# F to V converter problem

Discussion in 'Electronic Design' started by [email protected], Nov 11, 2008.

1. ### Guest

The frequency to voltage converter makes use of a LM324, the circuit
schematic can be found in (Page-4) http://www.fairchildsemi.com/an/AN/AN-140.pdf.

The LM324 is fed off a single supply and the decoupling is based on
the design found in (Figure 3, Page-2)
http://instruct1.cit.cornell.edu/Courses/ee476/ideas/singlesupply.pdf.

The circuit parameters chosen for the frequency to voltage converter
(corresponding to the circuit design parameter names used in the above
mentioned urls) were chosen to achieve the following performance:

1. Maximum frequency for max output with Vcc=9v is 1kHz
2. Response time is approximately 100ms
3. Circuit output approaches Low Level output of LM324
(approximately 5mV) as input frequency approaches zero

My question is as follows:

In my current circuit design, the circuit output is not behaving
correctly. As the input frequency approaches zero, the frequency to
voltage converter output is no where near the Low Level output of the
LM324. Further more, the output has unwanted ripples which i am trying
to eliminate. I am seeking help on the possible changes and/or
corrections required in order to achieve the desired circuit
performance as listed above. Thank you.

George.

2. ### James ArthurGuest

Offhand,

o The LM324 input impedances aren't matched, creating
an offset error, which is input to the integrator.

o The op-amp's inverting input can get dragged below
its common-mode range.

o The R1-C1 called out won't charge or discharge
fully or symmetrically for higher frequency inputs,
producing non-linearities.

o (d Vf/ dT) of D1-2 not compensated.

o Vout drifts between input pulses, due to dielectric
absorption in C2.

Anything else?

Cheers,
James Arthur

3. ### Guest

"As the input frequency approaches zero, the frequency to
voltage converter output is no where near the Low Level output of the
LM324."

By running this from a single supply you have changed your "zero" to
be at 1/2 the supply voltage. I would predict a zero frequency output
near 4.5 volts (Vcc = 9V) You could add another op-amp to do some
level shifting perhaps.

"Further more, the output has unwanted ripples which i am trying
to eliminate"

Do the ripples have a 100ms time constant? If so that is just the
nature of a F-V converter. You can reduce the ripple by increaseing
the time constant (averaging more of the input pulses) but this

George Herold

4. ### Guest

"No. Where would the 4.5 come from?"

Oh, He is running this single supply and his second link seemed to
imply that he had biasied the non-inverting input up to one half the
supply voltage. But this is just an assumption on my part. A real

BTW it is fairly simple to configure the ADVFC32 from analog devices
into a F-V converter.

George Herold

5. ### Rich GriseGuest

It must be mid-terms. I wonder if we'll ever see any of the latest batch
of googlies come back to check for answers, or do they expect them to just
show up on thier desk?

Thanks,
Rich

6. ### James ArthurGuest

Equally bad, when C1 pulses below ground its charge parcel
dumps partly through the op-amp's input diode rather than
getting integrated onto C2.

That can happen at every single pulse, since the
LM324 isn't fast enough to prevent it.

A resistor to GND would harden up the LM324's
output and make that less troublesome.

James Arthur

7. ### DanielGuest

Please be nice to this guy. He's a cadet engineer that I'm supervising
here at an electrical untility company and I told him that newsgroups
is a wonderful learning resource that he should take advantage of,
learn from....and eventually contribute to. I'll suggest he posts
values etc...many thanks on his behalf for the help so far.

Daniel

8. ### Guest

Gentlemen,

My apologies for not replying sooner, this is the first time i am
using this News Group
and certainly did not expect such quick replies. So thank you so much
I have sent out an email to all who posted comments to my question,
which contains
the latest circuit schematic and component values. Any further advice
and instructions
would be greatly appreciated.

George.

9. ### Ian MalcolmGuest

DONT take it to email unless you need confidentiality or to transfer a
binary file. You asked the question here and for every poster who
replied, there are typically between 10 and 100 users of this group who
have looked at this thread who might have some comment to make. Can you
put the schematic on a web page somewhere and post a link to it here?

Also many of the posters here DO NOT post with a valid email address to
avoid vast quantities of spam and will never get your message.

Usenet rule of thumb:

ONLY take it to email if invited to or you are already a friend of the
recipient.

10. ### Rich GriseGuest

Teach him to lurk, and to use the other side of google.

Good Luck!
Rich

11. ### Guest

Thanks for the 'Heads Up' Ian, point taken.

As i do not have a web-site of my own at the moment, shall look into
getting one up to make things easier.

George.

12. ### James ArthurGuest

We often post schematic in ASCII text, right here in s.e.d.

You can use a dedicated editor, like Andy's Ascii-Circuit V1.28

or just use plain old Notepad.

Here's your schematic, in ASCII: (view in Courier font)

| || +
.-. .---||----.
| | | || C2 |
R | | | ___ |
'-' C1 +--|___|--+
C | 74hc14 R1 | R2 |
|| | |\ ___ || D2 | |\ |
o--||----+---| >O--|___|--||-+---|<--+--|-\ |
|| | |/ || | | >---+--> output
| | .----|+/
.-. v | |/
R | | D1 - | LM324
| | | |
'-' === ===
| GND GND
===
GND

(created by AACircuit v1.28.4 beta 13/12/04 www.tech-chat.de)

Cheers,
James Arthur

13. ### Guest

Thanks James.

Could you please elaborate on a point you made earlier that "(d Vf /
Dt) of D1-2 not compensated".

I have rederived the design equation for the circuit output:

Vout = 4.5 x (1 - Z1 / Z2)

This equation assumes that Vin is a 9V peak-to-peak square wave coming
from the Schmitt Trigger
and the circuit supply voltage Vs is 9V DC. while Z1 is the equivalent
impedance of R1 & C1, Z2 is
the equivalent impedance of R2 & C2. Also assuming that i am using
single supply decoupling.

Doing this the old fashion way by bread-boarding it, the output is Vs/
2 regardless of the value of Vf.
I have not being able to resolve this issue.

Based on this design the new component values are as follows:

R1 = 10k
C1 = 0.1uF
R2 = 50k
C2 = 0.39uF

This gives the following dervied circuit performance @ 200Hz Vf:

response time = 20ms
Z1 = 18k
Z2 = 2k
Vout = 4.009V

The Schmitt Trigger i am using has a Negative Going Threshold of about
4V, thus, frequencies
below 200Hz will produce a high output on the Schmitt Trigger. This
will in turn be fed into another
Schmitt Trigger simply inverting the high output to low, which will
drive an LED that switches on
whenever the frequency is below 200Hz.

George.

14. ### James ArthurGuest

I can't. Thanks to the NYAG and the danger of someone
somewhere possibly seeing nakedness, my ISP no longer
provides abse.

James Arthur

15. ### James ArthurGuest

Decent.

I did a twin interleaved charge-dispenser
thing, to cut drift. Not worth the bother
in this case.

Cheers,
James Arthur

16. ### James ArthurGuest

It's a drift-over-temperature thing. Vf--the forward
voltage of D1 and D2--changes with temperature.

That will change the voltage placed across C1, and how much
of that voltage is input to the integrator U2.

Specifically, the effective voltage across C1 will increase
about 5mV/ºC. For a 5v supply, that's a drift of +0.1%/ºC.

You might not care.

Cheers,
James Arthur

17. ### James ArthurGuest

No. Consider the circuit as dispensing small packages of
charge from C1 at each input rising edge, and integrating
those in C2. R2 bleeds off that charge at a controlled rate.

If you prefer, as input pulses come more quickly, the
rectified pulses from C1 look more and more like a
continuous current being fed into an inverting stage
with a gain set by R2.

Cheers,
James Arthur

18. ### Guest

Ahh very nice. I didn't quite 'get' how the ripple cancels when you
described this before. But I see it now, (I think). You choose the
one shot pulse time to be 1/2 the period of the "zero ripple
frequency". This would imply that the maximum frequency is then
twice the "zero ripple frequency". Is that correct?

George Herold

19. ### ehsjrGuest

http://www.usenet-replayer.com/groups/alt.binaries.schematics.electronic.html

You have to wait (a day or 2 ?) then it can be seen.
Jim's is there, now.

Ed

Thanks!