PICOSTRAIN RDSON/Gain Compensation

RDSON Compensation

During the discharge of the capacitor the single resistors of the strain gage lie in series with the switching transistors of the PS021. The switch-on resistance (= Rdson) of these transistors as well as their temperature dependence and supply voltage dependence are added as a failure to the measurement result. The Rdson itself is in the range of 3 - 5 Ohm. The variation over temperature and voltage is in the range of 1 Ohm. It can be easily seen that this is bigger than the initial measurement signal of a 350 Ohm strain gage! The PICOSTRAIN measurement principle is enhanced only with the patented compensation method for the Rdson.

For compensation the switch is split into two transistors. For each SG resistor the capacitor is discharged 3 times, through transistor 1, through transistor 2 and finally through both transistors in parallel. The PS021 intrinsic DSP calculates the correction Rdson correction value from these three data.

GAIN Compensation

The comparator itself shows a significant propagation delay, which is added to the result. It introduces an additional, temperature dependent gain error. The absolute maximum of this error is in the range of 1% of the maximum measured value. PS021 is capable of measuring the comparator's propagation delay and therefore can compensate for the gain error.

This compensation takes 1 measurement (= 1 cycle time) per half-bridge. The measurement is done at the end of a normal measurement. For noise reduction the correction data can be averaged.

Using the PS021 gain compensation the gain drift over temperature is only about 1 ppm/K (0.0001 %/K) of full scale. Even for calibrated scales it is not necessary to do further compensation actions.

Schematic Diagram
(click to enlarge)

    Datasheet PS09 Vol. 2:
    CPU - Single-chip solution for strain gauges
    Datasheet PS09 Vol. 1:
    General Data and Front-end Description
    Datasheet PS09 Eval System:
    Evaluation Kit for PS09
    Datasheet PS09:
    DLC Evaluation Kit for PS09
    Datasheet PS081:
    Enhanced Single-chip solution for weigh scales
    Datasheet PS081 Eval System:
    Evaluation kit for PS081
    Datasheet PS021:
    PICOSTRAIN front-end for strain gauge sensors
    Datasheet PS021 Eval System:
    Evaluation system for PS021
    Datasheet ALCS350-V2:
    Load cell simulator
    Datasheet PicoProg081:
    Production Programmer for PS081
    Application Note 030:
    PS09: Using PICOSTRAIN with piezoresistive sensors
    Application Note 025:
    PS081: EMI Countermeasures for a Digital Load Cell
    Application Note 023:
    PS081: Design Guideline for Building a Solar Body Scale
    Application Note 022:
    PS081: Design Guideline for Building a Solar Kitchen Scale
    Application Note 021:
    General: Compensation of gain error for uncompensated load cells
    Application Note 018:
    PS08: Metrological investigations of PS08, Determining Zero Drift and Gain Drift
    Application Note 012:
    General: Strain gage wiring with PICOSTRAIN
    Product Report 0909:
    PS081: ESD reliability report
    White Paper 004:
    PS081: How to build Digital Load Cells with PICOSTRAIN conveniently
    White Paper 003:
    PS08: Millikelvin Resolution with only a few Microampere
    White Paper 002:
    PS08 / PS081: How to lower gain and offset drift of a load cell
    White Paper 001:
    PS08: Construction guideline for solar driven scales
    Screencasts PS09:
    38:11 min. webcast introduction into the PS09 evaluation kit
    Screencasts PS09:
    19:01 min. webcast shows how to do a correct temperature compensation of a load cell using PS09 using the load cell's Rspan
    Screencasts PS09:
    17:54 min. webcast shows how to do a correct temperature compensation of a load cell using PS09 using the PS09 internal temperature sensor
    Screencasts PS081/PS09:
    20:28 / 40:30 min. webcast introducing a new concept for digital load cells
    Screencasts PS08/PS081:
    13:36 min. webcast shows how to do a correct scaling of the HB0 measurement result
    Screencasts PS08-EVA-KIT:
    23:23 min. movie introducing the PS08 assembler software
    Software PS09:
    Software PS081:
    PS081-EVA-KIT, PICOPROG v2.0
    Software PS081:
    PS081-DLC-KIT, PICOPROG v2.0
    Software PSA021/PS021:
    PSA021 / PS021-EVA-KIT, COM Port
    Software PicoProg V3.0 (V2.0):
    Driver Installer Stand-Alone
    Software National Instruments:
    NI VISA runtime engine for Windows 7
    Software National Instruments:
    NI VISA runtime engine for Windows Vista
    Software National Instruments:
    NI VISA runtime engine for Windows XP
    PICOSTRAIN Measurement Principle

    The capacitor is charged to the supply voltage and then discharged through one of the SG resistors. The discharge time down to an arbitrary trigger level is measured with ultra-high precision using a TDC (Time-to-Digital Converter). The discharge time is in the range 100 µs. The TDC unit used have a typical single-shot resolution of less than 20 ps.

    This measuring process is repeated in time-multiplex with both resistors of a half-bridge, using the same capacitor and the same comparator. Calculating the ratio of the results will turn out the absolute values and temperature dependencies of the capacitor and the comparator.

    Additional patented circuits and algorithms inside the products compensate for further error sources like the switch-on resistance of the output drivers (Rdson) and the propagation delay of the comparator. The result is very precise, nearly free of gain errors and very stable with temperature. In total each single measurement is made of 8 discharge/charge cycles to solve this compensation task.

    Due to the measuring principle, does not need a full-bridge but a half-bridge is sufficient. The supply of the half-bridge is provided directly by the circuits. There is no need for a separate supply of the SG. Also the reference voltage is not required.

    Thanks to the pulsed drive easily controls the current through the whole system and, even more important, reduces the current consumption to re-markably less than comparable ADC systems.

    The measuring principle is showing a new approach to strain gage (SG) measurement. Contrary to the Wheatstone bridge, where the variation of resistance is transformed into a variation of voltage, solutions transfer it into a high-precision time interval measurement. For this purpose the SG resistors are connected to a capacitor, forming a low-pass filter.

    Measurement Principle
    (click to enlarge)

    PicoStrain Background : Measurement Task

    Metal strain gages (SG) change their value with mechanical deformation, especially a variation in length. The strain e designates the relative variation in length of the SG:

    Strain (e) = dL/L

    Common SG have a maximum strain of typical

    e(max) = 1000 µ (1000 x 10-6 or 0.1 %).

    The ratio of the resistance variation to the length variation is designated K-factor or strain gain.

    dR/R = K x dL/

    For metal SG the K-factor is typically of value 2. The maxim variation of the SG resistance is then given as:

    dR(max)/R = e(max) x K = 2000 ppm

    If the SG is connected in the manner of a Wheatstone bridge, this corresponds to a maximum signal output voltage of 2 mV/V. The resistance of common metal strain gauges is typically 350 Ohm or 1000 Ohm. The maximum variation in resistance and therefore the effective measurement range is within 0.7 Ohm to 2 Ohm. This small variation must be resolved according to the measurement task. The range of the resolution needed is very wide. It is between 10 ENOB (e.g. for pressure sensors) and 18 ENOB (e.g. calibrated scales). In the upper range the precision of the measurement has to be:

    Resolution : 2000 ppm/218 = 0.008 ppm eff.

    or 26.9 ENOB referenced to the full resistance.

    The typical measurement rates are in between

    2 - 8 Hz (e.g. scales) and

    4 - 10 kHz (e.g. fast pressure sensors).

    TypePart numberPackageRoHS compl.Shipping package
    PS09MNR 1783DiceYesWaffle pack
    PS09FNMNR 1840QFN40Yes
    PS09-EVA-KITMNR 1785PCBYes
    PS09-DLC-EVAMNR 1927PCBYes
    PS081MNR 1615DiceYesWaffle pack
    PS081FNMNR 1612QFN56Yes
    PS081-EVA-KITMNR 1525PCBYes
    PS021MNR 1002TQFP48YesTray
    PS021FNMNR 1001QFN48YesTray
    PSA21-STDMNR 984PCBYes
    PSA21-WSBMNR 985PCBYes
    PSA21mini-STDMNR 990PCBYes
    PSA21mini-WSBMNR 991PCBYes
    PicoProg Uni System V2.0MNR 1723PCBYes

    picostrain advantages


    Extremely low current consumption for your complete bridge sensor system

    The ultra-low power PICOSTRAIN measuring principle shows an extremely low overall current consumption of the total system. It is possible to reduce the power consumption by a factor 10 to 20 compared to conventional AD-Converter solutions.


    One solution suits to (nearly) all applications, from low cost up to highest precision

    PICOSTRAIN allows to build systems with 1.500 ... 150,000 peak to peak divisions (at 2 mV/V strain) and update rates up to 1 kHz. One and the same chip covers the requirements of many applications, just by variation of your software.


    Reduced overall system costs

    PICOSTRAIN reduces or saves the costs of the power supply. Power plugs will be replaced by batteries, batteries will be reduced in size. In many applications a single coin cell battery may last for 10 years under normal operation. The number of external components is far below the usual.


    New products that could not be done up to now

    There are new product options mainly due to the low current consumption and low operating voltage. First to mention are solar driven applications. Also 1.55 V silver-oxide batteries are possible although with some restrictions (e.g. re-programming). The measurement quality is impressive even at those low voltages and it might be difficult to find something comparable in the market.


    Improve your Quality and lower your production cost

    PICOSTRAIN offers unique possibilities to simplify the production of your bridge sensor and to improve its quality. The bridge offset can be simply adjusted by setting a register, gain drift can be reduced through software correction of Rspan. This method is more precise than with standard solutions. At the same time costly mechanical trimming, wich is still common practice, becomes redundant.