This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LM2917-N: LM2907/LM2917 variable reluctance sensor buffer to TTL

Part Number: LM2917-N

Hello.

We have a variable reluctance sensor that we would like to buffer to a TTL or other voltage level. The LM2907/2917 seems like a good candidate. On page 20 of the datasheet, figure 28 a circuit is described that outputs pulses at twice the frequency of the input frequency.

I assume the unit of the duty cycle is time and as our Vcc is 24V the duty cycle will equal C1 according to the formula. We will sample the pulse train with a minimum frequency of 9 MHZ and a maximum of 30 MHz. 

Assuming we sample the pulse train with 9 MHz we have to have a pulse width that is wide enough that the sampling device will detect it.

For example lets say the waveform has a frequency of 4.5 MHz and the duty cycle is 50%. I want my two-shot output duty cycle to be of the same time duration. Assuming Vcc = 24 that leaves me with a C1 of 11 uF. Is there any danger in using this large a capacitor?

In the example the capacitor is 500 pF which is way less and I reckon this would be used with an MCU to calculate the frequency using some interrupt service routine triggered by the pulses. We want to sample the waveform and do calculations afterwards.

The DAQ used is the NI 9401. The reason we want to use this and sample the waveform is to integrate it with other NI modules with which we sample other signal types.

Any suggestions/ideas?

Best regards,

David Kantzon

  • Hello David,

    I am just becoming familiar with the LM2907/LM2917 as it recently was moved to Precision Amplifiers from another group. That said, when I read the datasheet and the AN-162 Applications Note, I get the impression that it intended for low frequency applications from a few hertz to something on the order of 10 kHz. I don't find anything in these resources indicating that it is usable over a frequency range of 9 MHz to 30 MHz. They seem to be devoid of any timing/frequency information.

    I am going to try and find someone in TI who has experience with the device and find out what its usable bandwidth is.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • Hello Thomas,

    I know that the LM2907/2917 is intended for low frequency applications and I doubt we will have input frequencies over 1 kHz. Using the configuration on page 20 for two shot frequency doubler with a zener diode level setting that would leave me a pulse train of max 2 kHz with some duty cycle. With a supply voltage of 24 V the duty cycle will equal the value of the capacitor if I am not mistaken.

    I might have asked my question wrong but we will be sampling the waveform coming out of the 29x7 with a minimum sample rate of 9 MHz and a maximum of 30 MHz. The duty cycle cannot be too small for the ADC to detect it, neither can the capacitor be too large so as to have an acceptable rise time as the current source only outputs 180uA.
  • Hi David,

    The circuit on Pg. 20, Fig. 28, in the LM2907/LM2917 datasheet has a typo in the Pulse Width equation. It lists PW = (Vcc/2) (C1/12) - I couldn't understand where the 12 was coming from. However, when I viewed the two-shot circuit in AN-162 it showed the correct equation as PW = (Vcc/2) (C1/I2). It is listed correctly as shown in the diagram below from AN-162. You may have already have detected the error before I did.

    What pulse width are you attempting to attain? When I use C1 = 11 uF (the value you mentioned), Vcc = 24 V and I2 = 180 uA, then PW = 0.73 sec. Is that what you intend?

    Regards, Thomas

    Precision Amplifiers Applications Engineering