ADS1255 front end

Hi Paul,

Welcome to the TI E2E Forums!

My suggestions are:

• Use a resistor divider with the ADS1255's internal buffer enabled
• Or use the PGA281 in a gain of 1/8 in front of the ADS1255 (with buffer disabled)

I don't have my normal PC today, but I'll try to attach some schematics tomorrow.

Best Regards,
Chris

Hi Chris,

thank you for your prompt reply.

From ADS1255's datasheet I can see that input pins cannot manage negative signals.

Choosing pga setting = 1, input is given as +/-5V.

At pag. 2 and 3 it's written that both inputs voltage range are from GND to AVDD.

It is my understanding that I need to inject a positive DC offset on the resistor divider.

I have attached a schematic done with TINA, which gives on output 0v when input is -10V, 2.5V when it's 0, 5V when input is 10V.

Is it correct? 

What do you think about REF3333, is it a good choice ?

Regards,

Paul

8284.dc-offset.TSC

Hi Paolo,

For seismic applications, typically you would want to use a device that could handle bipolar supplies, such as the ADS1282, so that you don't have to perform a level-shifting operation...

If you want to use the ADS1255, then you'll either need to level-shift your input signals (as you mentioned)

OR

use a "pseudo-bipolar supply", where AVDD = 2.5V, AGND = DGND = -2.5V, and DVDD <= DGND + 3.6V, and then you would need to level shift the digital signals back to the nominal 0V to 3.3V range (for example).

Paolo Gerini said:

At pag. 2 and 3 it's written that both inputs voltage range are from GND to AVDD.

A word of caution, this is only the case if the buffer is OFF! Otherwise, with the buffer on, the maximum input voltage is AVDD - 2V.

When using a voltage divider, I would recommend buffering the inputs! This could also accomplished with an attenuator/level shifting amplifier prior to the ADS1255.

Is the input signal from a geophone, seismometer, etc...? What kind of load can it handle (does the input stage need high input impedance)? Here are the circuits I would suggest:

• Your TINA circuit would work. However, I would be concerned about the reference drift, resistor tolerances, op amp offset drift, and limited ADS1255 input range with buffer on. I would recommend using a REF50xx (such as the REF5033, or REF5025) instead; (I believe you wouldn't need the reference buffers if you used a REF50xx), and I would further increase the resistor divider attenuation to allow the use of the ADS1255 buffer.
  
  
  
• I think the best performance would be achieved with a resistor divider (with 0V common-mode voltage) connected to an ADS1282 with its integrated low-distortion PGA.
  
  
  
• You could also use the PGA281 to perform both the attenuation and level-shifting. See attached circuit: 5466.PGA281 (Gain = 0.344).TSC
  
  
  
• Alternatively, a fully-differential amplifier could perform a similar function to the PGA281, but would have a lower input-impedance and resistor tolerance/drift concerns.

Best Regards,
Chris

Hi Chris,

input signal comes from an accelerometer, supply voltage is +/-15v, outputs +/-10v and it's recommended to put a 100K resistor between signal and ground, that's all we can gather from datasheet.

Unfortunately ADS1281/2 wouldn't be a suitable solution, it would be above the system target price.

I found the pseudo bipolar supply solution very interesting, but I guess that this solution would be much more expensive than the other with PGA281, considering the cost of the negative supply rail and the level shifter ICs and in general the bom part count.

In terms of performance what would be the better choice (PGA281 or pseudo bipolar)?

I guess that using PGA281 solution internal buffer is no more recommended am I right?

Thank you very much,

Paolo

Hello Paul,

We have the same requirement, a +/- 10 volt range. We tried using a simple resistor divider but it did not work for inputs above about 0.8 volts differential. We think this is due to the strange requirement when the input buffer is enabled that the voltage to ground be within 0 to 3 volts for a 5 VDC VADD. If the buffer is disable you can get a +/- 5VDC span but then the input impedance is only 150k. We'd like to use the ADS1255 with the input buffer enabled but don't want a complex input level shifting scheme. 

We do not have a solution yet...

John Rogers

Hello Chris,

Hello Chris,

We looked at the ADS1282 and even designed a board but I think it would be too pricy for many of our customers. The ADS1255 is in our price range but we need to handle differential inputs in the +/- 10 volt range. We found that when the ADS1255 buffer is enabled, the voltage to ground is out-of-spec and simple voltage divider won't work, without some level shifting scheme. This level shifting sounds complicated and maybe can cause its own problems. 

What about running the ADS1255 with the internal buffer disabled and putting a voltage divider followed by a differential buffer/amp (X1) in front of the ADS1255 diff. inputs? We found that the ADS1255 can handle differential signals in the +/- 5 volt range if the internal buffer is disabled. So all we need is to attenuate by 2 and buffer so the input sensor sees a higher impedance.

Thanks

John Rogers