ADC boards that can be used with Real-Time Clocks of frequency 32.758 Khz

Hi, I am currently in possession of a real time clock of frequency 32.768 Khz and would like to integrate this with an ADC that would be inputting voltage signals. Most RTC modules that i come across are best suited for Raspbeery pi and arduino modules. May I know if other ADC boards can be used with RTC’s. I am currently in possession of a THDB ADA board from Altera. Would an RTC work with the same?

Best,

Sachin

Hello,

Thank you for contacting DigiKey’s Tech Forum. We are looking into this and will get back as soon as we can.

Thank you
Ryan

1 Like

This could mean different things, but I’m presuming that the intent is to try to digitize an analog signal in a time-stamped manner.

Arduinos, RasPis, and similar development platforms are popular (largely) by reason of their informally-standardized interconnect features, which make it possible for 3rd parties to build compatible functional modules or evaluation boards.

Development tools are not limited to these form factors however, and manufacturers of products such as FPGAs will often offer dev kits and accessory modules with forms very different from the popular ones mentioned.

While development tools generally do not plug-and-play across product ecosystems, this does not mean that their featured products cannot be used in conjunction; one simply needs to build the needed interconnections.

Programmable logic devices are about the most flexible processing devices available, and the sort your THDB-ADA board is targeted at could conceivably be interfaced to the majority of the ADC and RTC devices on the market. You will almost certainly need to build that interface yourself for any given combination of devices however.

1 Like

Hello sm9666,

Could you provide additional information about your application?

The THDB ADA board is true marvel with:

  1. Dual AD channels with 14-bit resolution and data rate up to 65 MSPS
  2. Dual DA channels with 14-bit resolution and data rate up to 125 MSPS

With a proper FPGA, and some ingenuity with RF circuit design, this hardware would be at home in a software designed radio.

For reference, the THDB-ADA board was designed to integrate with an advanced FPGA demo board such as the Altera DE2-15.

To my knowledge, this FPGA demo board does not feature a real time clock. However, there are a variety of real time clock modules that could be integrated into the design. One clock module suitable for FPGA integration is the Adafruit DS3231. This is a temperature compensated real time clock featuring an integral battery backup and a i^2C interface. You should have no trouble finding a I^2C Verilog or VHDL module for the FPGA. Developing an interface to program the module from within the FPGA will be challenging. Depending on your needs, you may want to integrate the real-time clock functionality into the associated microcontroller or PC device.

For a more detailed answer, we would need to know more about your application.

We look forward to hearing from you.

Best Wishes,

APDahlen

1 Like

Dear Rick and AP Dahlen,

Thank you for your responses.

Essentially I have a signal being read in from a photodiode that is connected to a Device under test, I’m using this photodiode to record the signal power however I’m more interested in the optical return signal that I would pick up from the photodiode itself which would have really incredibly low powers anywhere from 1mW to the Nano watts. This signal is being transmitted from a device that is placed on a vibration table with peak frequencies of upto 3Khz. I could use an Optical time domain reflect meter device to record the same , however the traditionally measured time averaged ORL would his important info of the peak levels and profiles of the actual temporal ORL. Hence why I would like to take the RTC route for a time resolved measurement. This signal from the Photodiode would the have to be integrated with a RTC on an adc board, where it can be fed onto a computer or an oscilloscope for analysis. Could you kindly help identify an ADC and a RTC that suit these requirements?

Best,

Sachin

Hi sm9666,

So, am I understanding that you are trying to get a time-stamped reading of an upto 3KHz signal? Do you need to know how the signal changes over time (in seconds, minutes, hours, etc) or are you looking to just get an accurate time base for taking ADC readings? If the latter, any FPGA board will already have an oscillator on board to provide an accurate time base (in MHz and with a little HDL code, almost any frequency you need below that).

I think you have some typos above which are making your explanation a little unclear.

Not sure what you are trying to say here. Also, for clarity, what is your definition of “ORL”?

Hello Sachin,

Let me start by explaining my understanding of you project:

  1. The DUT is on a vibration table.
  2. The vibration table has a mechanical frequency up to 3 kHz.
  3. The DUT is physically connected to an LED either optical or laser)
  4. The stationary side has a photodiode that detects the power of the DUT’s diode.
  5. A sensitive ADC is required to detect the variations in light intensity.
  6. A time stamp is required for the recording mechanism.

Please correct my understanding of your system.

If this interpretation is correct there are a few concerns:

  • Allowing for the possibility of mechanical resonance, the DUT may vibrate at a frequency considerably higher that 3 kHz. For example, the system may be nonlinear with harmonics above 3 kHz. By the Nyquist criteria, the ADC must sample at least two times higher than the highest harmonic. If we sample too low, aliasing will occur.
  • The time constant associated with the optical detector must be low relative to the frequency of vibration. If the system is slow to respond the measurement device will average the results.
  • With regards to the real time clock: while they are generally accurate in the long term, they are slow. Here “slow” is a relative term meaning slow relative to the desired sampling rate, even at a 6 kHz rate. Timing jitter will be difficult to resolve. Synchronization with the ADC will be difficult. Also, the resolution of a real time clock is generally in low millisecond range.

Please correct my understanding of the system and we may be better able to provide an answer. Also, please let us know your familiarity with microcontroller and FPGA programming.

Best Wishes,

APDahlen

P.S. I envision two options. One is a microcontroller system featuring interrupt-based timing. It may be useful to include ahigh performance oscillator for precision timing. The other option is FPGA based hardware.

Dear AP Dahlen,

Thank you for your response.

My knowledge with respect to fpga boards,specifically the DE2 board is restricted in only knowing to program which pins of the board would be used as input pins for say an ossciloscope and the pins that would be used as a trigger.

On a note that the RTC integration would drive the adc and incase the ADC’s SNR is lacking, an erbium dope amplifier would be used to raise the optical power up by 25dB.

I couldnt find a time constant associated with the photodiode but could find a response time of less than 1microsecond or 1us.

best,

Sachin

The frequency of the RTC at 32.768Khz would potentially be 10 times the highest frequency of the vibration table, essentially 5 times higher than the nyquist frequency.

Best,

Sachin

Dear David,

Essentially stick with the former wherein I would want see how the signal changes over time.

ORL is optical return loss. my apologies for the abbreviation,

I would like to know from the signal the time points at which the Optical return loss is at peak and compare that with respect to the frequency at which the vibration table was at at that particular point of time.

Sachin

Dear Sachin,

This is representative of the data that I understand you wish to receive.

I understand the sample time (T_S) is an important consideration. Likewise, I assume it’s important to minimize jitter which is to say, the sampling drumbeat should be stable.

Forgive me, I don’t yet see the importance of the absolute time (T_0) such as derived from a real time clock.

With that said, the DE2 FPGA, THDB board, and DS3231 would be a wonderful data acquisition tool. However, it would take a non-trivial FPGA experience to assemble such a system. Based on your experience and presumed critical timeline, you may be better off using a microcontroller.

You could program the micro to provide a reasonable drumbeat using an Interrupt Service Routine (ISR) As a starting point, the 4809 microcontroller featured on the Arduino Nano Every has a 10-bit ADC capable of 150 thousand sample per second. You could fill a data table and then print to the PC. Connecting the DS3231 is trivial as Adafruit should have a pre-built library.

You may find the video link below helpful as it showcases the capabilities of a micro, albeit one with more power and a steeper learning curve than the Arduino.

Best Wishes,

APDahlen

Hi sm9666,

I think you might be misunderstanding the function of a Real Time Clock. Their purpose is to keep track of the time of day, week, and sometimes year. They employ a 32.768KHz crystal to keep accurate time, but they are not typically used as a clock reference for timing data acquisition readings, though I suppose one could use it for that. As APDahlen covers, their typical jitter specs make them fairly unsuitable for such applications.

You will get better results using the oscillator that comes with FPGA boards, or if you choose to use a microcontroller, one with a MHz-range crystal or oscillator (not ceramic resonator) as a time base.

I take it you are from a university. If so, it’s possible that one or more of your labs may also have a data acquisition device (DAQ) on hand which might also be able to perform this function. We carry some, too, but they are not inexpensive. Here’s a sampling of some we carry, as an example. Keep in mind that, depending on the magnitude of your signal, you may have to amplify it to get to a more usable range. And, very low-level signals are not trivial to amplify without introducing significant noise into the signal.

1 Like