Simulating F2V-Converter (based on LM331) in LTSpice

Hello everyone,

Im trying to simulate the Frequency to Voltage Converter which is presented in the Datasheet of the LM331 from National Semiconductor. Here is a Link to the Datasheet. The circuit Im trying to simulate is on page 10 (Figure 7). Im trying to simulate this circuit in LTSpice.
I´ve attached the Circuit and the MakroModel of the LM331. However my Output does not show any frequency dependencie. Why? I just cannot figure out my mistake. Maybe one of you guys could have a quick look. I allready ajusted A1,A2,and A3 within the LM331 Model and set Vhigh to Vs. Which in my case is 5V. But that doesn´t seem to be the failure. I also set "Skip Initial operating point solution" but it did not help either.
Any Idea is highly apreciated, thanks in advance.

Best regards from Berlin,
Jan

C2 appears to be too small. Suggest you try 470nF instead or 470pF.

hi Jan,
You could consider using either of these two existing device models for the F2V.

Hi,

thank you Guys for your fast replies. Actually, Eric, the last 2h I just tried to use your suggested LM2907-8 design. But I guess my problem is a more fundamental one. I´d like to measure very low frequencies, meaning from 0Hz to max 40Hz (with high precision). And these F2V converters are not working linear in this area. Additionally I need a very clean output signal, no ripples etc. Any Ideas?

@crutschow: I changed c2 but it had no effect. Even with higher Frequencies the previously posted LTSpice circuit does not seem to work - even if it should.

Best regards,
Jan

Hi,

guess I have to use **broken link removed** piece of circuitry. Im trying to simulate this one as well. I got all models I need, just the 4518 is missing. Do you guys have any Idea on how to implement the 4518 in LTSpice? I allready searched the Yahoo!Group without any success.
My plan is to multiply my signal by 100. This should give sufficient results to use the F2V converter shown above.
By the way, the model I posted before does still not work and I have no clue why. So any Ideas are still much appretiated.

Best regards,
Jan

I haven't looked at the details of your circuit, but I have a suggestion: Instead of diode ORing your frequency sources, put them in series. Then they won't have a nonlinear output impedance.

The problem with measuring low frequencies with an F to V converter is that the filter time constants have to be very large,which makes the response to changes very slow.
Try counting the period instead, then taking the reciprocal. This requires a microcontroller.

By the way, the model I posted before does still not work and I have no clue why. So any Ideas are still much appretiated

Click to expand...

Shouldn't the right-hand end of R8 connect directly to the inverting input of the opamp?

Edit: BTW, trying to measure a frequency of 0 Hz will take for ever

Another problem with the sim is your use of an LT1001. Give it a negative supply below -6V (to avoid premature clipping) or replace it with a true ground-sensing single-supply amp (I couldn't find a suitable LTSpice model which worked as required here with both it's negative supply pin and the non-inverting input grounded). A lot of the so-called rail-to-rail in/out and single-supply opamps in LTSpice won't do what you want unless they have a negative supply voltage.