Re: [uCsimm] 115Kbaud Download fails because.... Add to FAQ's

From: Tom Walsh (tom@openhardware.net)
Date: Mon Nov 06 2000 - 21:30:03 EST


John Jeffers wrote:
>
> Well
>
> All I can claim is that I didn't build it and it does have more than enough
> current available from the adapter and at 9600 baud it always works and the
> correct part was built in.
>
> I have a 6 foot serial cord and this doesn't happen when the 10BT link is
> connected.
>
> However it seems that RT-Control doesn't say too much either way.
>

What do you see on the power bus when you look at it with a good
hi-speed scope? Can you see ringing or over/under shoot switching noise
on the supply lines? Just having "adequate" current available from a
supply is not the only thing you need, you also need a low series
resistance between the power source and the devices drawing upon that
power. For example: pushing 1 amp thru a 24 gauge wire can be done, but
the wire does have an inherent ohmic DC resistance and on current draw
peaks, you will drop voltage across that wire!

Let's not mention anything about the reactive (AC) resistance that is
also present within a supply... Wire also has an inductance & capacitive
property as well, a wire can be represented as a complex device
consisting of an inductor + capacitor + resistor. Generally you will
not see this reactive effect at low frequency power surges, but raise
the frequency (hook it up to a microprocessor circuit) and now you have
potential switching surges that reach well into the 100s of Mhz range
(remember your college calcus I course where the slope of the (sine)wave
determines the frequencies contained within that wave?).

You may indeed not have a problem with the supply, however, "never
assume". You may have a faulty PCB where there is a low-ohmic impurity
in the board itself that is causing a signal on the TCP wires to "suck"
current from an adjoining signal trace with the result being a change in
the rise / fall time of that signal. Moral? All else fails scrap the
board, but, don't rule the power supply out yet, "if you cannot find
what is wrong, prove what is right".

Regards,

TomW

-- 
Tom Walsh - WN3L - Embedded Systems Consultant
'www.openhardware.net', 'www.cyberiansoftware.com'
"Windows? No thanks, I have work to do..."
This message resent by the ucsimm@uclinux.com list server http://www.uClinux.com/



This archive was generated by hypermail 2b30 : Sun Apr 07 2002 - 00:01:39 EST