Re: blood pressure!


Chris wrote:
priyank wrote:
Hi Chris


I spent 5-6 hours trying to get the thing workiI am sorry to the forum
if you guys thought i was trying to get my homework done. Trust me that
was not the case. Its just that i am frustrated with this silly little
thing and i am not sure whats wrong with my circuit!
ng and get rid of noise and get clean signal!

I have now decided to chuck mpx2050gp out and get ASDX015D44R, its gona
cost me fortune but i wannt this thing working.

http://au.farnell.com/jsp/search2/browse.jsp?N=500001+401+1001501&No=50&Ns=PLS_SKU%7c0#results

http://www.farnell.com/datasheets/60901.pdf


Guys u wannt believe but in AUstralia i am not able to find mpx5050gp
and neither the amplier u suggested.

Dspace card does take in negative supply but then to I will surely get
rid of AD622 and get a single supply amplifier as per ur suggestion.

I have bought the LM324 as per the apllication notes and will get the
filter soldered on veroboard tonight.

WIth the filter i surely agree have lot of work needs to be done! I am
really confused about what filter would be best for my circuit!Ill
again go through the textbooks today

I have got the pictures of th output waveform which i am getting on the
scope(i am first trying to check circuit on oscilloscope) and ill try
post the link for pics within next 5 hours.(lol... it will atleast
prove im trying and not trying to get my homework done).

Chris i appreciate ur help and not even my best mates would have helped
me so much.
(sorry bout my spelling errors!)

cheers
priyank


priyank wrote:
Hi Chris


I spent 5-6 hours trying to get the thing workiI am sorry to the forum
if you guys thought i was trying to get my homework done. Trust me that
was not the case. Its just that i am frustrated with this silly little
thing and i am not sure whats wrong with my circuit!
ng and get rid of noise and get clean signal!

I have now decided to chuck mpx2050gp out and get ASDX015D44R, its gona
cost me fortune but i wannt this thing working.

http://au.farnell.com/jsp/search2/browse.jsp?N=500001+401+1001501&No=50&Ns=PLS_SKU%7c0#results

http://www.farnell.com/datasheets/60901.pdf


Guys u wannt believe but in AUstralia i am not able to find mpx5050gp
and neither the amplier u suggested.

Dspace card does take in negative supply but then to I will surely get
rid of AD622 and get a single supply amplifier as per ur suggestion.

I have bought the LM324 as per the apllication notes and will get the
filter soldered on veroboard tonight.

WIth the filter i surely agree have lot of work needs to be done! I am
really confused about what filter would be best for my circuit!Ill
again go through the textbooks today

I have got the pictures of th output waveform which i am getting on the
scope(i am first trying to check circuit on oscilloscope) and ill try
post the link for pics within next 5 hours.(lol... it will atleast
prove im trying and not trying to get my homework done).

Chris i appreciate ur help and not even my best mates would have helped
me so much.
(sorry bout my spelling errors!)

cheers
priyank

Hi, Priyank. OK. I'm assuming you're in an electronics course of some
kind, but I might be wrong. If you wouldn't mind, could you mention
whether you're in a trade school or college, you are an electronics,
computer programming or biology major, what's this class, and what
level class this is.

Here's the deal. If you put something together on a board and plug it
in, it *always* works. If flames are erupting from your board, if
power is hooked up backwards, if you've got everything miswired, it's
still working -- it's still doing exactly what you made it do. You may
have wired it up wrong, you may have hooked up power backwards, but
it's just obeying the laws of physics and the way the chips and
components are designed.

If your circuit isn't doing what you want it to do, you might not
understand what it's made to do. You might not have followed the
correct procedure to get it to operate properly. You may not really
understand what you want. But the problem is you, not the silicon or
wire or metal film or electrolyte and foil, even if there's smoke
pouring from your breadboard. Especially if there's smoke pouring from
the breadboard.

You should be able to cobble together something that works fairly
quickly. You have the sensor and the ICs right now to make something
happen now. An AD622 is a great IC -- it just happens to need dual
power supplies, which might not even be a problem for you. Wire up
your AD622 with the gain set resistor on a small piece of perfboard,
and mount it right next to the sensor (less than an inch away). Now
the venerable LM741 is easily capable of doing the job of the LM324 on
the other side of the 4-conductor wire (near your DAC board) if you
have dual power supplies. Look here:

|
| DAC Board
| .------------.
| Sensor Instr. | |
| Amplifier | |
|.--------. .----------. | |
|| 0-40mV | |Av = 1000 | | |
|| output | | | | |
|| | | o--o---------------o A |
|| | | | | | |
|| o---o | | | |
|| | | | | | |
|| o---o | | .---------. | |
|| | | | | | Filter/ | | |
|| | | | | |Amplifier| | |
|| | | | | | | | |
|'--------' '----------' '--o o--o B |
| | | | |
| '---------' | |
| '------------'
|
|
|
|
(created by AACircuit v1.28.5 beta 02/06/05 www.tech-chat.de)

This is easy.

Your basic cut 'n' paste circuit can be logically divided into blocks.
Look at block 1. It's the sensor. It gives a 40mV differential output
for 0-50kPa. OK. Disconnect the outputs from everything else, power
it up, and then just put a DVM across the inputs while you apply some
air pressure. If you see a millivolt signal, at least you know it's
kind of working. If not, you may likely have smoked it, either now or
before. If these are school samples, they may have been donated from a
broker who had problems with them and decided to get a tax writeoff for
the lot -- it's conceivable you got the one bad one in the lot. No
output? Get another one.

OK. Now look at block 2, the IA. You've chosen a resistor for a gain
of 100 (0-40mV to 0-4V). Does it work? You can reconnect the sensor
if you want and apply that pressure again, or you can just ground one
input and apply a millivolt signal to the other (if you don't have
anything handy, a "D" battery, a series resistor and a pot will make a
good one). Apply a 30mV input. Do you get a 3V output? Does it do
what you think it should do? If not, feel/smell for a hot IC, recheck
your wiring, look at every conceivable way you could have messed it up
before you conclude the IC is bad. If it is, you probably smoked it
before. Get another one. Use your meter. Use your head.

Now look at the filter section. It blocks DC. It lets AC pass. If
you don't have a function generator or any other lab equipment, just
hook up a low voltage transformer, and use a series resistor and a pot
to apply a 60Hz signal of about 2mV at the input of the filter. If you
wired it up right, you should see a large amplitude 60Hz signal at the
output. If you look at the app note, it says the Fig. 1 filter is
amplifying the millivolt signal up to tenths of a volt, so you should
see some good gain here.

Once you've gotten to this point, disconnect the sensor again, and
ground one side of the IA. Apply a 30mV DC voltage with a 1/4mA AC
signal riding on it (battery, transformer, series resistors will do
fine -- just apply your basic DC circuit theory you must have had
sometime before you got into this class). Look at the output of the IA
and the output of the filter amplifier. Do they work together?

This is a basic reality check on your stuff -- is it smoked, or does it
work. If I had your circuit on a board in the lab, I'd do this first.
With decent lab equipment, it wouldn't take more than a few minutes to
find all this out.

You've gotta walk that long road alone. And you've gotta use the tools
you have, including your head.

Git R Done
Chris

OK, Priyank. If you're saying you've got a week left, you don't have
time to play. Don't buy anything yet if you've got split supplies and
your DAQ board can handle negative voltage inputs. You don't have time
to wait for parts. Change over to the filter/amplification block shown
in the Freescale appnote, using your venerable 741. Put it all
together, as shown in the block diagram above. Place the AD622 right
next to the sensor. (By the way, use a couple of good high frequency
tantalums to bypass the AD622 power supply at the IC, otherwise you're
going to have noise problems big time.) Run through the above
preliminary checks to make sure the blocks are working in a gross
fashion, as shown above. Replace anything you smoked before. Get it
all to the point that you're ready to start.

Now that you've done that (shouldn't take more than an hour or two if
you've got access to split +/-10V supplies, but +/-12V or even a +/-15V
supply will work for your sensor, Freescale says +16V is abs. max.),
it's time to actually start working. Start by putting the BP cup on
your arm, closing the valve, pumping it up to over 200, and then
cracking open the valve so it will slowly decompress (should take about
30 seconds or so). Just feel what's happening with your arm as it
decompresses.

You should be able to feel the pressure of your blood pushing slightly
against the BP cup as it decompresses. The BP cup tries to compress
your arteries. As your heart pumps blood through your arm arteries,
they push back, trying to expand to accomodate the flow of blood. If
the BP cup is pressing too much, there isn't enough blood pressure
being generated by the heart muscle to open the arteries even at
maximum pressure. If the BP cup is pressing too little, there isn't
enough pressure to close the arteries at all, even inbetween beats.

That maximum point (the point where the artery is too compressed to
allow blood flow at all) is called the systolic pressure. The minimum
point (where the pressure is too small to cause constriction at all) is
called the diastolic point. Between those two points, the vessel is
pushing open during the high pressure points in the heartbeat cycle,
and closing in the lower pressure parts. This expansion and
contraction of your arm as blood is trying to flow is reflected in your
pressure sensor, although it's not a very reliable effect. It's a very
small effect (as shown in the appnote, about 1mm Hg to 3mm Hg, see
p.2). As Mr. Popelish noted above, this method isn't used in better
automatic sphygmomanometers, they'll use an audio pickup. It is useful
for the "free blood pressure" displays you see in local drugstores that
give our senior citizens something to obsess about while they're
waiting for their meds. And the "oscillometric method" is kind of
indestructible, although not extremely accurate.

But the effect *is* measurable. That's your job. If you want to do
this, you have to take responsibility for acquiring that signal, using
the sensor and the ICs and your DAQ board to do so. If they're working
properly, you do have the tools to do the job. If you have any lab
equipment (an oscilloscope would be nice, a DSO nicer, a function
generator would also be helpful) it will make the trip easier, but you
do have what you need. I could probably get it working with a cheapie
DMM, power supplies, and a transformer and handful of resistors and
caps, although I'd have a heck of a time troubleshooting or tweaking
it.

Again, the appnote filter alone will probably not be sufficient to do
the job well. You'll need either a programming algorithm or additional
filtering to get rid of power supply and line frequency noise.
Depending on the acquisition frequency of your board and your software
skillz, you may also want to add a peak detector to simplify the data
acquisition software to a minimum.

But you can work on that later. Get to the starting line first by
taking ownership of your project, really understanding what you need to
do, and putting together the tools you need to do the job.

Everything I've described here up to this point could easily be
prototyped by an engineering technician in an hour or two. It's not
that hard.

By the way, if you want more help, I need to know something about your
background to be able to answer usefully. Are you in a trade school or
university? What year are you? If university/college, is this the
senior project?

Git R Done
Chris

.

Loading