This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

PGA vs. VGA Selection

Other Parts Discussed in Thread: ADS7861, OPA335, LMH6522, PGA117, PGA116, PGA2500, PGA281, VCA821, LMH6518, PGA308, PGA204, PGA309, PGA280, PGA206, PGA205, PGA202, PGA203, PGA112, PGA113, INA2132

Hello Everyone,

It has always been a pleasure to work with TI parts. many thanks for all the designers.
Following are my queries and related project inputs for resolving the queries.
I shall be grateful to have them resolved.


Designing a Programmable (or Variable) Gain Control + Calibration block for ADS7861


ADC : ADS7861
Reference : 4.096 V reference (chip under selection)
Interface type : SPI
Input buffer : OPA335

Input Signals

Input 1 : 0 VDC to 100 VDC, Differential
Input 2 : 0 VDC to 100 VDC, Differential
Input 3 : 0 VDC to 10 VDC, Differential
Input 4 : 0 VDC to 10 VDC, Differential

Scaled down with a high impedance voltage divider and a buffer.


  1. What is the basic difference between a PGA and a VGA? On prima facie, both devices get a digital control input and adjust the gain of signal input.
  2. Considering above requirement whether to go for a PGA or for a VGA?
  3. Please confirm / deny my understanding about comparison:
    1. PGA would give inherent "Calibration" facility.
    2. VGA explicitly mentions "Differential" inputs.
  4. What is the difference between a Binary Gain and Scope Gain for a PGA? One can see the Binary Gain steps to be of the order 2^n. Is it the only difference?
  5. LMH6522 distinctly mentions "Differential input". Will a PGA also suffice the need?
  6. If PGA116 (or PGA117) be selected, is my understanding correct that no extra "Calibration" circuit block might be required?
  7. The selected ADS7861 has sampling rate of 500 kSPS. Will PGA116 (or PGA117) support it (as it will be the input block for ADS7861)?
  8. If answer to query 2 suggests use of a VGA, will VGA support this high speed of sampling?
  9. If a VGA is to be selected, which chip / chip-set would be useful for calibration?
  10. Does either of the devices support Differential Inputs? Idea is to feed the differential analogue input to PGA / VGA and to feed output of this stage to ADS7861.

Best regards,

Utpal Tembe

  • Hello Utpal,

    Before you read the "wall of text" below, please let me know the following information as ultimately you seem to need a device recommendation.

    • What is the required signal bandwidth of your system? You have selected an ADC with a 500kSPS sampling rate; does that mean your signal bandwidth is 250kHz? If not, what is it? I'd imagine that it's probably less since you have up to 100V inputs, and not many devices can output such high voltage at a high frequency.
    • What are the required input and output voltage and current requirements of your system? You mention high-impedance voltage attenuation at the inputs, but what are the exact inputs you expect?
    • What power supply voltages do you have available? Split supply or single supply?
    • Do you intend to have the OPA335 as a buffer for your input attenuator, for your ADC, or both?

    Here are the answers to your previous questions. This can be a bit confusing as there are several different groups within TI, and within the companies we've acquired, making these types of products with different naming conventions.

    1. In theory, there is not much difference between a PGA (programmable gain amplifier) and a VGA (variable gain amplifier). Both types of products will provide an adjustable signal gain. However, in practice there are some key differences:
      1. PGAs are usually digitally programmable, either through an SPI interface, I2C interface, or gain select pins. They are also usually designed for precision (lower bandwidth) applications, such as industrial sensing or audio. Examples would be the PGA281, PGA2500, and PGA116.
      2. VGAs, or VCAs (voltage-controller amplifier) can be digitally programmable (such as the LMH6518) or controlled by an analog voltage (such as the VCA821). These devices are usually high-speed, and intended for video, RF, driving high-speed ADCs, wide-bandwidth tunable active filters, and other similar applications.
    2. Whether your application requires a PGA, VGA, or VCA depends on a number of performance requirements.
    3. Many PGAs, such as the PGA11x family and the PGA30x family, have built-in calibration support. However this is not a rule for all PGAs. Likewise, all VGAs/VCAs are not necessarily differential input or output. Many PGAs are fully or partially differential, such as the PGA281, PGA204, and PGA308. If you look at a product's data sheet, a differential-input, single-ended output device will often be referred to as a programmable gain instrumentation amplifier.
    4. As you have guessed, binary gain steps means gain increments of 2^n. Scope gain means gain increments similar to what you'd find on an oscilloscope, such as 1, 2, 5, 10, etc. However, not all devices with binary or scope gain settings have all the same increments available.
    5. The LMH6522 is indeed a differential input device, as many of the high-speed PGAs or VGAs tend to be. The PGAs from our precision amplifiers portfolio with differential inputs are PGA202, PGA203, PGA204, PGA205, PGA206, PGA280, PGA281, PGA308, and PGA309.
    6. PGA112, PGA113, PGA116, and PGA117 have support for calibration through dedicated calibration input channels that are intended to be connected to a reference voltage. This allows you to take multi-point calibration measurements in order to compute corrected values after sampling by an ADC. This does not mean that the output of the PGA will change after calibration. Please see section 9.1.5 of the PGA11x family data sheet for more information. The PGA308 and PGA309 do support calibration of internal gain and offset blocks for use in specific bridge sensor signal conditioning applications.
    7. I don't believe the PGA116 or PGA117 would be suitable for your application, as you seem to require a programmable-gain instrumentation amplifier. Both of these devices only allow for a single-ended input signal.
    8. VGAs, in general, have high bandwidth and should support a 500kSPS sampling rate.
    9. I'd need more information to make a good recommendation for a PGA/VGA device. 
    10. Please see answer #5 for PGAs that support differential inputs. I can't list out the VGA/VCA devices, as they are supported by the High Speed Amplifiers group and dozens of part numbers meet this requirement.

    Best regards,

    Ian Williams
    Linear Applications Engineer
    Precision Analog - Op Amps

  • Many thanks Ian.

    It has really clarified my understanding.

    Following are the answers to your queries:

    What is the required signal bandwidth of your system?

    As you have correctly identified, the input analogue signal itself does not have a large bandwidth.
    The reason for selection of high sampling rate for a mere application requirement from end user.

    What are the required input and output voltage and current requirements of your system? You mention high-impedance voltage attenuation at the inputs, but what are the exact inputs you expect?

    By high-voltage input attenuation, I meant scaling down the input signal to suit PGA input range.
    The input analogue inputs are differential in nature.
    We have used a resistor divider for scaling down the input signal.
    Resistor values are so chosen as not to load input signal.

    What power supply voltages do you have available? Split supply or single supply?

    We have a liberty to go for single / dual supply.

    Do you intend to have the OPA335 as a buffer for your input attenuator, for your ADC, or both?

    The input section now does not have OPA335.
    The scaled down differential input is now fed to PGA.
    The reason for choosing PGA112 was availability of inbuilt calibration channel for ADC.

    The input signal path is as follows:

    Input ==> INA2132 ==> PGA112 ==> ADS7861

    Best regards,

  • Hello Utpal,

    Happy to help. As Pete and I are now helping you outside the forum, let's keep all further discussion in the e-mail chain.

    Best regards,

    Ian Williams