ADS5424-SP: Input Configuration

Hi Markus,

Is the input circuit requiring DC coupling or AC coupling?

Regardless, I would suggest looking at the LMH5401 differential amplifier.

This can be configured for a single-ended input and differential output to connect to the ADC.

However, I would be wise to put an anti-aliasing filter in-between. What is the BW requirement for your application?

Regards,

Rob

Hi Bob, thank you for your fast response.

For my application I need DC coupling.

Is it possible to realize an input impedance of at least 100kOhm with this amplifier?

I am not sure how to calculate the BW requirement. Actually the input signal changes its value every 100ns. So it looks like a staircase signal. This results in a sample rate of 10MSPS. I think the input filter should settle within 90ns to the new value, then the sample can be taken.

As the input signal span of (4.5V - 1.1V = 3.4V) is higher than the signal span of the ADC, which is 2Vpp, I need a gain that is lower than 1. Can this be realized with the LMH5401?

Do I need a 6.5-GHz amplifier for this application?

Best Regards, Markus

Hi Markus,

Is your signal is changing every 100nSec, then you might want to open up the sample rate, make it faster in order to acquire as the signal changes from one state to another. Therefore, the step or rise in change would be how to set the BW. At least, that is how I would look at it...

Thanks for the additional details. The LMH3401 would also be another candidate to look at for an amplifier, unfortunately, you need an amp that can do the single to diff conversion. And these both have pretty high BWs.

I will speak to the amp group and see if they have any other ideas and get back with you.

Regards,

Rob

Hi Rob,

for a better understanding here the representation of the signal:

The masterclock controls the changes of the analog signal. The sample clock of the ADC is synchronized to the masterclock. So we don't need to sample the edges of the signal.

For the calculation of the Bandwidth I tried the following:

Prior to taking the sample, the signal should settle to within 1 LSB. Using a 14-bit ADC this correspond to 0,006%. To give some margin the settling time is set to 90ns. Calculating tau for a RC low pass, the result is 9,26ns. This corresponds to a corner frequency of 17MHz.
Is this the required BW?

Regards, Markus

Hi Rob.

I will have a closer look to the LMH5401.
Following the detailed description for the THS4551 setting the gain to less than 1 is possible by chossing RG higher than RF. Is this also applicable for the LMH5401? E.g. RG = 2kOhm, RF = 1kOhm for a gain of 0.5?

The input impedance of the given circuit is fairly low. How can the input impedance be increased to more than 100kOhm? Will I need an additional OpAmp? What would be a suitable type?

=> Should I ask this questions in the amplifier forum?

Regarding the ADS5424-SP:

The datasheet states 30MSPS as the minimum sample frequency. Actually I need 10MSPS or maximum 12MSPS. Is it possible to use a clock frequency of 12MHz?

Regards, Markus

Hi Markus,

For your current amp questions, yes, the THS4551 description can be used with the LMH5401 by choosing a higher RG.

I also believe you will need an additional OpAmp to get the impedance up as well. The amplifier team can best help here. 

I will ask the amplifier team to respond this post. If needed they can move it over to the amplifier forum.

Regarding the ADS5424-SP, I will find a suitable replacement that can sample down to 12MSPS for you. Any other requirements for the converter that I need to know about?

Thx,

Rob

Hi Markus,

Please have a look at the ADC14155QML, this can sample down at 12MSPS.

5MSPS min, 155MSPS max.

Regards,

Rob

Hi Rob,

many thanks for your suggestion. We want to transmit the digital output via LVDS to an FPGA. Our selection for the LVDS-Serializer is a 3.3V type, so this works very well with the ADS5424.

Does TI offer an LVDS-Serializer for space apllications? I could not find one.

An important requirement demands an ENOB of 12 Bit. I know that the ADS5424 does not comply with this requirement, but it is quite close. We are thinking about operating the AS5424 with 40MSPS. Figure 9 in the datasheet shows the variation of the SNR with input frequency and sample rate. The result at 40MSPS is better than for 105MSPS. Correct?

Figure 9 shows the typical values. How can we assume the minimum value?

I do have one more question: How can we estimate the power consumption of the ADS5424 using a sampe rate of 40MSPS? Does the Analog supply current (IAVDD) depend on the sample rate?

Regards, Markus

Hi Jacob,

many thanks for the offer to simulate the circuit. When we are ready, I would like to contact you. How can I do that? In the e2E forum? In which section?

Is the LMH6559 suitable for space-applications?

Yes I need DC coupling. Above I have added an image of the analog signal and these are the parameters:

• The input signal is single ended with a signal range of 1.1V to 4.5V. This is the range of the signal that has to be converted. To give some margin a signal range of 0.9V to 4.7V shall be mapped to to full scale range of the ADC.
•  Voltages in the range 0V to 5V shall not cause malfunctions of the circuitry.

Preferably I am looking for a solution that works with a 5V single supply.

Regards, Markus

Hi Rob,

did you see me reply on Monday?

To make our decision on component selection, I especially need the information on power consumption:
How can we estimate the power consumption of the ADS5424 using a sampe rate of 40MSPS? Does the Analog supply current (IAVDD) depend on the sample rate?

Hi Rob,

many thanks for your suggestion. We want to transmit the digital output via LVDS to an FPGA. Our selection for the LVDS-Serializer is a 3.3V type, so this works very well with the ADS5424.

Does TI offer an LVDS-Serializer for space apllications? I could not find one.

An important requirement demands an ENOB of 12 Bit. I know that the ADS5424 does not comply with this requirement, but it is quite close. We are thinking about operating the AS5424 with 40MSPS. Figure 9 in the datasheet shows the variation of the SNR with input frequency and sample rate. The result at 40MSPS is better than for 105MSPS. Correct?

Figure 9 shows the typical values. How can we assume the minimum value?

I do have one more question: How can we estimate the power consumption of the ADS5424 using a sampe rate of 40MSPS? Does the Analog supply current (IAVDD) depend on the sample rate?

Regards, Markus

Hi Markus,

Sorry about that, missed your power question. Typically the DVDD supply will remain flat per the LVDS is a current based output and has a typical current draw at high or lower sampling rates.

The AVDD will scale with sample rate, but the closest I could find on this, was is in the commercial part datasheet. See attached.

I would have to do some investigation on a space grade serializer. I don't know of a product that we offer off hand. I will get back to you.

Figure 9 shows the ENOB/SNR of the part only at 105MSPS sampling and at various frequencies. If you want I can measure this for you in the lab under your conditions and get that data for you. However, i would have to track down the HW to set this up in the lab first. I may have to order a board from stock if you want the data from the space grade version.

Regards,

Rob

Hi Rob,

thanks for your response. For our project we need to use 28 ADCs and we do have a power budget of about 15W for them. So unfortunately it wont be possible to use the ADS5424.

Meanwhile I found the LM98640, LVDS is included an the power dissipation is fairly low. Unfortunately SNR/ENOB is not that good.

Is it possible to apply a differential analog input signal to the LM98640?

Shall I open a new post with my questions on the LM98640?

Best Regards, Markus