Creating An Activity Meter From A Pulse Width

[madhippiescientist]

I have a 556 servo driver circuit for a non-digital servo,
and it sends out about 50 pps centering around 1.5ms. I
also increased the resistance on the "edges" of the servo's
potentiometer (inside it's case), to make it more sensitive.
The timer varies per a 150 ohm coil connected with another
resistor between pins 6/7 and Vcc (the 555 pinout). The
servo offers excellent motion with little movement of the
magnet suspended inside of the 150ohm coil (it varies the
resistance by up to 50ohms). I am not working on making it
any more sensitive as I previously inquired! I want to do a
little twist though, to show it's receiving activity...

Servo Driver Circuitry

I would like to make a bar/dot meter (LM3915) that will light
more LEDs as the pulse widths change. I don't want it to
light according to position of the servo / duration of the width,
though if I have no other choices that would be better than
nothing. My idea is that if the seismometer starts to read a
signal (more sensitive than c 1/10 the servo's full range), the
LEDs will start to jump around like a VU meter, so it will grab
my attention from across the room. Again, ideally, it will light
more LEDs as the width's durations fluctuate (e.g., if it's at
1.5ms <center> and suddenly jumps to say 1.55ms or 1.45ms,
one LED would light, and then if it stays in that area the LEDs
would subside). This way the meter would actually be relaying
"activity" as compared to position. If I can't have it that way, I
would be willing to go with it displaying servo position (though
it's actually showing the width of the pulse as compared to the
differences between the string of pulses).

My problem. How can I integrate the above circuit link with my
coil/R4 between 6/7 and Vcc, by simply tapping into the servo's
signal line (I can't send power through or sense the coil's R,
because that would adversely affect the servo's sensitivity).
It's V is not going to be much or it may be up to the +V that the
servo gets?? (through a 7808), but it will be on for a very very
short duration and then off, 50 times a second. Any suggestions?

 

[ericgibbs]

I have a 556 servo driver circuit for a non-digital servo,
and it sends out about 50 pps centering around 1.5ms. I
also increased the resistance on the "edges" of the servo's
potentiometer (inside it's case), to make it more sensitive.
The timer varies per a 150 ohm coil connected with another
resistor between pins 6/7 and Vcc (the 555 pinout). The
servo offers excellent motion with little movement of the
magnet suspended inside of the 150ohm coil (it varies the
resistance by up to 50ohms). I am not working on making it
anymore sensitive as I previously inquired! I want to do a
little twist to show it's activity though...

Servo Driver Circuitry

I would like to make a bar/dot meter (LM3915) that will light
more LEDs as the pulse width changes.

hi,
You could use the 1V control pin [555] differential voltage to drive a LM3914

Like this image.

 

[madhippiescientist]

So you're speaking of pin5 "CV" on the 555?

 

[madhippiescientist]

Also...

Your mathematical electronic software shows values for
a range of voltage from 2.5 to 3.5v. How about the values
if I only want a range of say 2.75 to 3.25v, so it will be
more sensitive than the servo (I don't want it to take a
full swing of the servo to light the tenth LED, I want the
LEDs to practically all light up when activity is just beginning).

I was also wondering, the values you show for R3 and R4
(15 and 3 respectively), what are they, R in ohms or Kohms?
The software shows them as Vext Low and Vext High (pins
4 & 6), so what do I do with these pins?

Does the servo driver always vary the pin5 of the 555 no
matter which direction it moves? If not, I need to integrate
that to accommodate it being "center" (3.0v??), and maybe
build two meters, one inverse of the other with a lower V
range? Then I could gate corresponding outputs to (-V
LED#10 and +V LED#1 would light the same LED, etc.) ???

 

[ericgibbs]

hi,
The values of the voltages on the LM3914 can be set to suit your Vc[pin5/555]
The value of R4, which would be a Vpot in the final circuit would be connected to REFout pint 7 so that the 2.5V on pin4 could be adjusted to suit. R3 is not used.

The intention is that with a centre point of say 1.5mS == to 3V on Vc, LED #5 would light, as the 1.5mS period increased the the LED's 6,7,8,9,10 would light.
LED10 would indicate 2mS and LED1would indicate 1mS.
I would suggest BAR node for the LM3914, so for 1.5mS, LED's 1 to 5 would light.

If you want to go this way let me know and I will post a circuit diagram.??

 

[madhippiescientist]

I was thinking, no LEDs except "center" lit, as LED#1, and as
the pulse modulates (+ or -), the more LEDs would light. Even
If I did this cascading 2 meters, the visualization I'm aiming for
is each side of the range of LEDs lights in sequence. If I did it
with one meter, I would drive an LED on each side of nil, for
each step.

OOOOOOOOOXXOOOOOOOOO (would be inactive)
OOOOOOXXXXXXXXOOOOOO (would be a start)
OOOXXXXXXXXXXXXXXOOO (would be even more activity)

Any idears? I think I can figure out the cascading, I would just
need the values for a sort of "zero center meter", and use the
bar mode (without half of the meter being on when there's
no activity). Then set zero on the meter to the middle of the
reference range.

Again, you're speaking of the input originating from pin5 of the 555.
This DOES vary with the duration of the pulse it's sending out pin3??

 

[ericgibbs]

If I follow you correctly you are not asking for a pulse width/Vc voltage indicator but a rate of change indicator.??

This would be difficult to achieve in lighting the LED display pattern you have posted, also the LED's lit would be a very fast flash

 

[madhippiescientist]

"Rate Of Change" may be a good identification. I am aware
the LEDs would move very fast, such would be so as to grab
my attention before a big seismic jolt hit, so I could possibly
watch it's arrival on my servo stylus as compared to simply
looking at the graph afterwards. I thought this would be quite
difficult. If anybody could figure this out let me know. Back to
your circuit though, eric, if I do use it simply as a position
indicator (the voltage <position of servo> is displayed), I could
still do it that way, but in order for me to have it "warn" me of
possible ensuing heightened activity, I would need to make the
meter VERY sensitive, and center it's V so as to not show half
the meter lit up. Any ideas about two meters? One for the + &
one for the - , (meaning one will show 3.0 to 3.5v or less range),
the other will show below 3.0v, possibly by inverting the outputs so as
to drive a different common LED bus <Cathode>). This way, even
if they are all on (in the meter calibrated for up to 2.99v), only
those outputs that are off (much less than 2.99v) will trip a GP
transistor to indicate "on" and drive it's appropriate LED. So, I guess
I'm aiming at calibrating the low meter at say 2.75 to 2.99v, and the
upper meter at 3.00 to 3.25v. Then maybe I'll gate their outputs
so as to light 1 to 10 from the center out on each side of two 10 LED
bars (it doesn't need to specify which side is in motion).

I'm sure that would work, and I'll appreciate any feedback into that
idea you may have.

But I still have two Qs for U. Does pin5 have a center V which varies
(2.5 to 3.5v) depending on the pulsewidth that the 555's generating?

What would be ideal values for the meters, to cover a 0.25v range???
Or maybe even a smaller range (remember, 2.5 to 3.5v actually moves
the Servo about 270 degrees. If I could get the range down to maybe
a tenth of a volt, with the option of enlarging it via a set of pots),
then that would be the way I would want to go !!

 

[ericgibbs]

But I still have two Qs for U. Does pin5 have a center V which varies
(2.5 to 3.5v) depending on the pulsewidth that the 555's generating?

You said that the Vc control voltage to the 555 varies from 2.5 thru 3.5v, which corresponds to 1mS and 2mS.
If the Vc was connected to the input pin5 of a LM3914, set for 2.5Vlow to 3.5Vhigh, in DOT mode, at 'rest' ie;say 1.5mS then LED#5 would be lit

What would be ideal values for the meters, to cover a 0.25v range???
Or maybe even a smaller range (remember, 2.5 to 3.5v actually moves
the Servo about 270 degrees. If I could get the range down to maybe
a tenth of a volt, with the option of enlarging it via a set of pots),

It is possible to go to tenths of volts for the LED's
then that would be the way I would want to go !!

hi,
You could use low value resistors in series with the LED's and use PNP transistors to take off signals when the LED lights, which could alert you via say a beeper.
Have you considered using a LM339 quad comp and configuring a customised LED display/alarm.?

EDIT:
If you add a dual OPA for the Vc signal, one buffering and one inverting it would be possible to use two LM3914 and produce the LED pattern that you want.
Let me if you are interested and I will draw circuit.

 

[madhippiescientist]

I have built the meter, and taken readings off of pin5 of
BOTH 555s (astable & monoastable), but any fluctuation
(even full scale) of the servo's motion does little to nothing
re moving the meter. Either the meter needs to be much much
more sensitive, or I do need the LM339 comparator likely
with GP transistors to drive the LEDs...?

How about a schematic for the comparator/drivers (just one
and how to ladder them), the different resisitor values as you
see fit (upon escalation of signal), and maybe an amplifier to
increase the input V signal so as to make it hyper-sensitive??
Also, indicate which amp would be best ???
(and how to wire it / the formula for determining it's gain)

 

[ericgibbs]

I have built the meter, and taken readings off of pin5 of
BOTH 555s (astable & monoastable), but any fluctuation
(even full scale) of the servo's motion does little to nothing
re moving the meter. Either the meter needs to be much much
more sensitive, or I do need the LM339 comparator likely
with GP transistors to drive the LEDs...?

How about a schematic for the comparator/drivers (just one
and how to ladder them), the different resisitor values as you
see fit (upon escalation of signal), and maybe an amplifier to
increase the input V signal so as to make it hyper-sensitive??
Also, indicate which amp would be best ???
(and how to wire it / the formula for determining it's gain)

hi,
Are you able to measure the Vc voltage change on pin 5 of the 555's??
Can you post what the actual change is, over the range you want to indicate, let me know the steady Vc and how much it changes by.

The LM3914's can be made a lot more sensitive if required.

 

[madhippiescientist]

Will check. It may be a few days though, I'm out of town.

 

[madhippiescientist]

I'm back Eric!, so here it goes...

The pulsewidth generator (the monoastable..?) side of the 556 indicates;

Pin ID Reading/Note
Output w/ servo at center => 188mv, in operation from 182-194mv
Trigger Static at 4.99V
Control Static at 4.45v
Threshold adversely affects servo when meter is hooked up to it
Discharge adversely affects servo when meter is hooked up to it
(both of these means sensitivity is narrowed / "violently"
moves when it does)

The Output pin looks like what I would want to use, and with the centering
pot I use to center the servo, the reading varies +/- 1.0mv. Therefore, I
would like the range of the meter to be about 1.0mv itself, so I would get
signals from about a half a mv on either side, and I would (of course) need
to be able to adjust the meter's "centering" (for when I adjust the servo).
This sounds likely impossible, and I do have another idea of how to do all
of this, which may not be viable, so I'll wait and see what you think about
the 3914's total range of about a mv (or a very little bit more than that!).

Thanks.

 

[ericgibbs]

hi,
From a quick look at your +/-1mV change, I would suggest that it needs to be amplified to suit the LM3914.
An amplification of 100 would raise the signal change to +/-100mv
The amplifier OPA output could be set to a '0' centre voltage of 100mV, with a swing of 0V thru 200mV.
The centre '0' of +100mV would light LED #5 and less than 100mV would light LED's 4,3,2,1,,, greater than +100mV to 200mV would light LED's 6,7,8,9,10 as it increased.
The 0mV OPA output would require an ideal OPA whose output could go to 0V, most OPA's cannot when used with a signal supply voltage.

To overcome this OPA 0V out requirement, the OPA could be set for +100mV out which would represent -1mV [ +100mV after amplification] and +200mV would represent 0mV and 300mV == +1mV.
The thresholds of the LM3914 would be set as Vlow = 100mV and Vhigh = 300mV.

This may sound a little complicated , but in the actual circuit its quite easy to achieve.

If you need a circuit diagram let me know.?

EDIT:
Unzip this avi file.

 

[madhippiescientist]

What is "OPA" ??

It sounds quite complicated, so I would REALLY appreciate it
if you could help me with a schematic ! PLEASE !!!

 

[ericgibbs]

What is "OPA" ??

It sounds quite complicated, so I would REALLY appreciate it
if you could help me with a schematic ! PLEASE !!!

hi,
An OPA is an 'operational amplifier', its just an transistor amplifier with a defined performance specification, built in a single 8 pin IC.

I need to now more about the +/-1mV signal that needs amplifying.
Is it +/-1mV about 0V or is it a fixed voltage varying by +/-1mV.??

 

[madhippiescientist]

I figured after I wrote and left work, that you meant "OPA"
as op-amp, it's just I never heard it that abbreviated before...

I wish for the meter's full range to vary by 1mv, at approximately
the timer's "servo center" output (188mv), because the timer always
generates that reading when the servo is still (so, full range of
meter between 187.5 to 188.5mv). Using the 100:1 on an amp
makes this almost 18v...? Choose an amp that could do that !
I'm not that familiar w/ op-amps, so I don't know which one or
how I would wire it, would be best for this scenario.

I would also like to be able to adjust the op-amp's gain as well
(maybe by 10-20%). So, an 80:1 or 120:1 range would be nice.
Since I'm utilizing bar-mode, I'll likely set the meter at one LED
on, and just let the "motion" (the timer's output) drive the meter
on half of it's motion (since when the servo moves it goes back
AND forth). Sooo, a range as stated above, but I won't have to
center it's mid-scale to my servo center voltage, I'll simply center
the meter's bottom scale (LED#1) to the servo center voltage.
This way, I'll get a broader illuminosity over the range of 1mv,
and use less power, and keep the regulator cool as well (the 3914
is actually driving two paralleled LEDs on each output, so when
many of the lights are on for a while, it tends to get pretty hot).

Thanks !

 

[ericgibbs]

hi mhs,
OK, I will draw up a circuit over the weekend.

meter between 187.5 to 188.5mv). Using the 100:1 on an amp
makes this almost 18v...? Choose an amp that could do that !

That would be the case if the 188mV was not 'nulled' out by the OPA circuitry before the 'change' was amplified..

 

[madhippiescientist]

Aahhh, you're talking about amplifying the change!
See, it was quite complicated (at least for me).
I appreciate you drawing up a schematic. If you
don't hear from me, it isn't that I've lost interest
or anything, it will be because I won't have access
to the net for about two weeks (24 hours from now),
because my work is making me take all my paid leave
before the fiscal year ends in September. If I find
internet access while I'm away, I'll check for your
response! Though I'm not in a big hurry since I may
not see your next response until September 1st.

Thanks Bud !!

 

[ericgibbs]

Aahhh, you're talking about amplifying the change!
See, it was quite complicated (at least for me).
I appreciate you drawing up a schematic. If you
don't hear from me, it isn't that I've lost interest
or anything, it will be because I won't have access
to the net for about two weeks (24 hours from now),
because my work is making me take all my paid leave
before the fiscal year ends in September. If I find
internet access while I'm away, I'll check for your
response! Though I'm not in a big hurry since I may
not see your next response until September 1st.

Thanks Bud !!

hi,
Its important that the 180mV base voltage is constant, no drift or noise.
Before you get cut off, so to speak, can you post the circuit showing where exactly the 180mV +/-1mV is connected.?