From: Matthias Klumpp <Matthias.Klumpp_at_gmx.de>
Date: Wed, 14 Jul 2004 21:05:46 +0200

Hi Jörg,

# -----Ursprüngliche Nachricht-----
# >
# > +5V
# > |
# > +-------------------------------+
# > | |
# > +-+ +-+
# > | |22k | |4.7k
# > | | | |
# > +-+ +-+
# > +---------------------+ |
# > +-+ | |
# > | |12k | |
# > | | | |
# > +-+ | |
# > | | |
# > | | |
# > --- | |
# > |G |
# > ------ |
# > D| |S |
# > uC Out port--------+ +--------+-- AJB remote
# >
# > Like this ?
#
# Almost. You need to swap the values of the divider.

Yeah, I was recognizing this morning when I tried to simulate it in PSPice.

# And I forgot to mention you should have a pullup on the uC side, too.

Even so my PSPice simulation shows, that the output level is only going up
to 0,7V.
In the simulation I'm using a IRF150 (n-channel MOSFET) due to the reason I
only have
a demo version of it.

Is this why I do have that difference of the output level, or is there still
something wrong ?

# Hmm, the circuit is bidirectional, for open collector on both
# sides, which is not exactly what you want. Your uC serial output is a
plain driver.

Well bi-directional would not hurt.
The thing more important is, that I need a circuit that is working with both
Jukeboxes,
with and without internal pull-up (like Studio Players and Recorders).

# Forget what I said and simplify to this:
#
# +5V
# |
# |
# +-+
# | |4.7k
# | |
# +-+
# |
# |
# Schottky |
# uC Out port--------|<|---------+-- AJB remote
#
# The pullup won't hurt even if a box already has one. The diode will allow
# your uC to pull the pin low, but will prevent it from being
# driven with high output current to 5V. Instead it "gently" floats up via
the pullup. Use a
# Schottky because those have low voltage drop across, allowing for a
better "low" level.

Ok, I also did a PSPice simulation of this (with 1N4148 because the demo
doesn't have a Schottky).
The AJB remote level then toggels between high = 0,7V and low 0,612V ???

Well, I know, that these simulations are not very well done, but in previous
cases it worked
more or less like the final HW.

Let me show you another possible solution:
http://focus.ti.com/lit/an/slaa148/slaa148.pdf
or
http://focus.ti.com/docs/mcu/catalog/resources/appnoteabstract.jhtml?familyI
d=342&abstractName=slaa148

both pointing to a TI application report that describes how to interface 3V
devices by 5V devices.

The easiest way the describe (page 4) is via a simple voltage divider.
If you have a look into the calculation on page 4ff, they consider all kinds
of tolerances and finally get
to the result of a simple high ohmic voltage divider with 2x 1.2MOhm (also
avoiding current into the
device's protection diodes.

Wouldn't be that the really easiest way ?

THX

Matthias

_______________________________________________
http://cool.haxx.se/mailman/listinfo/rockbox
Received on 2004-07-14