LM331 Frequency to Voltage, 100kHz full-scale?

[gaubar]

I am having trouble converting frequency(0-100kHz) into a voltage signal using a LM331 N.

There is a basic schematic in the datasheet which is for 0-10kHz. I have successfully made this circuit, and it works fine for the range 0-10kHz. When trying to apply a frequency above this range (for example: 50kHz), the output is zero.

Has anyone completed a circuit using LM331 N for measuring frequency around 40kHz - 60kHz? Is it possible to tweak the 0-10kHz circuit to handle higher frequencies?

 

[schmitt trigger]

I find noteworthy that the datasheet provides examples of the device working up to 100 Khz in the V/F mode, but only shows examples up to 10 Khz in the F/V mode.

The probable cause is the differentiator network at the circuit's input. The RC constant must be made for the trigger pulse to be wide enough for reliably triggering. But I suspect that the period of a higher input frequency may approach the pulse width plus any internal settling.
If that is exceeded, the internal one-shot may never reset and the output will be zero.

 

[gaubar]

So probably, measuring 100kHz cannot be done with the LM331. I think they would have put an example circuit in the datasheet if it was possible.

Is there another way this measurement could be done? Are there any cheap/easily available devices that could measure the frequency and output a signal from say 0-5V/or 0-10V?
This might also be done using an ATMEGA/ATTINY microcontroller i suppose, any experience with that/link to some tutorial?

 

[schmitt trigger]

With a PLL, a lowly CD4046 will work all the way to 1.2 Mhz when powered by 10V.

Connect it as shown in figure 1 in the datasheet (no divider between pin 4 and 3).
The translated voltage you require is the VCO control voltage which is fed to pin 9. Since this is a very high impedance node, use a CMOS-input opamp to buffer it (unity gain). In the past I used a CA3140, but there are many modern amps that will also work.