I'm new to electronics and I'm hitting an issue beyond my skill set. I
should start by saying I'm not 100% sure signal reflection is the issue,
but it's my best guess.

I have a 3.3V clock signal that I'm trying to convert to 5V. The clock
speeds are sub 100 kHz. I'm using a TI level shifter, TXS0108E. Looking
at the signal on a scope, it's a clean signal on the input and output
side when I don't have anything connected on the output side.

When I hook up the clock signal on the output side, over an
approximately 1 meter wire the signal on the scope is full of noise. The
scope also says the frequency is much higher than one the input signal
should generate. I see the noise push through to the input side. I'm
assuming this is because the level-shifter is bi-directional.

I've tried a few failed solutions, like using a diode to try and block
out the reflections, but so far no luck.

Any advice on how to deal with this?

In article <yFLJN.29182$_a1e.6791@fx16.iad>, Dan <dan@nospam.invalid> wrote:

>I have a 3.3V clock signal that I'm trying to convert to 5V. The clock
>speeds are sub 100 kHz. I'm using a TI level shifter, TXS0108E. Looking
>at the signal on a scope, it's a clean signal on the input and output
>side when I don't have anything connected on the output side.
>
>When I hook up the clock signal on the output side, over an
>approximately 1 meter wire the signal on the scope is full of noise. The
>scope also says the frequency is much higher than one the input signal
>should generate. I see the noise push through to the input side. I'm
>assuming this is because the level-shifter is bi-directional.
>
>I've tried a few failed solutions, like using a diode to try and block
>out the reflections, but so far no luck.
>
>Any advice on how to deal with this?

The first thing I would suspect is external-noise pickup, possibly
from a local AM radio station, or switching noise from electronics in
the area (e.g. switching power supplies, DC to DC converters,
microprocessor or microcontroller circuits, etc.)

A meter of wire makes a pretty good antenna, and there's a lot of RF
noise and AM signal power floating around in the air these days.

You may be able to eliminate this (or reduce it to tolerable
proportions) by distributing the clock signal over a shielded coax
cable, rather than a bare wire.

Dan <dan@nospam.invalid> wrote:

> I'm new to electronics and I'm hitting an issue beyond my skill set. I
> should start by saying I'm not 100% sure signal reflection is the issue,
> but it's my best guess.
>
> I have a 3.3V clock signal that I'm trying to convert to 5V. The clock
> speeds are sub 100 kHz. I'm using a TI level shifter, TXS0108E. Looking
> at the signal on a scope, it's a clean signal on the input and output
> side when I don't have anything connected on the output side.
>
> When I hook up the clock signal on the output side, over an
> approximately 1 meter wire the signal on the scope is full of noise. The
> scope also says the frequency is much higher than one the input signal
> should generate. I see the noise push through to the input side. I'm
> assuming this is because the level-shifter is bi-directional.
>
> I've tried a few failed solutions, like using a diode to try and block
> out the reflections, but so far no luck.
>
> Any advice on how to deal with this?

Have you earthed the oscilloscope to the circuit you are testing?

--
~ Liz Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

On Mon, 18 Mar 2024 00:12:14 GMT, Dan <dan@nospam.invalid> wrote:

>I'm new to electronics and I'm hitting an issue beyond my skill set. I
>should start by saying I'm not 100% sure signal reflection is the issue,
>but it's my best guess.
>
>I have a 3.3V clock signal that I'm trying to convert to 5V. The clock
>speeds are sub 100 kHz. I'm using a TI level shifter, TXS0108E. Looking
>at the signal on a scope, it's a clean signal on the input and output
>side when I don't have anything connected on the output side.
>
>When I hook up the clock signal on the output side, over an
>approximately 1 meter wire the signal on the scope is full of noise. The
>scope also says the frequency is much higher than one the input signal
>should generate. I see the noise push through to the input side. I'm
>assuming this is because the level-shifter is bi-directional.
>
>I've tried a few failed solutions, like using a diode to try and block
>out the reflections, but so far no luck.
>
>Any advice on how to deal with this?

There's a possibility the fault could arise through improper use of
the level shifter, making it tip into self-oscillation when a load is
connected. This is not uncommon with voltage regulators as well. I
would suggest you strictly follow the recommendations on the data
sheet in regards to any ancillary components this device may require.

On Mon, 18 Mar 2024 00:12:14 GMT, Dan <dan@nospam.invalid> wrote:

>I'm new to electronics and I'm hitting an issue beyond my skill set. I
>should start by saying I'm not 100% sure signal reflection is the issue,
>but it's my best guess.
>
>I have a 3.3V clock signal that I'm trying to convert to 5V. The clock
>speeds are sub 100 kHz. I'm using a TI level shifter, TXS0108E. Looking
>at the signal on a scope, it's a clean signal on the input and output
>side when I don't have anything connected on the output side.
>
>When I hook up the clock signal on the output side, over an
>approximately 1 meter wire the signal on the scope is full of noise. The
>scope also says the frequency is much higher than one the input signal
>should generate. I see the noise push through to the input side. I'm
>assuming this is because the level-shifter is bi-directional.
>
>I've tried a few failed solutions, like using a diode to try and block
>out the reflections, but so far no luck.
>
>Any advice on how to deal with this?

Using a twisted pair for the signal might make sense. Running the
wire close to the system's metal container, or test bed ground plane
is also advised.

Where's the ground lead on the scope connected when you're
making these measurements? It can introduce gound loops.
If the noise/problem goes away when the the scope's disconnected
it's an indicator.

RL