Searching \ for '[EE] frequency to voltage converter (FVC)' 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/power.htm?key=voltage
Search entire site for: 'frequency to voltage converter (FVC)'.

Exact match. Not showing close matches.
PICList Thread
'[EE] frequency to voltage converter (FVC)'
2009\05\28@000635 by Bruno L. Albrecht [GMAIL]

picon face
Hey guys,

i've been trying the whole month to build an analogic frequency to
voltage converter, but got no good results. I can't use any black boxes
such as LM2907, LM331, LM555, etc. I should use only opamps, resistors,
capacitors, etc...

my input signal comes from a home-made encoder (one IR emitter spinning
and one IR receiver fixed on the side of the motor), so that it
generates an "impulses" in a random frequency. This frequency is what I
want to convert to a voltage or even another kind of signal, such as a
square or a sin waveform...

anyone got an idea?

(sorry for the bad English...)

cheers,
Bruno

--
 Bruno L. Albrecht
 Eng. da Computação/06
 Falker Automação Agrícola Ltda.
 http://www.falker.com.br

2009\05\28@025043 by Ruben Jönsson

flavicon
face
Hi Bruno,

> Hey guys,
>
> i've been trying the whole month to build an analogic frequency to
> voltage converter, but got no good results. I can't use any black boxes
> such as LM2907, LM331, LM555, etc. I should use only opamps, resistors,
> capacitors, etc...

I'm sure that there are a lot of people here that can help you but we have to
do a lot of guessing. You say you can't use a "black box" like a 555 but you
can still use an opamp which can also be considered a "black box". Perhaps you
can explain why you can't use certain parts and better define what a "black
box" is.

>
> my input signal comes from a home-made encoder (one IR emitter spinning
> and one IR receiver fixed on the side of the motor), so that it
> generates an "impulses" in a random frequency. This frequency is what I
> want to convert to a voltage or even another kind of signal, such as a
> square or a sin waveform...
>
> anyone got an idea?

Perhaps you can also explain why you can't use the frequency directly but still
can use a square or a sin waveform. What are you going to use the output signal
for and what will it be connected to?

{Quote hidden}

/Ruben
==============================
Ruben Jönsson
AB Liros Electronic
Box 9124, 200 39 Malmö, Sweden
TEL INT +46 40142078
FAX INT +46 40947388
spam_OUTrubenTakeThisOuTspampp.sbbs.se
==============================

2009\05\28@073449 by Bruno L. Albrecht

picon face
This is actually part of a speed controlled motor controller. Know what I
mean? I have to control the motor's voltage in order to keep its speed
constant. This means I'll have to compare the speed with some reference and
then raise or lower the motor's voltage to keep its speed as I want
(probably this reference will be adjustable). I should do all this only with
opamps, resistors, capacitors, etc.
By black boxes, I meant I can't use Frequency-to-Voltage ICs, I should do
one myself. And can't use any "digital devices", such as a microcontroller,
or a A/D converter...

Cheers,
Bruno

2009/5/28 Ruben Jönsson <.....rubenKILLspamspam@spam@pp.sbbs.se>

{Quote hidden}

>

2009\05\28@080017 by Thomas C Sefranek

flavicon
face
 ----- Original Message -----
 From: Bruno L. Albrecht
 To: Microcontroller discussion list - Public.
 Sent: Thursday, May 28, 2009 7:34 AM
 Subject: Re: [EE] frequency to voltage converter (FVC)


 This is actually part of a speed controlled motor controller. Know what I
 mean? I have to control the motor's voltage in order to keep its speed
 constant. This means I'll have to compare the speed with some reference and
 then raise or lower the motor's voltage to keep its speed as I want
 (probably this reference will be adjustable). I should do all this only with
 opamps, resistors, capacitors, etc.
 By black boxes, I meant I can't use Frequency-to-Voltage ICs, I should do
 one myself. And can't use any "digital devices", such as a microcontroller,
 or a A/D converter...

 Cheers,
  Bruno


Homework problem, lab project?


 *
 |  __O    Thomas C. Sefranek  .....tcsKILLspamspam.....cmcorp.com
 |_-\<,_   Amateur Radio Operator: WA1RHP
 (*)/ (*)  Bicycle mobile on 145.41, 448.625 MHz

ARRL Instructor, Technical Specialist, VE Contact.
hamradio.cmcorp.com/inventory/Inventory.html
http://www.harvardrepeater.org

2009\05\28@083033 by Bruno L. Albrecht

picon face
actually is the final project of a discipline...

On Thu, May 28, 2009 at 8:59 AM, Thomas C Sefranek <EraseMEwa1rhpspam_OUTspamTakeThisOuTarrl.net> wrote:

>  {Original Message removed}

2009\05\28@094557 by olin piclist

face picon face
"Ruben Jönsson" wrote:
> I'm sure that there are a lot of people here that can help you but we
> have to do a lot of guessing. You say you can't use a "black box" like
> a 555 but you can still use an opamp which can also be considered a
> "black box". Perhaps you can explain why you can't use certain parts

This smells like a homework assignment.


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

2009\05\28@095628 by Bruno L. Albrecht

picon face
not exactly...actually is just part of a major final project...
anyway...help?

On Thu, May 28, 2009 at 10:47 AM, Olin Lathrop <olin_piclistspamspam_OUTembedinc.com>wrote:

{Quote hidden}

>

2009\05\28@100219 by Eoin Ross

flavicon
face
Integrator circuit. This is only to get you facing in a direction for thought/research.

Thats all I'm willing to help on a school project. You should approach your professor/teacher and read your text books.

>>> "Bruno L. Albrecht" <@spam@brunolalbKILLspamspamgmail.com> 28 May 09 08:30:30 >>>
actually is the final project of a discipline...

On Thu, May 28, 2009 at 8:59 AM, Thomas C Sefranek <KILLspamwa1rhpKILLspamspamarrl.net> wrote:

>  {Original Message removed}

2009\05\28@101341 by Bruno L. Albrecht

picon face
i'm already trying the integrator...the problem is that the pulse that comes
from the encoder is almost an impulse! the capacitor doesn't charge
enough...

On Thu, May 28, 2009 at 11:01 AM, Eoin Ross <RemoveMEerossTakeThisOuTspamchemstation.com> wrote:

{Quote hidden}

> >  {Original Message removed}

2009\05\28@104329 by Harold Hallikainen

face
flavicon
face
A fairly common frequency to voltage converter, often called a "pulse
counting FM demodulator" consists of a monostable multivibrator followed
by a low pass filter. Since the DC component of the output of the LPF is
proportional to the duty cycle of its input, and the duty cycle of a fixed
pulse width stream is proportional to its frequency, you get frequency to
voltage conversion. This was frequently used for precision FM demodulation
in broadcast instruments (before it all went DSP). For a sample circuit
(largely using discrete transistors), see the FMM-1 at
louise.hallikainen.org/BroadcastHistory/index.php/Belar%20Electronics%20Laboratory
. I recently spoke with the designer of the FMM-1. He told me their later
monitor used a 74121 monostable instead of the discrete monostable and got
better performance with the 74121.

So, doing this with op amps, I'd make a monostable with one op amp, then
an LPF with another.

Harold



--
FCC Rules Updated Daily at http://www.hallikainen.com - Advertising
opportunities available!

2009\05\28@104719 by olin piclist

face picon face
Bruno L. Albrecht wrote:
> actually is the final project of a discipline...

Thought so.  So what have you done to figure this out yourself?


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

2009\05\28@104956 by olin piclist

face picon face
Bruno L. Albrecht wrote:
>> This smells like a homework assignment.
>
> not exactly...actually is just part of a major final project...

<sarcasm> Oh, that makes all the difference then. </sarcasm>


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

2009\05\28@105102 by olin piclist

face picon face
Bruno L. Albrecht wrote:
> i'm already trying the integrator...the problem is that the pulse that
> comes from the encoder is almost an impulse! the capacitor doesn't
> charge enough...

So what characteristic does the the true pulse have that the background
noise doesn't have?


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

2009\05\28@110038 by Tony Smith

flavicon
face
> This is actually part of a speed controlled motor controller. Know what I
> mean? I have to control the motor's voltage in order to keep its speed
> constant. This means I'll have to compare the speed with some reference
and
> then raise or lower the motor's voltage to keep its speed as I want
> (probably this reference will be adjustable). I should do all this only
with
> opamps, resistors, capacitors, etc.
> By black boxes, I meant I can't use Frequency-to-Voltage ICs, I should do
> one myself. And can't use any "digital devices", such as a
microcontroller,
> or a A/D converter...


DC motor?

Use PWM and measure the back EMF during the off periods.  A bit tricky
without a uC, I suppose.

Tony

2009\05\28@113109 by Harold Hallikainen

face
flavicon
face

> Bruno L. Albrecht wrote:
>> actually is the final project of a discipline...
>
> Thought so.  So what have you done to figure this out yourself?
>

Remember in the "real world," we are allowed to ask for ideas... Anyone
remember the movie "Shirlie Valentine?" In one scene, as I recall, she's
an elementary school student who apparently never knows the answer and is
never called upon. One day, she does know the answer and is called upon.
She gives the right answer, and the teacher says "someone must have told
you." She replies "How else would I know it?"

Harold

--
FCC Rules Updated Daily at http://www.hallikainen.com - Advertising
opportunities available!

2009\05\28@113615 by Michael Rigby-Jones

flavicon
face
> -----Original Message-----
> From: RemoveMEpiclist-bouncesspamTakeThisOuTmit.edu [piclist-bouncesEraseMEspam.....mit.edu] On
Behalf
> Of Bruno L. Albrecht
> Sent: 28 May 2009 15:12
> To: Microcontroller discussion list - Public.
> Subject: Re: [EE] frequency to voltage converter (FVC)
>
> i'm already trying the integrator...the problem is that the pulse that
> comes
> from the encoder is almost an impulse! the capacitor doesn't charge
> enough...


If encoder delivers a pulse for a fixed angle of rotation this could
never work anyway, since the integrated voltage would be the same
irrespective of the motor shaft speed.

You need to use a monostable multivibrator (AKA "one-shot") triggered by
the encoder prior to integration.

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\05\28@122035 by Bob Blick

face
flavicon
face

On Thu, 28 May 2009 10:56:25 -0300, "Bruno L. Albrecht"
<EraseMEbrunolalbspamgmail.com> said:
> not exactly...actually is just part of a major final project...
> anyway...help?

A single-ended differentiator followed by an integrator will do
frequency to voltage. To do stable motor speed control you will also
need at least one or two parts of a PID circuit, depending on how
sloppy/fast/accurate you need it to be. See Bob Pease's articles about
PID and "ball balancing on beam" for opamp PID circuits.

Cheers,

Bob

--
http://www.fastmail.fm - A no graphics, no pop-ups email service

2009\05\28@143652 by Herbert Graf

picon face
On Thu, 2009-05-28 at 10:51 -0400, Olin Lathrop wrote:
> Bruno L. Albrecht wrote:
> >> This smells like a homework assignment.
> >
> > not exactly...actually is just part of a major final project...
>
> <sarcasm> Oh, that makes all the difference then. </sarcasm>

Olin, enough. If you don't want to help don't, if you do, great, but
this is NOT constructive and is liable to start another flame war.

There is NOTHING wrong with asking for help in figuring something out,
even if it is a homework assignment.

2009\05\28@160831 by olin piclist

face picon face
Herbert Graf wrote:
>>>> This smells like a homework assignment.
>>>
>>> not exactly...actually is just part of a major final project...
>>
>> <sarcasm> Oh, that makes all the difference then. </sarcasm>
>
> Olin, enough. If you don't want to help don't, if you do, great, but
> this is NOT constructive and is liable to start another flame war.
>
> There is NOTHING wrong with asking for help in figuring something out,
> even if it is a homework assignment.

It's OK to ask for help with a homework assignment, and I've asked him some
particular questions to that end (which he hasn't replied to, at least yet).
However, trying to make a distinction between it being a project versus some
other type of homework is rather silly, which is what the comment above was
about.


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

2009\05\28@171711 by Herbert Graf

picon face
On Thu, 2009-05-28 at 16:09 -0400, Olin Lathrop wrote:
> Herbert Graf wrote:
> >>>> This smells like a homework assignment.
> >>>
> >>> not exactly...actually is just part of a major final project...
> >>
> >> <sarcasm> Oh, that makes all the difference then. </sarcasm>
> >
> > Olin, enough. If you don't want to help don't, if you do, great, but
> > this is NOT constructive and is liable to start another flame war.
> >
> > There is NOTHING wrong with asking for help in figuring something out,
> > even if it is a homework assignment.
>
> It's OK to ask for help with a homework assignment, and I've asked him some
> particular questions to that end (which he hasn't replied to, at least yet).
> However, trying to make a distinction between it being a project versus some
> other type of homework is rather silly, which is what the comment above was
> about.

Fine, I'll agree that "homework" and "project", for these purposes can
pretty much mean the same thing.

That said, whether or not YOU think making the distinction is silly, by
itself, doesn't belong on the list. Next time make it a private email,
if you wish.

Thanks, TTYL

2009\05\28@173338 by William \Chops\ Westfield

face picon face

On May 28, 2009, at 2:17 PM, Herbert Graf wrote:

> I'll agree that "homework" and "project", for these purposes can
> pretty much mean the same thing.

I dunno.  Asking for help on a Freq to Voltage converter that is part  
of a PROJECT to run a motor at constant speed seems to me to be a LOT  
different than having a homework problem that says "design a Freq to  
Voltage converter without magic chips" and posting that exact question  
to the net.  It's the difference between asking for help and asking  
for just the answer...

BillW

2009\05\28@180900 by olin piclist

face picon face
William Chops" Westfield" wrote:
> I dunno.  Asking for help on a Freq to Voltage converter that is part
> of a PROJECT to run a motor at constant speed seems to me to be a LOT
> different than having a homework problem that says "design a Freq to
> Voltage converter without magic chips" and posting that exact question
> to the net.  It's the difference between asking for help and asking
> for just the answer...

I didn't get the feeling that he was asking for the solution on a silver
platter, but I did think it was important to establish whether this was some
kind of assignment where the purpose was to learn about the solution.  Since
it is, hopefully nobody here will just tell him the answer or design a
circuit for him.  I have no problem leading him in the right direction,
however, and giving him general ideas how to solve the problem.

As for the solution, I see that a bunch of people gave the standard knee
jerk answer of low pass filtering a one-shot output.  While that is a valid
approach, it is a bit cumbersome to do just in analog electronics as his
assignment apparently dictates.  Then there is the issue of response time
versus ripple, which we don't have enough information to decide about, but
it will most likely need to be carefully considered.

I think there are easier ways to do this in analog considering a one-shot
chip is not allowed.  Think of creating a signal proportional to the time
between pulses instead of duty cycle.  No, I'm not going to spell out the
answer.  But if the OP comes back, answers the questions asked whether he
thinks they matter or not, and shows he's done some thinking I'm willing to
continue to give him clues to prod him into coming up with the idea himself.


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

2009\05\29@055546 by Michael Rigby-Jones

flavicon
face


> -----Original Message-----
> From: RemoveMEpiclist-bouncesEraseMEspamEraseMEmit.edu [RemoveMEpiclist-bouncesspam_OUTspamKILLspammit.edu] On
Behalf
> Of Olin Lathrop
> Sent: 28 May 2009 23:10
> To: Microcontroller discussion list - Public.
> Subject: Re: [EE] frequency to voltage converter (FVC)
>
>
> As for the solution, I see that a bunch of people gave the standard
knee
> jerk answer of low pass filtering a one-shot output.  While that is a
> valid
> approach, it is a bit cumbersome to do just in analog electronics as
his
> assignment apparently dictates

A one-shot can be implemented with a single op-amp, and the filter could
be purely passive design.  This is simple both in concept and in design,
what is cumbersome about it?  The only downside I could see is if you
want to control the motor over a very wide range of speeds, where a
fixed filter cut off frequency could be problematic.

If you want to get fancy with a V/F I guess you could have the encoder
output trigger a sample and hold on the output of an integrator
(integrating a fixed volateg), and then reset the integrator.  That
sounds like an overly complex scheme however.

Another option would be to implement a PLL and compare motor speed
against a reference frequency rather than voltage.  Again a more complex
system.

What alternative schemes do you have in mind?

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\05\29@120802 by olin piclist

face picon face
Michael Rigby-Jones wrote:
> A one-shot can be implemented with a single op-amp,

Remember, it has to put out a pulse of fixed width and reset to be ready to
do the next pulse in a reasonable time.  It also has to discriminate the
pulse from the noise.

> The only downside I could
> see is if you want to control the motor over a very wide range of
> speeds, where a fixed filter cut off frequency could be problematic.

At best your center speed will be at the 50% duty cycle output.  That still
leaves considerable ripple after a low pass filter or a slow response.  We
don't know how important these are since the OP refuses to provide more
meaningful information.

> What alternative schemes do you have in mind?

As I said, I don't want to just give out the answer to the OP.  What I have
in mind should be doable with two transistors and a few resistors and caps.
The output signal is inherently smoother than the PWM case, although some
filtering will likely be required.  You might add a opamp to buffer the
signal and allow for a little more filtering.

OK, I just drew out the circuit and it came to 2 transistors, 4 resistors,
and two capacitors, all jellybean parts.  Add another R-C or two and a opamp
if you want more filtering, but the original signal won't be too bad.


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


'[EE] frequency to voltage converter (FVC)'
2009\06\09@013526 by Bob Blick
face
flavicon
face
Olin Lathrop wrote:
>
> As I said, I don't want to just give out the answer to the OP.  What I have
> in mind should be doable with two transistors and a few resistors and caps.
> The output signal is inherently smoother than the PWM case, although some
> filtering will likely be required.  You might add a opamp to buffer the
> signal and allow for a little more filtering.
>
> OK, I just drew out the circuit and it came to 2 transistors, 4 resistors,
> and two capacitors, all jellybean parts.  Add another R-C or two and a opamp
> if you want more filtering, but the original signal won't be too bad.

Even if the original poster is long gone, I'm still interested in seeing
the circuit if you're willing to post it.

Thanks,

Bob

2009\06\09@072913 by olin piclist

face picon face
Bob Blick wrote:
>> OK, I just drew out the circuit and it came to 2 transistors, 4
>> resistors, and two capacitors, all jellybean parts.  Add another R-C
>> or two and a opamp if you want more filtering, but the original signal
>> won't be too bad.
>
> Even if the original poster is long gone, I'm still interested in seeing
> the circuit if you're willing to post it.

It's been so long since the drive-by poster asked the question that I'm not
sure of the original requirements anymore.  I think he has a motor with a
sensor that produces a pulse every rotation, and wants a analog signal
suitable as the speed feedback into a speed controller.

Most responses were along the line of making a signal proportional to
frequency by using a retriggerable one-shot and low pass filtering.  If I
remember right, the OP said the signal is a pulse with inherently low duty
cycle since it is produced by a sensor that is triggered once per
revolution.

My idea comes from two observations:

1 - You don't need a signal stricly proportional to frequency.  Linear is
nice, but mostly you need something monotonic with frequency that isn't too
non-linear.

2 - Polarity doesn't matter.  You can design the feedback circuit either
way.

My idea is to charge up a cap during the pulse off time, then reset it
during the pulse.  This produces roughly a sawtooth, although the rising
slope is exponential instead of linear.  This is fed into a peak detector to
yield a signal that is roughly proportional to, and definitely monotonic
with period.  It's not the same as frequency, but the stated purpose was to
control the motor to run at a fixed speed.  This can just as well be
accomplished by controlling it to a fixed period.

I'm at a customer site all day and can't draw up the circuit now.  I'll try
to remember to do it tomorrow when I'm back at Embed.  In the mean time I
just scribbled it out again on a piece of paper, and yup, it's 2
transistors, 4 resistors, and 2 caps.  One resistor can be eliminated if the
pulse is nice and clean, such as would come out of a normal logic gate.  I
sortof remember the OP said the pulse is from a 12V system directly from the
sensor, so I added a extra resistor that provides more noise immunity.

So you guys have one day to see if you can diagram what I described or even
do better.


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

2009\06\09@080200 by Michael Rigby-Jones

flavicon
face


> -----Original Message-----
> From: RemoveMEpiclist-bouncesTakeThisOuTspamspammit.edu [EraseMEpiclist-bouncesspamspamspamBeGonemit.edu] On
Behalf
> Of Olin Lathrop
> Sent: 09 June 2009 12:32
> To: Microcontroller discussion list - Public.
> Subject: Re: [EE] frequency to voltage converter (FVC)
>
> My idea is to charge up a cap during the pulse off time, then reset it
> during the pulse.  This produces roughly a sawtooth, although the
rising
> slope is exponential instead of linear.  This is fed into a peak
detector
> to
> yield a signal that is roughly proportional to, and definitely
monotonic
> with period.  It's not the same as frequency, but the stated purpose
was
> to
> control the motor to run at a fixed speed.  This can just as well be
> accomplished by controlling it to a fixed period.

Ahem, from my previous post:

"If you want to get fancy with a V/F I guess you could have the encoder
output trigger a sample and hold on the output of an integrator
(integrating a fixed voltage), and then reset the integrator."

The sample and hold in this case is a peak detector.  There would have
to be some way to reset the peak detector, a simple resistor to bleed
the cap might work, but would be a trade-off between bandwidth of the
control loop and ripple.

Add another transistor for a simple constant current source to improve
linearity.

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\06\09@092148 by olin piclist

face picon face
Michael Rigby-Jones wrote:
> Ahem, from my previous post:
>
> "If you want to get fancy with a V/F I guess you could have the encoder
> output trigger a sample and hold on the output of an integrator
> (integrating a fixed voltage), and then reset the integrator."
>
> The sample and hold in this case is a peak detector.  There would have
> to be some way to reset the peak detector, a simple resistor to bleed
> the cap might work, but would be a trade-off between bandwidth of the
> control loop and ripple.

I guess if you look at it sideways just right you can argue that what I said
is a implementation of what you said.  I doubt what most people reading your
post would have considered you meant a simple peak detector by sample and
hold.

Note that in either case the output is more directly a function of period
than frequency, so saying this is a "fancy" V/F converter is a bit
misleading since that would generally give the impression the output was
linear with frequency.

> Add another transistor for a simple constant current source to improve
> linearity.

Or just use the exponential signal in a small fraction of the power supply
voltage range.  However, again, even that would be linear with period, not
frequency.  I think the control system could be made to work with period
just fine, especially if the range of controlled speeds was reasonably
limited.  The OP only talked about a single speed if I remember right, which
would be no problem with a period signal.  But it's important to keep in
mind that the control outputs are more likely to be effecting speed linearly
than period, and therefore a high performance wide ranging speed controller
would be better suited with a frequency feedback signal than a period one.
That's the pure theory, but engineering is about getting to good enough,
which in this case can be achieved with a period given what little the OP
said.

The point is a truly "linear" input signal means proportional to frequency.
Therefore going thru much trouble to make the period signal linear is
missing the point.  In fact, the exponential nature of the period signal
with a simple R and C is actually more linear with respect to frequency.  A
misguided attempt to linearize the period signal by adding a current source
actually makes things worse.


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

2009\06\10@182504 by olin piclist

face picon face
part 1 1134 bytes content-type:text/plain; (decoded 7bit)

A diagram of the circuit I mentioned yesterday is attached.

The pulse is assumed to be asserted a small fraction of the time.  Q1 is
therefore off most of the time, allowing C1 to charge up.  Q2 is a emitter
follower and buffers the signal.  When the pulse comes along, Q1 is turned
on, discharging C1.  The R1/R2 voltage divider sets the pulse voltage needed
to turn on Q1.

Q2 and C2 form a peak detector.  Rising voltage on C1 is transferred to C2
quickly because of emitter follower Q2.  Falling voltages on C1 are not
directly transferred to C2 because Q2 only works in one directly.  The
voltage on C2 will eventually decay due to R4, which can be adjusted to be
much slower.

The R3*C1 time constant is set so that a reasonable output voltage results
from the desired period.  If a range of speeds needs to be controlled, then
the R3*C1 time constant should be such that C1 reaches maybe 2/3 of the
supply voltage at the longest period (slowest speed) of interest.

The output will have some ripple, but less than filtering a retriggerable
one-shot output at the same output time constant.


part 2 6627 bytes content-type:image/gif; (decode)


part 3 35 bytes content-type:text/plain; charset="us-ascii"
(decoded 7bit)

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