Resolve sub-millivolts with a PIC?

[mike]

Resolve sub-millivolts with a PIC?

This is a one-off project.

I want to measure the output voltage of a peltier device.
I have access to both wires. Voltage is under 10mV.
I'd like 7 bits or so of precision.
I said "precision". I don't care nearly as much about
the accuracy, I can calibrate that out.
Needs to have short term stability at fixed temperature
long enough to get from calibration cycle to measurement
cycle.
I want to cobble this onto an existing project that uses
a PIC16F627 to send data out the serial port.
It has no A/D converter. I really don't want to add
instrumentation amps to boost the signal anyway.

I've been looking into ways to generate a voltage
with the PWM, stack the peltier device onto that,
stuff it into the comparator...and "fix it in software".
I fear that I'll never get the system noise down to the
point where this is practical, but never hurts to ask.

Yes, I know that a couple of thermistors would be easier
to manage electrically, but much more difficult mechanically.
The Peltier device is just the right mechanical configruation
for what I want. And I have it in my hand, as do I the
pic16F627 system.

I'm not much interested in advice on using different devices
or technologies. I can think of lots of 'em. I can get what I want
with an IR thermometer and a calculator, but I'm bored with that.
I'm interested
in using what I already have to get me where I want to go.

A technique to measure small voltages would have general
applicability to other situations.

So, any clever ways to resolve sub-millivolt signals with a PIC
(without internal A/D) and not
much else?

Thanks, mike

 

[Joerg]

Resolve sub-millivolts with a PIC?

This is a one-off project.

I want to measure the output voltage of a peltier device.
I have access to both wires. Voltage is under 10mV.
I'd like 7 bits or so of precision.
I said "precision". ...

You meant "resolution"? Precision is usually something absolute

... I don't care nearly as much about
the accuracy, I can calibrate that out.
Needs to have short term stability at fixed temperature
long enough to get from calibration cycle to measurement
cycle.
I want to cobble this onto an existing project that uses
a PIC16F627 to send data out the serial port.
It has no A/D converter. I really don't want to add
instrumentation amps to boost the signal anyway.

I've been looking into ways to generate a voltage
with the PWM, stack the peltier device onto that,
stuff it into the comparator...and "fix it in software".
I fear that I'll never get the system noise down to the
point where this is practical, but never hurts to ask.

Yes, I know that a couple of thermistors would be easier
to manage electrically, but much more difficult mechanically.
The Peltier device is just the right mechanical configruation
for what I want. And I have it in my hand, as do I the
pic16F627 system.

I'm not much interested in advice on using different devices
or technologies. I can think of lots of 'em. I can get what I want
with an IR thermometer and a calculator, but I'm bored with that.
I'm interested
in using what I already have to get me where I want to go.

A technique to measure small voltages would have general
applicability to other situations.

So, any clever ways to resolve sub-millivolt signals with a PIC
(without internal A/D) and not
much else?

Have you looked into dual slope AD conversion? Essentially your PIC
(hoping it has at least a comparator ...), plus a couple resistors and a
cap. I think Microchip even has an app note about that. The resistors
set the scale factor.

 

[Jon Kirwan]

You meant "resolution"? Precision is usually something absolute

I think I was fine with his usage. He used the term to
assure the rest of us that he, himself, wasn't confused about
the idea of accuracy vs precision. Which was wise.

I don't like your phrasing, but I won't bicker. Others can
come to their own conclusions. But since you took issue,
I'll use that as a foil to just put out a few words and let
folks comment on their own understanding of each of them:

repeatability
resolution
precision
accuracy
correctness

Three of those can be considered independent concepts that
can stand alone (the other two are dependent entirely upon
developed relationships between particular pairs of those
three.)

Jon

 

[nospam]

I've been looking into ways to generate a voltage
with the PWM, stack the peltier device onto that,
stuff it into the comparator...and "fix it in software".
I fear that I'll never get the system noise down to the
point where this is practical, but never hurts to ask.

Sounds like you have been trying to invent a delta sigma converter.
If your input is stable you can average out noise. The comparator input
offset stability is probably the limiting factor.

 

[Spehro Pefhany]

So, any clever ways to resolve sub-millivolt signals with a PIC
(without internal A/D) and not
much else?

Thanks, mike

Not particularly clever, but that part has some rather crappy
comparators built in (10mV Vos maximum).

You could make a slope converter using a resistor as an almost
constant current source, a capacitor, and time transitions.
Probably a few resistors, a cap, a 4053 or even a 2N7002 and that's
about it.

I can't be bothered to think too much more about this, since it's such
an artificial constraint-- almost like a classroom exercise.


Best regards,
Spehro Pefhany

 

[Frank Buss]

Have you looked into dual slope AD conversion? Essentially your PIC
(hoping it has at least a comparator ...), plus a couple resistors and a
cap. I think Microchip even has an app note about that. The resistors
set the scale factor.

There is an application note about delta sigma AD conversion:

http://ww1.microchip.com/downloads/en/AppNotes/00700a.pdf

I wonder if dual slope woul be possible with passive components (besides
the PIC and the internal comperator), only. You do need at least one switch
for switching the input voltage you want to measure?

 

[mike]

I mentioned temperature only in the context of the voltage measurement
system stability.

Heat flow is exactly what I want to measure.

Both ends would have to be prettyI believe the surfaces represent the proper measurement conditions.
and you'd have to know one temperature, for the
Again, it's heat flow. I don't care about either temperature.

If you do want to measure a small temperature difference, Peltier
junctions can be nice - the Johnson noise level is low, and you don't
get into trouble with self-heating.

IIRR some calorimetry projects go further and use a feedback system to
inject enough current into a second Peltier junction to compensate for
the heat flowing the through sensing (zero current)Peltier junction at
the measured temperature difference, as well as through the
compensating Peltier junction.

I've been considering using the peltier device
to drive its own voltage to zero. The heat flow through the device should
be a function of the energy input required to achieve zero
open-circuit voltage...if I can successfully factor out
the efficiency/self heating of the device. Thermal time constants are
long enough that I should be able to multiplex the stimulus and measurement.

 

[mike]

There is an application note about delta sigma AD conversion:

http://ww1.microchip.com/downloads/en/AppNotes/00700a.pdf

I wonder if dual slope woul be possible with passive components (besides
the PIC and the internal comperator), only. You do need at least one switch
for switching the input voltage you want to measure?

Thanks for the link, but I was hoping for something more clever.
For example, I should be able to return the peltier device to
the reference instead of ground. That way, the whole counter resolution
can be spread over the maximum input voltage plus the comparator offset
plus the loading effects on the reference.
Maybe parallel two more outputs across the peltier device so I can PWM
some energy into it to drive its voltage to zero.
I haven't figured it out yet, but my gut tells me something interesting
can be done with the 10-bits of the pwm/ccp module.

 

[mike]

You meant "resolution"? Precision is usually something absolute

I knew that would be an issue. I agonized over it before changing the
subject to "resolve...". I think "precision" works in this context
as one could include the noise issues that degrade the precision
from the number you'd expect to get from the resolution alone.

 

[Joerg]

There is an application note about delta sigma AD conversion:

http://ww1.microchip.com/downloads/en/AppNotes/00700a.pdf

That is a very good one.

I wonder if dual slope woul be possible with passive components (besides
the PIC and the internal comperator), only. You do need at least one switch
for switching the input voltage you want to measure?

Can't find the Microchip example back, I think it was call PIC
multimeter or something like that. Very simple, just the uC and some
discretes.

 

[Joerg]

Thanks for the link, but I was hoping for something more clever.
For example, I should be able to return the peltier device to
the reference instead of ground. That way, the whole counter resolution
can be spread over the maximum input voltage plus the comparator offset
plus the loading effects on the reference.
Maybe parallel two more outputs across the peltier device so I can PWM
some energy into it to drive its voltage to zero.
I haven't figured it out yet, but my gut tells me something interesting
can be done with the 10-bits of the pwm/ccp module.

PWM is the way to go, but your VCC for the uC must be super precise if
you want to do that without any external parts. If it ain't then I'd
consider a TLV431 or LMV431.

 

[Frank Buss]

Can't find the Microchip example back, I think it was call PIC
multimeter or something like that. Very simple, just the uC and some
discretes.

Maybe this one?

http://ww1.microchip.com/downloads/en/AppNotes/00512e.pdf

They don't mention "dual slope", but of course it is the dual slope
concept. But you can't measure an input voltage with it, only resistance,
so something for converting small voltages to a resistance is required,
maybe a FET?

 

[Jamie]

I mentioned temperature only in the context of the voltage measurement
system stability.



Heat flow is exactly what I want to measure.

Both ends would have to be pretty


I believe the surfaces represent the proper measurement conditions.
and you'd have to know one temperature, for the



Again, it's heat flow. I don't care about either temperature.



I've been considering using the peltier device
to drive its own voltage to zero. The heat flow through the device should
be a function of the energy input required to achieve zero
open-circuit voltage...if I can successfully factor out
the efficiency/self heating of the device. Thermal time constants are
long enough that I should be able to multiplex the stimulus and
measurement.

It almost looks like you're looking for a kelvin bridge input?

 

[Joerg]

Maybe this one?

http://ww1.microchip.com/downloads/en/AppNotes/00512e.pdf

They don't mention "dual slope", but of course it is the dual slope
concept. But you can't measure an input voltage with it, only resistance,
so something for converting small voltages to a resistance is required,
maybe a FET?

You can also do it for voltage measurements, without needing an external
mux. I can't find the Microchip note back but here is the one from NXP:

www.nxp.com/documents/application_note/AN10187.pdf

 

[Frank Buss]

You can also do it for voltage measurements, without needing an external
mux. I can't find the Microchip note back but here is the one from NXP:

www.nxp.com/documents/application_note/AN10187.pdf

Ok, but you'll need an internal mux But this is possible with the
PIC16C62X, too.

 

[Joerg]

Ok, but you'll need an internal mux But this is possible with the
PIC16C62X, too.

Ok, but so far the uCs I had to deal with had a mux in front of the
comparator. A uC without one wouldn't be all that useful to an analog
guy like me

Essentially that's just about the only way to zero out the comparator
input offset which is usually quite horrid in uC. Well, back to my new
toy now, this one just arrived:

http://www.mastechpowersupply.com/f...nction-signal-generator-sfg-2120/prod_58.html

So far (except for a rather serious amplitude setting glitch) it's been
quite impressive.

 

[mike]

You can also do it for voltage measurements, without needing an external
mux. I can't find the Microchip note back but here is the one from NXP:

www.nxp.com/documents/application_note/AN10187.pdf

Thanks, but I've been woefully inadequate at getting my point across.

The methods suggested are charge balance measurements.

A sigma-delta converter is a big bucket. There's a hole in the bottom
and the size of the hole represents the quantity being measured.
You sit at the top and teaspoon more water in to keep the water level
constant. The number of teaspoons per unit time has to equal the
volume of water running out the bottom.

Similarly, a single slope converter lets the water level fall to
a threshold and you dump a whole lot of water in the top to bring the
level instantly back to the top threshold. Same concept, bigger
spoon and you only do it once/cycle. Time measurement vs count...
potato...potato...

These are both current controlled devices. You're measuring the current
produced thru the resistor from the input to the reference voltage for
sigma-delta
and the current charging the cap in the slope converter.

Neither of these are very helpful when you want your input voltage
dynamic range to be half the offset spec of the comparator.

In the sigma-delta case, it's like trying to keep the level in a
5-gallon bucket constant by pouring from a peanut butter
jar into a straw sealed into the top of the bucket without a funnel.
Some of the water gets in, some doesn't. The "slop" in the system
exceeds the quantities being measured.

What I'm looking for is a CLEVER topology that gets rid of the slop.
It's the funnel in the above analogy. You're more likely to find
such an item in a patent disclosure than in "A/D for dummies".

The obvious "funnel" is a voltage to current converter that gets
the signal into the proper paradigm for these types of converters.
Might even be able to do it with a modified current mirror.
The nice thing about current is that the voltage on the cap keeps
changing until it reaches the threshold...even if the threshold ain't
exactly where you expect it. You can calibrate out systematic
errors. You can't do nothin' if the voltage never gets to the threshold.

Somebody asked for numbers. The numbers I gave in the first post
are what I want. 10mV full scale and 1% resolution.
But I did dig out a real voltmeter. I wrapped the peltier device
in cellulose...aka a paper towel, stuck it into the plastic lid
of a spray can and set it on the desk. After a LONG time,
the voltage did settle to about 100nanovolts. That's the best this
old meter can do. I was surprised it did anywhere that well as I
took zero precautions to balance the thermals in the interconnect.
I can see significant voltage transient if I flex the coax that connects it.

If I take it out of its cocoon, the voltage is all over the place.
Stick it on a window with 14F inside/outside temperature differential,
voltage goes to 100mV or so and dribbles down to about 2mV. It's the
2mV I want to measure. my 1% resolution is worth 5% at 20% of FS.

All seems consistent with my initial estimates.

The back story is this...
I got a new gas furnace and a bunch of insulation added to the house.
I've been graphing heat input.
Consistent with my minimalist approach, I used a flapper on
a microswitch sitting on a vent and plugged into the serial port
of a Palm IIIc...and fixed it in software.

I've been seeing significant anomalies that I can't explain by outside
air temperature fluctuations.
Consistent with my lazy approach, the measurement system was a bottom-up
design...who am I kidding, it wasn't designed, it was evolved...
I had no idea what I was measuring. After I did the math,
I discovered that the system has a RESOLUTION of 17 BTU. I don't claim
that it's anywhere near that accurate, but it does resolve well.
Turns out that I can see when the computer goes into standby or when
I'm watching TV mirrored in gas consumption.

By this time, I'm into measuring losses. My windows are specified
as R3.125. But I didn't measure anywhere near that high. And when
I added R5 foam on the outside the improvement was more than I expected
from R8.125/R3.125. That's where all the peltier device stuff started.

So, I have no idea how they measure the insulating capabilities of
windows, but the numbers seem optimistic.
This is an interesting reference:

http://www.omega.com/manuals/manualpdf/M1331.pdf

It's the operator/application manual for their IR thermometer
tweaked to measure building heat loss.

If you sandwich the window between two metal plates and measure R, you get
a small number. If you measure ambient air temperature in still air
on both sides, you get a much bigger number. If you use the method
outlined in the link, you get a third number.

There are all kinds of issues with boundary layers, convection, wind
etc. on both sides
of a window. I started looking into those effects. And I need some
repeatable method to measure it.

Why do I want to do this? Because I can...and I'm bored.

Still looking for that clever way to resolve small voltages with
no hardware.

 

[Frank Buss]

http://www.mastechpowersupply.com/f...nction-signal-generator-sfg-2120/prod_58.html

So far (except for a rather serious amplitude setting glitch) it's been
quite impressive.

Nice, but 25 MHz? Isn't this like audio frequency for you?

 

[Frank Buss]

These are both current controlled devices. You're measuring the current
produced thru the resistor from the input to the reference voltage for
sigma-delta
and the current charging the cap in the slope converter.

Neither of these are very helpful when you want your input voltage
dynamic range to be half the offset spec of the comparator.

I don't know much about analog electronics, but IIRC the idea of the dual
slope converter is to cancel out any inaccuracies of the comparator,
because it is a ratiometric measurement.

I've done a quick amateurish test: A 12 V Battery powered LM393: At
inverting input a voltage devider with two 10 k resistors and at the
non-inverting input two 33 k resistors and a 2 k potentiometer for setting
the raw offset, and a 100 ohm potentiometer in series for fine tuning. Some
100 nF and 1 uF at the inputs to ground, to avoid oscillating and noise. At
the output a LED and a 1 k resistor to +12 V.

Then I was able to switch on the LED with very small movements of the 100
ohm potentiometer. Sometimes the LEDs blinks, faster if I move the hand
some cm near or away the circuit, so maybe more shielding and better setup
is required. But I measured the difference of the potentiometer when it
switched the LED on and the calculated voltage change was below 0.2 mV
(hystereses was much larger with my setup).

I think the resolution depends on the voltage gain and the short-term
stability. The LM393 has a minimum voltage gain of 50 V/mV, so you could
measure with a resolution of 0.1 mV, if you need 5 V output. Is it possible
to cascade two comparators without problems?

The PIC16F627 datasheet and the comparator has +/-10 mV maximum input
offset voltage, but I didn't found anything about the gain. But maybe for
high precision dual slope measurement you'll need an external comparator
anyway.

 

[TheJoker]

Nice, but 25 MHz? Isn't this like audio frequency for you?

The brand says it all.

Sounds like "Qwacko". Bwuahahahahahahaha!