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.

LM231: LM231 F-V schematic review~

Part Number: LM231
Other Parts Discussed in Thread: LM358, OPA313

Hi Sir

We need to test the frequency of the crystal oscillator, One design is count by MCU, Another design is LM231。

About LM231,Could you help to check the whole schematic? (some questions about 2-pole Filter, OPA U112B, Could we use LM358 for this Filter) 

Because Figure 19 in the datasheet shows a negative voltage output, but we need to meet the ADC input spec:0<Input<5V, Could you help to check the schematic carefully

for this portion?.

Another question:
The test clock is 32.768Khz(From crystal oscillator)


Which design is better (Design1 or Design 2):


Design1 Without Frequency division(U119) :
>> Fin(Pin6)=4096Hz Rf=100K(R416)
Vout=4096Hz*2.09V*(100Kohm/14Kohm)*6.8K*0.01uF=4.158V


Design 2 (Frequency Factor=2^3=8):
>>Fin(Pin6)=32.768KHz Rf=12.4K(R416)
Vout=32768Hz*2.09V*(12.4Kohm/14Kohm)*6.8K*0.01uF=4.125V

LM231 footprint is PDIP, It is unpopular, and has few designs using it.

Why does Ti Still keep it, Could we use LM231/331 in a new design? (I am worried about if it will be EOL)

And does Ti have Plan to re-design a new footprint (For SMT)? I only can find another design about Freq-Voltage on the ADI website is AD650.

Foxconn  Qin

  • Why would you want to test the frequency of a ready made crystal oscillator?? These oscillators are unimpeachable and I see absolutely no need to test anything.

    Kai

  • Hi kai

    Actually, It is a test item to verify the frequency of crystal oscillators in DUT.

    What we design is a test fixture in the product line, in order to read the oscillator frequency from PCBA,

    Our customer asks us to check the frequency of the crystal oscillator.

    The test spec is [30Khz,34Khz], we need to find the solution, I know the easy way is MCU count the pulse.

    I would like to know if LM231 can also be used in the design for this function

    thanks

    Qin

  • Hi Qin,

    "We need to test the frequency of the crystal oscillator, One design is count by MCU, Another design is LM231."

    You mention the oscillation frequency test spec is 30 kHz, 34 kHz. Although specified having a full scale frequency the precision of the LM231 in V-to-F, or F-to-V falls off rapidly above 10 kHz. See the Typical Characteristics graphs in the datasheet Section 7.1, Figures 1, 2 , 7, 8 and 12. It is evident that the highest accuracy is obtained at 10 kHz and less, and there are no assurances of the what level of performances can be achieved at 30 to 34 kHz as an F-to-V.

    It seems evident to me that a digital counter function performed by an MCU could provide much higher accuracy frequency conversion accuracy than applying the LM231 at 30 kHz. If you think you want to pursue the LM231 solution I suggest placing a decade divider after the 32 kHz oscillator to bring the frequency down to 3.2 kHz which the LM231 can more accurately handle. The low-pass filter from the datasheet Figure 19, Precision Frequency-to-Voltage Converter schematic could be replaced with a different low-pass topology. It appears to be a form of a low-frequency integrator. I expect a Sallen-Key (SK) active low-pass could be applied there instead. It does not invert the input signal.

    No doubt the PDIP package's days are numbered with few exceptions. The LM231 in the DIP package has been available for decades and was a popular package type when the LM231 was introduced decades ago. I do not know of any plans to offer the LM231 in a different package type and can't disclose such information on the open e2e forum.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • Hi Thomas

    Great! 

    U Said: I suggest placing a decade divider after the 32 kHz oscillator to bring the frequency down to 3.2 kHz 

    For CD4040, The divisor is odd. We can bring the frequency down to 2.048K(32.768K/16),Does it work?

    So what is the Cut off Frequency of SK low-pass? 

    Could we supply a +5V to OPA  without negative voltage?

    What is your suggestion for a costdown-OPA?

    Thanks, Qin

  • Hello Qin,

    Sure, I think you can divide the 32 kHz crystal frequency down by 16, to 2.048 kHz as long as your method of collecting the frequency information accounts for the division.

    The filter topology used in the LM231 Figure 19 circuit is not completely familiar to me, but looks most like a multiple feedback (MFB) low-pass topology where the output impedance at pin 1 in conjunction with C3 (100 nF) are parts of the filter. The purpose of the filter is to remove ripple from the voltage output  so that a more constant DC level is had which corresponds to a particular frequency.

    When I do some examination of the filter I conclude the cutoff frequency is around 79 Hz. It is difficult to know for sure because the impedance at pin 1 isn't provided and that would be needed to determine the filter's characteristics. It does appear that the idea is to use a low cutoff frequency to remove as much output ripple as possible. However, the filter cutoff frequency will affect the response time so a compromise is needed between ripple and response time.

    I located an LM231/331 applications report report that I was looking for. The report is about applying the LM231/331 as a frequency to voltage converter. It has a very complete section about the filter and offers different options. You can find it here:

    https://www.ti.com/lit/an/snoa734b/snoa734b.pdf

    Regarding the single supply, +5 V op amp, none will swing negative or completely all the way to 0 V at the output with a single supply. That is because there is some finite output resistance associated with the output transistors. There will be some minimum voltage across the output transistor sinking or sourcing current and that prevents the output from swinging all the way to the supply rails such as +5 V and 0 V. The best you can do is to use an op amp whos output swings as close to the supply rails as possible. The general purpose OPA313 you have identified swings close to the rails and has low cost.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • Hi Thomas

    Thank you for your help and patience