TI E2E Community
Precision Data Converters
Precision Data Converters Forum
2.5V REF can be optional for ADS79xx?
I am in the midst of prototyping a couple of boards using the ADS7951.
Starting on page 34 of the data sheet, the example circuits mark the REF5025 part as "o/p" which implies that it is an optional part. The text states that the chip "can operate with an external 2.5V ± 10mV reference."
I just want to clarify what the options are. While the data sheet says it "can" operate with a part like the REF5025, it's not precisely clear about whether the chip can operate without it. Besides 2.49V to 2.51V external, is there any kind of internal reference at all? Can the pin be tied to 5V instead, like the analog supply? Are there really any usable options for a 12-bit AD?
Better text might be to say that the reference is not optional, but rather it must be used, or something equivalent or similar.
One part I have been using is the Vishay Si9183DT-25, which provides 150 mA of reference current at 2.5V for one-tenth the cost of the REF5025, so I am wondering if anyone has used this ... or would it compromise the accuracy of the 12-bit converter?
The "o/p" was meant to convey "output", not "optional". I have no idea why someone felt compelled to even put that there, but sorry for the confusion. The ADS7951 does not have an internal reference so the use of an external reference voltage is NOT optional - you have to supply a 2.5V reference in order for the device to work properly. You cannot tie the reference pin to 5V, as this is outside the specified range for the reference - see the electrical tables.
The Vishay part you reference is not a voltage reference per se but a low-dropout regulator - they are not the same thing. The initial accuracy of the Vishay regulator is 2.5% over temperature, while the REF5025 is 0.1%. The output voltage noise on the REF5025 is 7.5uVp-p, while the regulator is 300uVrms - which is almost 2mV peak-to-peak! To maintain any kind of accuracy, the noise must be much lower than that. A regulator does not usually make a good A/D reference!
Thanks. That makes perfect sense.
As a followup question, if I also wanted to use the REF5025 output as a virtual ground, I imagine that I'd need to place a voltage follower in between since the REF5025 only sources/sinks 10mA.
I have 8 capacitively-coupled op-amp buffers ahead of the ADS7951, and I am already using 2.5V as a virtual ground, so it seems that the REF5025 would be handy for that, but might need a little help. Sure, the steady-state current should be 0, but I just don't think the REF5025 can feed 9 loads. Instead it would seem better to connect the REF5025 directly to the ADS7951, but buffered before being used elsewhere. Any suggestions for a voltage follower op-amp? I'm not quite sure which forum to ask this particular question.
You're right, you probably don't need all that much current but the transient currents may be high - and they may be enough to cause the interference with the A/D. I'd agree that you could hook the REF5025 directly to the ADS7951, but use a buffer to create your virtual ground. The choice of op amp here depends upon how critical the 2.5V value is to the inputs - perhaps you can tolerate a fair amount of offset, but perhaps not. I can only offer the experience that the OPA350 is what we usually use as a reference buffer - to buffer it to ADCs which have a high transient input current on their reference pins. Its low wideband output impedance makes it ideal for that purpose, and it can handle driving pretty hefty capacitive loads. Since you'll be wanting to bypass that virtual ground node, I'm sure, to keep noise low, it seems that the OPA350 might be a good fit here as well.
The samples from the AD feed an FFT where I'm ignoring several bins at the bottom, certainly the DC component, so I expect that my application is highly tolerant of offset. Bandwidth shouldn't be a factor since 2.5V is DC. That leaves PSRR and EIN as the important factors. I'll scan around the various TI op-amps for less-expensive options than the OPA350, assuming there are other op-amps with similar low-noise performance.
Well, yes, 2.5V is DC, but the transient load you may place on the output of the op amp means you probably want something with a little bandwidth. That's what makes the OPA350 ideal for driving the ADC reference pin. The demands placed on the op amp for your virtual ground are probably nowhere near as much as driving an ADC reference pin, so something like a good ol' TL072 might do you just fine...
Thanks again. You've completely answered my question about the ADS7951, so I have moved my followup questions to the Precision Amplifiers forum under Selecting voltage-follower for low noise and low cost
All content and materials on this site are provided "as is". TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with regard to these materials, including but not limited to all implied warranties and conditions of merchantability, fitness for a particular purpose, title and non-infringement of any third party intellectual property right. TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with respect to these materials. No license, either express or implied, by estoppel or otherwise, is granted by TI. Use of the information on this site may require a license from a third party, or a license from TI.
TI is a global semiconductor design and manufacturing company. Innovate with 100,000+ analog ICs andembedded processors, along with software, tools and the industry’s largest sales/support staff.