| |
VK2ZRG > TECH 21.12.05 05:10l 78 Lines 3633 Bytes #999 (0) @ WW
BID : 1898_VK2ZRG
Read: DL1LCA GUEST OE7FMI
Subj: Re: A puzzling fault
Path: DB0FHN<DB0FOR<DB0SIF<DB0EA<DB0ACC<DB0GOS<DB0RES<DK0WUE<7M3TJZ<HS1LMV<
I0TVL<VK2TGB<VK2IO<VK2WI
Sent: 051221/0215Z @:VK2WI.#SYD.NSW.AUS.OC #:1622 [SYDNEY] FBB7 $:1898_VK2ZRG
From: VK2ZRG@VK2WI.#SYD.NSW.AUS.OC
To : TECH@WW
VK2ZRG/TPK 1.83d Msg #:1898 Date:21-12-05 Time:1:46Z
Andy GM7HUD wrote :-
>Regard RS232 signalling levels derived from charge pump circuits....
>
>Dont for get that RS232 specs demand that the signal is less than -3V ot
>greater than +3V. So you dont really need much voltage swing.
>
>In theory that is. In practice you may need a bit more "oomph" if you try
>slewing the lines at high speeds (115kbps etc.)
>
>Talking of charge pumps, devices like MAXIM's MAX232 series normally
>use 10uF caps in the charge pump, but you can nearly always get away with
>0.1uF caps. The voltage available is lower but it means you can build the
>charge pump circuit on a smaller board.
Yes, 10 uF or so is the usual value Andy. That's what I would have used
if I was designing the circuit. I raised the charge pump frequency to 80 kHz
and now have -3.8 and +7.3 volts supply to the MC1488. The TNC is working OK
now. The 0.1 uF caps are really a bit too small, even for a 14C88 and load.
I do have some small tantalum caps that I'll try in the circuit before I
go off to the rip-off merchants to get a 14C88.
>As a final point, companies like MAXIM, Linear Technology and Dallas are
>a delight to deal with. They have comprehensive websites that dont restrict
>access to any of their data and best of all you can place small orders
>directly with the company rather than having to find a distributor who
>#will still process small orders. Chips like the MAX232 are easy to find at
>hobby suppliers but the more esoteric devices like the MAX3232 (3.3v not 5v)
>is harder to get. Likewise if you need one in SMD to fix something.
>This is when companies like the 3 above shine, they'll ship you 1 or 2 often
>with no carriage charges in many cases. Of course you can get free samples
>but dont be a cheapskate and abuse such services!
>It's quite amusing that now in 2005 we are still mucking about with RS232.
>Think about it... a standard from the middle 60s when +/=15V supplies were
>common. Nowadays we have a CPU often running at 1.5V which needs gubbins to
>shift its IO levels to 3.3v or 5V. Then we shift those signals to something
>around +/- 10V for RS232 signalling. At the remote end another chip drops
>them down to 5V or 3.3V followed finally by shifting to core levels.
>On battery powered gizmos... that's a lot of expense to achieve a slow
>error prone signalling scheme. That's why USB is just sooooooooo much better!
>73 de Andy GM7HUD
I am aware of the policy of Maxim re samples. Their thinking is probably
that if they give away samples, then circuit designers are likely to use
their product in a production run of many units; and they'll recoup the
cost of the samples.
But would they supply a 14C88 by snail mail, gratis, for Joe Ham to fix
his TNC, with no possibility of selling him a large order?
USB may be lots better Andy, but my TNC is around 13 years old and uses
RS 232 which is quite good enough for packet radio. It's well made with
sockets for the 14C88 and the 74HC14 that is used in the charge pump circuit.
So it was easy enough to plug in another 1488 to fix it rather than buying
a new one.
Speaking of easy fixes, I've fixed a few PC monitors by simply replacing
a fuse that had failed. I just ignored the usual lie that "there are no
user serviceable parts inside" and opened up the monitor case. I wonder how
many monitors have gone to land fill because of a failed fuse?
73s from Ralph VK2ZRG@VK2WI.#SYD.NSW.AUS.OC
/ack
Taglines Version 1.00, by Colin Coker G4FCN
C Program run, C Program Crash...ReWrite in Pascal!
Read previous mail | Read next mail
| |
