Searching \ for '[EE] High Resolution ADC?' 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/io/atod.htm?key=adc
Search entire site for: 'High Resolution ADC?'.

Exact match. Not showing close matches.
PICList Thread
'[EE] High Resolution ADC?'
2009\02\13@190837 by solarwind

picon face
Hey all, does anyone know of any high resolution ADC chips? I'm trying
to build a simple electrocardiograph as a fun experiment and I need a
good ADC chip to connect to my PIC. I don't think the 10 bit ADC
onboard is good enough, correct me if I'm wrong.

--
solarwind

2009\02\13@193747 by Bob Blick

face
flavicon
face
On Fri, 13 Feb 2009 19:07:56 -0500, "solarwind"
<spam_OUTx.solarwind.xTakeThisOuTspamgmail.com> said:
> Hey all, does anyone know of any high resolution ADC chips? I'm trying
> to build a simple electrocardiograph as a fun experiment and I need a
> good ADC chip to connect to my PIC. I don't think the 10 bit ADC
> onboard is good enough, correct me if I'm wrong.

10 bits is plenty - you need to do a frontend amplifier (with isolation,
please!) with adjustable (either manual or automatic/digital) gain.

Check out some of the online ciscuits for EKGs and just hook your PIC
into the place where the chart recorder would go.

Did I mention isolation? The simplest ones use optoisolated amps with
batteries.

Cheers,

Bob

--
http://www.fastmail.fm - Email service worth paying for. Try it for free

2009\02\13@200538 by solarwind

picon face
On Fri, Feb 13, 2009 at 7:37 PM, Bob Blick <.....bobblickKILLspamspam@spam@ftml.net> wrote:
> 10 bits is plenty - you need to do a frontend amplifier (with isolation,
> please!) with adjustable (either manual or automatic/digital) gain.
>
> Check out some of the online ciscuits for EKGs and just hook your PIC
> into the place where the chart recorder would go.
>
> Did I mention isolation? The simplest ones use optoisolated amps with
> batteries.
>
> Cheers,
>
> Bob
>

Thanks! I can understand isolation when connected to a computer, but
why when using batteries?

http://www.eng.utah.edu/~jnguyen/ecg/ecg_index.html

That's one I found...



--
solarwind

2009\02\13@201114 by Vitaliy

flavicon
face
solarwind wrote:
> Hey all, does anyone know of any high resolution ADC chips? I'm trying
> to build a simple electrocardiograph as a fun experiment and I need a
> good ADC chip to connect to my PIC. I don't think the 10 bit ADC
> onboard is good enough, correct me if I'm wrong.

2^10 = 1024, why would you need more steps?

Depending on what your experiment is about, you may be better off using the
8-bit mode so you wouldn't have to worry as much about dealing with noise.

Vitaliy

2009\02\13@201651 by solarwind

picon face
I still have to learn about analog circuits, ADC, ADC sampling
frequency, ADC sampling resolution, ADC voltage range, analog signal
amplification, analog filters (low pass/high pass, that sort of thing)
and so on. If you have any information whatsoever on this, I would
much appreciate it. Already did some research and found some good
sites to learn.

2009\02\13@202616 by solarwind

picon face
On Fri, Feb 13, 2009 at 8:09 PM, Vitaliy <spamspamKILLspammaksimov.org> wrote:
> 2^10 = 1024, why would you need more steps?

This is the thing - I don't know much about analog circuits, but I
want to learn. en.wikipedia.org/wiki/Analog-to-digital_converter
the wikipedia article says that the voltage resolution is equal to
overall range divided by the discrete intervals (2^10 in this case)?
But I don't know what the voltage range of the ADC in the PIC is.

> Depending on what your experiment is about, you may be better off using the
> 8-bit mode so you wouldn't have to worry as much about dealing with noise.

Question: how does using 8 bit mode (for the ADC?) help with dealing
with noise? I'm planning to make some sort of a bandpass filter to
filter out high frequency noise.

--
solarwind

2009\02\13@202751 by solarwind

picon face
Found a cool page here: www.sensorsmag.com/articles/0701/24/main.shtml

2009\02\13@212637 by Vitaliy

flavicon
face
"solarwind" wrote:
> On Fri, Feb 13, 2009 at 8:09 PM, Vitaliy <.....spamKILLspamspam.....maksimov.org> wrote:
>> 2^10 = 1024, why would you need more steps?
>
> This is the thing - I don't know much about analog circuits, but I
> want to learn. en.wikipedia.org/wiki/Analog-to-digital_converter
> the wikipedia article says that the voltage resolution is equal to
> overall range divided by the discrete intervals (2^10 in this case)?
> But I don't know what the voltage range of the ADC in the PIC is.

I don't know what PIC you're using, but this information is specified in the
datasheet (IIRC, the maximum voltage for the ADC is something like 0.7VDD
for an 8-bit PIC).

So what you do, is you determine the range you're dealing with "in the real
world", and then you scale this voltage to fit the range your ADC can
handle. In simple terms, if ADC range is 0..5V, then every step is:

5,000 mV / 1024 = 4.88 mV

If you're measuring 0..100V, you just use a voltage divider to get this down
to 0..5V. In your case, you would amplify the weak signals to get them into
the ADC range.


>> Depending on what your experiment is about, you may be better off using
>> the
>> 8-bit mode so you wouldn't have to worry as much about dealing with
>> noise.
>
> Question: how does using 8 bit mode (for the ADC?) help with dealing
> with noise? I'm planning to make some sort of a bandpass filter to
> filter out high frequency noise.

On a recent project, we were sampling the output of a pot, using a 12-bit
ADC. Pots are inherently noisy, so the numbers would jump by hundreds,
despite the filters we put on. Additional filtering (averaging) had to be
done in software.

Lower resolution would do the rounding for you. Or you could just throw away
the lower bits.

Vitaliy

2009\02\13@213330 by Bob Blick

face
flavicon
face
solarwind wrote:
>
>
> Thanks! I can understand isolation when connected to a computer, but
> why when using batteries?

If the whole thing, including your display, is run from batteries and
not connected to anything else, you don't need isolation.

But if it has connections to any other device, it should be isolated.
And then you need to power the amplifiers from an isolated source, hence
batteries in the preamp.

>
> www.eng.utah.edu/~jnguyen/ecg/ecg_index.html
>
> That's one I found...

So instead of connecting it to the sound card, connect it to a PIC and a
graphic LCD? Sure, sounds good.

High resolution is not needed. Look at the vertical resolution of your
monitor, if it's 1200 pixels, that's just over 10 bits. So 10 bits is
more than enough for a graphic LCD or a window on a monitor. You use the
preamp to scale the signal to a usable voltage.

Cheers,

Bob

2009\02\13@215427 by solarwind

picon face
On Fri, Feb 13, 2009 at 9:24 PM, Vitaliy <EraseMEspamspam_OUTspamTakeThisOuTmaksimov.org> wrote:
> I don't know what PIC you're using, but this information is specified in the
> datasheet (IIRC, the maximum voltage for the ADC is something like 0.7VDD
> for an 8-bit PIC).
>
> So what you do, is you determine the range you're dealing with "in the real
> world", and then you scale this voltage to fit the range your ADC can
> handle. In simple terms, if ADC range is 0..5V, then every step is:
>
> 5,000 mV / 1024 = 4.88 mV
>
> If you're measuring 0..100V, you just use a voltage divider to get this down
> to 0..5V. In your case, you would amplify the weak signals to get them into
> the ADC range.

Yes, I think the voltages on the skin near the heart are in the range
of 1 - 5 mV so I need to amplify them a LOT.

> On a recent project, we were sampling the output of a pot, using a 12-bit
> ADC. Pots are inherently noisy, so the numbers would jump by hundreds,
> despite the filters we put on. Additional filtering (averaging) had to be
> done in software.
>
> Lower resolution would do the rounding for you. Or you could just throw away
> the lower bits.

Ok, I think I'm going to throw away a few of the lower bits and use a
bandpass filter.

--
solarwind

2009\02\13@221926 by solarwind

picon face
Also, what is the maximum sampling frequency that the PIC can do?

2009\02\13@224302 by Bob Blick

face
flavicon
face
16F series figure 40uS per channel. If you are just doing one channel
you can shave some off the track-and-hold time. That's just off the top
of my head, the datasheet has better numbers but takes some
headscratching to sort out all the phases of the conversion process.

Cheers, Bob

solarwind wrote:
> Also, what is the maximum sampling frequency that the PIC can do?

2009\02\13@224658 by Vitaliy

flavicon
face
"solarwind" wrote:
> Also, what is the maximum sampling frequency that the PIC can do?

Datasheet is your friend.


2009\02\13@230542 by solarwind

picon face
On Fri, Feb 13, 2009 at 10:42 PM, Bob Blick <bobblickspamspam_OUTftml.net> wrote:
> 16F series figure 40uS per channel. If you are just doing one channel
> you can shave some off the track-and-hold time. That's just off the top
> of my head, the datasheet has better numbers but takes some
> headscratching to sort out all the phases of the conversion process.
>
> Cheers, Bob

I believe 40 us is more than enough speed to measure heart beat. Thanks!

--
solarwind

2009\02\14@013634 by Rich

picon face
The Linear Tech LTC1605 is a really good 19-bit ADC.  I have used it on
various projects.  It is only 100 Ksps but if you don't need more than that
it works great.

{Original Message removed}

2009\02\14@022840 by solarwind

picon face
On Sat, Feb 14, 2009 at 1:35 AM, Rich <@spam@rgrazia1KILLspamspamrochester.rr.com> wrote:
> The Linear Tech LTC1605 is a really good 19-bit ADC.  I have used it on
> various projects.  It is only 100 Ksps but if you don't need more than that
> it works great.

Thanks for mentioning the company. I looked at their other chips. They
seem to have a lot of options...
--
solarwind

2009\02\14@023334 by solarwind

picon face
I'm going to request a few samples from them.

2009\02\14@025039 by solarwind

picon face
www.cisl.columbia.edu/kinget_group/student_projects/ECG%20Report/E6001%20ECG%20final%20report.htm

Found another good site. It says electrode voltages will be in the
range of 1 - 5 mV and the atrial systole will take about 100 ms at 75
bpm. So, if I want to show a heart beat at 225 bpm (3 * 75), that'll
be about 33 ms or about 30 Hz. So ya, 100K samples/s sounds good
enough to me. And the PIC takes about 40 us to do a conversion which
is about 25 KHz. Sounds good to me. So all I need is a PIC and some
amplification circuitry I guess.

2009\02\14@093554 by olin piclist

face picon face
solarwind wrote:
> Hey all, does anyone know of any high resolution ADC chips? I'm trying
> to build a simple electrocardiograph as a fun experiment and I need a
> good ADC chip to connect to my PIC. I don't think the 10 bit ADC
> onboard is good enough, correct me if I'm wrong.

I'm not that familiar with electrocardiograph signals, but I expect 10 bits
is way more than enough for the signal itself.  If a high resolution is
commonly used, this is probably to account for a large dynamic range between
different setups as apposed to the need for high resolution of the signal
itself once it scaled into range.  Since this is a one-off experiment, you
can probably compensate for the unknown range of the input signal with a
adjustable gain amp, essentially a volume control for the signal.  Keep in
mind that nurses and doctors routinely read these waveforms off a 3 inch
high display from 5 feet away.  Even 8 bits would be enough for that.

If you really need high resolution, check out the Microchip sigma-delta
A/Ds.  These can be 18 bits or more at up to 60Hz.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2009\02\14@094016 by olin piclist

face picon face
Vitaliy wrote:
> 2^10 = 1024, why would you need more steps?
>
> Depending on what your experiment is about, you may be better off
> using the 8-bit mode so you wouldn't have to worry as much about
> dealing with noise.

Throwing out low bits is not a good way to deal with noise.  A much better
way is to read the 10 bit A/D significantly faster than needed, then apply
several poles of low pass filtering in software.  After that you can
truncate to 8 bits if that's all you need and using a single byte for
readings becomes useful.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2009\02\14@095002 by olin piclist

face picon face
solarwind wrote:
> This is the thing - I don't know much about analog circuits, but I
> want to learn.
> en.wikipedia.org/wiki/Analog-to-digital_converter
> the wikipedia article says that the voltage resolution is equal to
> overall range divided by the discrete intervals (2^10 in this case)?
> But I don't know what the voltage range of the ADC in the PIC is.

The PIC A/D makes a number proprotional to where the input voltage is within
the Vref- to Vref+ range.  For now keep things simple by not using separate
Vref inputs.  This means the A/D input range will be ground to Vdd.  Use a
5V regulator to make Vdd and you'll be all set for the purposes of your
experiment.

>> Depending on what your experiment is about, you may be better off
>> using the 8-bit mode so you wouldn't have to worry as much about
>> dealing with noise.
>
> Question: how does using 8 bit mode (for the ADC?) help with dealing
> with noise?

Short answer: Not very well.  See my other post.

> I'm planning to make some sort of a bandpass filter to
> filter out high frequency noise.

You only need a low pass filter for that.  Since this is inherently a AC
signal, you do need one high pass filter stage somewhere, but that is really
just to block DC.  In practise, your circuit will need to block DC in the
front end to keep the DC from overwhelming later stages.  Then it needs to
add Vdd/2 DC offset at the end to keep the signal centered in the A/D input
range.  Your firmware will need to remove this offset, probably right up
front by only passing the difference signal to the rest of the system.
Technically that's a filter, but your "real" filter will probably be two or
three poles of low pass with the rolloff frequency carefully chosen.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2009\02\14@095333 by olin piclist

face picon face
solarwind wrote:
> Ok, I think I'm going to throw away a few of the lower bits and use a
> bandpass filter.

Don't do that.  Apply the filter, *then* throw away extra low bits you don't
need.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2009\02\14@095342 by olin piclist

face picon face
solarwind wrote:
> Also, what is the maximum sampling frequency that the PIC can do?

RTFM!


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2009\02\14@095619 by olin piclist

face picon face
solarwind wrote:
> I believe 40 us is more than enough speed to measure heart beat.

I would probably set up a 1mS periodic interrupt and read the A/D at that
rate, then apply the low pass filtering.  You'll be sampling a lot more
information than you really need, but the extra information allows a better
filtering job.  Then you can decimate and truncate the information after the
filter if you want to.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2009\02\14@104025 by Vitaliy

flavicon
face
Olin Lathrop wrote:
>> Depending on what your experiment is about, you may be better off
>> using the 8-bit mode so you wouldn't have to worry as much about
>> dealing with noise.
>
> Throwing out low bits is not a good way to deal with noise.  A much better
> way is to read the 10 bit A/D significantly faster than needed, then apply
> several poles of low pass filtering in software.  After that you can
> truncate to 8 bits if that's all you need and using a single byte for
> readings becomes useful.

I suggested using 8-bit mode in the interest of keeping the software simple.
I don't have direct experience with capturing cardiograms, but I've seen
several student projects and it looked very trivial. Based on solarwind's
description, he just needs to capture the pulse (the peaks), he probably
doesn't care about the waveform.

Vitaliy

2009\02\14@113610 by solarwind

picon face
On Sat, Feb 14, 2009 at 10:38 AM, Vitaliy <KILLspamspamKILLspamspammaksimov.org> wrote:
> I suggested using 8-bit mode in the interest of keeping the software simple.
> I don't have direct experience with capturing cardiograms, but I've seen
> several student projects and it looked very trivial. Based on solarwind's
> description, he just needs to capture the pulse (the peaks), he probably
> doesn't care about the waveform.
>
> Vitaliy

Actually, I want to try and get the best waveform I can. The guy in
the link I posted got a really nice waveform. Just getting the peak to
measure bpm would be a lot simpler I believe. My treadmill has two
plates on each handlebar where you can hold on to measure your heart
rate.


--
solarwind

2009\02\14@130839 by Sean Breheny

face picon face
I have not made an EKG device myself, but I've read about it and while
I think it is within
the "doability" range of a beginner, it is not easy. The back end (ADC
and display) is the easiest part.
The hardest part is the instrumentation amp and leads. The minimum is
considered to be three leads,
two of them are balanced inputs to the instrumentation amp and the
third is a ground reference so that
the common-mode range of the amp is not exceeded. Usually the two
sense leads are placed on either side of the heart
and the reference lead is placed on the side of the abdomen or on the leg.

Attaching these leads to the person is not easy. If you just take wire
and touch it to the skin, you will
get lots of noise from the varying amount of contact with the skin.
Even with the professional
adhesive "button" contacts, patient movement often results in noise
which is called "artifact"

Also, you will probably need to reject 60Hz (or whatever the power
line frequency is in your country) very strongly.

I would guess that they have a bandpass response with a very low
bottom cutoff (perhaps 0.1Hz) and an upper cutoff
of perhaps 15Hz, along with a notch filter for power line frequency if
the rolloff of the main filter is not fast enough to
very highly attenuate 50/60Hz.

As others have said, isolation is considered essential if this is
going to have any connection to the mains power. Even if
the only connection to mains is to ground the person, it would be bad
to do so without more than 1 Meg ohm of resistance between
the person and ground. This is because hard-grounding someone makes
them MUCH more likely to be killed if they touch
something which is at a large voltage with respect to ground (like the
hot side of the AC line or the case of equipment which has a lot of
leakage current).

One option for isolation would be to have a battery-powered wireless
module which does the measurement and digitizing, and then sends the
data to a PC.
You could also do the same thing but use optocouplers instead of RF.

You could also use a laptop PC, make sure that it is NOT connected to
the AC line in any way but instead running from its batteries, and
then connect your
device to the laptop.

If you want a real hard-wired connection to a line-powered PC, you
will have to get an isolated instrumentation amp AND then make sure
that you don't do
anything which defeats the isolation.

Sean


On Sat, Feb 14, 2009 at 10:38 AM, Vitaliy <RemoveMEspamTakeThisOuTspammaksimov.org> wrote:
> I suggested using 8-bit mode in the interest of keeping the software simple.
> I don't have direct experience with capturing cardiograms, but I've seen
> several student projects and it looked very trivial. Based on solarwind's
> description, he just needs to capture the pulse (the peaks), he probably
> doesn't care about the waveform.
>
> Vitaliy
>
> -

2009\02\14@134040 by John Coppens

flavicon
face
On Sat, 14 Feb 2009 13:08:37 -0500
Sean Breheny <spamBeGoneshb7spamBeGonespamcornell.edu> wrote:

> Also, you will probably need to reject 60Hz (or whatever the power
> line frequency is in your country) very strongly.

And don't forget, as good as the input amplfiers have to be (milivolts),
they hay to resist shocking the hart back into functioning.

I remember readlng an interesting article in a (very old, when they still
published it) "Hewlett Packard Journal", where all the components are
described in detail.

John

2009\02\14@134858 by Sean Breheny

face picon face
Perhaps they have to be tolerant to that in professional use, but I
hope that solarwind is not planning on doing home do-it-yourself
defibrillation!!!

On Sat, Feb 14, 2009 at 1:39 PM, John Coppens <TakeThisOuTjohnEraseMEspamspam_OUTjcoppens.com> wrote:
{Quote hidden}

> -

2009\02\14@135131 by Vitaliy

flavicon
face
Sean Breheny wrote:
>I have not made an EKG device myself, but I've read about it and while
> I think it is within
> the "doability" range of a beginner, it is not easy.

The projects I've seen, were done by biotech students. I was told that the
problem of getting/conditioning the signal has a precanned solution.

An engineer we hired in 2007 used EKG to control a robot arm, for his
student project, I guess I can ask him.

Vitaliy

2009\02\14@135625 by solarwind

picon face
On Sat, Feb 14, 2009 at 1:48 PM, Sean Breheny <shb7EraseMEspam.....cornell.edu> wrote:
> Perhaps they have to be tolerant to that in professional use, but I
> hope that solarwind is not planning on doing home do-it-yourself
> defibrillation!!!

Lol, no, I just thought it would be fun to do some work with analog
stuff and record my heart's work throughout a 24 hour period. Would be
nice to see how the heart beats during various stages of sleep, waking
up, eating, playing counterstrike, exercising, talking on PICLIST
(stressful) and so on :P
--
solarwind

2009\02\14@142754 by Vitaliy

flavicon
face
"solarwind" wrote:
>> I suggested using 8-bit mode in the interest of keeping the software
>> simple.
>> I don't have direct experience with capturing cardiograms, but I've seen
>> several student projects and it looked very trivial. Based on solarwind's
>> description, he just needs to capture the pulse (the peaks), he probably
>> doesn't care about the waveform.
>>
>
> Actually, I want to try and get the best waveform I can. The guy in
> the link I posted got a really nice waveform.

I see. What would you be looking for, besides the heart rate?


> Just getting the peak to
> measure bpm would be a lot simpler I believe. My treadmill has two
> plates on each handlebar where you can hold on to measure your heart
> rate.

Yes, getting the heart rate sounds like a very trivial problem to solve. I'm
wearing a watch with a built-in heart rate monitor.

Vitaliy

2009\02\14@155533 by John Coppens

flavicon
face
On Sat, 14 Feb 2009 13:55:44 -0500
solarwind <EraseMEx.solarwind.xspamgmail.com> wrote:

> > Perhaps they have to be tolerant to that in professional use, but I
> > hope that solarwind is not planning on doing home do-it-yourself
> > defibrillation!!!
>
> Lol, no, I just thought it would be fun to do some work with analog
> stuff and record my heart's work throughout a 24 hour period. Would be
> nice to see how the heart beats during various stages of sleep, waking
> up, eating, playing counterstrike, exercising, talking on PICLIST
> (stressful) and so on :P

;-) just saying. Read somewhere that static voltage increased about 100V
per meter height.

Anyway, I'm suspecting that measuring in all those situations will be
near to impossible. Generally they lay the patient down - mostly on a
table which has some conductive surface (below the plastic), to improve
noise and balance.

I'd be interested in seeing if anything reasonable can be plotted while
sitting up, in front of computer and keyboard  which are actively
irradiating pulses throughout the spectrum. (not to mention CRTs. Though
LCDs do they best too)

John

2009\02\14@155854 by Sean Breheny

face picon face
I am under the impression that most heart-rate monitors do not acquire
the signal electrically but by looking for changes in the reflectance
of the skin at near IR frequencies. As the heart beats, the pressure
pulse in the arterioles and capillaries causes increased quantities of
blood momentarily, which causes (I think) more absorption of certain
IR wavelengths.

Sean


On Sat, Feb 14, 2009 at 2:26 PM, Vitaliy <RemoveMEspamEraseMEspamEraseMEmaksimov.org> wrote:
> Yes, getting the heart rate sounds like a very trivial problem to solve. I'm
> wearing a watch with a built-in heart rate monitor.
>
> Vitaliy
>
> -

2009\02\14@160052 by Sean Breheny

face picon face
Are you talking about atmospheric electricity here? That wouldn't
apply indoors and it would be greatly affected by the body's
conductivity anyway.

I personally have had an EKG while standing so I know that it is not a
problem, per se. I think it is usually done lying down for comfort, to
minimize movement, and because a full EKG requires, I think, 10 leads
(to map out which region of the heart each particular pulse is coming
from).

Sean


On Sat, Feb 14, 2009 at 3:54 PM, John Coppens <RemoveMEjohnspam_OUTspamKILLspamjcoppens.com> wrote:
{Quote hidden}

> -

2009\02\14@160846 by Alan B. Pearce

face picon face
>I'm not that familiar with electrocardiograph signals, but I
>expect 10 bits is way more than enough for the signal itself.
>If a high resolution is commonly used, this is probably to
>account for a large dynamic range between different setups
>as apposed to the need for high resolution of the signal
>itself once it scaled into range.

One of the problems you get is muscle artifacts that are much larger than
the heartbeat signal, which cause significant varying offsets. It is now
some 35 years since I had dealings with designing electrocardiograph gear,
so don't have any experience with doing any digital processing of it.

In our case the isolation was achieved by radio link.

2009\02\14@161831 by solarwind

picon face
Ok so I'll try to make a bpm monitor first and forget about the
waveform. So first step is to make something that will just capture
your heart's bpm. Any ideas guys?

2009\02\14@163837 by Alan B. Pearce

face picon face
>I remember readlng an interesting article in a (very old,
>when they still published it) "Hewlett Packard Journal",
>where all the components are described in detail.

Yes, that article came out at about the time I was involved in heart monitor
development. that makes it mid to late '70s.

Someone posted a link here some time back to an archive of HP journals.

Right, found the HP Labs link, and Sept 1977 looks like it will give you
some ideas.
http://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/hpjindex.html

Its not the one I was look for, but will give you some ideas.

2009\02\14@164605 by Alan B. Pearce

face picon face
>Anyway, I'm suspecting that measuring in all those situations
>will be near to impossible. Generally they lay the patient
>down - mostly on a table which has some conductive surface
>(below the plastic), to improve noise and balance.

The unit I worked on was designed to allow the patient to be mobile. We had
a 'leaky feeder' consisting of unterminated 300 ohm ribbon cable run through
the facility as the receiving aerial, and the standard trick when the system
was demonstrated was for one of us to wear the transmitter while wandering
around. System seemed to work real well.

2009\02\14@174357 by Rich

picon face
If speed is not important a voltage to Frequency converter can be configured
to 18 or 20 bits.

----- Original Message -----
From: "Olin Lathrop" <RemoveMEolin_piclistTakeThisOuTspamspamembedinc.com>
To: "Microcontroller discussion list - Public." <EraseMEpiclistspamspamspamBeGonemit.edu>
Sent: Saturday, February 14, 2009 9:36 AM
Subject: Re: [EE] High Resolution ADC?


{Quote hidden}

> --

2009\02\14@174408 by Vitaliy

flavicon
face
Sean Breheny wrote:
>> Yes, getting the heart rate sounds like a very trivial problem to solve.
>> I'm
>> wearing a watch with a built-in heart rate monitor.
>
> I am under the impression that most heart-rate monitors do not acquire
> the signal electrically but by looking for changes in the reflectance
> of the skin at near IR frequencies. As the heart beats, the pressure
> pulse in the arterioles and capillaries causes increased quantities of
> blood momentarily, which causes (I think) more absorption of certain
> IR wavelengths.

The one in my watch definitely acquires the pulse using electrical means. I
don't know the exact mechanism, but basically it's using the base of the
watch as one of the electrodes, and a metal button on the face as the #2
electrode.

Vitaliy

2009\02\14@180139 by Jinx

face picon face
> heart-rate monitors do not acquire the signal electrically but by
> looking for changes in the reflectance of the skin at near IR

A very early electronics project of mine that could have been saleable
was a collection of biofeedback sensors. Biofeedback was all the rage
in the 80s. Based on the 8-bit ADC080x, with signals optically coupled
to and processed and graphically displayed by a Commodore 64. The
screen was split horizontally, with moving sliders (copied from the
original Star Trek medical lab) up top and moving graphs below

The pulse-rate monitor was a finger clamp with an IR beam through
the fingertip. Worked very well. Probably still does, in storage

Although it didn't have an ECG, it did have a three-connection EMG
for measuring muscle activity. Placed properly it could be used as an
approximate ECG, producing on the screen what looked just like you'd
expect. The EMG was the most sensitive part of the project, needing
good supply, screening etc. I don't recall the components off-hand but
they weren't anything too fancy

2009\02\14@181346 by Tony Smith

flavicon
face
> On Sat, Feb 14, 2009 at 1:48 PM, Sean Breheny <RemoveMEshb7KILLspamspamcornell.edu> wrote:
> > Perhaps they have to be tolerant to that in professional use, but I
> > hope that solarwind is not planning on doing home do-it-yourself
> > defibrillation!!!
>
> Lol, no, I just thought it would be fun to do some work with analog
> stuff and record my heart's work throughout a 24 hour period. Would be
> nice to see how the heart beats during various stages of sleep, waking
> up, eating, playing counterstrike, exercising, talking on PICLIST
> (stressful) and so on :P


It's fascinating when you discover your sleeping heart rate can vary between
40 & 180 BPM.

Tony

2009\02\14@182713 by Jinx

face picon face
> Yes, getting the heart rate sounds like a very trivial problem to solve

Something like this is the 'traditional' method - light beam and PIC
processing

http://web-ee.com/schematics/medical/heartbeat-monitor/

> The one in my watch definitely acquires the pulse using electrical
> means. I don't know the exact mechanism, but basically it's using
> the base of the watch as one of the electrodes, and a metal button
> on the face as the #2 electrode

Much as the heart-rate monitors that use a chest strap do, it would
measure the potential changes across the body as the heart goes
through its phases. My guess would be that there would be a time
difference in this signal between a wrist and a finger on the opposing
hand. Two arm's lengths of nerve travel perhaps

2009\02\14@193558 by John Coppens

flavicon
face
On Sat, 14 Feb 2009 16:00:48 -0500
Sean Breheny <shb7STOPspamspamspam_OUTcornell.edu> wrote:

> Are you talking about atmospheric electricity here? That wouldn't
> apply indoors and it would be greatly affected by the body's
> conductivity anyway.

Well - it was a long time ago (memory failing, not the static
electricity ;-) I suspect that walls and conductivity would effectively
form a faraday's cage. On the other hand synthetic fibres can easily add
a couple of hundreds of volts (or more, when it's dry).

Thanks for the link to the HP journals. I learned a lot from those
journals back then.

John

2009\02\14@225804 by solarwind

picon face
This is everything I found so far:
solar-blogg.blogspot.com/search/label/Heart

More... (looser matching)
- Last day of these posts
- In 2009 , 2010 only
- Today
- New search...