SAMPLE MEASUREMENT SYSTEM user manual R.Schouten TUD TN-VS-QT 5-8-97 (rev 31-08-00) CONTENTS [1] System concept & description [2] Units description 1 Sample Signal unit 2 Sample Data unit 3 User Data unit 4 Battery unit [3] Modules description 1 Sample Signal unit modules: 1 S1: Iso-Voltage Source 2 S2: Summing Module 3 S3: Voltage Source 4 S4: Current Source 5 Matrix Module 6 M1: Current Measurement 7 M2: Voltage Measurement 2 Sample Data unit modules: 1 D1: option 2 D2: Dual Source DAC 3 D3: Dual Measure ADC 4 D4: uController 5 D5: Hex BiasDAC 3 User Data unit modules: 1 U1: option 2 U2: Dual Source ADC 3 U3: Dual Measure DAC 4 U4: Opto/RS232 module [4] Manual & Remote Control 1 the manual control panel 2 remote control hardware setup 3 remote control commands and software [5] Performing measurements 1 system setup & connections 2 Performing I-V measurements 3 Performing V-I measurements [6] User operated calibrations & checks [7] Special applications & options [8] Specifications 1 System concept & description R.S. System concept: The Sample Measurement System (SMS) is designed to enable high quality measurements on very small nanometerscale devices located in a cryogenic environment. Design priority was the minimalisation of noise and interference signals on the wires connected to the measured device (sample). The SMS contains the electronics, switching, shielding, filtering, supply and isolation functions to operate as a complete remote source-, measure- and bias-system. The system is physically divided in a userpart and a samplepart only interconnected by fibers for data/control exchange. The SMS does not perform measurements itself, commercial equipment like lock-in amplifiers, generators, voltmeters can be connected to the userpart of the SMS (isolated from the sample). A computer can be connected to the userpart of the SMS by means of a serial (in-out) port. The computer can set biasvoltages, gain, source ranges and reads battery status and signal overloads. If this computer also controls and reads the already mentioned commercial equipment (GPIB-bus) a complete measurement could be made and registrated in a software environment. Manual control of the SMS is also possible for all the source and measure settings, only the biasvoltages need to be set by computer. System description: The SMS is designed for measuring I-V and V-I curves with additional biasvoltages in a signal bandwidth of d.c. to 20kHz. All circuits connected to the measurement are battery operated and fiber isolated. Noise and interference are minimised by electronic design, mechanical design, filtering and shielding up to frequencies of more then 10GHz. For further reduction cryogenic filters and signal dividers could be used. The SMS consists of four boxes; a Sample Signal unit, a Sample Data unit, a battery unit and a User Data unit. Figure 1 shows a basic setup. The Sample Signal unit is located as close as possible to the sample.The Sample Data unit can be placed at max 1.5m from this and the User Data unit is located as far away as possible (fiberlength typ. 3m from the Sample Data unit). The wires from the sample enter the Sample Signal unit and are filtered and then distributed to the (analog) sources, amplifiers and biasvoltages in this unit. Here the signals are transformed to volt-level and leave the unit through a second filterport. Via a shielded multicable these signals enter the Sample Data unit. This unit is containing the shielded and filtered digital electronics to create the biasvoltages and circuits to perform the transfer of the data and control signals via fibers to the User Data unit. On the front of the User Data unit the connections to commercial equipment and computer are made. Both Sample Signal unit and Sample Data unit are battery powered, uninterruptible exchange of battery units is possible. 2 Description of system units R.S. 2.1 The Sample Signal unit (SSU) The SSU is containing the sample wire-filters & switches, the sources, amplifiers and BiasDAC outputs. The BiasDACs itself are not located in the SSU (DACs being a digital device) but the outputs are filtered and divided or summed here in the summing module. The SSU is the signal conditioning part of the SMS, user generated source signals enter as a 1 Volt(typ) level and are transformed to the desired excitation current or voltage. Measured currents or voltages are transformed to a 1 Volt level and then send to the Sample Data unit. All inputs on this 1 Volt level are differential to prevent ground loops. The unit is housed in a shielded 19-inch case (-60dB at 1GHz) containing only analog circuits generating no clock signals or transients . All gain and mode control is done using special shielded bistable relais causing no heating effects like thermocouples or drift. The relais control lines & (battery)power are filtered and rise-time limited before entering the unit. The absence of manual controls on this unit minimizes the mechanically induced noise. Every wire entering the SSU from the multicable to the Sample Data unit is filtered in filterport 1 (100kHz LPF) on the back of the case. Every sample wire is filtered in filterport2 (pi-filters -80dB at 10GHz). Filterport2 is located on the side of a copper box placed inside a u-metal box called the matrix module (inside the SSU). The matrix module offers the user the possibility to select any of the 40 sample wires for source/measure/bias operation. 2.2 The Sample Data Unit (SDU) This unit contains all the (battery powered) digital electronics that is needed to operate the SSU and to transfer the signals via optical fibers to the User Data unit. The ground of the SDU is galvanically connected (via filterport1) with the SSU and thus with the sample (via filterport2), the 19-inch case containing this unit should therefore be mounted isolated from mains ground. Every digital module in the SDU is individually shielded and the lines entering/leaving each module are filtered. The BiasDAC module is isolated (optocouplers) and seperated from the u-controller setting its values. All BiasDAC outputs are tranferred in shielded cables in the overall shielded multicable to the SSU summing unit. On the front of the SDU an operating panel [4.1] offers manual controls for all source/measure modules in the SSU. For remote computercontrol [4.2] enable/disable switches are present. On the back of the SDU the modules have their frontpanels described further in this manual [3.2]. Also present on the back of the SDU is the supply panel with the input connectors for battery units. Two connectors (1,2) are for one main battery, offering uninteruptable operation by plugging in a fresh battery before removing the old one. Two connectors (A,B) are for optional floating BiasDACs (DAC 1-8 , DAC 9-16). The battery mode for the BiasDACs can be selected by a switch on the battery panel. 2.3 The User Data unit (UDU) R.S. This unit contains all the (mains powered) digital electronics needed to transform the signal-fibers from the SDU to 1 Volt signals for the commercial equipment and the electronics to translate the control-fibers to RS232 signals for the computer. To prevent mains and/or computer interference entering the system both parts have a seperate mains supply with shielded transformer. The ADC inputs are differential to prevent ground loops. 2.4 The battery unit The SMS can be operated with minimal 1 battery unit (main battery) containing 2 rechargeable batteries powering the SSU and the SDU. It is possible to have the computer read the output voltages of the main battery unit and when needed a second (charged) battery umit can be plugged in before the first is decoupled, giving uninterupptable operation. Expected operation time for 1 battery unit is 48 hours. As an option it is possible to use seperate battery units for BiasDAC 1-8 and/or BiasDAC 9-16. This lowers the noise level of the BiasDACs and gives the user galvanically isolated BiasDAC-groups to prevent ground-loops .These additonal battery units output voltages cannot be measured by the computer, only a battery-good bit is given. However, operation time of a battery on the BiasDAC module exceeds more then 10 times that of the main battery so its sufficient to start with 1 fresh battery unit to do a full measurement session. Metal battery-enclosures are used in combination with shielded cables. These enclosures are connected to the grounds of the SSU and SDU and thus to the ground of the sample so mounting them should be done free of connections to mains ground. The fuses on the battery units prevent excessive currents (>100A) in case of short circuit. All battery unit connectors & switches are found on the back of the Sample Data unit (SDU). Commercial battery-chargers are used, which can remain connected to a battery unit even if it is fully charged (indication on charger). Do not try to use a power supply or other charging method as this could cause degrading or overheating of the batteries. 3 Modules description R.S. Modules are removable cassettes that are placed in the (19-inch) units of the SMS. Each module contains a specific function like current source or biasDAC. The modules can only be placed in a specific unit and in one more specific place (slot) , the modules are identified by a letter-number code on their front. Slots are identified by a letter-letter code in the units Sample Signal unit modules: default slot: 1 S1: Iso-Voltage Source Sa 2 S2: Summing Module Sb 3 S3: Voltage Source Sc 4 S4: Current Source Sd 5 Matrix Module (fixed, only removable for service) 6 M1: Current Measurement Ma 7 M2: Voltage Measurement Mb Sample Data unit modules: 1 D1: option Da 2 D2: Dual Source DAC Db 3 D3: Dual Measure ADC Dc 4 D4: uController Dd 5 D5: Octal BiasDAC De User Data unit modules: 1 U1: option Ua 2 U2: Dual Source ADC Ub 3 U3: Dual Measure DAC Uc 4 U4: Opto/RS232 module Ud 3.1 The Sample Signal unit modules (fig.2) The modules can be divided in source (S) and measurement (M) type with the exception of the Matrix module that contains the samplewire filters/switches and connection nodes for patch cables. All S&M modules contain analog signalconditioning electronics having (lemo) connectors on the front that can be connected to the 40 nodes (MCX connectors) on the matrix module by means of patch cables. Routing of the (1V) control-signals for the S&M modules (via the back-plane) is user-defined in the Summing module S2. 3.1.1 S1, the Iso-Voltage Source module This is a module containing an isolation amplifier to generate a galvanic isolated biasvoltage from a userselected source ( BiasDAC, SourceDAC or external source). This selection can be made in the summing module S2. The Iso-Voltage source has a floating output isolated with >10Gohm and 1pF from the other electronics. This output can only drive high impedance loads (100k or more) and the circuit for this is powered by a 9V-battery in the module. Battery status can be manual and computer-checked. Battery life is minimal 1 year. Gain can be set to 1 or 2. Applications for this module are: -Solving ground-loop problems between gate-control lines. -Solving dc-voltage potential problems between gate-control lines. -Enlarging the voltage swing of a BiasDAC (select gain=2) 0..4V will be 0..8V, +/-2V will be +/-4V Larger voltage swings can be obtained by series connection of this unit to a second BiasDAC (using a special patch-cable) giving 0..12V, +/-6V It should be noted that drift and noise of this module exceed that of the BiasDACs so it should only be used when the problems above occur. location: SSU, slot Sa,Sc,Sd input: back: x1 (from S*-V1) routing in S2 (Summing module) output: front: Vout (lemo pin 4/2) output: SDunit: Battery ok indicator control: SDunit: Gain, Battery check 3.1.2 S2, the Summing Module This is a module containing no electronics, but it offers the user the possibility to patch controlsignals for S&M modules, to filter or sum gate voltages and some other options. All the BiasDACoutputs from the Sample Data unit enter here, on a internal printed circuit board the user can sum,divide and filter these before going to the output (lemo) connectors on the front. Also selections for the sources controlling slot Sa/Sc/Sd and Ma/Mb are made here. Provisions for powering a cryogenic pre-amp can be made here. The direct connect connectors (lemo) on the back of the Signal unit (on filterport1) enter here and can be routed. This unit can be tailored for the specific user or application. (you can modify your personal module and plug it in when measuring) input : back: all BiasDACs,SourceDacs output: back: Sa-V1, Sc-V1, Sc-V2, Sd-V1, Sd-V2, MV1, MV2 back: in/out: back: direct connect connectors (2xlemo pin 4/2) from filterport1 in/out: front: direct connect connector (1xlemo pin 4/2,1/3) output: front: Supply-Out for cryogenic preamp front: BiasDAC 1-4, 5-8, 9-12, 13-16 (4xlemo pin 1/2/3/4, connector-shell is DAC-ground) control: none 3.1.3 S3, the Voltage Source module This module contains the voltage source needed for sample excitation when performing an V-I measurement. To prevent ground-loops the output is galvanically isolated (>10Gohm,1pF). It has two inputs: x1 (sweep) and x0.01 (modulation) connected to the S*-V1 and S*-V2 lines of the slot. The input signals are summed inside the module. The Range switch (V/V) on the SDunit divides the output down to the selected level. The function x1/x0.1 switch divides the inputs (both sweep and modulation) by 1 or 10. The last option has two advantages: -the switching is not done on the sample-connected part so it could be performed DURING a measurement to enlarge dynamic range. -an extra decade can be selected beyond the lowest V/V Range setting without degrading noise/drift performance because the SourceADC/DAC noise/drift is dominating. The Output mute switch short-circuits the output, this gives a possibility to do a zero-check or even a computercontrolled-auto-zero. To prevent possible contact-resistance of this switch degrading the exact zero level it is advisable to also switch Range x0.1 on. (the remaining output offset will then be < 1uV ) input : back: x1(from S*-V1), x0.01(from S*-V2) routed in S2 output: front: Vout (1x lemo pin 4/2) control: SDunit: Range 1m/10m/100m V/V (Gain 1/2/3) Range x1/x0.1 (Function 1) Ouput on/mute (function 2) 3.1.4 S4, The Current Source module This module contains the current source needed for sample excitation when performing an I-V measurement. It has two inputs: x1 (sweep) and x0.01 (modulation) controlled from the User Data unit inputs 1 (x1) and 2 (x0.01). The input signals are summed inside the module. The Range switch (A/V) on the SDunit sets the output circuit to the selected level. The Range x1/x0.1 switch divides the inputs (both sweep and modulation) by 1 or 10. The last option has two advantages: -the switching is not done on the sample-connected part so it could be performed DURING a measurement to enlarge dynamic range. -an extra decade can be selected beyond the lowest A/V Range setting without degrading noise/drift performance because the SourceADC/DAC noise/drift is dominating. The output symmetric/single switch selects two ouput modes: -single: pin 4 is the high impedance ouput node delivering the selected current, pin 2 is connected to ground (classic current source). -symmetric: pin 4 is the high impedance output node delivering the selected current, pin 2 is a low impedance node at exactly the opposite voltage level (-Vout) of pin 4. The output voltage is symmetrical to ground. Applications are: preventing a common mode voltage across the sample, preventing a common mode voltage on the amplifier (M2) , doubling the maximum voltage output swing of the current source. The symmetric mode has a slightly higher noise level. The Output mute switch disconnects the output of the electronics from the high impedance output node (pin 4) a passive resistance to ground remains, its value being 2/(A/V Range setting) so 10nA/V gives a 200MEG resistance to ground. This gives a possibility to do a zero-check or even a computercontrolled -auto-zero. (the only offset contribution will be the input bias current of the sense opamp being <1pA) When the output voltage exceeds 2..4 Volt (depends on loadresistance) an overload indication on the SDunit occurs (also computer-readable) Selecting the symmetric mode gives 4..8 maximum output voltage. input: back: x1(=SourceDAC1), x0.01(=SourceDAC2) output: front: Iout (1x lemo pin 4/2/3) single 4=Iout / 2=ground / 3=ground symm 4=Iout / 2=-Vout / 3=ground SDunit: Overload indication control: SDunit: Range 10m/1m/1m/100u/10u/1u/100n/10n A/V Range x1/x0.1 Output symmetric/single Ouput on/mute 3.1.5 The Matrix Module The sample wires from the dillution refrigerator or cryostat enter in this module (via a short u-metal flexible tube on the back of the SSunit). The wires are distributed in a copper box to PI-filters (filterport 2) and leave the box to the sample switches on the front of the unit. The sample switches select on-open-ground for each pair of sample wires. In the "on" position the sample wires are connected to the matrix pins offering patch cable connections to Source/ Measure/ Bias circuits. The module has a u-metal enclosure and removable cover to prevent magnetic coupling in the distribution wiring inside of it. The copper box inside is mounted to the brass ground-reference plate supporting the SSunit. The u-metal tube to the sample is fixed to the matrix module. Dismounting the matrix module from the SSunit is a delicate operation (this is a warning). 3.1.6 M1, The Current Measurement module This module converts an input current to a voltage that is presented at output 1 of the User Data unit (module U3, assuming input select D3 is set to internal see [3.2.3]). The module has 2 input nodes, one (pin 2) is connected to ground, the second (pin 4) has a low impedance. The design of this module is tailored to the specific application which shows in : Input noise is bandwidthlimited to prevent excess noise on the sample. Capacitive loading on the input (PI-filters) causes no instability. Input offset is low an can be trimmed (even remote) to < 2uV. The user can make some trade-offs between input noise, output noise*, input resistance and bandwidth (see also specs [8.1.6]). -The absolute value of the integrated input noise is lowering with rising V/A setting, for a given V/A setting its lowest on the "Low Noise" setting. -The ABSOLUTE value of the output noise in the measurement bandwidth is rising with rising V/A setting but the equivalent input noise is lowering (by a square root relation), for a given V/A setting its lowest on the "Low Noise" setting. -Input resistance is lowest for the "Low Rin" setting and ten times higher for the "Low Noise" setting. Input resistance Rin on "Low Rin" = (V/A setting) / 1E4. Example: Rin is 10k on the 100M V/A setting and Low Rin mode. -Bandwidth is lowering with rising V/A setting ,for a given V/A setting the "Low Rin" reduces the input timeconstant giving a faster response. As an option the "Postgain x100" setting could be used to create a higher bandwidth by reducing the V/A setting while maintainimg the resulting overall gain. * output noise can be transformed to an equivalent input noise but this is not the same as the input noise really present at that node !! output noise: how noisy is the measurement (averaging helps) input noise : what is on the sample (nothing helps). the output noise also changes with sample resistance. The "Input on/zero" switch short-circuits the input on the zero setting This gives a possibility to do a zero-check for the input offset VOLTAGE (or even a computercontrolled autozero for this). This is important because the remaning offset voltage will be imposed on the sample. When switched to "Input zero" input offset voltage will be amplified by 1E3 (Low-Noise) or 1E4 (Low Rin). The resulting output voltage can be trimmed to zero on the module itself (a trimmer on the front of the cassette) or via a user selected Bias-DAC. A check for the remaining input offset CURRENT can be done by using the "Mute" setting of the Voltage Source module S3. A description of the controls: The Low Rin / Low Noise switch has applications already described, it can be changed during a measurement as it does not results in switching at the sample-connected input stage. The Postgain x1/x100 switch is needed to prevent the following Measure-ADC-DAC noise to dominate. Postgain is needed when the output signal has a large dynamic range and the V/A setting cannot be changed during the measurement because it changes the input stage conditions. Its also possible that Postgain is needed when in the highest V/A setting (1G) the output signal is still to low. The V/A switch sets the conversion factor of the module. input: back: remote offset (from summing module) front: manual offset trim front: I-in (lemo, pin 4=in pin 2=ground) output: back: V-out (to measure ADC1-DAC1) control: SDunit: Range 1G/100M/10M/1M V/A Input on / zero Low Rin / Low Noise Postgain x1 / x100 Options (non-standard, see [7]) : A functionally identical module optimised for high-bandwidth (instead of low-noise) could be used (option M1b). An optimised lower noise module (with fixed 1G V/A setting) can be inserted in M*. This unit (option M1c) improves results (up to 15x) when measuring on sample impedances below 10Mohm or when measuring at frequencies above 10Hz . 3.1.7 M2, The Voltage Measurement module This module contains a differential voltage amplifier presenting its output voltage at output 2 of the User Data unit (module U3, assuming input select D3 is set to internal see [3.2.3]). The input stage of the amplifier is formed by the gates of FET-opamps, there are no input resistors added. Input impedance is very high (>10Gohm) also in "ac-coupling". As a consequence the input will slowly drift away when not connected, resulting in a overload signal at the measure ADC-DAC 2 input. To prevent this, a short-circuit connector should be placed on the input when not in use. All functions including "coupling ac/dc" can safely be operated during a measurement as switching is done after the input stage. input: front: manual offset trim1 trim2 front: V-in (lemo, pin 4=+in pin 3=ground pin 2=-in) output: back: V-out (to measure ADC-DAC 2) control: SDunit: Range 1/10/100/1000/10.000 V/V coupling ac/dc Options (non-standard, see [7]) : Two optimised VOLTAGE pre-amps modules can be inserted in M*. A ultra-low bias (3fA), high impedance (1Tohm) differential FET-module (option M1d) or a low voltage noise differential FET-module (1.9nV-1kHz / 2.5nV-1Hz) (option M1e) are possible. 3.2 The Sample Data unit modules (fig.3) All modules in the Sample Data unit are placed on the back of this unit to create room for the frontlocated control panel [4.1] that is controlling the modules of the Sample Signal unit, the battery panel on the back is described in [2.2]. 3.2.1 D1, Option No plans up to now. 3.2.2 D2, Dual Source DAC This module forms (with module U2) a galvanic isolation between the user and the samplepart while transferring the user-generated excitation signals to the source modules. The module offers two output signals (+/-1Vtyp) that are a copy of the signals connected at the inputs of the Dual Source ADC (U2) in the User Data unit. For this these two modules have to be connected via one fiber which transports the digitised values of the source signals. A standard 18 bit commercial audio link is used (SP/DIF), giving dc to 20kHz bandwidth (500Hz when Bandwidth Select "low" is set). An important aspect involved with the oversampling and digital filtering in these kind of systems is the (fixed) delay time between "in" (U2) and "out" (D2). The resulting delay (1.7mS) is higher then can be expected from the 20kHz bandwidth. For Lock-In amplifiers etc. this will cause no problems as the delay value is fixed (measured phasejitter at 30Hz was < 0.0001deg). input: front: Opto in, fiber from module U2 output: front: Receive error, LED indication for fiberdata errors output: back: Voltage outputs 1 and 2 to source modules in SSU levels: +/-1Vtyp, +/-2Vmax controls: Bandwidth select High/Low High= 20kHz Low= 500Hz (-6dB/oct slope) giving noisereduction 3.2.3 D3, Dual Measure ADC This module forms (with module U3) a galvanic isolation between the sample and the userpart while transferring the amplified measurement signals. The module offers two inputs, the signals presented here (typ +/-1V) are digitised and send to the Dual Measure DAC (U3) in the User Data unit. For this these two modules have to be connected via one fiber which transports the digitised values of the measure signals. A standard 18 bit commercial audio link is used (SP/DIF), giving dc to 20kHz bandwidth (500Hz when Bandwidth Select on U3 "low" is set). An important aspect involved with the oversampling and digital filtering in these kind of systems is the (fixed) delay time between "in" (D3) and "out" (U3). The resulting delay (1.7mS) is higher then can be expected from the 20kHz bandwidth. For Lock-In amplifiers etc. this will cause no problems as the delay value is fixed (measured phasejitter at 30Hz was < 0.0001deg). The Input-Select switches on the module offer the possibility to measure one or two external signals instead of those coming from the current and voltage measurement modules in the SSU (intern). For example a temperature could be measured and send (see [7]). The offset of the ADC can be trimmed to <100uV using a calibration command send to the uController, the procedure is described in [6]. On power-up the ADC performs this calibration, but it is advisible to repeat this once after a warm-up time of >10 minutes. input: front: extern 1, extern 2, differential in +/-1Vtyp +/-3Vmax connector:lemo, 4=+ / 2=- input: back: Voltage inputs 1 and 2 from measure modules M1 and M2 in the SSU output: front: Opto out, fiber to module U3 in the UDU control: Input Select intern/extern 1 Input Select intern/extern 2 3.2.4 D4, The uController module This module handles the communication between the Personal Computer (PC) of the user and the measurement system. To do this two fibers have to be connected between this module and module U4 of the User Data unit. The hardware setup and the commands are presented in [4.2]/[4.3] . To prevent excessive digital signal generation all actions of this module are started on request of the user (PC). The module functions are: -setting the Bias-DACs (D5) to values received from the PC reading the Bias-DACs polarity-switches -(When the switch of the corresponding module is set to REMOTE:) Controlling all settings (gain etc.) of the amplifier and source modules in the SSU (S1,S3,S4,M1,M2) by means of the PC. -Reading the overload-indication-latch, the overloadsignals are latched to prevent from missing one and having a corrupted measurement without knowing. On a PC command the latch contents is send and the latch is cleared. -Reading main battery voltages (pos/neg). reading additional BiasDAC batteries status. reading Iso-Voltage Source internal battery status. Switching power on/off for the Source-DAC (powercontrol 1) Switching power on/off for the Measure-ADC (powercontrol 2) (giving a total of 80% reduction in main battery power consumption). -Reading the status of the Measure-ADC input select switches. Reading the status of the Source-DAC bandwidth select switches. Forcing a user-requested offset calibration routine for the Measure-ADC and the Source-ADC. input: front: Opto in from the User Data unit (module U4), fiber containing the serial signal from the PC-RS232 port (comport) back: Bias-DACs Polarity-bits All overloadsignals Main battery voltages Bias-DACs optional battery A and B statusbits Iso-Voltage Source internal battery statusbit Switch-position-bits from MeasureADC and SourceDAC output: front: Opto out to the User Data unit (module U4), fiber containing the serial signal to the PC-RS232 port (comport) Send/Receive LED for checking fibercommunication back: Bias-DAC control-bits (16 DACs max) Control bits for remote controlling the Source and Measure modules in the Sample Signal unit. Power control lines 1-2-3 (corresponding to the LEDs on the control panel) that have the following functions: 1 = Source-DAC power-on (LED=on) 2 = Measure-ADC power-on (LED=on) 3 = Measure-ADC activated (LED=on) , standby (LED= continuous off), calibration (LED=1second off) While 1+2 off saves 80% battery power M-ADC and S-DAC need a warm-up time of 10 minutes. Standby saves less power but needs no warm-up. Calibration is advised after warm-up. control: On the front there are two openings for ballpoint-operated-switches (emergency only): "Power-Down", when the uController is suspected from generating interference this switch puts it in a sleep-mode that can only be ended by: "Reset", used for leaving Power-Down mode or used when the uController is suspected from errors. 3.2.5 D5, Octal BiasDAC This module contains 8 (optional 16) 16bit DACs for creating gate-voltages to control the sample. To protect the sample the polarity of the DACs can be manually set (remote-read only), each group of 4 DACs has a polarity switch on the front. DAC values are send to the uController module (D4) by the user PC [4.3] ,the uController sets the DACs via optocouplers . The DAC-output signals leave this module and the SDU through a shielded cable and go to the Summing module (S2) in the SSU. In this module the user can place additional summing, filtering and dividing for the BiasDAC-voltages and connect the results to the (lemo) connectors on the front. From there they can be patched to the matrix unit. The BiasDACs have a common ground being the main-battery supply unless one of the following changes is made: -A second battery unit is connected to "A" on the battery panel to give DAC 1-8 a floating supply (set DAC-battery-switch to "A-floating") -If DAC 9-16 are present: A third battery unit is connected to "B" on the battery panel to give DAC 9-16 a floating supply (set DAC-battery-switch to "B-floating") -The ISO Voltage Source (S1) is used to isolate one DAC, selected in the Summing module. Output impedance of the DACs is 1kohm (1%). The output voltage range of the DACs is 0..4V , 0..-4V, -2..+2V depending on the setting of the polarity switch. The DAC values to send remain 0..65535 (16bit) ,the corresponding output voltage can be calculated by (software) reading the polarity switch status. A higher output voltage can be obtained by using the Iso-Voltage Source [3.1]. It is important to notice that the output voltage CHANGES (up to 8V) when the polarity switch is operated, this is also why polarity can not be software controlled to prevent errors destroying the sample. On power-up the uController checks the Polarity switch and loads a value in the DACs corresponding to zero volt output for that setting (this happens also when battery A or B is changed). 3.3 The User Data modules (fig.4) All modules in this unit are mains powered (switch on the back) using two shielded transformers separating the connected user PC from the user measurement equipment part. 3.3.1 U1, Option No plans up to now. 3.3.2 U2, Dual Source ADC This module forms (with module D2) a galvanic isolation between the user and the samplepart while transferring the user-generated excitation signals to the source modules. The module offers two differential inputs (+/-1Vtyp), the signals presented here are digitised and send to the Dual Source DAC (D2) in the Sample Data unit. For this these two modules have to be connected via one fiber which transports the digitised values of the source signals. A standard 18 bit commercial audio link is used (SP/DIF), giving dc to 20kHz bandwidth (500Hz when Bandwidth Select "low" is set on D2). An important aspect involved with the oversampling and digital filtering in these kind of systems is the (fixed) delay time between "in" (U2) and "out" (D2). The resulting delay (1.7mS) is higher then can be expected from the 20kHz bandwidth. For Lock-In amplifiers etc. this will cause no problems as the delay value is fixed (measured phasejitter at 30Hz was < 0.0001deg). In a standard connection setup (see [7] for options) input 1 is send to the x1 input of the sources, input 2 is send to the x0.01 input of the sources. This could be used for sweep and modulation purposes. Using seperate transmission, plus attenuation afterwords, for the modulation signal reduces noise and digitising artefacts . Manual offset calibration for this ADC is possible by pressing the calibrate switch on the back of the unit, the ADC inputs have to be grounded for this. Although this calibration is automatically performed on power-up it is advisible to repeat this once after a warm-up time of >10 minutes (the procedure is described in [6]). A (user) software controlled calibration is also possible, although there is no direct control from the uController to this unit. When the Measure ADC in the SDunit is receiving a calibration command from the user PC (via the uController [4.3]) the Measure DAC in the UDunit detects this via a short signal mute and automatically sends a calibration bit to the neighbouring Source ADC. As a result both ADCs are calibrated using one command. input: front: 2x BNC (differential: outer=50ê to ground inner= 1Mê to ground) signal: +/-1Vtyp +/-2Vmax output: front: LED indication for input overload (pos/neg) on each channel. output: front: Opto out, fiber to module D2 in the Sample Data unit. controls: back: calibration command bit from U3 (see [6]) back: Calibration switch (manual) on the back of the unit. 3.3.3 U3, the Dual Measure DAC This module forms (with module D3) a galvanic isolation between the user and the samplepart while transferring the measured signals to the User Data unit. The module offers two output signals (+/-1Vtyp) that are a copy of the signals connected at the inputs of the Dual Measure ADC (D3) in the Sample Data unit. For this these two modules have to be connected via one fiber which transports the digitised values of the source signals. A standard 18 bit commercial audio link is used (SP/DIF), giving dc to 20kHz bandwidth (500Hz when Bandwidth Select "low" is set). An important aspect involved with the oversampling and digital filtering in these kind of systems is the (fixed) delay time between "in" (D3) and "out" (U3). The resulting delay (1.7mS) is higher then can be expected from the 20kHz bandwidth. For Lock-In amplifiers etc. this will cause no problems as the delay value is fixed (measured phasejitter at 30Hz was < 0.0001deg). The overload LEDs on the front panel function as an indication for overload at the measure ADC inputs. input: front: Opto in, fiber from module D3 output: front: Receive error, LED indication for fiberdata errors output: front: BNC Voltage outputs 1 and 2 (Rout= 100ohm) representing the output signals from measure modules in the SSU levels: +/-1Vtyp, +/-3Vmax output: front: LED indication for overload (pos/neg) on each channel. control: Bandwidth select High/Low High= 20kHz Low= 500Hz (-6dB/oct slope) giving noisereduction 3.3.4 U4, the Opto/RS232 module This module forms a galvanic isolation between the user and the samplepart while transferring control signals between the uController module (D4) and the PC RS232 port. For this the two modules (D4,U4) have to be connected via two fibers and the PC RS232 port (comport) has to be connected to U4 using a standard serial cable. A separate mains supply is used for this digital unit to prevent computer interference entering the source-ADC or measure-DAC in this unit. 4 Manual & Remote control Manual control is possible for all functions except BiasDac-value setting and power-control. When operating the switches the uController is not involved and, when suspected from interference, could be removed (or set in sleep mode [3.2.4]). Remote control can be enabled/disabled for every module in the Sample Signal unit to prevent software errors corrupting the sample. As an example the current source could be set fixed to 0..100nA while the Voltage measurement could be remote controlled. 4.1 The Manual Control Panel (fig.5) All switches on this panel operate one or more relais in the modules of the SS unit. The panel is divided in three sections: Power, Source control, Measurement control. Every source and measurement module on this panel has a switch setting labelled "Remote" ,only at this setting computer control is enabled. Before switching to this setting read [4.3] first. The power section contains LEDs showing: -the presence of the main battery (pos and neg) the actual voltage can be measured by software [4.3]. -the status of the BiasDAC supply. When the LEDs are not lit supply voltage is too low (module might still work) or supply batteries are not connected or the Bias-DAC-Supply Coupling switch (Battery panel) is in the wrong position. Depending on the setting of this switch supply A and B can be the main battery or a seperate one. -The status of the power control lines, described in [3.2.4], 1 = Source-DAC power-on (LED=on) 2 = Measure-ADC power-on (LED=on) 3 = Measure-ADC activated (LED=on) , standby (LED= continuous off), calibration (LED=1second off) In a normal operating situation all power-LEDs should be lit. The Source section starts with the Iso-Voltage-Source (S1) used for creating an isolated Biasvoltage. The selectable BiasDAC is chosen in the Summing module (S2). The switch functions and module applications are described in [3.1.1]. The Voltage-Source (S3) [3.1.3] is used for V-I measurements and is controlled by the input voltages at Source-ADC1 (sweep) and Source-ADC2 (modulation) located in the User Data unit. The Current-Source (S4) [3.1.4] is used for I-V measurements and is also controlled by the input voltages of Source-ADC1 (sweep) and Source-ADC2 (modulation) located in the User Data unit. A +/-1V signal at ADC1 gives 100% of the selected source range value, for the ADC2 input this is 1% . To prevent excessive noise or digitalisation effects it is good practice to set the range of the source module in use for an equivalent input signal level of 100mV...1V. Clipping occurs at +/-2V input level. The current source overload LED indicates a voltage compliance limitation, the sample resistance is too high for the given current. The described signal flow of Source-DAC1 and 2 that is also shown on the panel can optionally be changed [7], when this is done it is advised to show this on the panel (sticker with routing). The Measure section contains an I-V converter called Current Measurement [3.1.6] and a differential amplifier called Voltage Measurement [3.1.7] When the input select switches on the Measurement ADC (D3) are set to intern the measure modules feed their output signals to Measure ADC1 and 2 as shown on the panel. The Measure-ADC and DAC overload LEDs indicate clipping by ac (+ and -) or dc (+ or -) for each channel. The resulting output voltages can be monitored at the Measure-DAC outputs in the User Data unit. To prevent excessive noise or digitalisation effects it is good practice to raise the gain of the measurement module in use until an output signal level of 100mV...1V is obtained. Clipping occurs at +/-3V output level. 4.2 Remote control hardware setup The Sample Data unit contains a uController module [3.2.4] that is connected to the Opto/RS232 module [3.3.4] in the User Data unit via 2 fibers. The Opto/RS232 module can be connected to the serial communication port (comport) of a Personal Computer using a standard serial connection cable. The PC comport should be set to send and receive on a baudrate of 19.2k, wordlength 8bits, parity check on, even parity. The uController module contains a send/reveive LED to check transmission. To reduce digital interference and to prevent PC loading, commands send to the uController are not echoed back to the PC in the default setting (echo mode can be enabled, see [4.3]). A good way to check valid transmission on starting up is to ask for the uController version, copy the string "UCVER;" to the comport and a string "SMS VERSION 2.0 6-07-1997" (or higher) should be returned. A check on this number is also advisable to prevent future (different) developments in uController operating systems to conflict with PC software based on earlier versions. During a measurement transmission can also be checked by asking for the battery voltage (which is also usefull then). 4.3 Remote control commands and software Remote control can be enabled/disabled on the control panel (SDunit) WARNING !!: when switching to "remote" the settings of the corresponding module are changed to those present in the control memory (last send remote commands). To prevent sudden unwanted settings it is advised to send a full set of commands for the modules in use (to update memory) BEFORE switching to remote. On power-up control memory resets to the default values listed at the end of this section. Software commands are listed on the two following pages. Commands are consisting of a ASCII string that has to be send to the serial comport of the PC. This action can be performed by almost all programs and languages (Basic,Pascal,C,Labview etc.). Existing data-aquisition software could be updated with the given commands. Not all commands need to be implemented. For a minimal setup only a BiasDAC-control program is needed, a battery check routine is advised. The remaining functions can be manual controlled. 5 Performing measurements 5.1 Start up & connections Yesterday: Start with checking for a fully charged battery-unit and make sure to have a second one already connected to the charger (charging could take 24 hours). Today: Set all the switches on the Matrix-module (in the SSU) to a safe (ground) setting. Verify that the modules in the SSU are the appropriate versions and are placed in the preferred slots, standard: module S1 > slot Sa, S2>Sb, S3>Sc, S4>Sd, M1>Ma, M2>Mb . Optional modules (like:M1a) can have other functions (see [7]). The Summing module (S2) contains all the user-selected control routing, so this module should be replaced or checked. Set the two source switches on the control panel (SDunit) in the "mute" position. Switch mains power on at the User Data unit. Connect a full battery unit to the main battery connector of the battery panel (SDunit) and switch on the power switch next to it. After this control transmission can be tested by checking the battery voltage using the PC ( 6.5V for a full battery). For optimal dc-performance the Dual Measure-ADC (D3) and the Dual Source-ADC should get a calibrate command after a warm up of at least 10 min, for this all four inputs need to be grounded, see [6]. Signal transmission can be tested by connecting the Voltage Source output to the Voltage Measurement input and the Current Source output to the Current Measurement input (using 2 lemo-lemo cables). Set all control panel tumblerswitches in the upright position. Set the Voltage Source to 100mV/V, Voltage Measurement 10V/V, Current Source to 1uA/V, Current Measurement 1MV/A, . 1V on input 1 of the Source ADC (U2) should give 1V on output 1 (U3) 1V on input 2 should give 1V on output 2 (U3). To reduce noise the bandwidth select on both DAC modules (D2,U2) can be set to "low" when using only frequencies below 500Hz. 5.2 Performing I-V measurements Although detailed specifications are given in [8] and module descriptions are given in [3], some general remarks can be made about a setup using the Current Source / Voltage Measurement combination. Gain: A nominal 1V signal should be applied at the inputs of U2. The desired excitation current can be set by the A/V switch, in combination with the x1 / x0.1 switch. The Voltage Measurement module gain should be set to result in a typical 100mV...1V signal on the output of U3. Bandwidth: This is mainly determined by the Sample and lead resistance R in combination with the capacitance C of the PI-filters in the matrix module. The resulting bandwidth will be 1/2ãRC (C=3nF for the PI- filters or 300pF for the optional low capacitance lines [7]). Noise: For high values of sample resistance ( > 1/Current Source range) the noise current of the source will dominate. For lower resistance values the voltage noise of the Voltage Measurement module is dominant. For low gain settings ( <100x ) the ADC-DAC noise dominates. Offset: For high values of sample resistance ( > 1/Current Source range) the offset current of the source will dominate, at the highest values (>100Mohm) the input bias current of the Voltage measurement also can become significant. For lower resistance values the offset voltage of the Voltage Measurement module is dominant. Interference: Operating the A/V switch or the "mute" switch results in relais-switchimg at the sample-connected output nodes. The relais are optimised for low signal measurement applications but some effect may be noticable when operated during a measurement. All other switches for both modules are free from this effect and could be operated during a measurement. A simulation program is available (?) to estimate gain, bandwidth and noise for all settings and options (R.Schouten). 5.3 Performing V-I measurements Although detailed specifications are given in [8] and module descriptions are given in [3], some general remarks can be made about a setup using the Voltage Source / Current Measurement combination. Gain: A nominal 1V signal should be applied at the inputs of U2. The desired excitation voltage can be set by the V/V switch, in combination with the x1 / x0.1 switch. The Current Measurement module gain (V/A) in combination with the PostGain (x1/x100) should be set to result in a typical 100mV...1V signal on the output of U3. Gain from the Current Measurement module can be influenced by its finite input resistance described in [3.1.6]. Bandwidth: This is mainly determined by the finite input resistance of the Current Measurement module in combination with the capacitance of the PI-filters in the matrix module. At the most sensitive range setting and using low noise mode signal bandwidth is lowest at 30Hz for M1a. At all other settings a higher bandwidth is obtained [8]. Optional modules offer higher bandwidths up to 20kHz at the cost of higher noise levels. Noise: The Voltage Source is unlikely to be dominant, at the highest (100mV/V) setting the noise from the controlling ADC-DAC can be noticed, for the lower ranges a passive resistive divider is used (1k-100-10 ohm). Noise contribution from the Current Measurement module is depending on several settings, described in [3.1.6]. Offset: The Voltage Source contributes less than: 1mV x(V/V setting)x(attenuator) The Current Measurement can be trimmed to less than 2uV. Interference: Operating the V/V switch, the V/A switch or the "mute" and "zero" switches results in relais-switchimg at the sample-connected in/output nodes. The relais are optimised for low signal measurement applications but some effect may be noticable. All other switches for both modules are free from this effect. A simulation program is available(?) to estimate gain, bandwidth and noise for all settings and options (R.Schouten). 6 User operated calibrations & checks 8 Specifications All specifications are valid after at least 15 minutes warm-up at 20..25 celcius. All non-lineairity, gain-error ,offset and bandwidth spec's are related to a 2Vpp signal level at the in/outputs of the User Data Unit. Noise levels are measured using battery supply. System performance: I-V measurement Current Source S4 + Voltage Measurement M2 Bandwidth (Hz) +/- 20% : { }=Rload +/-10% 10mA 1mA 100uA 10uA 1uA 100nA 10nA x1 20k 20k 10k 1k* 100* 10* 1* {100} {1k} {10k} {100k} {1M} {10M} {100M} x10 20k 20k 15k 10k 1k* 100* 10* {10} {100} {1k} {10k} {100k} {1M} {10M} x100 20k 20k 20k 15k 10k 1k* 100* {1} {10} {100} {1k} {10k} {100k} {1M} x1E3 20k 20k 20k 15k 10k 5k 1k* {0.1} {1} {10} {100} {1k} {10k} {100k} x1E4 20k 20k 20k 15k 10k 5k 1k {0.01} {0.1} {1} {10} {100} {1k} {10k} notes: * A high value sample resistance in combination with the 3nF PI-filters (Matrix module) results in a low effective bandwidth [5.2], this limit is not set by the electronics, but by the I-V concept, a V-I measurement gives a higher bandwidth for this. -The attenuator setting x1. x0.1 has no influence on bandwidth Output offset (mV) : { }=Rload +/-10% 10mA 1mA 100uA 10uA 1uA 100nA 10nA x1 <1 <1 <1 <1 <1 <1 <1 {100} {1k} {10k} {100k} {1M} {10M} {100M} x10 <1 <1 <1 <1 <1 <1 <1 {10} {100} {1k} {10k} {100k} {1M} {10M} x100 <2 <2 <2 <2 <2 <2 <2 {1} {10} {100} {1k} {10k} {100k} {1M} x1E3 <5 <5 <5 <5 <5 <5 <5 {0.1} {1} {10} {100} {1k} {10k} {100k} x1E4 <50 <50 <50 <50 <50 <50 <50 {0.01} {0.1} {1} {10} {100} {1k} {10k} notes: Output noise (uV/ûHz in signal bandwidth)) : { }=Rload +/-10% 10mA 1mA 100uA 10uA 1uA 100nA 10nA x1 <1 <1 <1 <1 <1 <2 <5 {100} {1k} {10k} {100k} {1M} {10M} {100M} x10 <1 <1 <1 <1 <1 <2 <5 {10} {100} {1k} {10k} {100k} {1M} {10M} x100 <1 <2 <2 <2 <2 <5 <15 {1} {10} {100} {1k} {10k} {100k} {1M} x1E3 <10 <10 <10 <10 <10 <20 <60 {0.1} {1} {10} {100} {1k} {10k} {100k} x1E4 <50 <50 <50 <50 <50 <70 <150 {0.01} {0.1} {1} {10} {100} {1k} {10k} notes: S1 Iso-Voltage Source