Help designing a low-noise TIA

Hi all,

Again, thanks for the conversation. I can't admit I have followed it all. Here is what my current take-aways are.

John Larkin has suggested a differential amplifier for the voltages generated by the rotating and compensating coils.

Jan Panteltje has suggested a transistor amplifier to use the transistor as a current amplifier.

Wit3rd has suggested a charge amplifier, but one that uses a switch to keep the feedback shorted to make sure the capacitor does not send the op-amp to the rails, then open the switch when we make a measurement.

It's not clear if any of these designs will help me with my initial concern of 60 Hz noise. Suggestions for that have included unplugging everything around the measurement system and running it at night (all good ideas), building an electrostatic shielding box around the experiment, adding a 60Hz notch filter on the input, and finally digitally removing 60Hz, 180 Hz, and 300Hz in the signal after we have collected the data.

Have I missed anything vital?

Thanks!

 

[John Larkin]

On Sat, 08 Jun 2019 15:01:45 GMT, Jan Panteltje
<pNaOnStPeAlMtje@yahoo.com> wrote:

On a sunny day (Sat, 08 Jun 2019 07:08:12 -0700) it happened John Larkin
jjlarkin@highlandtechnology.com> wrote in
a5gnfelsbbsa0b1au9mh0uca0dtrm09i9o@4ax.com>:

On Sat, 08 Jun 2019 06:59:37 GMT, Jan Panteltje
pNaOnStPeAlMtje@yahoo.com> wrote:

On a sunny day (Fri, 07 Jun 2019 16:47:59 -0700) it happened John Larkin
jjlarkin@highland_snip_technology.com> wrote in
gmtlfepv8gamvmfbsl7g09qsmhj3dq822h@4ax.com>:

On Fri, 7 Jun 2019 19:41:40 -0400, "Carl"
carl.ijamesXYZ@ZYXverizon.net> wrote:

"Joseph Gwinn" wrote in message
news:0001HW.22AB175D0056A5727000016402EF@news.giganews.com...
While we may not be able to put a battery in the rotating coil, it occurs
to
me that we could put a zero-bias JFET preamp in the coil, power from
outside
via a twisted pair that also carries the amplified signal. Feed from a
constant voltage source and sense the current draft. This way, capacitance
in
the twisted pair has little effect. And the twisted pair cancels magnetic
fields from powering the JFET.

It's Friday and I'm bored waiting for my pizza to arrive - Why not put a
ring of solar cells around the center of the disc that supports the coil,
with a ring of stationary high intensity LED's above them, to power the coil
and on-board preamp, and then put an LED on the axis with a photodiode above
it to get the signal out. No moving contacts, no long leads. Put the solar
cells on the bottom and the output LED on the top of the disc to minimize
any crosstalk.

I was thinking of using a supercap to power a diffamp. If the
amplified signal comes off the spinning coil via a pair of slip rings
or equivalent, the same pair of wires could be used to charge the
supercap, probably through the ESD diodes of the diffamp.

A non-contact way to get the signal off would be great, but requires a
lot more parts.

1 laser diode in center disc shining up?

= photo doide receiver

^ laser diode with simple drive electronicspowered by lithium cell
======== disk
=== bearing
[ | ] break
=== bearing
|
[ ] spring or rubber band driving mechanism, no electric motors and fields
| \0 /
deserted island O experimenter \/ \/drinks O hula girls
///////////////////////////////////////////////////////////////////////////////


Are you proposing that we instrument the angular rotation of hula
girls? That might require an extended series of field experiments.

I once left a woman in a volcano crater on Maui. We weren't getting
along.

Oops, she still there?

No, she later hiked out on her own. Never date lawyers.

Now this..
Take a 74HC4046, center frequency 27 MHz, sweep 1 - 50 MHz for 1 to 5 V makes 10 MHz per volt.
take the 200 nV, and put it into VCO in,
this cause a frequency shift of 2e-7 * 1e7 = 2 Hz.
Small antenna on output 4046.

Tune to 27 MHz with my CB narrow band FM receiver.
Has about +- 5 kHz for 1Vpp out to headphone jack?

So 2 / 5000 volt .4 mV....

Maybe you can already hear that! -46dB

If you add the TIA or something with a gain of 100 at the TX side.
then you get 40 mV....
Clipping is no problem...

This also solves the 'how do I get the signal out' problem.
Somehow I think HC4046 can be configured more sensitive?
Need to read up on the datasheet and what I did in the past with it,
whole lot of those chips in projects.

The CB set runs on battery, no mains needed.
It has, on NBFM, very good SNR.

Would also work with my Tecsun PL600 PLL shortwave receiver, pick any frequency not in use.

I guess he could spin a diffamp and a VCO (4046, or a purchased RF
VCO) and couple the signal out RF-wise or magnetically or
electrostatically or something.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[Joseph Gwinn]

On Jun 8, 2019, ithacacollegephysics@gmail.com wrote
(in article<8042e7dc-1519-44c9-a403-c40b75a1e19e@googlegroups.com&gt:

Hi all,

Again, thanks for the conversation. I can't admit I have followed it all.
Here is what my current take-aways are.


John Larkin has suggested a differential amplifier for the voltages generated
by the rotating and compensating coils.

Jan Panteltje has suggested a transistor amplifier to use the transistor as a
current amplifier.

Joe has suggested a JFET at the coil, powered via the same twisted pair as
carries the amplified signal back to the instrumentation. The JFET turns the
voltage signal into a current signal, which is turned back into an amplified
voltage signal well away from the rotor. .

Wit3rd has suggested a charge amplifier, but one that uses a switch to keep
the feedback shorted to make sure the capacitor does not send the op-amp to
the rails, then open the switch when we make a measurement.

It also sounds like a chopper amplifier would help greatly. The electronic
ones chop at 20 KHz or above, so low-pass filtering is needed to prevent
aliasing et al. Given a signal in the Hertz, a simple RC filter should
suffice. This will greatly slow amplifier offsets from causing drift into the
rails, making drift mitigation or cancellation easier to accomplish.

..

It's not clear if any of these designs will help me with my initial concern
of 60 Hz noise. Suggestions for that have included unplugging everything
around the measurement system and running it at night (all good ideas),
building an electrostatic shielding box around the experiment, adding a 60Hz
notch filter on the input, and finally digitally removing 60Hz, 180 Hz, and
300Hz in the signal after we have collected the data.

As for power-line noise, if one locks the sampling window to the local
power-line frequency such that an integral number of whole power-line cycles
will fit into the sampling window, the power line frequency and all its
harmonics will be profoundly suppressed. It is not required that the window
be synchronized to the power-line fundamental, but that will work. This
filter is widely implemented in digital multimeters - look for the ability to
set the sampling aperture in NPLCs (Number of Power Line Cycles). Handheld
instruments use a crystal clock to make the aperture a multiple of either 50
Hz or 60 Hz cycles - a tenth of a second works for both, but does not adjust
for the actual current power frequency, which varies by a Hertz or two.
Benchtop instruments of high precision will synchronize to the actual power
line cycle, this being essential to achieve the needed degree of cancellation
of power-line interference to support such precision. For DC measurements, I
often set NPLC to 100, and live with the 5-second settling time.

And you will need the electrostatic shields, which can be made of copper
window-screen fabric soldered to copper skeleton wires with a big soldering
iron such as those used when fabricating copper gutters.

Joe Gwinn

 

[Joseph Gwinn]

On Jun 8, 2019, ithacacollegephysics@gmail.com wrote
(in article<326636f8-2b6c-465b-a939-561a584cf4b6@googlegroups.com&gt:

Hi all,

I lot of people are suggested putting the measurement circuit directly on the
rotating coil, which is an interesting idea, but has me worried on a couple
counts. First, the coil is very heavy (2 kg) and spins very quickly (6000
rpm), and all the mass is at the edges of the disk (20 cm in diameter) so
balancing that is critical.

I assume that the wheel is dynamically balanced in two planes using a
commercial balancing system then.

The trick is to keep the electronics near the axis of rotation, connected to
the coil using a radial twisted pair (discussed later). Radial so there is no
contrary acceleration signal there. TP to reduce interference.

..

Second, the large coils are very sensitive magnetic field detectors
(unfortunately -- that's my whole problem!). So I am worried as the current
flows in the circuit on the coil I will measure that current and create some
sort of unintended infinite feedback.

I think that it was you that mentioned that the radial breathing of the coil
caused a signal due to the variation in the area of the coil through which
the earth´s magnetic field could couple. I have not worked out the details,
but it seemed to me that if one could arrange for two counter-rotating coils
in parallel and almost touching, wired such that the electron deceleration
signals sum, the interference from area variation would cancel. I phrased
this in terms of oscillating coils (in a reply to whit3d?), but the principle
is the same.

..

We used to have the speed sensor on the shaft below the coils. It was down
about 10 cm below the coils. It is a little photodiode and laser that senses
an alternating black and white strip of paper. But as the current turns on
and off in the photodiode, I could see that signal in the coil clear as day.
We had to move the motor and speed sensor 1 m away from the coils to make it
go away. So, I am worried that even a small current right on top of the coil
will create enough magnetic field for my coil to pick it up.

This should have worked, and not required a one-meter separation (versus 0.1
meter). I´d venture that the wiring was such that there was a large open
loop of wire supplying the led. If one used twisted pairs to carry the led
drive and photo currents, magnetic effects would be sharply reduced. If
it´s really that sensitive, then use optical fiber to carry light to and
from the paper chopper wheel on the shaft. (One gets a far larger signal if
the chopper wheel is built for use in transmission.)

One can purchase integral 3-channel rotation sensors with considerable
angular rotation using components such as HEDS 9140 from Broadcom. It´s
best to buy the code wheel from Broadcom as well. These components are widely
used in motion-control systems. This approach will allow the tracking of
shaft angle versus time in great detail during a braking event.

Given the sensitivity of the apparatus, all circuits carrying significant
current should be done with twisted pairs. Voltage only circuits will not
cause magnetic interference, but can pick up magnetic signals, and so should
also be twisted. And shielded, to control electrostatic interference.

..<https://en.wikipedia.org/wiki/Twisted_pair>

Joe Gwinn

 

[John Larkin]

On Sat, 8 Jun 2019 07:49:31 -0700 (PDT),
ithacacollegephysics@gmail.com wrote:

Hi all,

Again, thanks for the conversation. I can't admit I have followed it all. Here is what my current take-aways are.

John Larkin has suggested a differential amplifier for the voltages generated by the rotating and compensating coils.

Jan Panteltje has suggested a transistor amplifier to use the transistor as a current amplifier.

Wit3rd has suggested a charge amplifier, but one that uses a switch to keep the feedback shorted to make sure the capacitor does not send the op-amp to the rails, then open the switch when we make a measurement.

It's not clear if any of these designs will help me with my initial concern of 60 Hz noise. Suggestions for that have included unplugging everything around the measurement system and running it at night (all good ideas), building an electrostatic shielding box around the experiment, adding a 60Hz notch filter on the input, and finally digitally removing 60Hz, 180 Hz, and 300Hz in the signal after we have collected the data.

Have I missed anything vital?

Thanks!

The electronics should not be a fundamental limit to the measurement;
the physics is bad enough, what with the motion, ambient fields,
centrifugal growth, thermoelectric and other effects in coupling the
nanovolt signal...

The best signal conditioning would be two voltage-input diffamps, one
per coil, one spinning.

May as well lowpass some as soon as possible, even before the diffamp,
to keep junk from railing the amp and wrecking any integration.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[John Larkin]

On Sat, 8 Jun 2019 07:10:39 -0700 (PDT),
ithacacollegephysics@gmail.com wrote:

Hi all,

I lot of people are suggested putting the measurement circuit directly on the rotating coil, which is an interesting idea, but has me worried on a couple counts. First, the coil is very heavy (2 kg) and spins very quickly (6000 rpm), and all the mass is at the edges of the disk (20 cm in diameter) so balancing that is critical.

Second, the large coils are very sensitive magnetic field detectors (unfortunately -- that's my whole problem!). So I am worried as the current flows in the circuit on the coil I will measure that current and create some sort of unintended infinite feedback.

Any electronics would have to be along the spin axis, tall and skinny,
probably potted, somewhere on the drive shaft. With a bit of care (and
bandwidth limiting) it wouldn't couple into the coil. Some physicist
could do the math!




>We used to have the speed sensor on the shaft below the coils. It was down about 10 cm below the coils. It is a little photodiode and laser that senses an alternating black and white strip of paper. But as the current turns on and off in the photodiode, I could see that signal in the coil clear as day. We had to move the motor and speed sensor 1 m away from the coils to make it go away. So, I am worried that even a small current right on top of the coil will create enough magnetic field for my coil to pick it up.

--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[Jan Panteltje]

On a sunny day (Sat, 08 Jun 2019 08:39:49 -0700) it happened John Larkin
<jjlarkin@highlandtechnology.com> wrote in
<7flnfel7ttk2bk59f05vtas5o4n2105qt0@4ax.com>:

On Sat, 08 Jun 2019 15:01:45 GMT, Jan Panteltje
pNaOnStPeAlMtje@yahoo.com> wrote:

On a sunny day (Sat, 08 Jun 2019 07:08:12 -0700) it happened John Larkin
jjlarkin@highlandtechnology.com> wrote in
a5gnfelsbbsa0b1au9mh0uca0dtrm09i9o@4ax.com>:

On Sat, 08 Jun 2019 06:59:37 GMT, Jan Panteltje
pNaOnStPeAlMtje@yahoo.com> wrote:

On a sunny day (Fri, 07 Jun 2019 16:47:59 -0700) it happened John Larkin
jjlarkin@highland_snip_technology.com> wrote in
gmtlfepv8gamvmfbsl7g09qsmhj3dq822h@4ax.com>:

On Fri, 7 Jun 2019 19:41:40 -0400, "Carl"
carl.ijamesXYZ@ZYXverizon.net> wrote:

"Joseph Gwinn" wrote in message
news:0001HW.22AB175D0056A5727000016402EF@news.giganews.com...
While we may not be able to put a battery in the rotating coil, it occurs
to
me that we could put a zero-bias JFET preamp in the coil, power from
outside
via a twisted pair that also carries the amplified signal. Feed from a
constant voltage source and sense the current draft. This way, capacitance
in
the twisted pair has little effect. And the twisted pair cancels magnetic
fields from powering the JFET.

It's Friday and I'm bored waiting for my pizza to arrive - Why not put a
ring of solar cells around the center of the disc that supports the coil,
with a ring of stationary high intensity LED's above them, to power the coil
and on-board preamp, and then put an LED on the axis with a photodiode above
it to get the signal out. No moving contacts, no long leads. Put the solar
cells on the bottom and the output LED on the top of the disc to minimize
any crosstalk.

I was thinking of using a supercap to power a diffamp. If the
amplified signal comes off the spinning coil via a pair of slip rings
or equivalent, the same pair of wires could be used to charge the
supercap, probably through the ESD diodes of the diffamp.

A non-contact way to get the signal off would be great, but requires a
lot more parts.

1 laser diode in center disc shining up?

= photo doide receiver

^ laser diode with simple drive electronicspowered by lithium cell
======== disk
=== bearing
[ | ] break
=== bearing
|
[ ] spring or rubber band driving mechanism, no electric motors and fields
| \0 /
deserted island O experimenter \/ \/drinks O hula girls
///////////////////////////////////////////////////////////////////////////////


Are you proposing that we instrument the angular rotation of hula
girls? That might require an extended series of field experiments.

I once left a woman in a volcano crater on Maui. We weren't getting
along.

Oops, she still there?

No, she later hiked out on her own. Never date lawyers.

Yes, she would have sued you after hare death...

Now this..
Take a 74HC4046, center frequency 27 MHz, sweep 1 - 50 MHz for 1 to 5 V makes 10 MHz per volt.
take the 200 nV, and put it into VCO in,
this cause a frequency shift of 2e-7 * 1e7 = 2 Hz.
Small antenna on output 4046.

Tune to 27 MHz with my CB narrow band FM receiver.
Has about +- 5 kHz for 1Vpp out to headphone jack?

So 2 / 5000 volt .4 mV....

Maybe you can already hear that! -46dB

If you add the TIA or something with a gain of 100 at the TX side.
then you get 40 mV....
Clipping is no problem...

This also solves the 'how do I get the signal out' problem.
Somehow I think HC4046 can be configured more sensitive?
Need to read up on the datasheet and what I did in the past with it,
whole lot of those chips in projects.

The CB set runs on battery, no mains needed.
It has, on NBFM, very good SNR.

Would also work with my Tecsun PL600 PLL shortwave receiver, pick any frequency not in use.


I guess he could spin a diffamp and a VCO (4046, or a purchased RF
VCO) and couple the signal out RF-wise or magnetically or
electrostatically or something.

Just after posting that, I realized I HAVE a much better receiver:
https://www.ebay.com/itm/272411458376
very good stability,
all you need to type in Linux is
rtl-fm -f <frequency> | play -t raw -r 24k -es -b 16 -c 1 -V1 -
to _hear_ the effect,
and if you want to log it at the same time
rtl-fm -f <frequency> | tee -a logfile.raw | play -t raw -r 24k -es -b 16 -c 1 -V1 -

That USB stick will go as low as 20 MHz, as high as 1.7 GHz,
but 74HC4046 does not go that high, so 20 MHz or so is nice.
Then you can also run my spectrum analyzer, it uses that stick:
http://panteltje.com/panteltje/xpsa/index.html
and see the sidebands .. where possible interference is.
If the guy has a laptop with Linux, that is all you need for playing and logging,
batteries...

Now I am curious, just for fun,
maybe in the coming week I will set that up
simple transistor amp, 4046 with a piece of wire as antenna,
that rtl-sdr stick as NBFM receiver, then
apply a 200 nV 1 kHz square wave to the input, see what we can hear in audio
and see on the analyzer.

Antenna should be in the middle of the coil,
else you get the moving away moving closer sort of Doppler.
One could use capacitive coupling too, 2 metal plates, these sticks are extremely sensitive.

Cracked it?

Ticket to Maui ;-)

 

[John Larkin]

On Tue, 4 Jun 2019 12:28:49 -0700 (PDT),
ithacacollegephysics@gmail.com wrote:

Hi all,

I am posting to this forum at the suggestion of George Herold from Teachspin, he says you are the people to help me out. I understand a decent amount about circuits, but not enough to design the low-noise TIA I would like to build.

We're trying to create a modern version of the Tolman-Stewart experiment https://en.wikipedia.org/wiki/Stewart%E2%80%93Tolman_effect that was one of the first proofs that electrons inside metals carry the current.

In this experiment, a coil of wire is spun to high speeds and then braked rapidly. The electrons keep moving and create a small pulse of current. Originally, Tolman and Stewart used a ballistic galvanometer to act as a charge amplifier and integrate the current to find the total charge.

I’d like to use a TIA to convert the small current pulse into a voltage, then record that voltage as a function of time. The problem is that the coil acts as a giant antenna and picks up all sorts of unwanted noise, so I’d like to get rid of that noise. In particular, it is really good at finding 60 Hz signals in the room.

Right now we’re using an OPA 140 with 1 GOhm and 10 pF as a feedback resistor and capacitor in parallel. We attach the coil (about 200 Ohm resistance, 500 mH inductance) to one input and put the other input across 200 Ohms to ground. The large amplification leads to huge amplification of the noise, and it is hard to see our signal. We expect the current pulse to be 1 nA of current, almost square wave in shape, and it should last the duration of the braking, about 0.5 seconds.

Any suggestions appreciated!

--Matt Sullivan
Ithaca College Physics

Consider this:

Take two identical flat coils very close together (which you have
already) and connect them with a torsional coupling. Set up an angular
oscillation, one against the other; the mechanical drive is trivial.
The electron-slosh potentials add (twice the signal!) and most
everything else subtracts.

It's easy to put diffamps (and MEMS accelerometers?) on the coils,
easy to get the wires out.

At, say, 3 Hz, you get over 20,000 acceleration peaks per hour, 250K
overnight, with no drama, all serene and quiet. Signal average.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[whit3rd]

On Saturday, June 8, 2019 at 8:03:51 AM UTC-7, Jan Panteltje wrote:

On a sunny day (Sat, 8 Jun 2019 07:10:39 -0700 (PDT)) it happened
ithacacollegephysics@gmail.com wrote in
326636f8-2b6c-465b-a939-561a584cf4b6@googlegroups.com>:
So, I am worried that even a small current right on top
of the coil will create enough magnetic field for my coil to pick it up.

Circuit in mu-metal box?

I'm thinking that motion (even just vibration) of the box will be an issue.
Small circuit, low current, attention to the opamp bias source/sink wiring,
and a capacitor rather than a battery (batteries commonly have nickel or
steel cases) can at least minimize that issue, so it could calibrate out.

The whole circuit (op amp, capacitors, analog switch) would have to weigh a gram,
but not much more. Less, if one goes chip-on-board.

 

[whit3rd]

On Saturday, June 8, 2019 at 7:49:36 AM UTC-7, ithacacoll...@gmail.com wrote:

It's not clear if any of these designs will help me with my initial concern of 60 Hz noise. Suggestions for that have included ...
Have I missed anything vital?

Three items that aren't on the list:

(1) gating the sensor to an exact multiple of 1/60 second (alas, that only
compensates 60 Hz currents into non-active loads, though)

(2) there's a possibility that q/m for charge carriers in different metals
is different. So, you could do a one-rotor experiment with counter-wound
dissimilar metals as the coils, and get much better rejection (same temperature,
glued adjacent wires, so very good geometry matching). Bill Sloman
has some experience with matched paired windings.

(3) a remote island, with pleasant weather, and all cellphones etc.
in a box far from the experimental pavilion. Also, comely attendants bearing
chilled beverages...

 

[Joerg]

On 2019-06-08 07:49, ithacacollegephysics@gmail.com wrote:

Hi all,

Again, thanks for the conversation. I can't admit I have followed it
all. Here is what my current take-aways are.

John Larkin has suggested a differential amplifier for the voltages
generated by the rotating and compensating coils.

Jan Panteltje has suggested a transistor amplifier to use the
transistor as a current amplifier.

Wit3rd has suggested a charge amplifier, but one that uses a switch
to keep the feedback shorted to make sure the capacitor does not send
the op-amp to the rails, then open the switch when we make a
measurement.

It's not clear if any of these designs will help me with my initial
concern of 60 Hz noise. Suggestions for that have included
unplugging everything around the measurement system and running it at
night (all good ideas), building an electrostatic shielding box
around the experiment, adding a 60Hz notch filter on the input, and
finally digitally removing 60Hz, 180 Hz, and 300Hz in the signal
after we have collected the data.

Have I missed anything vital?

One other thing is vital. Reduce the gain in the TIA, big time. There is
always a chance of noise getting in especially since you want to run
this during the busy times when your students are there.

When the amp rails it's almost all over, can't use the signal anymore.
So the TIA should only have as much gain as needed to get you over the
non-man-made nosie floor (Johnson noise, noise in the opamp, et cetera).

Now filter the heck out of the signal, preferably using a digital system
such as a laptop run on its battery. The other nice side effect would be
that your students will also learn about how to use a nowadays simple
tool such as a laptop for signal pickup and filtering. Most universities
have LabView. Us "regular folk" who don't have access to academia often
use more pedestrian methods

--
Regards, Joerg

http://www.analogconsultants.com/

 

[whit3rd]

On Sunday, June 9, 2019 at 7:47:50 AM UTC-7, Joerg wrote:

... the TIA should only have as much gain as needed to get you over the
non-man-made nosie floor (Johnson noise, noise in the opamp, et cetera).

Now filter the heck out of the signal, preferably using a digital system
such as a laptop run on its battery.

It sounds good, BUT the laptop has bunches of DC/DC converters between the active
stuff and the battery. You want a small laptop (tablet, maybe?) with 'airplane mode'
and you want it rather far distant from the active region. Distance should
be large compared to the separation of rotating and stationary coils.
Try to find a recent enough chromebook that the display backlight isn't CCFL type.

The easy 'filter' is just to gate the integration of signal over an exact multiple of
1/60 second, and that doesn't take any computing. it won't help much, though,
against the high voltage transformers for a fluorescent-tube backlight.

 

[Joerg]

On 2019-06-09 16:10, whit3rd wrote:

On Sunday, June 9, 2019 at 7:47:50 AM UTC-7, Joerg wrote:

... the TIA should only have as much gain as needed to get you over the
non-man-made nosie floor (Johnson noise, noise in the opamp, et cetera).

Now filter the heck out of the signal, preferably using a digital system
such as a laptop run on its battery.


It sounds good, BUT the laptop has bunches of DC/DC converters between the active
stuff and the battery. You want a small laptop (tablet, maybe?) with 'airplane mode'
and you want it rather far distant from the active region. Distance should
be large compared to the separation of rotating and stationary coils.
Try to find a recent enough chromebook that the display backlight isn't CCFL type.

I bet most modern laptops are LED backlit. Even my more than 10 year old
Samsung NC-10 is. However, they might employ cheap PWM for brightness
control.

Generally laptops are quiet when not fed by their little power brick. My
Gammatech Durabook (semi mil-spec) is quiet while on mains power but
that's a different quality of engineering. I have found sources of very
faint noise where fancy analyzers from Standford Research and others
have failed, using a laptop.

The easy 'filter' is just to gate the integration of signal over an exact multiple of
1/60 second, and that doesn't take any computing. it won't help much, though,
against the high voltage transformers for a fluorescent-tube backlight.

Yes, that, and VFD drive in buildng elevators, A/C systems in the
ceilings, and so on.

Main thing is, don't cram all that gain into the very first stage with
no filtering whatsoever. It's not going to work in situations like Matt's.

--
Regards, Joerg

http://www.analogconsultants.com/

 

[Jan Panteltje]

On a sunny day (Sat, 08 Jun 2019 08:39:49 -0700) it happened John Larkin
<jjlarkin@highlandtechnology.com> wrote in
<7flnfel7ttk2bk59f05vtas5o4n2105qt0@4ax.com>:

I guess he could spin a diffamp and a VCO (4046, or a purchased RF
VCO) and couple the signal out RF-wise or magnetically or
electrostatically or something.

OK, so I grabbed the soldering iron and made a simple 4046 VCO at about 25 MHz.
C 100 pF, R 10k trimpot near minimum, pin 9 at 2 V or so,
5V supply, antennas 1 inch pieces of wire, distance about 1 feet (30 cm duh)
http://panteltje.com/pub/4046_vco_test_setup_IMG_6990.JPG
board in the middle, there is also other stuff on that board, but disconnected
Supply 9V in, also tried lipo battery 11,1V in, regulator LM317,
lots of decoupling caps,
still some crap on the receive side.
http://panteltje.com/pub/4046_carrier.gif
Lots of other stuff running so looks like supply pickup,
added more decoupling caps..
This will work for data transfer if enough input signal.
Needs a low pass before the pin 9 VCO in, to avoid sidebands all over the place.

Receiver is a rtl-sdr stick...
I will think about this a bit and then add a TIA or maybe transistor pre-amp.

One could also frequency modulate a xtal oscillator, stability would be much better,
http://panteltje.com/pub/xtal_carrier.gif
no TX antenna, just the crystal on the other board.

As to SNR... more signal is better.

Anyways, melted some lead...

 

[George Herold]

On Saturday, June 8, 2019 at 10:49:36 AM UTC-4, ithacacoll...@gmail.com wrote:

Hi all,

Again, thanks for the conversation. I can't admit I have followed it all.. Here is what my current take-aways are.

John Larkin has suggested a differential amplifier for the voltages generated by the rotating and compensating coils.

Jan Panteltje has suggested a transistor amplifier to use the transistor as a current amplifier.

Wit3rd has suggested a charge amplifier, but one that uses a switch to keep the feedback shorted to make sure the capacitor does not send the op-amp to the rails, then open the switch when we make a measurement.

It's not clear if any of these designs will help me with my initial concern of 60 Hz noise. Suggestions for that have included unplugging everything around the measurement system and running it at night (all good ideas), building an electrostatic shielding box around the experiment, adding a 60Hz notch filter on the input, and finally digitally removing 60Hz, 180 Hz, and 300Hz in the signal after we have collected the data.

Have I missed anything vital?

Thanks!

Hi Matt, I'd say first thing is to figure out 'what' the 60 Hz is.
Is it electrostatic or magnetic? Then see if you can figure out where
it is coming from. I was looking at pickup from a pair of coils
in my lab. Mostly electrostatic, till I turned the florescent light
off. Then it looked to be magnetic... (has some directional/angle
dependence. But no real sign of 60Hz B-field.. but mostly what I'm
guess are gunk from switch mode power supplies. The gunk
synced with the line, but much higher frequency. The only way
I see 'real' 60 Hz B-fields in my lab is with a coil inside a metal
box.

Re Notch filter: I'm not sure about this. I've never used a notch filter
but you are looking at a transient effect... and the notch may not really
help much. (I'd look at the transient/ step response of your notch.)
A two pole 100ms time constant filter does a good job at killing 60 Hz
stuff in our optical pumping. I don't know if that is too slow for you.

Good luck, try some stuff, and get back to us in a few months. :^)

George H.

 

[Jan Panteltje]

On a sunny day (Tue, 11 Jun 2019 10:11:45 GMT) it happened Jan Panteltje
<pNaOnStPeAlMtje@yahoo.com> wrote in <qdnup7$odi$1@dont-email.me>:

On a sunny day (Sat, 08 Jun 2019 08:39:49 -0700) it happened John Larkin
jjlarkin@highlandtechnology.com> wrote in
7flnfel7ttk2bk59f05vtas5o4n2105qt0@4ax.com>:

I guess he could spin a diffamp and a VCO (4046, or a purchased RF
VCO) and couple the signal out RF-wise or magnetically or
electrostatically or something.


OK, so I grabbed the soldering iron and made a simple 4046 VCO at about 25 MHz.
C 100 pF, R 10k trimpot near minimum, pin 9 at 2 V or so,
5V supply, antennas 1 inch pieces of wire, distance about 1 feet (30 cm duh)
http://panteltje.com/pub/4046_vco_test_setup_IMG_6990.JPG
board in the middle, there is also other stuff on that board, but disconnected
Supply 9V in, also tried lipo battery 11,1V in, regulator LM317,
lots of decoupling caps,
still some crap on the receive side.
http://panteltje.com/pub/4046_carrier.gif
Lots of other stuff running so looks like supply pickup,
added more decoupling caps..
This will work for data transfer if enough input signal.
Needs a low pass before the pin 9 VCO in, to avoid sidebands all over the place.

Receiver is a rtl-sdr stick...
I will think about this a bit and then add a TIA or maybe transistor pre-amp.

One could also frequency modulate a xtal oscillator, stability would be much better,
http://panteltje.com/pub/xtal_carrier.gif
no TX antenna, just the crystal on the other board.

As to SNR... more signal is better.

Anyways, melted some lead...

Update, (after thinking)
To get better center frequency stability and eliminate interference,
I thought 'Oh I can PLL it to a 25 MHz xtal'.
That is the usual system for a FM transmitter, done it before...
used 4046 PLL as demodulator too, in a translation system.
And then
'sh..t I have that, and no need as I have this:
http://panteltje.com/panteltje/pic/fm_pic/
just was not using it, but it works 100%,
has have good SNR on FM, can be changed to mono, wrote the software..
is stable, and any FM radio can get the signal within a range of several meters.

So the link to the coil issue is solved
Also it is legal here, very low power FM.

So, now for that input amp...
thing wants a few hundred mV, maybe those TCL274 as TIA,
or as amp after the first transistor.

 

[Jan Panteltje]

On a sunny day (Tue, 11 Jun 2019 14:28:41 GMT) it happened Jan Panteltje
<pNaOnStPeAlMtje@yahoo.com> wrote in <qdodr0$fiq$1@dont-email.me>:

On a sunny day (Tue, 11 Jun 2019 10:11:45 GMT) it happened Jan Panteltje
pNaOnStPeAlMtje@yahoo.com> wrote in <qdnup7$odi$1@dont-email.me>:

On a sunny day (Sat, 08 Jun 2019 08:39:49 -0700) it happened John Larkin
jjlarkin@highlandtechnology.com> wrote in
7flnfel7ttk2bk59f05vtas5o4n2105qt0@4ax.com>:

I guess he could spin a diffamp and a VCO (4046, or a purchased RF
VCO) and couple the signal out RF-wise or magnetically or
electrostatically or something.


OK, so I grabbed the soldering iron and made a simple 4046 VCO at about 25 MHz.
C 100 pF, R 10k trimpot near minimum, pin 9 at 2 V or so,
5V supply, antennas 1 inch pieces of wire, distance about 1 feet (30 cm duh)
http://panteltje.com/pub/4046_vco_test_setup_IMG_6990.JPG
board in the middle, there is also other stuff on that board, but disconnected
Supply 9V in, also tried lipo battery 11,1V in, regulator LM317,
lots of decoupling caps,
still some crap on the receive side.
http://panteltje.com/pub/4046_carrier.gif
Lots of other stuff running so looks like supply pickup,
added more decoupling caps..
This will work for data transfer if enough input signal.
Needs a low pass before the pin 9 VCO in, to avoid sidebands all over the place.

Receiver is a rtl-sdr stick...
I will think about this a bit and then add a TIA or maybe transistor pre-amp.

One could also frequency modulate a xtal oscillator, stability would be much better,
http://panteltje.com/pub/xtal_carrier.gif
no TX antenna, just the crystal on the other board.

As to SNR... more signal is better.

Anyways, melted some lead...


Update, (after thinking)
To get better center frequency stability and eliminate interference,
I thought 'Oh I can PLL it to a 25 MHz xtal'.
That is the usual system for a FM transmitter, done it before...
used 4046 PLL as demodulator too, in a translation system.
And then
'sh..t I have that, and no need as I have this:
http://panteltje.com/panteltje/pic/fm_pic/
just was not using it, but it works 100%,
has have good SNR on FM, can be changed to mono, wrote the software..
is stable, and any FM radio can get the signal within a range of several meters.

So the link to the coil issue is solved
Also it is legal here, very low power FM.

So, now for that input amp...
thing wants a few hundred mV, maybe those TCL274 as TIA,
or as amp after the first transistor.

So.. after a night sleep...
And thinking about the FM modulator and audio like bandwidth low noise thing
a quick search with google for
'microphone preamplifier integrated circuit 80dB'
found this as first hit:
http://www.thatcorp.com/press34.shtml

I downloaded the datasheet from
http://www.thatcorp.com/1510-1512_Audio_Preamplifier_ICs.shtml

This chip is made for 200 Ohm dynamic microphones, so low impedance inductive source.
It had decent noise specs, and 60 dB gain (1000x voltage gain).
So a 200 nV signal results then in 200 uV.
Add an other normal opamp for an other 1000 x voltage gain, and we have 200 mV,
enough to drive the FM modulator module.
You could add any form of clipping and filtering for mains frequencies in that chain,
or go digital and write to SDcard or EEPROM locally (on the coil, no need for link)
but better to stay away from all that power stuff and use the deserted island solution,
students will like that much better,
combine with a survival training, and you will even be ready for Agent Orange's provoked wars.
If you go digital here is a very good noise filter:
http://panteltje.com/panteltje/newsflex/download.html#humfilter
it a comb filter, so also attenuates the harmonics.
Raspberry Pi?
PIC writing to SDcard? space for thousands of runs.

Any ideas?

 

[George Herold]

On Wednesday, June 12, 2019 at 5:29:07 AM UTC-4, Jan Panteltje wrote:

On a sunny day (Tue, 11 Jun 2019 14:28:41 GMT) it happened Jan Panteltje
pNaOnStPeAlMtje@yahoo.com> wrote in <qdodr0$fiq$1@dont-email.me>:

On a sunny day (Tue, 11 Jun 2019 10:11:45 GMT) it happened Jan Panteltje
pNaOnStPeAlMtje@yahoo.com> wrote in <qdnup7$odi$1@dont-email.me>:

On a sunny day (Sat, 08 Jun 2019 08:39:49 -0700) it happened John Larkin
jjlarkin@highlandtechnology.com> wrote in
7flnfel7ttk2bk59f05vtas5o4n2105qt0@4ax.com>:

I guess he could spin a diffamp and a VCO (4046, or a purchased RF
VCO) and couple the signal out RF-wise or magnetically or
electrostatically or something.


OK, so I grabbed the soldering iron and made a simple 4046 VCO at about 25 MHz.
C 100 pF, R 10k trimpot near minimum, pin 9 at 2 V or so,
5V supply, antennas 1 inch pieces of wire, distance about 1 feet (30 cm duh)
http://panteltje.com/pub/4046_vco_test_setup_IMG_6990.JPG
board in the middle, there is also other stuff on that board, but disconnected
Supply 9V in, also tried lipo battery 11,1V in, regulator LM317,
lots of decoupling caps,
still some crap on the receive side.
http://panteltje.com/pub/4046_carrier.gif
Lots of other stuff running so looks like supply pickup,
added more decoupling caps..
This will work for data transfer if enough input signal.
Needs a low pass before the pin 9 VCO in, to avoid sidebands all over the place.

Receiver is a rtl-sdr stick...
I will think about this a bit and then add a TIA or maybe transistor pre-amp.

One could also frequency modulate a xtal oscillator, stability would be much better,
http://panteltje.com/pub/xtal_carrier.gif
no TX antenna, just the crystal on the other board.

As to SNR... more signal is better.

Anyways, melted some lead...


Update, (after thinking)
To get better center frequency stability and eliminate interference,
I thought 'Oh I can PLL it to a 25 MHz xtal'.
That is the usual system for a FM transmitter, done it before...
used 4046 PLL as demodulator too, in a translation system.
And then
'sh..t I have that, and no need as I have this:
http://panteltje.com/panteltje/pic/fm_pic/
just was not using it, but it works 100%,
has have good SNR on FM, can be changed to mono, wrote the software..
is stable, and any FM radio can get the signal within a range of several meters.

So the link to the coil issue is solved
Also it is legal here, very low power FM.

So, now for that input amp...
thing wants a few hundred mV, maybe those TCL274 as TIA,
or as amp after the first transistor.

So.. after a night sleep...
And thinking about the FM modulator and audio like bandwidth low noise thing
a quick search with google for
'microphone preamplifier integrated circuit 80dB'
found this as first hit:
http://www.thatcorp.com/press34.shtml

I downloaded the datasheet from
http://www.thatcorp.com/1510-1512_Audio_Preamplifier_ICs.shtml

Hmm, I wonder about the input bias current and offset. ~1 uA
=~ 200uV, x1000 is 200mV... or is that not a problem?

JL's low noise int amp had a bunch of bias current too...
so maybe that isn't a problem.

George H.

This chip is made for 200 Ohm dynamic microphones, so low impedance inductive source.
It had decent noise specs, and 60 dB gain (1000x voltage gain).
So a 200 nV signal results then in 200 uV.
Add an other normal opamp for an other 1000 x voltage gain, and we have 200 mV,
enough to drive the FM modulator module.
You could add any form of clipping and filtering for mains frequencies in that chain,
or go digital and write to SDcard or EEPROM locally (on the coil, no need for link)
but better to stay away from all that power stuff and use the deserted island solution,
students will like that much better,
combine with a survival training, and you will even be ready for Agent Orange's provoked wars.
If you go digital here is a very good noise filter:
http://panteltje.com/panteltje/newsflex/download.html#humfilter
it a comb filter, so also attenuates the harmonics.
Raspberry Pi?
PIC writing to SDcard? space for thousands of runs.

Any ideas?

 

Hi everyone,

Thanks for all the suggestions. Again, I definitely did not follow them all. My current plan is to try the TIA or the integrator, or just the difference amplifier. I will also take heed of all the suggestions and try and reduce the noise as much as I can by turning things off, running on a laptop, and running at night. I will let you know how it goes!

--Matt.

 

[Jan Panteltje]

On a sunny day (Wed, 12 Jun 2019 05:49:54 -0700 (PDT)) it happened George
Herold <gherold@teachspin.com> wrote in
<9fc32656-f716-43e5-86c8-99b12f4f8248@googlegroups.com>:

On a sunny day (Tue, 11 Jun 2019 10:11:45 GMT) it happened Jan Panteltje
pNaOnStPeAlMtje@yahoo.com> wrote in <qdnup7$odi$1@dont-email.me>:
So.. after a night sleep...
And thinking about the FM modulator and audio like bandwidth low noise thing
a quick search with google for
'microphone preamplifier integrated circuit 80dB'
found this as first hit:
http://www.thatcorp.com/press34.shtml

I downloaded the datasheet from
http://www.thatcorp.com/1510-1512_Audio_Preamplifier_ICs.shtml

Hmm, I wonder about the input bias current and offset. ~1 uA
=~ 200uV, x1000 is 200mV... or is that not a problem?

I think for a reasonable temperature range subtracting a few hundred mV and setting zero before the measurement
is not a problem.

JL's low noise int amp had a bunch of bias current too...
so maybe that isn't a problem.

One could look for better mike pre-amps chips.
Also can we use AC ? We will see a pulse AFAIK?

I see that 1512 on ebay, 2 for 20 $, plus shipping from US source:
https://www.ebay.com/itm/262773847594