DPRG List  

[DPRG] Over dioded?

Subject: [DPRG] Over dioded?
From: Dale Wheat dale at dalewheat.com
Date: Tue Dec 28 11:20:22 CST 2010


Thanks for looking at this for me.  I'm leaning towards keeping the protection diodes, but moving the LEDs a little farther to the right on the schematic, so that they, too, enjoy the protection afforded by the other diodes.


From: jdrumm9015 at aol.com 
Sent: Tuesday, December 28, 2010 9:39 AM
To: dale at dalewheat.com ; dprglist at dprg.org 
Subject: Re: [DPRG] Over dioded?

  The inherent "diode-ness" will only work if you knew what the other end was referenced to.  The protection diodes work well since one end is connected to ground, you know what the reference is.  When the other end drops one forward biased diode drop (0.6-0.7 volts Si) then the diode conducts and protects the line from negative transients.

  The diodes in series with the data are referenced to the base of the transistors (Q1, Q2, Q3).  Is that diode forward biased or not?  What would a negative transient see on that line?  Maybe it depends on the logical value on the other end of the series diode, or maybe it depends on if the transistor base-emitter is forward biased.  I believe a negative transient is almost guaranteed to blow out the base-emitter junction of the transistors, Q1, Q2, or Q3 since the series diode on state is somewhat nebulous given the unknown state of the base-emitter on state.

  Just my two cents.

John Drummond

PS:  The series resistor (high tech battery charger circuit) might give you all the protection you need anyway since it will limit any transient current flow, you still might be able to eliminate the protection diodes, but not because of the series LEDs.

-----Original Message-----
From: Dale Wheat <dale at dalewheat.com>
To: dprglist <dprglist at dprg.org>
Sent: Mon, Dec 27, 2010 6:31 pm
Subject: [DPRG] Over dioded?

I'm designing a new RS-232 to TTL adapter for the NXP LPC Cortex-M0 and 
Cortex-M3 processors.  It uses some of the modem control lines to reset the 
chip and put it into bootloader mode.  I'm using very simple 
resistor-transistor level shifting to constrain the incoming signals to 
typically a 0V-3V range (or whatever is supplied on the Vcc pin).

My original design works well, even with "barely compliant" USB to RS-232 
adapters.  I went and got all fancy by adding some LEDs inline with the 
incoming signals, instead of adding an additional driver transistor for each 
signal.  Using high-brightness LEDs, I can just see the LEDs twinkling when 
bits are flowing to and from the chip.

My question is this:  With the LEDs in series with the incoming signals, do 
I still need the protection diodes (D1, D2, D3) to shunt potential negative 
levels to ground?  Or will the inherent diode-ness of the LEDs prevent this 
from ever happening?

Here is a PDF of the schematic-in-progress:


Any suggestions are appreciated!


Dale Wheat
(972) 486-1317 

DPRGlist mailing list
DPRGlist at dprg.org


DPRGlist mailing list
DPRGlist at dprg.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://list.dprg.org/pipermail/dprglist/attachments/20101228/4098cb02/attachment.html

More information about the DPRG mailing list