TINA/Spice/LM231: lm231

Hello Ian,

Thank you very much.  I have a small seismograph setup in my basement.  I am very interested in using telemetry to provide communications between the seismometer in a more remote place and my computer.  The seismometer is an Iris design called the AS-1.  The seismometer consists of a vertical sensing pendulum with a horseshoe magnet on the end of the pendulum and a coil that sits inside the magnet.  Any motion on the part of the pendulum or the coil will generate a very small signal.  I would like to place the seismometer in a more remote location.  The output of the coil would go to an amplifier consisting of a low-pass filter and high gain amplifier.  

I am looking for a fairly simple and straight forward circuit both for the transmitter and the receiver.  The output of the amplifier is + - 10 Volts.  The frequency is almost at a DC level with fluctuations in the voltage when there is any kind of vibration from the seismometer.  I thought about using a CD4046 for the transmitter (modulator) and another CD4046 for the receiver (demodulator).  I am not sure what audio frequency to use.  I was thinking of staying between 300 Hz to 3000 Hz.  I know that there is a lot of information on the Internet about the CD4046.  I had some trouble with understanding the graphs and the equations for the minimum and maximum frequency did not seem to work out with the values of the timing resistor/capacitor combination.  While searching for information about simple telemetry circuits, I cam across information on the LM331.  

The LM331 seemed to be a better choice because of better linearity and the design is more straight forward.  I was hoping to use Multisim (student version 13) to experiment with the values of the timing resistor and capacitor.  I found a circuit in TI's data sheet on the LM331/ LM231.   I am considering two circuits in the Texas Instruments "LMx31x Precision Voltage-to-Frequency Converters".  The modulator circuit is shown as figure 16 on page 14.  The demodulator is shown as figure 18 on page 15 or another circuit, figure  19 on page 16.  

I am admitting that I am more of a hobbyist than an engineer.  I have a strong desire to learn about designing communication circuits.  I had some trouble understanding some of the graphs about linearity.  The equation for determining the values for Rt and Ct seem to be easy enough.  I am ready to order the parts needed for the two circuits, figure 16 and figure 18.  I would like to have a better understanding of the circuits.  

I would like to hear any advice that you have.  Are the LM331 or LM231 good choices for analog telemetry.  What would be a good frequency to pick for a quiet signal (DC level) and for the quick perturbations that an earthquake would cause.  Are there better choices than the LM331 or LM231.  I would like to keep the frequency range in the voice band (300 Hz to 3000Hz).  I am not sure what kind of transmitter or receiver to  use.  I thought of using a couple of old RF modules from Abacom for the transmitter and receiver.  I also thought of using a cable pair with an audio isolation amplifier to prevent ground loops due to each end having their own power supply. 

I saw adds for Tina in Everyday Practical Electronics (EPE).  EPE has both a hobbyist version and a student version.  How does Tina compare with Multisim Student Version 13.  From what you said, I think that Tina would be very useful in my effort to practice and understand communication circuits.  

Thank you again,

Richard Phillips

Hi Richard,

It certainly appears that the LM231/LM331 could be employed in a simple telemetry system, but we have no experience in setting it up for that particular application. Keep in mind that the LM231/LM331 isn't a wide bandwidth device and that a top frequency of 10 kHz or a few tens-of-kilohertz would be the upper limit. It would seem that would be sufficient for the very low frequencies associated with sesmic activities.

An alternate solution would be to employ an A-to-D converter on the sensor side, and a D-to-A on the receiving side. The transmission from the sensor side to the receiver side would be digital and much less susceptible to noise and interference. I expect that this approach would result in much higher measurment accuracy and speed than a V-to-F/ F-to-V soultion.

I searched the TI resources to see if we could provide anything else on the subject of telemetry. Here is one resource that I found. It may be of use to you:

www.ti.com/.../slyt571.pdf

Regarding Multisim Spice and TINA Spice it comes down to understanding what is needed and what they provide. TINA-TI is free and is very easy to learn and start using in tens-of-minutes. However, there are limits in terms of the device models and analysis capabilities it provides as compared to the for-purchase TINA versions. I would say for many applications the TINA-TI version will do everything that is needed. We use it for complex simulations daily. The for-purchase versions provide additonal analysis capabilities, additonal models, PC board layout, etc.

We don't have Multisim Spice here so I haven't used it, but from the NI website it appears to be a very nice, very capable simulator. The student version appears to be quite powerful and the cost is low.

Regards, Thomas
Precision Amplifiers Applications Engineering

Hello Richard,

The LM331 was commonly used for telemetry. The key is to use good, stable capacitors and resistors and not push the linearity limits.

Also look at the LM565 and LM566. These are VCO and PLL blocks that were commonly used to transmit DC to audio-band information over various links. They are obsolete now, but they are widely available on the surplus market.

What you are trying to do is similar to the "IRIG Channel 13 Demodulator" in the LM566 datasheet. Before digital, telemetry links were analog - so they could be transmitted via radio, phone line or recorded on audio tape. Each "channel" was a different carrier frequency. But bandwidth is very limited.

Years ago, the USGS around here had remote seismometers up in the hills with low power UHF FM radio links (~420MHz) back to Menlo Park. They output a steady tone around 1kHz that was modulated by any movement. I could pick them up on a scanner and a good antenna. Of course, these have been replaced by GPS & LTE based telemetry now, but the concept is the same as to what you are trying to do.

As Thomas eluded to, there is a limit to the modulation bandwidth. The receiver will have limitations on how far it can track the frequency deviation. The 1kHz tone the USGS used only had less than a few hundred Hz of usable signal bandwidth. Most likely a pendulum is not going to move too fast, so a few hundred Hz of BW may be enough. Looking at the AS-1 rig, it is not going to move that fast unless you kick it...so a low bandwidth may be okay for you.

Dynamic range (vertical resolution) is probably more critical for you. Since you have limited modulation range, to get more "resolution" you would need to crank up the modulation (Hz per volt). That also cranks up the drift and noise and decreases the maximum amplitude.

I tend to agree with Thomas - I think you would be better-off digitizing with a good 16-24bit or greater A/D at the seismometer and sending it digitally via WiFi. Sounds like a good Arduino (or LaunchPad) project. A good starting point would be the wireless speaker Arduino projects - since you are just digitizing low frequency audio. Google around, I bet someone has already done it...