount both parts in the same holes on 
    the board, and the trimmer will be above the level of the
    resistor beside it so there shouldn't be any problem with space.
    (If you can't afford the added board height, you probably can 
    mount the trimmer on top of the board through the holes and 
    put C1 and R2 on the bottom of the board.  In the "old days"
    we just used fixed capacitors (without the trimmer) and it 
    was close enough - my worst transmitters drifted a couple of
    seconds over the course of an 8 hour event, and that was partly
    because I didn't use the same value capacitors each time I built
    the board.

       So the real reason to chose the original or revised version
    of the circuit is the programming of the PIC controllers that you
    have.  If they are programmed for the new circuit, move the one 
    lead on the board.  If they are programmed for the original board,
    then use it as is.  (The variable capacitor is unrelated to the 80m
    change, and just allows fine adjustment of the timing.)  Basically,
    the difference between the two is which output line from the controller
    is used to flash the monitor LED.  The controllers can be reprogrammed
    if necessary by some of the folks on this list.  And you only care
    about the functional difference between the two if you are using
    the ON7YD transmitter boards.

       Of course, if you want a compact board, then the FAR Circuits
    one isn't the best.  (I had the same problem with my 80m receiver
    boards.)  For that you'll need to do them on pref board (which
    works fine) or get a new board designed.

       Oh, and I strongly recommend using sockets for the PIC chips
    rather than soldering them in  - that way you can reprogram them
    if needed, and swap chips around to help debug the boards if you
    have a problem.  (That resolves whether the problem is in the
    controller chip or the board wiring.)

       Good luck!

         - Dale



k5atm at comcast.net wrote:
> 
> I am at work so I don't have my fox controller info in front of me but I wanted to respond before I forgot.  I am doing so from memory so things might be a little fuzzy.
> 
> The PCBs that FAR circuits sold back when Albuquerque put on the first event (and I suspect are the same PCBs they are selling now) has a pretty deadly flaw that the newer version of the fixes.
> 
> The "old" version when used with our 80 meter transmitter could key up the final with the oscillator un-keyed -- which would make the final transistors catch on fire pretty quickly.
> 
> The problem as I recall had to do with how the LED was engineered into the circuit and how the LED was accessed by the firmware when the firmware detected an illegal DIP switch setting.  At power on, if the firmware would see, for example, a fox set as MO5 with a 3 minute cycle time it would start flashing the LED.  Problem is the same PIC line that flashes the LED also keys up the 80 meter final.
> 
> We discovered this a week or so before the event and had to make some quick changes.  We did a little PCB rework and I modified the firmware.  The easiest thing for us to do at the time was to re-route one of the PIC I/O lines used to access the DIP switch and use it to drive the LED (or maybe it went to the 80 meter transmitter).  I then modified the firmware to decode the DIP switch a little differently (fewer delayed start options) and modified the firmware to control the 80 meter transmitter in a safe way.  I also modified the firmware to change the xmit speed.  The change increased the amount of time we transmitted energy.  We wanted to make it a little easier to grab the 80 meter signal.
> 
> The hardware mod was not difficult.  I think I only had to cut one or two traces and solder on a couple of jumpers. I can provide details on the hardware changes and provide source / hex files for the new software if that is of interest.
> 
> After discovering the problem I contacted FAR circuits and suggested they either update their PCB or at lease issue a warning.  Looks like they did neither.
> 
> The old firmware will not work with the new design and the new firmware will not work with the old design.
> 
> The original fox source was written in assembly and a free assembler can be downloaded from Microchip.  It is pretty easy to modify the source and re-generate a HEX file.
> 
> The new firmware is written in CCS C.  CCS does not offer a free version of their compiler.  To make changes, one would either have to buy a CCS C compiler or port the code to High-Tech C (they do offer a free version of their 16F84 compiler).
> 
> I have the CCS C compiler so I could compile / make changes if that is needed.
> 
> 
> I hope I remembered right and this all makes sense.
> 
> -Mike, K5ATM
>
---
[This E-mail Scanned for viruses by Online Northwest]