# Can anyone help me to use 4046 as frequency to voltage converter

#### likhithablr

Joined Jan 28, 2018
1
Hello guys, I am new to AAC. I am working on a small sensor based project.
I am using a hall effect sensor (461C), which only respond to one polarity, thus producing a square wave. My supply is only a single rail 5V supply, so my square wave has amplitude of 5V. A magnet is attached horizontally to a motor and brought close to the sensor, which produces square waves of varying width according to the supply to the motor(which is between 0-3V)
Since it is a analog project, I have to convert frequency to analog dc voltage. On trying out different circuits such as peak detected, voltage doubler circuits, I was advised to use 4046 for FCV, since frequency that I get form my sensor is very small only between 0-100Hz.
Can I ask you to advice me one how to use 4046 as FCV and if possible a circuit diagram can help me.

Thank you in advance for all suggestions and help offered.

#### shteii01

Joined Feb 19, 2010
4,644
I looked at the datasheet, I don't think 4046 will work for this purpose.

#### AlbertHall

Joined Jun 4, 2014
12,157
From figure 9c in the datasheet the maximum ratio of maximum frequency to minimum frequency is about 80 - so from 100Hz down to about 1.2Hz.

#### Alec_t

Joined Sep 17, 2013
13,153
You could use the Hall sensor to trigger a monostable circuit generating fixed-width pulses, then integrate the pulses with an RC filter.
A monostable can be made with a CD4538, for example, or with gates having an inverting function (e.g CD40106, or CD4093).

#### AlbertHall

Joined Jun 4, 2014
12,157
The LM2907 might do the job but the problem still arises at the 0Hz end of the scale. The chip generates pulses then a low pass filter smoothes the pulses into a DC level. For very low frequencies this filter would need a long response time and so the output voltage will always lag changes in the input frequency. That problem will generally apply to such circuits including post #4.

#### Alec_t

Joined Sep 17, 2013
13,153
Yes, the filter is the problem. The only way I can see of avoiding one would be to use a micro (or whatever) to measure and process the pulse period.

#### AlbertHall

Joined Jun 4, 2014
12,157
Yes, the filter is the problem. The only way I can see of avoiding one would be to use a micro (or whatever) to measure and process the pulse period.
Yes, I thought of that. It does still get difficult when the period is multiple seconds, 1S, 10S...

#### ericgibbs

Joined Jan 29, 2010
16,383
hi lik,
I agree with the other posters its the low frequency filtering thats a problem.

For ref this PDF may help.
E

#### Attachments

• 326 KB Views: 16

#### Sensacell

Joined Jun 19, 2012
3,066
At this point, the \$64,000.00 question is what you need to do with the resulting analog signal?

#### Alec_t

Joined Sep 17, 2013
13,153
Shaft speed sensors on a car are faced with the same low frequency problem and overcome this by use of a multi-toothed reluctor wheel. This results in multiple sensor pulses per shaft revolution. Could you do something similar with your setup?

#### ian field

Joined Oct 27, 2012
6,536
Hello guys, I am new to AAC. I am working on a small sensor based project.
I am using a hall effect sensor (461C), which only respond to one polarity, thus producing a square wave. My supply is only a single rail 5V supply, so my square wave has amplitude of 5V. A magnet is attached horizontally to a motor and brought close to the sensor, which produces square waves of varying width according to the supply to the motor(which is between 0-3V)
Since it is a analog project, I have to convert frequency to analog dc voltage. On trying out different circuits such as peak detected, voltage doubler circuits, I was advised to use 4046 for FCV, since frequency that I get form my sensor is very small only between 0-100Hz.
Can I ask you to advice me one how to use 4046 as FCV and if possible a circuit diagram can help me.

Thank you in advance for all suggestions and help offered.
That would only use a small part of the 4046, and its not really designed for that application.

there were various voltage/frequency chips used in mode detection in early VGA multisync monitors - most application notes show examples of either way conversion.

The NE567 might be a candidate, but its a version of a PLL that I don't know much about. AFAICR: its listed as a tone decoder - a handy key word to search with.

#### Alec_t

Joined Sep 17, 2013
13,153
For my own interest I've come up with an analogue method of generating a voltage proportional to the pulse period. It updates within one pulse period and does not need a low-pass filter. However, since this is a homework thread, I think we need to let the TS come back with his own ideas before offering any further hints/guidance.

#### DickCappels

Joined Aug 21, 2008
9,317
Alect_t's post immediately above reminded me that this thread is in Homework Help. I checked and the thread starter has not been on the site since I made the post immediately above his, which included a detailed schematic. That post has been deleted. Thank you @Alec_t . I hope you can show us your other circuit sometime.

#### Alec_t

Joined Sep 17, 2013
13,153
I hope you can show us your other circuit sometime.
A few clues: sawtooth, analogue gate, sample-and-hold. I expect you can take it from there .

#### DickCappels

Joined Aug 21, 2008
9,317
Yep

#### sarahMCML

Joined May 11, 2019
151
Post just caught my eye out of interest. No need for further updates, thanks.

Regards,

Sarah.