ADC recommendation, ADS8509 multiple channels alternative

Hi Carlos!

Welcome to the forum!  Thank you for considering the ADS1259 or the ADS8509!  We'll go over your options here and get back to you as soon as possible.

Hi Carlos,

We don’t have a one-chip solution for this project, but we can certainly help you find the right combination of ICs for it.

We are currently debating between programmable signal attenuation and/or amplification to fit the ranges. In terms of converters, we are currently considering multi-channel ADCs with internal MUX in order to have the MUX always synchronized with the converter.

Since the ADS1259 (24bits, delta-sigma, 14KSPS) and ADS8509 (16bits, SAR, 250KSPS) are very different converters, we would like to know a few more details about the project.

Must the resolution be at least 16 bits? Can it be less than 16 bits? What is the noise requirement?

How much time can the converter take to converter all four channels? Or, what is the slowest rate for the conversion of all four channels?

For example, for the ADS1259, new data from all channels can be read about every 2ms(500SPS) or more. For the ADS8509, new data from all four can be read about every 20us(50KSPS) or more.

Other important considerations are the speed and type of input signal.

Typically, the signal that arrives to the convert should be buffered, and in order to select the right buffer amplifier it is important to know the signal type.

Will the input signals come directly from four sensors or from four instrumentation amplifiers?
If it comes directly from sensors, what type of sensors?
Will it be DC or AC?
If AC, what bandwidth?

Once we receive the answers to these questions, we will be able to narrow down combinations of IC for your project.

Thanks,

Hi Rafael,

Thanks in advance for your support.

The main problem we are dealing here is that we don't have a fully signal specs due to the fact that the board should can be used with a very wide set of sensor, we are talking of 4-20mA sensors (in 2, 3, 4 wires versions), analog voltage output sensor (single ended, differential, unipolar, bipolar, ...  in ranges +-10 +-5 +-2.5 ....) . For the same reason, there is no a minimum conversion rate (but should be in the range of a few ms), the signal will come directly from the sensor (not to the ADC but to the board). Most sensors we have work with are DC but some of them work in AC and the way to obtain the sensor value is to measure the AC frequency of the output signal.

For this reason, and knowing that this design option depend on us, we can decide the following specs:

Maximum conversion time for all channels: 2-10 ms

Accepting unipolar, bipolar single ended or differential inputs in ranges +-10 +-5 +-2.5 0-2.5 0-5 0-10.

Maximum AC signal frequency: 10-100 khz.

Accept 4-20 mA.

After having review a lot of products from your catalog, we have almost resign to the fact that configuration jumpers will be needed, in order to accept the full range of signals that we want to provide. If there is another solution, please, point me in the right direction. We have seen ADCs with integrated PGA's but in order to only use an ADC in each board we think that using a "self design" (and self is between "" because the design is based on your "Using fully differential op amps as atenuators" document) driven by jumpers to accomodate all the signal input ranges to a suitable -2.5 +2.5 differential signal is the best option. I'll attach a TINA-TI file with the circuit in mind.

3652.InputAccomodate.rar

For the 5 2.5 and 1.25 "sources" at the bottom of the diagram, ref 5050 ref 5025 and ref 3312 IC's have been choosed, but we need to assure that they can provide enough current for it's use in 4 inputs.

If an ADC with no internal 2.5 reference is choosen, a different 5025 can be added if needed, and through an opa350 buffered to the reference input of the ADC (and also to the Vocm of the THS4521 used in the accomodate circuit. 

One additional question is if the circuit attached can be considered a "buffer" to the multiplexer of the ADC or not.

Thanks again for your attention, 

Carlos Leyva Guerrero

Hello Carlos,

The THS4521 is a good driver, and it should be able to drive many differential ADCs.

Besides, we have two different options so far for this project with differential inputs drivers. In order to make an input single ended, the negative input of the differential input can be grounded.

1) 16-bit ADS8331

Attenuation Input Drivers (no resistors required):
2 x INA2143 (G=0.1), 100k input impedance

Multiplexer breakout:
The ADS8331 gives access to the mux output

Gain selector with hardware/software setting for multiple input ranges:
PGA205 connected between the ADS8331_MUX_OUT and ADS8331_ADC_IN

Analog to Digital Converter:
ADS8331

2) 14-bit TLC3578

G=1,2,4,8 Input Driver:
4xPGA205 hw/sw setting for multiple input ranges

Multiplexer + ADC:
The TLC3574 MUX is directly connected to the ADC input

Regards,

hello! I'm sebastian electronic engineering student, I'm from Argentina! ads8509 dac and have a need to make an entry that has the option of choosing different voltages 0-5, 0-10, + -5, + -10, through jumpers.
could help me with the circuit.
apologize for my English, first time I write in the forum.

Greetings thanks.

Hi Sebastian,

Welcome to the forum! Page 7 of the EVM user guide (URL below) has the EVM schematic that shows how you can use jumpers to configure the device for the various input ranges.

http://www.ti.com/lit/ug/slau140a/slau140a.pdf

Best Regards,

Harsha