Searching \ for '[EE] easiest way to create and couple noise into a' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=easiest+way+create
Search entire site for: 'easiest way to create and couple noise into a'.

Exact match. Not showing close matches.
PICList Thread
'[EE] easiest way to create and couple noise into a'
2007\01\31@163224 by Peiserma

flavicon
face
Most circuits have some inputs and outputs (external wires) connecting
them to the real world. I want to create electrical noise and subject an
unsuspecting PIC and its inputs to noise.

I'd love to have a gizmo that can throw some noise into the system in
the lab. It doesn't need to be fancy, just an effective noise-maker. I'm
thinking a bank of relays with long leads to the relay coils (no coil
suppression of course). The leads could then be wrapped around the
inputs to the circuit. I also have in my lab a sample 120V DC motor; I
could run that directly from rectified 120V AC, and switch rapidly
on-off with one of the aforementioned relays...

Does anyone have a favorite way of doing a quick bench test like this in
their lab? Something you've maybe constructed that you'd like to share?


2007\01\31@170021 by Marcel duchamp

picon face
A common method is a relay with N.C. contacts wired in series with the
coil.  It 'self-oscillates'. Then the supply voltage to the coil can be
wound around the circuit under test.

Sometimes it helps to ground the noise-generator to the circuit.

One tester I made used a two pole relay and capacitor setup as a flying
capacitor.  The charge side was line voltage through a variac feeding a
 voltage multiplier then to a large resistor charging the capacitor.
When the relay was made to toggle, the cap 'flew' from the charge side
to the circuit under test side.  By varying the size of the cap and
including a series resistor to apply the zap, as well as the variac to
control the cap voltage, you can sneak up on the circuit being tested.
This could produce spikes from zero up to a few thousand volts.  I
connected this setup one at a time directly to each circuit node that
went off the board.

Of course, it is a good idea to monitor the circuit with a scope to know
that you are really injecting noise and also how much of a spike you are
making.

spam_OUTpeisermaTakeThisOuTspamridgid.com wrote:
{Quote hidden}

2007\01\31@171610 by Richard Prosser

picon face
One thing someone has suggested in the past (on this list ?) was a
file and a car ignition coil. The connection is made through the file
by running a wire along it. Struck me it was likely to be as bad as it
gets. Probably pay to fit a spark gap to the coil output.

RP

On 01/02/07, .....peisermaKILLspamspam@spam@ridgid.com <peisermaspamKILLspamridgid.com> wrote:
{Quote hidden}

> -

2007\01\31@171952 by David VanHorn

picon face
An old ratty drill, with universal (brushed) motor works well.

I built an impulse tester years ago that I use in a much more controlled
manner, but it's a pretty lethal device.  For small discharges, the main
unit charges through a 40 meg high voltage resistor into a 100pF capacitor
made from circuit board material, and that cap is discharged through a 1500
ohm resistor into the circuit.

For the harsh stuff, I use the main unit directly, with about a 1" air gap,
discharging it's main caps.
The main unit is an encapsulated DC-DC converter with 0-30,000V output,
charging into three 0.01uF 30kV low inductance caps.  The discharges are
loud enough to require hearing protection.  The charge stored in the caps
and residual inductance gives me a few cycles at about 2 MHz, with a
radiated power of about a million watts, give or take.  Small wire loops can
be used to get a larger magnetic field.  It puts audible pulses into radio
and wired phone systems for quite a distance, and the charged caps are quite
lethal.  I only use it with an assistant present, who knows CPR, and I
strictly obey the "one hand" rule.

Basically, start the target system, and the sparker, and bring them closer
and closer till the target has an upset.  Apply whatever you think will
help, and try again, noting the difference in distance before the upset.

2007\01\31@191326 by Mark Scoville

flavicon
face
My favorite way to make lots of electical noise is to use a large motor
contactor. The trick is to complete the circuit to the contactor with a
loose wire and a file. Drag the loose wire across the file and the contactor
will chatter like crazy - it's not very scientific, but it does make lots of
EMI.


-- Mark

{Original Message removed}


'[EE] easiest way to create and couple noise into a'
2007\02\01@004826 by John Chung
picon face
Just apply AC and DC together......... It really
depends on the noise range that you want to generate.

John


--- .....peisermaKILLspamspam.....ridgid.com wrote:

{Quote hidden}

> --

2007\02\01@010255 by Ian Rozowsky

flavicon
face
We've recently had great success using a 10kV neon sign transformer, and
two pieces of wire in a "Jacobs Ladder" configuration. We ran an
insulated wire from the base of one of the HV wires, and coiled it above
the unit under test. Each time the Jacobs Ladder arced, a nice noise
pulse was generated. These noise pulses would often hang our system.
This set up allowed us to evaluate various different methods of
recovering from noise related hang-ups.

The obvious caveats apply when working with any HV
systems!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Ian Rozowsky
R&D Director
Centurion Systems (Pty) Ltd.
Box 506 Cramerview 2060 Gauteng South Africa
+27-11-699-2434
http://www.centsys.co.za

Cerno et Prodo!



{Original Message removed}

2007\02\01@025806 by Jinx

face picon face
Put it next to my mother

2007\02\01@090424 by Peiserma

flavicon
face
piclist-bounces@mit.edu wrote:
> For
> small discharges, the main unit charges through a 40 meg high
> voltage resistor into a 100pF capacitor made from circuit
> board material, and that cap is discharged through a 1500 ohm
> resistor into the circuit.

do you use a probe to physically touch the inputs, or do you get an air discharge at these voltages? Sounds like you built your own ESD gun!

> For the harsh stuff, I use the main unit directly, with about
> a 1" air gap, discharging it's main caps.
> The main unit is an encapsulated DC-DC converter with
> 0-30,000V output, charging into three 0.01uF 30kV low
> inductance caps.

holy $#@& I'm impressed :) Where did you get the transformer from? I assume you either did some rewinding of old units, or you used more than one in series?

A scaled-down version of this unit would still be impressive to have on the bench. And perhaps call into question my sanity with my co-workers !

> Basically, start the target system, and the sparker, and
> bring them closer and closer till the target has an upset.
> Apply whatever you think will help, and try again, noting the
> difference in distance before the upset.

yes, this is the sort of bench-test i had in mind. Maybe not quite as lethal as you version though. Have to ask: got any pictures of this device?

2007\02\01@091044 by Peiserma

flavicon
face
piclist-bounces@mit.edu wrote:
> We've recently had great success using a 10kV neon sign
> transformer, and two pieces of wire in a "Jacobs Ladder"
> configuration.


that also sounds intriguing. I'll be looking into this, too. Doesn't sound like it needs too many parts, either

> We ran an insulated wire from the base of one of the HV wires, and coiled it above the unit under test.

just an open-loop wire? So it acted as an antenna?

2007\02\01@142547 by Richard Prosser

picon face
`I think I tried something using a LOPT from a TV at one stage. Just
drove it from a MOSFET and a Sig gen. (& a bench power supply). Could
get quite useful sparks & managed to blow my voltmeter in the process.

RP

On 02/02/07, EraseMEpeisermaspam_OUTspamTakeThisOuTridgid.com <peisermaspamspam_OUTridgid.com> wrote:
{Quote hidden}

> -

2007\02\01@154529 by Paul Anderson

face picon face
On 1/31/07, KILLspampeisermaKILLspamspamridgid.com <RemoveMEpeisermaTakeThisOuTspamridgid.com> wrote:
> Most circuits have some inputs and outputs (external wires) connecting
> them to the real world. I want to create electrical noise and subject an
> unsuspecting PIC and its inputs to noise.
>
>
This may not really apply to your goal, but I remember reading about
this idea in an application note from Analog Devices.  You make a stub
antenna that's a quarter wavelength at the frequency you want to check
for noise, and solder it to one of the device leads.  Then expose to
the RF, and it will dump the RF in the exact predictable spot.  You
desolder the antenna and solder it to different pins to get an idea of
the response to injecting the noise into different pins.  I thought it
was quite clever.



--
Paul Anderson
VE3HOP
spamBeGonewackyvorlonspamBeGonespamgmail.com
http://www.oldschoolhacker.com
"May the electromotive force be with you."

2007\02\01@184054 by Bob Axtell

face picon face
Paul Anderson wrote:
> On 1/31/07, TakeThisOuTpeisermaEraseMEspamspam_OUTridgid.com <RemoveMEpeisermaspamTakeThisOuTridgid.com> wrote:
>  
>> Most circuits have some inputs and outputs (external wires) connecting
>> them to the real world. I want to create electrical noise and subject an
>> unsuspecting PIC and its inputs to noise.
>>
>>
>>    
> This may not really apply to your goal, but I remember reading about
> this idea in an application note from Analog Devices.  You make a stub
> antenna that's a quarter wavelength at the frequency you want to check
> for noise, and solder it to one of the device leads.  Then expose to
> the RF, and it will dump the RF in the exact predictable spot.  You
> desolder the antenna and solder it to different pins to get an idea of
> the response to injecting the noise into different pins.  I thought it
> was quite clever.
>
>
>
>  
I have a standard design that works pretty well for me.  When I have an
RF noise problem, I create
a reliable source by using a coil of #24AWG stranded wire, a 24V AC
wallwart, and a 24V AC relay.

I wire the relay as a buzzer (the coil is in series with its NC-COM
contact path) and the wire to one
of the coil contacts is coiled into a loop about 3" in diameter. I
usually use about 30 turns to have a
good RF radiator.

This arrangement generates a noise source that floods AM/FM  heavily and
sometimes interferes with
cable TV.

When placed next to a PIC with ICSP (using 22K pullup to MCLR pin) this
setup will cause repeated
WDT resets at about 1/2" or closer.  Unless you have placed
ESD-protection devices or tiny caps (10pF+)
on PIC pins with leads longer than a couple of inches, you will
interfere with PIC operation most of the time.
This proves conclusively that a WDT is critical to restart ESD-affected
projects; without WDT, the PIC
may simply lock up. I always use a WDT except when debugging.
Crystal-timed designs are very sensitive,
especially if the crystal-to-PIC leads are more than 1/2".

I use this on all items I design because I design stuff that is near RF
signals of some strength much of the time.
Saves a lot of grief later.

Having said this about PICs, I want to add that PICs are LESS sensitive
than every other uP I've dealt with.

--Bob


2007\02\02@075724 by Peiserma

flavicon
face
piclist-bounces@mit.edu wrote:
> I have a standard design that works pretty well for me.  When
> I have an RF noise problem, I create a reliable source by
> using a coil of #24AWG stranded wire, a 24V AC wallwart, and
> a 24V AC relay.

Thanks for sharing that. I'm going to build one of these,
sounds perfect and not quite as lethal!

I must confess after I googled Jacobs Ladder, I knew one of
those must be built also. Ordered a 12kV neon transformer
off ebay yesterday. Nice weekend project.

> Having said this about PICs, I want to add that PICs are LESS
> sensitive than every other uP I've dealt with.

The one i'm dealing with is not a pic, and appears rather
sensitive. It fails both 61000-4-2 and 61000-4-4. This noise
maker should help verify changes before re-testing.

Thanks for everyone's suggestions.

More... (looser matching)
- Last day of these posts
- In 2007 , 2008 only
- Today
- New search...