Don Bradbury wrote:

Don

These devices only implement a single RS485/RS422 transmitter and receiver.
Thus except for reducing the number of IC types they are not
particularly useful for this application.
However they do operate over a 3V to 5V power supply range.
They are also only available in SMT packages which can be inconvenient
for prototyping.

Bruce

Hi Bruce,

That's true for the 3160 which can be switched between two RX and two TX
of RS232 or one RX and one TX of RS485.
The 3162 does both functions simultaneously.

I had a look at the MAX3535E recently, which has a isolated RS485
transceiver with capacitive coupling in it, did anyone of you use those
to break ground-loops? I'm especially interested if this kind of
"capacitor-isolation barrier" adds any significant noise to the RS485
lines, which might couple into sensitive signals running in parallel, or
if I should just go the traditional optocoupler-route.

Greetings from Germany,

Chris

I am using the LTC1535 in an application right now, and if you need RS-485,
it is pretty neat. It includes the hardware to drive a small transformer
that will provide you with an isolated 5V floating supply too.

However, I am MUCH more impressed with the Analog Devices ADuM5241ARZ which
not only has two isolated digital channels, but also the power transformer
is built-in, and the whole device fits in an SO-8 package. You can get 5V at
10mA out of it. Analog calls this series "Digital Isolators"

This is absolutely truly amazing. I am using one in one of my products right
now. The Analog Device part is not RS-485 compatible, it is just a digital
isolator (TTL in, TTL out), but with the built-in 5V power supply, you can
drive any RS-485 driver. For RS-485, you will need the 4 port version of the
chip (for the Tx Enable)

The app note says it runs at 300 MHz. I have not checked the device with the
spectrum analyzer yet, but with a 100 MHz scope, I can't see anything.

They now have special versions for I2C and even a new one that provides
isolated gate drive to power MOSFETs.

Both devices are specified to sustain 2500VAC for one minute (UL
requirement).

Didier KO4BB

Internal Virus Database is out-of-date.
Checked by AVG.
Version: 7.5.519 / Virus Database: 269.22.5/1358 - Release Date: 4/3/2008
6:36 PM

Didier Juges wrote:

Didier

Any idea of the jitter of the Analog Devices transformer coupled isolators?

Bruce

Bruce,

That should be easy enough to check. I believe they use the 300MHz clock
gated, so the jitter is probably at least equal to the clock period, and
probably more if they use envelop detection. I have not come around to doing
that yet (all I know is that they have no problem getting my 9600 baud
serial data across :-) I will check with the scope first, then if it looks
good, I will put the 5370 on it.

The LTC part uses capacitive coupling instead of transformers, but they do
not say if they use an RF carrier, or simply on/off pulses. I could check
that one too.

I will let you know.

Didier

Internal Virus Database is out-of-date.
Checked by AVG.
Version: 7.5.519 / Virus Database: 269.22.5/1358 - Release Date: 4/3/2008
6:36 PM