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

Exact match. Not showing close matches.
PICList Thread
'[EE] Pulsemeter schematic'
2009\02\14@195257 by Benjamin Grant

flavicon
face
Hi all,
I'm trying to design a pulsemeter for the developing world. The
requirements are that it is low cost(<$8 is the target but I have
already conceded that is not going to happen), relatively robust and
reusable. The probe I'm using is composed of velcro and some feltlike
material. It doesn't keep out ambient light. I'm essentially using
pulseoximetry technology but only the IR led, not the red LED, as i'm
uninterested in 02 concentration. Anyway, I'm pulsing the IR led at
200 hz, measuring the signal with the LED on and LED off, and
subtracting out the ambient light. I'm a biomedical engineering
student but analog design is certainly not my strength.  This is
unpaid work for an NGO, so you're not helping me get an A or anything,
so any help would be much appreciated. The problems that have been
brought to my attention by more qualified engineers are

I posted the schematic
here<wiki.duke.edu/download/attachments/10715668/Pulsemeter.bmp>
1.) the nominal output voltage of all my opamps is 0, and a single
supply op amp can't drive the voltage all the way down to it's
negative rail. To fix this issue, it was recommended i power all the
op amps with -.3V on their negative rail.  I tried this and my signal
just goes down to -.3 V, so the same problem is met.    Any
suggestions on addressing this issue?
2.) Even if the actual output is 300 mV away from the negative rail,
because of the bias current and the voltage offset of the opamps, the
last amplification stage with gain of 3300 is likely to create a
situation in which the voltage is again driven to the nonoperating
region of the opamp. I actually don't know much about calculating the
theoretical effect of maximum bias current, voltage offset, etc. and
any reference on the subject would be great.
3.) I'm using an LED display(3, 7 segment displays), which likely will
use more power than I'd like, as this device needs to be run on a
battery.  However, LCD screens were not feasible if using a low cost
pic, and needing a low cost screen. I am strobing the LED digits, so
I'm only powering one at a time.  This part has been completed and
appears to work, but any power saving advice would be appreciated.

As this device needs to run off a battery, utilizing single supply
solutions is preferable for sure.  The output of this circuit will go
into the pic16f684, assuming i can keep the program size small enough
to fit on this pic.  Thanks in advance for any suggestions regarding
modifying this schematic to work properly. Also, the first part of the
circuit was taken from an appnote from TI. However, the resistor values I
chose, so the resistor values might be poor but the transimpedance amplifier
circuit(consisting of the first two opp amps) should work in theory. The
filtering and gaining was done by me, so they're an area of larger concern.

Ben

2009\02\14@201859 by Matt Pobursky

flavicon
face
On Sat, 14 Feb 2009 19:52:54 -0500, Benjamin Grant wrote:
{Quote hidden}

Some basic advice. I've done probably a dozen commercial battery powered
medical designs. None of them had quite your budgetary constraints but a
few were close.

1) Always run off as low a power supply voltage as you can with as slow a
clock as you can.

2) Run the electronics off 2 or 3 alkaline cells, size appropriate to
expected battery life. Use a 3V native PIC like a J or K series device.
They have some very low priced J series parts with NanoWatt technology that
will help you in your design. You can get one with enough port pins to
direct drive the LCD and still be very low cost.

3). Dump the LED displays, they are power hogs. You can do a port driven
LCD display quite easily with a PIC. 3V Static drive LCDs are cheap and
easy to come by. If you are actually going to produce these in volume they
don't cost much to have a custom part made.

4. You probably want some sort of frequency reference (crystal or ceramic
oscillator) for the PIC unless 1% or so accuracy is OK for your design.

5. Low power CMOS analog components. There are many many inexpensive parts
out there now that will run down to 1.8V.

6. Work very hard on your finger probe to minimize the current draw. This
will (or should) be the largest power drain in your design. You should
probably be able to get the entire circuit current drain down to about 1-
2mA average for the design with some careful component choices and power
management. If you use a reflective emitter-detector pair like someone else
posted a link to (Jinx or Tony?) you can have very simple electronics and
cheap, lower power analog front end with standard CMOS op-amps.

Matt Pobursky
Maximum Performance Systems

2009\02\14@203547 by A K

flavicon
face


Benjamin Grant wrote:
{Quote hidden}

In addition to what Matt said, try to use reasonable values for your
components.  0.1pf in the integrator is too small to put in a real-world
macro circuit.  Also for your ambient subtraction you could consider a
lock-in amplifier type topology.  Finally you could bias the analog
ground above the power ground by 0.7V with a diode.  The -0.3V
saturation you're seeing is probably a result of your integrator saturating.

2009\02\14@203746 by Roger, in Bangkok

face
flavicon
face
Let me poke my nose in here for a quick comment from the 3rd world.  Forget
about alkaline batteries in our real-world.  Even though available in most
places, they are at least 10-12 times the cost of a good alkaline.  Fact is
that even in the best of environments they will use the cheapest zinc
chloride or carbon zinc ... and yes, they will remain in use until they
leak, so please come up with a clever cleanable/replaceable battery
compartment:-)

Interesting project, is this going to commercial or open?

Regards/Roger, in Bangkok

On Sun, Feb 15, 2009 at 8:18 AM, Matt Pobursky <spam_OUTpiclistTakeThisOuTspammps-design.com>wrote:

{Quote hidden}

2009\02\15@084136 by olin piclist

face picon face
Benjamin Grant wrote:
> I posted the schematic
> here https://wiki.duke.edu/download/attachments/10715668/Pulsemeter.bmp

While I've certainly seen worse, you need to go back and clean up your
schematic.  Neatness counts whether you think it should or not.  You've got
lines going right thru caps and opamps, redundant part designators,
gratuitous dots in random places, and writing so small as to be nearly
unreadable.  You've already blown your one opportunity to make a good first
impression.  This schematic says "Eh, who cares, good enough for a bunch of
chumps I'm going to ask for free advice.".  And no, I don't want to hear you
blame it on whatever too you are using.  It's your responsibility to both
use the right tool and learn how to use it properly.  It would also be
better if you exported the schematic as a PDF file instead of a image.  That
way we can scale it to actually read some of the text.

As for the circuit:

1 - U is acting like a pure integrator.  It will accumulate any DC offset
from ground indefinitely, including its own offset voltage.  The output of U
won't take long to peg one way and stay that way.  Since that gets used as
the DC level a other U amplifies about, you likely won't get what you
intend.

2 - It looks like OP1 is a photodiode you are trying to use in leakage mode.
That by itself is fine, but I don't see any DC biasing.

3 - I don't see what purpose you think U serves at all.

4 - Q, Q, and Q are NPN transistors with their emitters tied to the positive
rail.  It looks like they are supposed to be PNP.

If you're confused about which U I'm talking about, remember that you gave
them all the same name (duh).


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

2009\02\15@134706 by solarwind

picon face
On Sat, Feb 14, 2009 at 7:52 PM, Benjamin Grant <.....bdg4KILLspamspam@spam@duke.edu> wrote:
> Hi all,
> I'm trying to design a pulsemeter for the developing world. The
> requirements are that it is low cost(<$8 is the target but I have
> already conceded that is not going to happen), relatively robust and
> reusable.

Hey, I messaged you off list, but I'm going to reply here as well. I'm
trying to do the same thing you are and have done a lot of research on
the topic already. There are several ways you can go about this I
think: optical where the light passes through the finger, optical
where light reflects off the finger, electrical where you measure the
voltages generated by the heart by placing electrodes on the chest, or
similarly, measure voltages across two fingers. The electrical way
seems to be the most accurate but I don't know if it's more/less
expensive and by how much.

--
solarwind

2009\02\16@002149 by M. Adam Davis

face picon face
You only need 22 pins to drive a 3 digit LCD display directly from the PIC:

http://www.ubasics.com/driving_static_lcds

But if that's too many pins you can use some very cheap ICs:

http://www.ubasics.com/Jinxs_static_LCD_drive_solution

-Adam

On Sat, Feb 14, 2009 at 7:52 PM, Benjamin Grant <bdg4spamKILLspamduke.edu> wrote:
{Quote hidden}

> -

2009\02\16@004403 by M. Adam Davis

face picon face
Oh, I should mention:

The LCD likely only needs to run in the range 0 to 299.  You can tie
segments together on the 2 digit so you only need three lines (one
displays the segments common to both 1 and two, the other two display
segments different between the two).

This reduces your needed pincount to 18 pins.

0-199 needs 16 pins.

Further, if you can live with a resolution of 5 beats per minute (ie,
65, 70, 75, 80) so the last digit only displays 0 or 5, then you need
only three pins for the last digit, resulting in 0-195 needing 12 pins
(0-295 14 pins).  This is only 2 more pins than you are currently
using for the display.

The fun aspect?

One of the pins is the LCD common - it has to switch from high to low
30-100 times a second - I imagine you can combine that with your
transmission LED.

That still leaves one needed pin which, with care, can be used for
both the A/D measurements and the LCD.  It'll require a bit more care
with the software, but you've got to fiddle with that front end
anyway, so you might consider this as well.

No need to go to a larger PIC or use external chips, and you get
substantial power savings.  Of course, you're limited to a resolution
of 5bpm, and a maximum of 195.

Going to a slightly larger chip, though, is not going to be
substantially more expensive, and depending on what other
manufacturers are using in the PIC line it may be cheaper to go with a
larger chip.  However, a simple shift register is very cheap, so there
are options other than going to a larger chip as well, without trying
to combine pin functionality.

-Adam

On Mon, Feb 16, 2009 at 12:21 AM, M. Adam Davis <.....stienmanKILLspamspam.....gmail.com> wrote:
{Quote hidden}

>> --

2009\02\16@044705 by Tamas Rudnai

face picon face
Also there are PICs with built in LCD driver for static and mux glasses.
16F917 is ok for 2.5 digits for example, or 946 higher number of
digits/segments.

Tamas

On Mon, Feb 16, 2009 at 5:21 AM, M. Adam Davis <stienmanspamspam_OUTgmail.com> wrote:

{Quote hidden}

2009\02\16@050358 by Marco Genovesi

picon face

Not really what you want, but perhaps some useful tips from this old
article, a TSL230R-Based Pulse Oximeter that measures pulses and oxygen
content:

www.circuitcellar.com/library/print/1204/Bachiochi173/index.htm
http://www.circuitcellar.com/library/print/0105/Bachiochi174/index.htm

For a low-count parts project, you could also try to use a photodiode or
phototransistor tied directly to the PIC (no Op-Amp) but I suspect that the
IR light intensity transmitted through the finger is very weak and would be
difficult to have enough Hz frequency measurement.

regards
Marco




{Original Message removed}

2009\02\16@100157 by PAUL James

picon face

Ben,

Do you know, or know of, a man at Duke University named Peter Malin?  He
is associated with the SAFOD project
Which deals with seismic activity on the San Andreas Fault in California
near where James Dean was killed.


Jim
 

{Original Message removed}

2009\02\17@001742 by Benjamin Grant

flavicon
face
Hi all
Sorry if you felt that my schematic was insulting. I will rework it soon and
fix my errors as well as replacing the LEDs with an LCD display. I have
worked on the schematic a little bit - cleaning up wires and labeling parts
Solarwind, all those methods are possible, but reflection/absorption
techniques are the most straightforward, reliable techniques.  Again Olin,
this is not my area of expertise and I will try to get my schematic more up
to par with your standards, but your argument that I'm trying to use "a
bunch of chumps for free advice" fails in this instance, as I'm a chump
voluntarily spending a large amount of time trying tot develop a nonprofit
product. Please, if in anyway you think I am using you to help those in
need, feel free to filter any messages from me, I will not be the least bit
offended.  I will, however, relabel my parts and straighten my lines.
Admittedly, not assigning new names to parts was a very poor mistake, and
I'm sorry I overlooked it. I am looking at using the PIC16F723 because it
meets the nanotechnology specs recommended by others and the lumex
LCD-S301C31TR 3 digit LCD display.  While the lumex datasheet says it
requires 5 V, other sources show that it can be used at 3 V with slightly
lower contrast, so I believe running the PIC at 3 V will be feasible.  It
has 24 pins but 2 are unnecessary(decimal points) so that leaves me with 3
I/O pins from the pic for controlling the LED and receiving the input. The
schematic found here
<http://web.archive.org/web/20080103013653/focus.ti.com/docs/solution/folders/print/330.html>is
what I'm basing the key components of my circuit off of. It recommends <1 pF
feedback capacitance and has a similar set up for the photodiode as me,
however, posters here have criticized both of these in my schematic. So I'm
wondering if the reason that TIs schematic is fundamentally sound while mine
is not is because they utilized an auto-zero op amp with exceptionally low
voltage offset. Olin pointed out that any DC offset is going to be
integrated and will result in a signal that is not what I would expect.
Furthermore, he pointed out(correctly) that I had a NPNs in my schematic
where a PNPs belonged. I've since fixed that. My cleaner schematic is shown
here in bmp
<wiki.duke.edu/download/attachments/10715668/Pulsemeterrev.bmp>or
emf <https://wiki.duke.edu/download/attachments/10715668/PulseOx.emf>. I
will have my next revision in PDF format. Hopefully this schematic is easier
to read than my previous one. If anyone has suggestions as to why the TI app
note is suitable in their case but not mine and any circuit specific advice,
I'd greatly appreciate it. Thanks a lot

2009\02\17@121731 by olin piclist

face picon face
Benjamin Grant wrote:
> Sorry if you felt that my schematic was insulting.

Not insulting, but it showed lack of attention to detail and sloppiness on
your part.  Why should anyone else take your project seriously if you don't?
Besides sloppy people invariably make bad engineers, so helping them with
engineering is pointless.

>  My cleaner schematic is shown here in bmp
> https://wiki.duke.edu/download/attachments/10715668/Pulsemeterrev.bmp

Somewhat cleaner.  You didn't fix all the things I mentioned, including the
line thru C2 and gratuitous dots on the net connected to pin 13 of LED3 and
by C2.  This the last time I will comment on your schematic if you don't fix
it.

As for the circuit, U1 is still a pure integrator, C2 is still way too
small, U3 still serves no purpose, there is still no bias for PD1 and other
parts of the analog circuit.  You did change the NPN transistors to PNP, but
I don't notice anything else addressed that I mentioned last time.


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

2009\02\17@140719 by Forrest W Christian

flavicon
face
Just a couple of comments...

Being a digital guy, I would be much more inclined to take the output of
the OP950 and run it through a single opamp to get it in the correct
voltage range for an ADC input, and then stick it into an ADC on a
suitable PIC.   A lot of the newer pics should have plenty of horsepower
for this, and should be within pennies of the cost of the PIC you have
listed.  The only question is what the signal looks like out of the
OP950.  But if this will work, your parts count goes down to the OP950
(and LED), a single opamp with preferrably r-r outputs, a couple of
resistors, and a PIC, oh and a display:

Qty 1000 prices:
OP950 - .50
IRLED - .10
Opamp - .50
Resistors - .10 (this is rounded way up)
LCD - 1.00
PIC - 2.50  (you can probably get away with much less - this is a "high
end" dspic)
      --------
       4.70

Not counting the PCB, assembly, and mechanicals.

-forrest


2009\02\17@162937 by Benjamin Grant

flavicon
face
Hi Forest,
that would be great to do it that way. I mean the OP950 just outputs a
current signal, so are you suggesting to use 1 single opamp to be a
transimpedance amplifier that outputs the voltage in a reasonable range and
then process it digitally?
Ben

On Tue, Feb 17, 2009 at 2:08 PM, Forrest W Christian <KILLspamforrestcKILLspamspamimach.com>wrote:

{Quote hidden}

> -

2009\02\17@181053 by Forrest W. Christian

flavicon
face
Benjamin Grant wrote:
> that would be great to do it that way. I mean the OP950 just outputs a
> current signal, so are you suggesting to use 1 single opamp to be a
> transimpedance amplifier that outputs the voltage in a reasonable range and
> then process it digitally?
>  
That is roughly correct.   Although I would probably bias the OP950 so
it effectively was a voltage source, not a current source, then amp it
up with an opamp.

The question will be whether you have enough bits on the pic's ADC
(which is why I suggested a part with a 12 bit adc) to be able to handle
the dynamic range which is needed.

-forrest

2009\02\18@185804 by Benjamin Grant

flavicon
face
Hi,
I updated the schematic
<https://wiki.duke.edu/download/attachments/10715668/PulseLCD.bmp>. I
removed the integrator, as it's functionality depended on expensive
auto-zeroing opamps and unreasonable capacitance values. I have a 4 hZ
single pole filter here now, but I believe I'll replace it with a 2 pole
filter. Otherwise, are there any thoughts on the updated schematic? Oh,
also, i replaced the LED display with an LCD display for power saving
purposes
Ben

On Tue, Feb 17, 2009 at 6:10 PM, Forrest W. Christian <RemoveMEforrestcTakeThisOuTspamimach.com>wrote:

{Quote hidden}

> -

2009\02\19@072046 by olin piclist

face picon face
Benjamin Grant wrote:
> I updated the schematic
> <https://wiki.duke.edu/download/attachments/10715668/PulseLCD.bmp>.

Why are you running the opamps on only 1.5V when the PIC is running on 3V?
This doesn't make sense.

R2 is shorted out.

The whole circuit around PD1 and U1 is a mess.  It looks like you copied it
from somwehere then made a few modifications without first understanding how
it works.  Do you understand how PD1 is used to sense light?  Explain its
characteristics when it is in dark and light.  Then explain how you extract
a signal making use of those characteristics and then amplify the signal to
something useable.

Once again I don't see why U2 is needed the way it's used.

U3 very likely doesn't do what you want.  Did you really intend for such
very large hysteresis?  I didn't think so.

LEDs operate on current thru them.  This is difficult to achieve when only
one lead is connected.

Look up "bypass capacitor".

The 16F72 has no internal oscillator.

> Oh,
> also, i replaced the LED display with an LCD display for power saving
> purposes

You did?  Then why is it called LED-3D?  If this is a LCD, is it a bare
glass or is it a module with driver chip?  There are a lot more lines going
to it than I would expect for a module with integrated chip.  However, you
can't just drive a bare glass like that.  They need to be driven with AC
such that the average DC stays 0.  If this is a bare glass, it is best if
you use one of the PICs intended for driving them, such as the 16F9xx.


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

2009\02\19@075536 by Forrest W. Christian

flavicon
face

I'm going to make an even more radical suggestion.. which just occured
to me.

This application is *screaming* for a Cypress PSoC.

I am about 99.9% sure you can throw away all of the opamps and the pic,
and replace it with a single PSoC.

What is a PSoC?  Well, imagine a PIC-like processor, with programmable
digital and analog blocks, which can be configured at will, and
reconfigured at will.

Or let me put it this way....  Imagine being able to wire that sensor up
directly to an input pin on the PSoC, and the psoc having internal
analog blocks which can be configured as amplifiers of any gain, or
bandpass filters, or any one of several other types of analog
circuits.   Once you have it filtered, the analog signal can then be
routed (still internal to the chip) to a counter or similar.  If one way
doesn't work, you can try a different one... without changing the
hardware itself - the signal routing is all programmable through the
(free) Cypress software.

I am 99.9% sure you could do your application with the IR sensor you
have, a single bias resistor, and a PSoC - not counting the display.  
The most expensive PSoC you might need is only $3.90 in qty 100.  You
will likely be able to do it with a smaller PSoC...depending on how many
blocks you need, you might find you only need a $2 (or less) PSoC.

But I agree with Olin:  it appears you are missing some basic
understanding of what is going on with the circuit.  Have you tried
hooking this stuff up on a breadboard and a Scope and understanding what
the signal looks like with a live, human, pulse?

The reason I ask, is that although the PSoC solution I am suggesting is
going to permit you to build and test many options, it does require some
knowledge of analog and digital system design - although I guess you
might be able to use the simplified tools for this application.  If you
are starting with incorrect assumptions, it is going to make working
with the PSoC more difficult.

If you want to go the PSoC route, you should probably start with a
CY3210-PSoCEval1 , which is available from Mouser and Digikey, among others.

This is an interesting project to me, I'd like to hear more about what
you are really expecting to accomplish, and what you personally are
trying to get out of it - so I can perhaps help you in the right
direction.  If it's a learning experience for you, then that is one
thing.  If you are in a position to actually get a whole bunch of these
made and distributed, but don't really care about figuring out how it
would work, then I may suggest a different path than trying to figure it
out.  But if you could let us know a bit more about the project and your
role in it, I (or others) might be able to help in a more constructive way.

-forrest

2009\02\19@090029 by M. Adam Davis

face picon face
That's an excellent suggestion - I keep thinking PSoC's ought to have
more usage, but I just don't see them used as often.  This appears to
be a perfect application for one.

-Adam

On Thu, Feb 19, 2009 at 7:54 AM, Forrest W. Christian
<spamBeGoneforrestcspamBeGonespamimach.com> wrote:
>
> I'm going to make an even more radical suggestion.. which just occured
> to me.
>
> This application is *screaming* for a Cypress PSoC.

2009\02\19@092110 by olin piclist

face picon face
M. Adam Davis wrote:
> That's an excellent suggestion - I keep thinking PSoC's ought to have
> more usage, but I just don't see them used as often.  This appears to
> be a perfect application for one.

A PSoC may be a good solution for volume production, especially if space is
tight.  However the OP doesn't understand his circuit.  He will need
considerable time breadboarding it and looking at various points with a
scope.  This will be much easier to do if everything is exposed with
discrete parts on a breadboard that can be easily rewired.  Telling the OP
to use a PSoC at this point is just going to add to his confusion.

Get it working first, then once you know what the circuit should be and
understand it, convert to PSoC if still desired.  a PSoC would be the
physically smallest solution, but not necessarily the lowest cost one.


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

2009\02\19@094052 by PAUL James

picon face
Looks like u2 is a low pass filter that is AC coupled to the next stage.


-----Original Message-----
From: TakeThisOuTpiclist-bouncesEraseMEspamspam_OUTmit.edu [RemoveMEpiclist-bouncesspamTakeThisOuTmit.edu] On Behalf
Of Olin Lathrop
Sent: Thursday, February 19, 2009 6:22 AM
To: Microcontroller discussion list - Public.
Subject: Re: [EE] Pulsemeter schematic

Benjamin Grant wrote:
> I updated the schematic
> <https://wiki.duke.edu/download/attachments/10715668/PulseLCD.bmp>.

Why are you running the opamps on only 1.5V when the PIC is running on
3V?
This doesn't make sense.

R2 is shorted out.

The whole circuit around PD1 and U1 is a mess.  It looks like you copied
it from somwehere then made a few modifications without first
understanding how it works.  Do you understand how PD1 is used to sense
light?  Explain its characteristics when it is in dark and light.  Then
explain how you extract a signal making use of those characteristics and
then amplify the signal to something useable.

Once again I don't see why U2 is needed the way it's used.

U3 very likely doesn't do what you want.  Did you really intend for such
very large hysteresis?  I didn't think so.

LEDs operate on current thru them.  This is difficult to achieve when
only one lead is connected.

Look up "bypass capacitor".

The 16F72 has no internal oscillator.

> Oh,
> also, i replaced the LED display with an LCD display for power saving
> purposes

You did?  Then why is it called LED-3D?  If this is a LCD, is it a bare
glass or is it a module with driver chip?  There are a lot more lines
going to it than I would expect for a module with integrated chip.
However, you can't just drive a bare glass like that.  They need to be
driven with AC such that the average DC stays 0.  If this is a bare
glass, it is best if you use one of the PICs intended for driving them,
such as the 16F9xx.


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

2009\02\19@111517 by Benjamin Grant

flavicon
face
U2 is a buffer between a low pass filter and the gain stage.. The op amps
are powered at -1.5 V and 1.5 V(those sum to 3 V, we're running off two 1.5
V batteries). The static LCD screen can indeed be run by a pic, as
seen here<www.ubasics.com/driving_static_lcds>.
Yes, I didn't connect a resistor to the LED or ground it, i'll fix that,
it's a 560 ohm resistor and the back end is shorted with ground. The circuit
shown in my first schematic I've built and breadboarded and looked at the
signal many times. In fact, the signal was correct and we were able to get a
heart beat out of it, but there was some artifactual noise causing extra
"beats" every once in a while.

On Thu, Feb 19, 2009 at 9:39 AM, PAUL James <James.PaulEraseMEspam.....colibrys.com> wrote:

> Looks like u2 is a low pass filter that is AC coupled to the next stage.
>
>
> {Original Message removed}

2009\02\19@111605 by olin piclist

face picon face
PAUL James wrote:
> Looks like u2 is a low pass filter that is AC coupled to the next stage.

Yes, but it has a gain of one and isn't performing any impedence buffering
either.  The filter functions could be just passive parts between the two
other opamp stages.


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

2009\02\19@112325 by Benjamin Grant

flavicon
face
Yes, I agree, I'm going to remove U2, it serves no purpose. Secondly, I do
understand a transimpedance amplifier, as seen in stage one, and I'm sorry
for drawing an extra wire to ground, but once removed as in the
updated schematic
<https://wiki.duke.edu/download/attachments/10715668/PulseLCD.bmp>I don't
believe there are large problems with the transimpedance amplifier.

On Thu, Feb 19, 2009 at 11:17 AM, Olin Lathrop <EraseMEolin_piclistspamembedinc.com>wrote:

{Quote hidden}

> -

2009\02\19@113546 by Mark Rages

face picon face
On Thu, Feb 19, 2009 at 10:22 AM, Benjamin Grant
<RemoveMEbenjamin.grantEraseMEspamEraseMEduke.edu> wrote:
> Yes, I agree, I'm going to remove U2, it serves no purpose. Secondly, I do
> understand a transimpedance amplifier, as seen in stage one, and I'm sorry
> for drawing an extra wire to ground, but once removed as in the
> updated schematic
> <https://wiki.duke.edu/download/attachments/10715668/PulseLCD.bmp>I don't
> believe there are large problems with the transimpedance amplifier.
>

You cannot get -1.5V, 0V, 1.5V and 3V from a pair of cells -- that's
4.5 V total.  Something is wrong with your grounds or power supply.
Maybe add the cells to the schematic to show how they are connected.

Regards,
Mark
markrages@gmail
--
Mark Rages, Engineer
Midwest Telecine LLC
RemoveMEmarkragesspam_OUTspamKILLspammidwesttelecine.com

2009\02\19@115134 by Benjamin Grant

flavicon
face
sorry, you're right that was very dumb. I actually just meant to offset
analog ground. Like this
<https://wiki.duke.edu/download/attachments/10715668/PulseLCD.bmp>to avoid
rail problems I was seeing before.
Ben

On Thu, Feb 19, 2009 at 11:35 AM, Mark Rages <RemoveMEmarkragesTakeThisOuTspamspamgmail.com> wrote:

{Quote hidden}

> -

2009\02\19@135621 by Jinx

face picon face
<https://wiki.duke.edu/download/attachments/10715668/PulseLCD.bmp>

Just curious - as far as I can remember, that picture is the first one I've
seen that "builds" from the bottom up when loading. How odd

A couple of observations on the IR LED. Not familiar with the 16F72,
does pin 1 have any output capability ? And 560R seems high. At only
3V supply the transmitted power won't be much, although perhaps you
don't need much in this application. More commonly, eg remote control
and data links, IR LEDs are driven with short pulses, and low resistance
to the supply rail. This extra amperage gives the IR range and penetration.
Note the average wattage of high-current pulses and the heating spec of
the LED

2009\02\19@150528 by Bob Ammerman

picon face
You probably shouldn't leave the decimal point segments of the LCD floating.
You can connect them to COM.

-- Bob Ammerman
RAm Systems

{Original Message removed}

2009\02\19@151459 by Michael Rigby-Jones

picon face


> -----Original Message-----
> From: piclist-bouncesSTOPspamspamspam_OUTmit.edu [spamBeGonepiclist-bouncesSTOPspamspamEraseMEmit.edu] On
Behalf
> Of Benjamin Grant
> Sent: 19 February 2009 16:23
> To: Microcontroller discussion list - Public.
> Subject: Re: [EE] Pulsemeter schematic
>
> Yes, I agree, I'm going to remove U2, it serves no purpose. Secondly,
I do
> understand a transimpedance amplifier, as seen in stage one, and I'm
sorry
> for drawing an extra wire to ground, but once removed as in the
> updated schematic

I'm not convinced you need to use the diode in leakage mode in this
application.  Typically you do this when you need to get maximum speed
out of the diode, since you can put a high and constant bias across it
(using a TIA) to reduce it's junction capacitance.  As you are only
requiring a maximum of a few hundred Hz bandwidth, operating in the
photovoltaic mode could reduce complexity, you get less noise and you
don't have to worry about dark current for starters.

Regards

Mike

=======================================================================
This e-mail is intended for the person it is addressed to only. The
information contained in it may be confidential and/or protected by
law. If you are not the intended recipient of this message, you must
not make any use of this information, or copy or show it to any
person. Please contact us immediately to tell us that you have
received this e-mail, and return the original to us. Any use,
forwarding, printing or copying of this message is strictly prohibited.
No part of this message can be considered a request for goods or
services.
=======================================================================

2009\02\19@161142 by cdb

flavicon
face


:: The static LCD screen can indeed be run by a pic, as
:: seen here<http://www.ubasics.com/driving_static_lcds>

Something you might want to consider as you need low costs, is instead
of using two batteries to get a split supply, especially for the
static LCD, you could use two equal value resistors about 470 - 560
ohms in series across the power supply and connect the LCD common to
the midpoint of the series connection.

I haven't checked out the op-amps, but if they are available as a quad
pack, after experimentation, these might be cheaper and the spare
op-amp could be used in the famous supply splitter configuration if
you still need a split supply. This might enable you to use a small
lithium 3v battery.

Colin
--
cdb, KILLspamcolinspamBeGonespambtech-online.co.uk on 20/02/2009

Web presence: http://www.btech-online.co.uk  

Hosted by:  http://www.1and1.co.uk/?k_id=7988359







2009\02\19@212532 by Benjamin Grant

flavicon
face
Colin,
excellent points -- both of which I agree with and actually had been trying
to determine what i should use the 4th op amp for(if i change u2 to an
active filter).  The quad package is quite a bit cheaper indeed and would be
nice to use.  The only thing I have to consider it is it would be most
appropriate for the batteries to be replaced onsite, and the availability of
lithium 3 V batteries may be low. I'll have to check with people i know in
the regions of interest(mainly Tanzania and neighboring countries as well as
central america).

Jinx - no I need to fix the pin1 and I also need to maneuver some pins to
enable a spot for an external oscillator. Although, perhaps I should
reconsider the pic so as not to need an external oscillator. I guess an
external oscillator enable higher precision, but greater precision is likely
unnecessary and perhaps not cost effective.  Also, I will be pulsing the
LED, and the 560 ohm was from the 5 V design -- perhaps I will decrease this
resistance even more than proportional to allow for a higher power signal.

Thanks!

On Thu, Feb 19, 2009 at 4:10 PM, cdb <EraseMEcolinspamEraseMEbtech-online.co.uk> wrote:

{Quote hidden}

> -

2009\02\19@212923 by Benjamin Grant

flavicon
face
"I'm not convinced you need to use the diode in leakage mode in this
application.  Typically you do this when you need to get maximum speed
out of the diode, since you can put a high and constant bias across it
(using a TIA) to reduce it's junction capacitance.  As you are only
requiring a maximum of a few hundred Hz bandwidth, operating in the
photovoltaic mode could reduce complexity, you get less noise and you
don't have to worry about dark current for starters."

Mike - as my analog ground is 1.5 V for this circuit , I'm going to remove
that resistor between analog ground and the diode to use it in photovoltaic
mode. Most applications for pulseoximetry run in pholtovoltaic mode, thanks
for pointing this out.
Ben

2009\02\23@130618 by Gerhard Fiedler

picon face
Jinx wrote:

>  <wiki.duke.edu/download/attachments/10715668/PulseLCD.bmp>
>
> Just curious - as far as I can remember, that picture is the first one I've
> seen that "builds" from the bottom up when loading. How odd

<http://en.wikipedia.org/wiki/BMP_file_format>

"Pixels are stored "upside-down" with respect to normal image raster
scan order, starting in the lower left corner, going from left to right,
and then row by row from the bottom to the top of the image.
Uncompressed Windows bitmaps can also be stored from the top row to the
bottom, if the image height value is negative."

Gerhard

2009\02\23@141443 by Vasile Surducan

face picon face
On 2/19/09, Benjamin Grant <spamBeGonebenjamin.grantspamKILLspamduke.edu> wrote:
> sorry, you're right that was very dumb. I actually just meant to offset
> analog ground. Like this
> <https://wiki.duke.edu/download/attachments/10715668/PulseLCD.bmp>to avoid
> rail problems I was seeing before.

Well, not bad for a collective design.
:)
However I can't understand why have you added a filter (R3C3) when the
same effect was adding a smaller capacitor C1 in parallel on R1 ( R1C1
approx equal with R3C3).
The second curiosity is why you need a derivator (R4C4) ? The current
to voltage converter U1 has very small offset. U2 it's a repeater so
again it has a small offset...
So, why?

Vasile

2009\02\23@160114 by Jinx

face picon face
> <http://en.wikipedia.org/wiki/BMP_file_format>

Interesting

The OP might like to know that the 866kB BMP is quite readable as
a 15kB 50% x 50% dimension-reduced PNG. Or 40kB as a GIF

I believe IrfanView is free and can create GIF, not sure about PNG

As long as it's not JPG. JPGed Line drawings look terrible and frequently
important details are lost


2009\02\23@161542 by Benjamin Grant

flavicon
face
thanks jinx - I use gimp and can save it in PNG... jpeg's are horrible, I'd
prefer to use SVG personally, but can't export as that. Will indeed change
to PNG though for readability. Thanks
Ben

On Mon, Feb 23, 2009 at 5:01 PM, Jinx <.....joecolquittspam_OUTspamclear.net.nz> wrote:

{Quote hidden}

> -

2009\02\26@010311 by Benjamin Grant

flavicon
face
changed to png
https://wiki.duke.edu/download/attachments/10715668/PulseLCD.png for ease of
loading. Also, Olin, care to elaborate on the nature of the hysteresis
likely to occur on U3?

2009\02\26@075656 by olin piclist

face picon face
Benjamin Grant wrote:
> changed to png
> https://wiki.duke.edu/download/attachments/10715668/PulseLCD.png for
> ease of loading. Also, Olin, care to elaborate on the nature of the
> hysteresis likely to occur on U3?

Look carefully at the feedback around U3.  It would also help if you
described what you intend U3 to do.


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


'[EE] Pulsemeter schematic'
2009\04\02@013159 by Benjamin Grant
flavicon
face
sorry for not responding for so long.
wiki.duke.edu/download/attachments/10715668/PulseOx_rev3.pngcontains
my latest schematic. I admittedly have yet to breadboard the
revision, but plan to do so on friday and any insight before building would
be great.  I'll explain what I intend to do at least. U1 is just a
transimpedance amplifier.  U2 is used to enable 2nd order filtering with a
cut off of about 4 hz(4.16 i believe). The next two stages are to provide
approximately 20 gain to allow for signal to be in usable range. Thanks in
advance
Ben

On Thu, Feb 26, 2009 at 8:56 AM, Olin Lathrop <TakeThisOuTolin_piclist.....spamTakeThisOuTembedinc.com>wrote:

{Quote hidden}

> -

2009\04\02@014723 by solarwind

picon face
On Thu, Apr 2, 2009 at 5:31 AM, Benjamin Grant <TakeThisOuTbenjamin.grantKILLspamspamspamduke.edu> wrote:
> sorry for not responding for so long.
> wiki.duke.edu/download/attachments/10715668/PulseOx_rev3.pngcontains
> my latest schematic. I admittedly have yet to breadboard the
> revision, but plan to do so on friday and any insight before building would
> be great.  I'll explain what I intend to do at least. U1 is just a
> transimpedance amplifier.  U2 is used to enable 2nd order filtering with a
> cut off of about 4 hz(4.16 i believe). The next two stages are to provide
> approximately 20 gain to allow for signal to be in usable range. Thanks in
> advance
> Ben

That page can not  be found.

2009\04\02@020236 by Grant Tudor

picon face
Drop the "contains" after the .png

https://wiki.duke.edu/download/attachments/10715668/PulseOx_rev3.png<https://wiki.duke.edu/download/attachments/10715668/PulseOx_rev3.pngcontains>

On Thu, Apr 2, 2009 at 4:46 PM, solarwind <.....x.solarwind.xspamRemoveMEgmail.com> wrote:

{Quote hidden}

> -

2009\04\02@020445 by Grant Tudor

picon face
Although when I tried the link in my message it put the "contain" back in
the URL so you may have to manually edit the URL in your browser

On Thu, Apr 2, 2009 at 5:02 PM, Grant Tudor <spamBeGonegrant.tudor@spam@spamspam_OUTgmail.com> wrote:

{Quote hidden}

>> --

2009\04\02@021538 by Benjamin Grant

flavicon
face
yeah sorry sorry

https://wiki.duke.edu/download/attachments/10715668/PulseOx_rev3.png


On Thu, Apr 2, 2009 at 2:04 AM, Grant Tudor <RemoveMEgrant.tudorEraseMEspamspam_OUTgmail.com> wrote:

{Quote hidden}

2009\04\02@021843 by cdb

flavicon
face


:: The next two stages are to provide
:: approximately 20 gain to allow for signal to be in usable range

You are aware that you have a total gain of approx 450 (26.6dB per
stage)

Colin
--
cdb, spamBeGonecolinEraseMEspambtech-online.co.uk on 2/04/2009

Web presence: http://www.btech-online.co.uk  

Hosted by:  http://www.1and1.co.uk/?k_id=7988359







2009\04\02@024337 by Benjamin Grant

flavicon
face
yeah sorry sorry I said that incorrectly.. i wanted a total gain of between
400 and 500.. i just meant each was around 20. Sorry.. thanks
ben

On Thu, Apr 2, 2009 at 2:18 AM, cdb <colinspamBeGonespambtech-online.co.uk> wrote:

{Quote hidden}

> -

2009\04\02@051620 by Marechiare

picon face
> I'm trying to design a pulsemeter

> I posted the schematic

Wrong approach, the right one would be:

---
"I got a couple of IR parts. When I place them around my finger and
feed the first part with 10ma, the second with 1ma, the signal on the
scope(1 MOhm input), attached to the second part, is about 0.1mV AC
1Hz over 10mV AC 50Hz over 2V DC. That 0.1mV 1Hz AC is my pulse. How
the hell on the earth could I filter the pulse signal out and show its
frequency on a display?"
---

Don't do any schematic untill you try out the analog probe and get its
signal specs. When the specs are figured out, add stages step by step,
Olin is right regarding the system-on-a-chip.

Good Luck.

2009\04\02@074911 by olin piclist

face picon face
Grant Tudor wrote:
> Although when I tried the link in my message it put the "contain"
> back in
> the URL so you may have to manually edit the URL in your browser

Or you could do it, since you're the one asking for a favor.  Duh.

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

2009\04\02@075543 by olin piclist

face picon face
Benjamin Grant wrote:
> https://wiki.duke.edu/download/attachments/10715668/PulseOx_rev3.png

I'm not so sure about how the sensor is hooked up and the U1 circuit.  You
are holding the sensor at 0V and measuring current it produces.  That might
work depending on the sensor.  A more typical approach is to measure reverse
leakage current.

Are you really sure U2 is stable the way you have it hooked up?  I didn't
check, but you need to.

U3 and U3 (arg, same component ID again) each have a gain of 21, for a
combined gain of 450.  Note that this is also DC gain, which is not a good
idea.  Every millivolt offset on the input turns into half a volt offset on
the output.  You should AC couple U3 and U3.

Next time check the schematic before asking others to look at it.  There is
no excuse for sloppiness like two U3 designator, especially since this was
pointed out last time.


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

2009\04\02@121555 by Benjamin Grant

flavicon
face
grant tudor is not me... he's not asking for help that's why he wrote what
he did.  Otherwise thank you

On 4/2/09, Olin Lathrop <.....olin_piclist@spam@spamEraseMEembedinc.com> wrote:
{Quote hidden}

> -

2009\04\02@175606 by Benjamin Grant

flavicon
face
"I got a couple of IR parts. When I place them around my finger and
feed the first part with 10ma, the second with 1ma, the signal on the
scope(1 MOhm input), attached to the second part, is about 0.1mV AC
1Hz over 10mV AC 50Hz over 2V DC. That 0.1mV 1Hz AC is my pulse. How
the hell on the earth could I filter the pulse signal out and show its
frequency on a display?"
---
Again, I've built all of this, i know the ouptut from the probe. I've
already built earlier versions of this schematic, just seeing if there were
any possible pitfalls before building. AC coupling is a good point. Also,
what are you referring to as "first part" and "second part"?

Don't do any schematic untill you try out the analog probe and get its
signal specs. When the specs are figured out, add stages step by step,
Olin is right regarding the system-on-a-chip.

On Thu, Apr 2, 2009 at 5:16 AM, Marechiare <.....marechiareRemoveMEspamgmail.com> wrote:

{Quote hidden}

> -

2009\04\02@223908 by cdb

flavicon
face
You might want to check out the datasheet for the Analog Devices
AD795, this op-amp is often used as a photodiode pre-amp, and figures
42 and 46 might have just what you're looking for.

Colin
--
cdb, .....colinSTOPspamspam@spam@btech-online.co.uk on 3/04/2009

Web presence: http://www.btech-online.co.uk  

Hosted by:  http://www.1and1.co.uk/?k_id=7988359







2009\04\02@231240 by Benjamin Grant

flavicon
face
thanks.. i'll check it out. I fixed the schematic with olin's suggestions..
including bypass capacitors, and AC coupled gain stages.
https://wiki.duke.edu/download/attachments/10715668/PulseOx_rev3.png

On Thu, Apr 2, 2009 at 10:39 PM, cdb <colinEraseMEspam@spam@btech-online.co.uk> wrote:

{Quote hidden}

> -

2009\04\02@233021 by cdb

flavicon
face
Have you tried connecting the anode of PD1 to a negative potential
rather than positive - I think you'll see it be more sensitive. The
only problem with photoconductive mode is that you will always have a
small amount of current measurable. Something you might need to
calibrate out in software.

The datasheet I referred to, also has some PCB suggestions concerning
guard traces around pin 2 or 3, which will minimise any noise coming
from the resistors/IC.

Colin
--
cdb, spamBeGonecolinKILLspamspam@spam@btech-online.co.uk on 3/04/2009

Web presence: http://www.btech-online.co.uk  

Hosted by:  http://www.1and1.co.uk/?k_id=7988359







2009\04\03@065101 by olin piclist

face picon face
Benjamin Grant wrote:
> thanks.. i'll check it out. I fixed the schematic with olin's
> suggestions.. including bypass capacitors, and AC coupled gain stages.
> https://wiki.duke.edu/download/attachments/10715668/PulseOx_rev3.png

Don't put capacitors in series with the power leads!!!

U3 and U4 together now have a gain of 100 where you previously had 450.
Which one is right?

Your AC coupling high pass filters roll off at .16Hz.  Does it really need
to be that low?


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

2009\04\03@065524 by olin piclist

face picon face
cdb wrote:
> Have you tried connecting the anode of PD1 to a negative potential
> rather than positive

Actually, he's holding 0 volts accross it.

> The only problem with photoconductive mode

He is doing the opposite, using photovoltaic mode.

> is that you will always have a
> small amount of current measurable. Something you might need to
> calibrate out in software.

No, it will never get to the software.  The AC coupled gain stages will
guarantee that.  The only issue with a steady current is that the first
stages are not biased in the middle and effectively limit the range in one
direction.  In any case, a photodiode completely in the dark with 0 volts
accross is will produce no current.


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

2009\04\03@083831 by Marechiare

picon face
> Again, I've built all of this, i know the ouptut from the probe.

Why not tell the community the number of microVolts/microAmps of
signal change that correspond to human pulse? Without the number all
the schematics are irrelevant speculations. There is no such thing as
generic circuit.

2009\04\03@113300 by Bob Axtell

face picon face
I guess I'm just not getting this. No offense is intended here, but
you might consider getting a medical consultant before you waste too
much time. What I have seen so far indicates that
you might be in over your head.

Some folks here have tried to be gentle. I will be less so, because I
am in a bad mood.

$8 for a medical device? If the raw parts cost just $1, the device
will HAVE to sell for at
least $25 just to cover the cost of insurance. So now you will need to
include another
$15 MORE of parts to cover the need of an insurance company to feel
safe in covering an
$8 device (few will).

Medical items cost a LOT of money because innovation is NOT what
drives that industry.
What drives it is INSURANCE. If you are unable to convince an
insurance company that
it works PERFECTLY, they won't cover it, and you are sunk. Examples:
PIC's 10-bit A/D
won't cut it, but the 18-bit MCP3421 WILL cut it, but that triples the cost.

Did you know that there are few electronic parts that are even ALLOWED
to be designed into a medical product? National Semiconductor makes
wonderful parts, but National will
NOT sell parts to you if they will be used in a medical device; and
believe me, they ALWAYS find out. The people that WILL sell you parts
will sell you the expensive, super-tested version, not the cheapy
50-cent one. And the list goes on and on.

I have designed a couple of very simple medical devices in years past,
and frankly, I'd
never do it again.

--Bob A

On Mon, Feb 16, 2009 at 10:16 PM, Benjamin Grant <bdg4spam_OUTspam@spam@duke.edu> wrote:
{Quote hidden}

>

2009\04\03@121624 by Benjamin Grant

flavicon
face
Bob,. I've done extensive work in the developing world. A few points
a.) this is to be manufactured by the nonprofit I am volunteering for, and
not sold, but donated. They(businessmen, doctors, and administrators) chose
$8 as what they saw the maximum price that would make it feasible.  b.)
You're speaking of the US medical community. The hospitals in Tanzania are
hospitals you need to visit before you could understand, but I assure you
they don't care about the brand of components used. c.) This isn't a device
for profit and I'm not worried about the insurance companies because...
largely there aren't any there to worry about.
ben

On Fri, Apr 3, 2009 at 11:32 AM, Bob Axtell <spamBeGonebob.axtell@spam@spamgmail.com> wrote:

{Quote hidden}

2009\04\03@142448 by Bob Axtell

face picon face
Good answer. I keep forgetting that the Piclist is global.

Have a great day.

--Bob

On Fri, Apr 3, 2009 at 9:16 AM, Benjamin Grant <spamBeGonebenjamin.grantspam_OUTspamRemoveMEduke.edu> wrote:
{Quote hidden}

2009\04\04@041813 by Nate Duehr

face
flavicon
face

On Apr 3, 2009, at 9:32 AM, Bob Axtell wrote:

> Medical items cost a LOT of money because innovation is NOT what
> drives that industry.
> What drives it is INSURANCE. If you are unable to convince an
> insurance company that
> it works PERFECTLY, they won't cover it, and you are sunk. Examples:
> PIC's 10-bit A/D
> won't cut it, but the 18-bit MCP3421 WILL cut it, but that triples  
> the cost.


Aviation and avionics are the same.  A $200 radio in another other  
market will balloon into a $2000 radio when it has to be put in an  
aircraft's radio panel.

:-)

--
Nate Duehr
EraseMEnateRemoveMEspamSTOPspamnatetech.com




2009\04\06@001619 by Benjamin Grant

flavicon
face
So even I admit that I deserve to be called an idiot for my myopic decision
on the pic to use. It appears as though MPLAB says it's not supported
utilizing picstart plus programmer. I'm using the PIC16F724. Does this mean
I have to choose a new pick? Again, I recognize this was poor planning on my
behalf and you're welcome to remind me of that, but if you have any advice
the doesn't require finding a new pic that'd be great! Best,
Ben

On Sat, Apr 4, 2009 at 4:18 AM, Nate Duehr <RemoveMEnateKILLspamspamTakeThisOuTnatetech.com> wrote:

{Quote hidden}

> -

2009\04\06@023511 by Heinz Czychun

flavicon
face
Hi Benjamin,

On 3-Apr-09, at 12:16 PM, Benjamin Grant wrote:

{Quote hidden}

       With this sort of market in mind,

       Have you considered the TSL230 as a replacement for the opto-
analogue portion of your project? It's a light to frequency  
converter. Although likely more expensive, than the analogue  
components, TAOS maybe persuaded to give you a break for humanitarian/
advertising considerations. It would move the problem of pulse rate  
detection into the firmware realm.

       Circuit Cellar's Jeff Bachiochi has a nice writeup available on-line  
on his design.
        http://www.circellar.com/library/print/0105/Bachiochi174/
index.htm

 Also the TSL235, smaller, cheaper, less programable.

Heinz Czychun









2009\04\06@072726 by olin piclist

face picon face
Benjamin Grant wrote:
> So even I admit that I deserve to be called an idiot for my myopic
> decision on the pic to use. It appears as though MPLAB says it's not
> supported utilizing picstart plus programmer.

I heard some museum was missing one.  You really should return it.

> I'm using the
> PIC16F724. Does this mean I have to choose a new pick?

That looks to be a rather unremarkable 16F PIC with 40 pins.  I don't
remember exactly what you're doing, but there are probably lots of PICs that
can do the job.  However, the 16F724 seems to be real, so there shouldn't be
a problem with it.


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

2009\04\06@082928 by Wouter van Ooijen

face picon face
>> supported utilizing picstart plus programmer.
>
> I heard some museum was missing one.  You really should return it.

Why? I still use it for OTP chips like the 16C745.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2009\04\06@112349 by olin piclist

face picon face
>>> supported utilizing picstart plus programmer.
>>
>> I heard some museum was missing one.  You really should return it.
>
> Why? I still use it for OTP chips like the 16C745.

You should return those to the museum too.

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

2009\04\06@113010 by Benjamin Grant

flavicon
face
haha I've been using Duke's equipment and for whatever reason that's what
they have. Anyway, I was going to buy a programmer for myself -- any
recommendations Olin? I was looking at pickit2 but if you have other
recommendations I'm open to them

On Mon, Apr 6, 2009 at 11:26 AM, Olin Lathrop <RemoveMEolin_piclistspam_OUTspamembedinc.com>wrote:

{Quote hidden}

> -

2009\04\06@115004 by Wouter van Ooijen

face picon face
Olin Lathrop wrote:
>>>> supported utilizing picstart plus programmer.
>>> I heard some museum was missing one.  You really should return it.
>> Why? I still use it for OTP chips like the 16C745.
>
> You should return those to the museum too.

I am (at least partly) customer-driven. There are some (mostly
hobby-level) designs out there that use the 16C745, maybe because it was
for some time the only USB-PIC (is that correct?), maybe because it is
the only 14-bit core USB PIC. People build those designs, and buy chips
for them (some programmed, some unprogrammed). So it makes perfect
(economic) sense for me to keep my PICkit2.

There are also some minor irks with the pickit2 (with chips that have a
timer configured to the programming pins or something like that), so
that is an additional reason to keep the pickit2 as 'reference' programmer.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2009\04\06@115105 by Michael Hagen

flavicon
face
PicStart+, usually some on eBay?
I have had 2 of them for years.
Have bought SMD adaptors for some Pics.

{Original Message removed}

2009\04\06@124909 by olin piclist

face picon face
Benjamin Grant wrote:
> I was looking at pickit2 but if you have other
> recommendations I'm open to them

I'm just a tad biased to my own programmers, but for your situation a
PicKit2 is probably the right answer for develpment.  Once you get into
production you'll need something better, which is where my programmers
become attractive.


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

2009\04\06@142908 by Benjamin Grant

flavicon
face
alright well I'll try out a pickkit2 for now but i'll be sure to look into
yours if I ever have a real product. Best,Ben

On Mon, Apr 6, 2009 at 12:51 PM, Olin Lathrop <olin_piclistspamspamembedinc.com>wrote:

{Quote hidden}

> -

2009\04\06@232653 by Benjamin Grant

flavicon
face
"Why not tell the community the number of microVolts/microAmps of
signal change that correspond to human pulse? Without the number all
the schematics are irrelevant speculations. There is no such thing as
generic circuit."
Fair enough. My main problem is the original signal from the probe is really
annoying me. At first I thought perhaps I didn't have a strong enough LED to
get through the finger(actually that might have been a possibility) so I
bought a more powerful one. The problem is I can't see the unamplified
signal alone as it's in the uA range. Furthermore, if you don't put a finger
in between the LED and the sensor then it of course rails, which is too be
expected. So this time I pulsed a sin wave through the led, placed my finger
in between the LED and sensor and measured the output at each stage. The 4
hz filter works well(i changed the sin wave input frequency from .5 - 10 hz)
and the final output looked fine. However, when I put in a DC signal to the
LED I basically get 0 signal coming out of the final stage.   It's not
suprising it's centered at 0 as the gain stages are AC coupled... however
there's absolutely no apparent modulation due to blood flow.  It's really
frustrating because I don't have a commercial pre-amplifier to amplify the
output of the first stage. Conversely using a sin wave through the IR led
essentially as a signal generator has verified the gain stages and filter,
so I'm not sure as to the root of the problem.

On Mon, Apr 6, 2009 at 2:29 PM, Benjamin Grant <spam_OUTbenjamin.grantspam_OUTspamspam_OUTduke.edu>wrote:

{Quote hidden}

>> --

2009\04\07@100013 by Marechiare

picon face
> The problem is I can't see the unamplified
> signal alone as it's in the uA range.

Get the professional gear to see it. If you want to make billions
people happy, you have to get one.

> however there's absolutely no apparent
> modulation due to blood flow.  It's really
> frustrating because I don't have a commercial
> pre-amplifier to amplify the output of the first stage.

See above.

> Conversely using a sin wave through the IR led
> essentially as a signal generator has verified
> the gain stages and filter, so I'm not sure as to
> the root of the problem.

Start with the root, the rest makes sense after you solve it.

2009\04\07@102514 by Mike snyder

picon face
reading this in gmail popped up http://www.semedicalsupply.com/cms-50hhc.htm

Maybe you can get some ideas from their implementation?

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