CD74HC7046A: CD74HC7046A

Hi Itamar,

For the simplest low pass filter, you only need 2 components:

There are essentially two factors in calculating this:

(1) Time required to stabilize the output

(2) Output voltage ripple

Ideally, the filter would take zero time to update the output to the correct voltage, and that voltage would be a perfect DC value.  Of course, this isn't really possible. You can spend a long time with the math trying to optimize both of those items, but I find that to be a waste of time for a first order filter.

I usually recommend choosing a low pass filter corner frequency of ~ 2 orders of magnitude less than the desired operating frequency.  In this case, the corner frequency would be at 0.5 Hz.

The formula for the corner frequency of a first order low pass RC filter is: f_c = 1/(2*pi*R*C)

So, we can rearrange that formula:

R*C = 1/(2*pi*f_c) = 1/pi = 0.3183

To get close to that value, we need pretty big R and C values.  I'd start with a 1uF capacitor, since those are extremely common and easy to get.

R*(1e-6) = 0.3183

R = 0.3183e6 = 318.3 kohm

Selecting the nearest 1% resistor value gives 324 kohm.

So, the final first order low pass filter design would be

R3 = 324 kohm

C2 = 1 uF

Going back to our 2 criteria, note that this will settle in ~5RC = 1.62 s, which is a relatively long time for a circuit.

The output voltage ripple will be ~76 mV at 5V input, which I would consider to be very good.

Hi

Thanks a lot for the information it's much clearer now.
Unfortunately I used values according to the calculation, I get the desired frequency 500khz but I do not have a lock on the 50hz.
I tried testing with a potentiometer without any noticeable improvement.
Is there a solution to the problem?
What is the stability of the VCO?
I attached pictures of the signals (the file name is according to the pin number).

(pin-1, pin-13 , pin-4 )

Best Regard

Itamar  attached pictures of the signals (the file name is according to the pin number)

Hi Itamar,

It looks like the output of the VCO is exactly what we'd expect (500 kHz). Can you show the two inputs to the phase comparator? 

Also, that last image of the VCO output looks quite odd - like there's an issue with the ground of the device or probe. It should be reaching ~0V on every cycle. I would recommend double checking all ground connections and make sure you have a decoupling capacitor. If that all checks out, you might try another unit to see if this one has been damaged.

Hi Emrys

Indeed, I get 500KHZ but I would expect the PV2-OUT will be in high impedence and the Ld will be in constant high level.
I used decoupling capacitors in all the components.
The problem with VCO output was because the scope.
I attached the two inputs you requested.
Of course the Comp_in is at 50HZ.
Thank you very much for your help.

Best Regard

Itamar

When you take scope shots of logic signals, they generally need to be on the same display (2 channel or more) -- unless you're just looking at signal integrity. Otherwise you can't see the related timing and the data isn't very helpful. I can understand that we don't all have 8 channel scopes, but it's extremely hard to troubleshoot a logic circuit if you can't see the inputs/outputs on the same display with the same time scale.

From my experience, I've found that the PC2 comparator is the most prone to errors and can get into a state where it's 'stuck' changing between hi-z and either high or low depending on the timing of the signal inputs.

It looks like your circuit is doing what it's supposed to though - the signal is ~50 Hz. If you can view the COMP_IN, SIG_IN and PC2_OUT signals on the same scope shot, you can follow the logic diagram in the datasheet and figure out what's going on.

I apologize for my late response I was on a weekend break.
I attached the two inputs: COMP_IN, SIG_IN.
I did not attached the PC2_OUT because the measurement affects the signal. Logically the circuit works fine .I expected the difference between the phases to be zero and this is not the case, even though I get the desired frequency

What happens when you connect the PC2 output to the scope? I wouldn't expect much change for that node with a high-impedance scope probe.

I attached two pictures of the measurement of the signals.
The first with Prob X10 and the second with Fet Prob (Referring to the photo in the previous email)
You can see that there is an effect of the probe, also you can see that the circuit logic works correctly

I understood . Thanks for all the help