wrong frequency of LC based generator with LM311

I have a problem with the frequency of LC meter based on comparator (a very popular solution, similar to this one: http://www.vk6fh.com/vk6fh/lc_meter_vk6fh.htm). I have recreated generator part in Proteus (see figure, where I have increased the values of L and C by 1000 to speed up the simulation). According to eq. $$\f=\frac{1}{2\pi\sqrt{LC}}\$$, frequency should be 503Hz, but what I have is 437Hz. I have the same relative discrepancy in the real circuit (of course, frequencies are then in hundreds of kHz). Although there are many similar solutions in Internet, no one observed this problem, and I did not find the answer. Please, help.

• I think that it is a simulation "error" nothing more. – G36 Apr 28 at 16:26
• I'm sure the sim is telling you the truth and that you have pushed a flaky design too far in trying to make it work with C1 at 1 uF. After all, the general idea is that C2 should be miles bigger than C1 and, the resonance point should be miles higher than the RC cut-off of R1 and C3. – Andy aka Apr 28 at 16:51
• I have changed back L1 to 100uH and C1 to 1nF (1sec of the simulation lasts 20 min now). These are the same values as in the real circuit. I've got f=460kHz, but from the equation it should be 503kHz. So, the problem must be somewhere else. – piotr Apr 29 at 6:40