ADS1255: Using one crystal / voltage reference for multiple ADS1255/ADS1256

Hi Patrick,

Welcome to the TI E2E Forums!

Yes, you can use a single clock source to operate all 6 ADS1255/6 ADCs. The first ADC can be clocked with a crystal, or crystal oscillator, and then its CLKOUT pin can be routed to the CLKIN pin of the next ADC, and so on...

If you do operate the clocks like this, then I would recommend adding a ~50 Ohm series resistor on each of the CLKOUT signals. The resistor helps to slow down the fast clock edges, and I've found that it helps to reduce noise coupling (from the clock) in this configuration.

Regarding the reference voltage, I would avoid using a single reference source for all six ADCs. The reference inputs connect to an internal switched-capactor circuit, so sharing a single reference source could result in a lot of reference noise or reference droop. If possible, I would try to buffer the reference to each ADC, or at least provide local decoupling capacitors to each ADC to try to keep transient currents from coupling between devices.

Hi Patrick,

For the power supply, I would make sure that the LDO (or LDOs) that is powering all of these devices is able to supply more than enough current to every device. When first powering up the system, the LDO will need to drive much higher currents than normal to charge up all of the decoupling capacitors on the power rail; therefore, I would design the power supply with some additional margin to handle transient currents (such as power-up) and not only the minimum required current to operate all devices while in steady state. If the LDO is sized properly and decoupling is provided close to each ADC, then it should be fine to use a single supply source. Just be sure to use a very wide trace or a power plane to supply current to all of these devices to reduce the effects of common-impedance coupling between devices that are sharing a supply.

The DAC8532 is a lower power-device and will work off of the same 5V supply. Just follow best practices for layout and decoupling and I don't think you will have any issues with sharing the supply between multiple devices.

I would avoid any inductive filtering on the supply rails! This may seem counter intuitive; however, transient currents on the supply rails will interactive with inductive filters and will result in large voltage spikes due to the presence of these inductors. Also, choking off the current to an IC may result in odd behavior, especially during power-up. I would recommend placing local decoupling capacitors near each ADC and avoid adding any inductance on their supplies. If you think additional filter is still required, then I would advise adding small series resistors on the supply in place of any inductors.

Thee LM285D-2.5 is a very lower power reference. As such it will not provide the best precision, nor is the output voltage of this reference buffered. Since you are designing a system with a very high resolution ADC, I would advise using a precise and buffered reference source, such as the REF5025 or any of the  REF6x25 devices. The REF6250 I believe is one of our more accurate references, but the REF5025 is probably sufficient and is offered at a much lower cost.

Hi Patrick,

Here is a list of about 9 different LDOs that I would would work well, depending on your input voltage:

http://www.ti.com/power-management/linear-regulators-ldo/products.html#p451max=1;3&p634max=5;125&p634min=-37;5&p1128=0;50.3&p1130=0.08;4.3951702&p236typ=0.0005;2&p1811=0;25.0397744&p1498=Catalog&p1530=32.8;90&p985typ=0;429

NOTE: The ADS1256's sampling frequency is up at about 1.92 MHz, so look for an LDO with a high PSRR around this frequency range, since the digital filter passband will repeat at multiples of the sampling frequency.

The ADS1256 does not  allow for reference voltages larger than 2.6V, so I would not recommend using a REF5050 with the ADS1256. Since the REF5025 is already buffered, you do NOT need to add an OPA350 between the REF5025 and ADS1255 (the REF5xxx reference had not been released when the ADS1255 datasheet recommended adding a buffer to the external reference source). The TRIM/ND capacitor on the REF5025 is optional but it does help to reduce the reference noise. Ceramic capacitors are fine, that is what I typically use.

However, it's fine to use the REF5050 with the DAC8532. I do not know the answer to the DAC8532 headroom question... I would recommend starting a new thread on E2E to ask specific questions about the DAC, then you'll get someone with more expertise with that device answering your DAC questions.

For the input RC filters, you can't go too wrong as long as you keep the RC components small. Large RC values take a long time to settle and will result in a settling error. I recommend using high-quality C0G type capacitors, which means the capacitor cannot be any larger than 120 nF (in a 1206 package) or 15 nF (0603). The resistor should not exceed 10 kOhm (though it is better to keep it less than 5 KOhm) because the thermal noise of the resistor will start to degrade the resolution of the ADC.

Hi Patrick,

It shouldn't be a problem to measure +/- 72 mV. Will one side of your shunt resistor be connected to ground (e.g. VINN = 0V)?

The Excel calculator is complaining because you entered a +144mV differential voltage (VINP = 0.072, VINN = -0.072), which would be too large for the gain of 64 V/V range.