The Larkin front end

[Eeyore]

If you're referring to capacitive imbalances turning common-mode stuff
into differential, that's a different issue.

No different in effect from inductive imbalance caused by off the shelf ferrite beads doing
that too.

In the case of load
cells, with ballpark 200 ohm drives and 0.1 uF caps, and a deep 50/60
Hz notch in the delta-sigma adc response, it's not an issue at all; do
the math. The ferrites are dead shorts at these frequencies. With a 25
mV full-scale load-cell signal, I can weigh things to 5-digit noise
and linearity with *unshielded* leads.

Line rejection is a trivial issue here, something obvious that simple
math can handle. RF rectification is far more insidious, less
predictable, and more dangerous (a little AC noise is nothing compared
to hard railing the ADC) and what you see on the schematic is just
part of the story.

Dealing with RF was entirely my point. I've wound bifilar chokes to keep the balance
correct on differential inputs.

Another thing I don't understand is why you insist on calling me
stupid when you disagree with me. I've never called you stupid.

You're the only one here to use the word !

Graham

 

[Eeyore]

The only complication being that it does.

In the AM band where most of the problems lie ? Around 1MHz ?

Graham

 

[Jim Thompson]

On Wed, 08 Nov 2006 07:33:06 +0000, Eeyore
[snip]

You're not as smart as I once thought.



Consider the inherent rejection of common mode signals in a differential front end. It
doesn't work so well if the signals are different sizes ( rather obviously ).

I'll use 1% caps here too.

Graham

If you're referring to capacitive imbalances turning common-mode stuff
into differential, that's a different issue. In the case of load
cells, with ballpark 200 ohm drives and 0.1 uF caps, and a deep 50/60
Hz notch in the delta-sigma adc response, it's not an issue at all; do
the math. The ferrites are dead shorts at these frequencies. With a 25
mV full-scale load-cell signal, I can weigh things to 5-digit noise
and linearity with *unshielded* leads.

Line rejection is a trivial issue here, something obvious that simple
math can handle. RF rectification is far more insidious, less
predictable, and more dangerous (a little AC noise is nothing compared
to hard railing the ADC) and what you see on the schematic is just
part of the story.

Another thing I don't understand is why you insist on calling me
stupid when you disagree with me. I've never called you stupid.

John

You're very polite, John.

Eeyore/Graham IS stupid AND ignorant ;-)

...Jim Thompson

 

[Joerg]

Hello Rich,

Hey, there's something that'd never occurred to me - if the core is loose,
it will actually become like a little motor, right?

Never seen one spin

But often, upon diagnosing some EMI problem at a client, I found shards
in the base of the system. It's amazing what a rough flight or a truck
barreling along a washboard road can do to electronics. Goes back to the
first law of Murphy: If there is anything that can break it will.

 

[Jim Thompson]

Hello Rich,


Never seen one spin

But often, upon diagnosing some EMI problem at a client, I found shards
in the base of the system. It's amazing what a rough flight or a truck
barreling along a washboard road can do to electronics. Goes back to the
first law of Murphy: If there is anything that can break it will.

I did some testing on Kodak's shake "table" in Rochester. The had a
simulation profile called "Truck".

We were warned not to stand on the "Table", it would shake your
kidneys loose :-(

Particularly disturbing to me was watching the necks of CRT's with a
strobe, and seeing them FLEXING!

...Jim Thompson

 

[John Larkin]

On Wed, 08 Nov 2006 07:33:06 +0000, Eeyore
[snip]

You're not as smart as I once thought.


If you lowpass filter both
sides effectively, you're way better off, even if the lowpass
filtering isn't exactly the same on both sides. The transistors in the
front-end of an opamp are essentially square-law RF rectifiers. If you
have 100 mv of RF, they will rectify it big-time. If you have 1 mv,
they barely will at all. 100^2 = 10000, and 10000:1 is usually
considered to be a pretty big improvement.

Consider the inherent rejection of common mode signals in a differential front end. It
doesn't work so well if the signals are different sizes ( rather obviously ).

I'll use 1% caps here too.

Graham

If you're referring to capacitive imbalances turning common-mode stuff
into differential, that's a different issue. In the case of load
cells, with ballpark 200 ohm drives and 0.1 uF caps, and a deep 50/60
Hz notch in the delta-sigma adc response, it's not an issue at all; do
the math. The ferrites are dead shorts at these frequencies. With a 25
mV full-scale load-cell signal, I can weigh things to 5-digit noise
and linearity with *unshielded* leads.

Line rejection is a trivial issue here, something obvious that simple
math can handle. RF rectification is far more insidious, less
predictable, and more dangerous (a little AC noise is nothing compared
to hard railing the ADC) and what you see on the schematic is just
part of the story.

Another thing I don't understand is why you insist on calling me
stupid when you disagree with me. I've never called you stupid.

John

You're very polite, John.

Eeyore/Graham IS stupid AND ignorant ;-)

...Jim Thompson

Well, at least he does let his emotions interfere with his thinking.

John

 

[Jim Thompson]

On Wed, 08 Nov 2006 08:28:34 -0800, John Larkin
[snip]

You're very polite, John.

Eeyore/Graham IS stupid AND ignorant ;-)

...Jim Thompson

Well, at least he does let his emotions interfere with his thinking.

John

Sno-o-o-o-ort ;-)

...Jim Thompson

 

[John Larkin]

In the AM band where most of the problems lie ? Around 1MHz ?

Graham

The guts of most ICs still work fine at 1 MHz, so little rectification
takes place. And wiring and pcb traces don't resonate at low
frequencies like this. High-Q resonances often make EMI sensitivity at
some frequencies tens of times worse than the broadband sensitivity.

Our company is in a wooden building in sight of the Twin Peaks
broadcast tower and not far from the Mt San Bruno antenna farm. If we
scope things, we see all sorts of stuff, the biggest usually around 1
MHz, a big AM station. But if we look at things like diffamp outputs
using a bandlimited scope - the rectified stuff - we don't usually see
audio, which is what we'd get from detecting AM... we see what's
obviously video! And when hand-waving or grabbing cables shifts DC
offsets, ferrite beads will usually help a lot.

http://www.chamoismoon.com/Resources/Murals/SF-Sunset-TP.jpg


Beads are low-inductance, very lossy inductors. They come as close to
the magical human finger damping effect as you can get in a
surface-mount part.

John

 

[John Fields]

You're very polite, John.

Eeyore/Graham IS stupid AND ignorant ;-)

...Jim Thompson

 

[Rich Grise]

A ferrite bead won't do that.

The only complication being that it does.[/QUOTE]

It sounds to me like we're comparing lossy low-pass filters to common-mode
chokes here. Isn't that kind of like apples and tomatoes?

Thanks,
Rich

 

[Joerg]

Hello Jim,

I did some testing on Kodak's shake "table" in Rochester. The had a
simulation profile called "Truck".

There are various grades when simulating "truck". There is the smooth
glide along Interstate 80, then there is the not so smooth rumble from
Shannon Airport to Ballinsloe, and then there is the driver who tries to
find the "speed of least shaking" on a sastrugi track through the Sahara
desert.

We were warned not to stand on the "Table", it would shake your
kidneys loose :-(

Particularly disturbing to me was watching the necks of CRT's with a
strobe, and seeing them FLEXING!

After finishing a hi-rel module they showed me the impact tables where
it'll all be tested. Now I knew where that regular ka-bummmmph sound was
coming from. They raised the unit several feet and then let it drop onto
a concrete pad. Then again. And again. Every once in a while someone
would pour another layer. The concrete slowly became a pillar that drove
itself into the ground.

 

[Joerg]

Hello John,

Our company is in a wooden building ...

Wooden? Oh man. I hope it has at least a light-weight upper floor and
roof, for when the next big one hits. I remember a picture you posted
with a scarily mounted lifting crane.

... in sight of the Twin Peaks
broadcast tower and not far from the Mt San Bruno antenna farm. If we
scope things, we see all sorts of stuff, the biggest usually around 1
MHz, a big AM station. But if we look at things like diffamp outputs
using a bandlimited scope - the rectified stuff - we don't usually see
audio, which is what we'd get from detecting AM... we see what's
obviously video! And when hand-waving or grabbing cables shifts DC
offsets, ferrite beads will usually help a lot.

During an EMI job on an ultrasound machine they didn't really believe we
had shortwave radio interference. Until I asked them for the register
address that could change the Doppler CF and fiddled with it. The sound
of Big Ben was echoing through the engineering lab and the BBC World
Service news came out of the speakers...

http://www.chamoismoon.com/Resources/Murals/SF-Sunset-TP.jpg

Once I saw that from a plane taking off. Only the top stuck out after we
punched through the thick cloud layer. I could still bite myself that I
had left the camera in the baggage bin.

Beads are low-inductance, very lossy inductors. They come as close to
the magical human finger damping effect as you can get in a
surface-mount part.

Yes. You want molasses-like behavior when fighting EMI.

 

[Eeyore]

The guts of most ICs still work fine at 1 MHz, so little rectification
takes place. And wiring and pcb traces don't resonate at low
frequencies like this. High-Q resonances often make EMI sensitivity at
some frequencies tens of times worse than the broadband sensitivity.

Our company is in a wooden building in sight of the Twin Peaks
broadcast tower and not far from the Mt San Bruno antenna farm. If we
scope things, we see all sorts of stuff, the biggest usually around 1
MHz, a big AM station. But if we look at things like diffamp outputs
using a bandlimited scope - the rectified stuff - we don't usually see
audio, which is what we'd get from detecting AM... we see what's
obviously video! And when hand-waving or grabbing cables shifts DC
offsets, ferrite beads will usually help a lot.

http://www.chamoismoon.com/Resources/Murals/SF-Sunset-TP.jpg

Beads are low-inductance, very lossy inductors. They come as close to
the magical human finger damping effect as you can get in a
surface-mount part.

Indeed, their lossiness is very handy.

Have you ever tried intentionally injecting modulated RF into a front end btw ?

Graham

 

[Eeyore]

Considering you've never seen me !!!!

Can you still pull 20 somethings ?

Graham

 

[Eeyore]

It sounds to me like we're comparing lossy low-pass filters to common-mode
chokes here. Isn't that kind of like apples and tomatoes?

It depends on the application.

There's no reason the ferrite in a wound choke can't be lossy too.

Graham

 

[APR]

Can you still pull 20 somethings ?

Graham

Are we talking weight (stone) ????

 

[martin griffith]

Considering you've never seen me !!!!

Can you still pull 20 somethings ?

Graham

20 what, pushups? ....


martin

 

[Jim Thompson]

Hello Jim,


There are various grades when simulating "truck". There is the smooth
glide along Interstate 80, then there is the not so smooth rumble from
Shannon Airport to Ballinsloe, and then there is the driver who tries to
find the "speed of least shaking" on a sastrugi track through the Sahara
desert.



After finishing a hi-rel module they showed me the impact tables where
it'll all be tested. Now I knew where that regular ka-bummmmph sound was
coming from. They raised the unit several feet and then let it drop onto
a concrete pad. Then again. And again. Every once in a while someone
would pour another layer. The concrete slowly became a pillar that drove
itself into the ground.

I was told at Ford 40 years ago that our packaging had to stand a
"boxcar" test.

Huh?

It was explained to me: Shove the shipping container off of a boxcar
deck height onto the pavement below ;-) Talk about "ka-bummmmph"!

...Jim Thompson

 

[Glen Walpert]

On Wed, 08 Nov 2006 08:28:34 -0800, John Larkin

... With a 25
mV full-scale load-cell signal, I can weigh things to 5-digit noise
and linearity with *unshielded* leads.

The load cell based weigh scales used for mixing semiconductor gasses
at Praxair have a 100 kg capacity with 1 mg resolution - 8 digits. The
plant manager claimed that they were the most accurate scales in
existence. I didn't get the linearity spec, but I would be surprised
if it was worse than 7 digits, as linearity is the only thing that
counts in this application and cost is nearly irrelevant. Some
processes require dopants mixed to concentrations as low as 200 ppb
(in multiple dilution steps), where once mixed it is impossible to
verify mixing accuracy with any existing instrument. If you can come
up with a mass spectrometer that can measure dopant and impurity
concentration with a few ppb accuracy (as opposed to the few tens of
ppb now available) you could probably sell at least one to every
semiconductor gasses plant in existence. All 3 of them IIRC .

Have you looked at the TI ADS1232/4 - looks pretty good for moderate
accuracy load cell weighing, as does the older and more available
AD7730 (I am still waiting for my ADS1232 eval board to arrive,
ordered several months ago). I intend to put it in the load cell with
only a few cm of strain gage leads inside the shielded load cell
enclosure, with filtering then only required on the power leads to a
precision linear regualtor also inside the load cell enclosure.

 

[Eeyore]

Are we talking weight (stone) ????

Years of age.

Graham