KPETERSEN@SIMTEL20.ARPA (08/31/85)
From: Keith Petersen <W8SDZ@SIMTEL20.ARPA>
This is relayed from a local packet machine:
Posted: Sun Aug 25, 1985 6:00 AM GMT Msg: OGIF-2241-3677
From: TCLARK
To: docs
Subj: TAPR TNC2 Hints 'n Kinks
Now that many of you have TNC2's coming, here are a few minor things you
might want to watch out for:
(1) I'm not sure if the artwork got fixed for the "production" run, but on
the "beta" test boards, there is one trace error. On the DB25 RS-232
connector, there is a trace on the back side going to pin 16. It
should be going to pin 6. An easy way to fix it is to cut the trace to
pin 16 on the back side, and then run a small jumper wire from the
plated-thru hole near the DIP switch to pin 6.
(2) If you have trouble in the demod calibration exercise, the problem may
be with noisy transitions from U8 [U8 is use ONLY in the calibration
steps, not in normal operation]. I found that a 0.1 bypass between
pins 3 & 4 helped a lot.
(3) Unless you use all CMOS parts, the 7805 regulator runs a bit warmer
than I'd like to see. In later versions of the TNC2 board, the
regulator will bolt to one of the top rails in the cabinet. For the
first batch of boards, I'd suggest the following modification to
improve the heat conduction from the 7805 to the aluminum cabinet:
a. Drill a small hole (to clear the 4-40 or 6-32 screw in step b.) in
the bottom of the case to line up with the mounting screw for the
7805 regulator. this should be 1.32" from the front lip of the
case and 0.55" from the side of the case.
b. Find some 4-40 or 6-32 threaded (preferably aluminum) spacers and
cut off a 0.30" piece. Use this threaded spacer as the mounting nut
for the 7805 on the back side of the board with a short screw going
thru the tab on the top side of the board. After you slide the
board into the cabinet, use a second short screw to "bond" the
spacer to the cabinet.
(4) Also on the topic of heat control, I'd advise mounting R80 "in the
air" about 0.3" above the board. On my board, the leads are at about
the same height above the board as are the leads on the adjacent C12
electrolytic capacitor.
(5) Since N2WX added in the time-of-day clock software, the frequency
accuracy of the xtal oscillator is more critical. I'd suggest putting
a small 10-50 pf trimmer capacitor in C47 in lieu of the 22 pf "blue"
capacitor provided with the kit.
(6) Keep the leads to the electrolytic capacitors C3, C6, C7, C8, C10, C16
and C18 as short as possible, or you may have trouble with them
hitting the rails on the top of the aluminum cabinet.
(7) Although headers are (probably) not provided, JMP1 will be necessary
if you try to use hardware handshaking (e.g. on a W0RLI BBS). I'd also
suggest installing JMP3 header -- the plated-thru hole on the trace
going to the + end of C12 carries a lot of current and is a potential
"fuse"; the JMP3 header will insure a positive jumper between the top
and bottom of the board.
(8) Be very careful to trim the tabs on the top of the push-button switch
and power connector or else they can short to the rails on the top of
the case.
(9) I have a mod that gives a PTT indication by changing the intensity of
the power LED. Is anybody interested?
73, Tom