This document is a reference manual for the AIOUSB library. It describes all the API functions, as well as global variables and constants. It also provides some guidance as to how to use AIOUSB. This document is intended to serve both the Linux and Windows implementations of AIOUSB.
The API functions are grouped by category, and within each group they are sorted in alphabetical order.
Everything is described in this frame, so the best way to navigate around the document is by using the menu at left.
Where appropriate, Linux- or Windows-specific notes are included.
Names of files, program statements, functions, variables and constants are highlighted using a different text style than the regular body text.
Instead of replicating the same information dozens of times, commonly used variables and constants are described in a single place.
Below is an example of a minimalist C++ program for Linux that demonstrates how to properly initialize AIOUSB, query the bus for devices, query an individual device for its product ID and name and then terminate use of AIOUSB. If AIOUSB and libusb are properly installed, you should be able to copy this sample program from this document, paste it into a file named test.cpp and compile it using the command shown below. This program uses the first ACCES device it finds on the bus. A "real" application would probably be looking for devices of a particular type, which can be determined from the product ID.
/* * compile with: g++ test.cpp -laiousbcpp -lusb-1.0 -o test */ #include <aiousb.h> #include <stdio.h> using namespace AIOUSB; int main( int argc, char **argv ) { unsigned long result = AIOUSB_Init(); if( result == AIOUSB_SUCCESS ) { const int MAX_NAME_SIZE = 20; char name[ MAX_NAME_SIZE + 2 ]; unsigned long productID , nameSize = MAX_NAME_SIZE; result = QueryDeviceInfo( diFirst , &productID, &nameSize, name, NULL, NULL ); if( result == AIOUSB_SUCCESS ) { name[ nameSize ] = 0; printf( "Found a device with product ID %#06x and name '%s'\n" , ( unsigned ) productID, name ); } else printf( "Error '%s' querying device\n" , AIOUSB_GetResultCodeAsString( result ) ); AIOUSB_Exit(); } // if( result ... return ( int ) result; } // main()
The above sample program can be ported to Windows by doing the following:
Omitting the statement using namespace AIOUSB;
Changing the result code names from AIOUSB_* to ERROR_*
Removing the calls to AIOUSB_Init() and AIOUSB_Exit()
Removing the call to AIOUSB_GetResultCodeAsString() and printing the result code as a number instead of a string
Using a different command line - one appropriate for Windows - to compile the program
Clears the streaming FIFO, using one of the method codes below.
Analog input, buffered digital I/O; Linux, Windows
DeviceIndex - a standard device index
Method - the method to use when clearing the FIFO. May be one of:
CLEAR_FIFO_METHOD_IMMEDIATE (0) - Clear FIFO as soon as command received (and disable auto-clear)
CLEAR_FIFO_METHOD_AUTO (1) - Enable auto-clear FIFO every falling edge of DIO port D bit 1 (on digital boards, analog boards treat as 0)
CLEAR_FIFO_METHOD_IMMEDIATE_AND_ABORT (5) - Clear FIFO as soon as command received (and disable auto-clear), and abort stream
CLEAR_FIFO_METHOD_WAIT (86) - Clear FIFO and wait for it to be emptied
Linux: Use the named constants listed above.
Windows: Use the numeric values shown in parenthesis.
A standard result code
Sets the streaming block size.
Analog input, buffered digital I/O; Linux, Windows
DeviceIndex - a standard device index
BlockSize - the streaming block size you wish to set. For DIO streaming, this will get rounded up to the next multiple of 256. For A/D streaming, this will get rounded up to the next multiple of 512.
A standard result code
Closes handles and clears records of unplugged devices.
All products; Linux, Windows
A standard result code
Reads data from the custom programming area of the device EEPROM.
All products; Linux, Windows
DeviceIndex - a standard device index
StartAddress - number from 0x000 to 0x1FF of the first custom EEPROM byte you wish to read from.
DataSize - pointer to a variable holding the number of custom EEPROM bytes to read. The last custom EEPROM byte is 0x1FF, so StartAddress plus DataSize can't be greater than 0x200.
Data - pointer to the start of a block of bytes to fill with data read from the custom EEPROM area.
A standard result code
Writes data to the custom programming area of the device EEPROM.
All products; Linux, Windows
DeviceIndex - a standard device index
StartAddress - number from 0x000 to 0x1FF of the first custom EEPROM byte you wish to write to.
DataSize - number of custom EEPROM bytes to write. The last custom EEPROM byte is 0x1FF, so StartAddress plus DataSize can't be greater than 0x200.
Data - pointer to the start of a block of bytes to write to the custom EEPROM area.
A standard result code
Searches the bus for an ACCES device with the specified serial number. In the unlikely event that multiple devices have the same serial number, GetDeviceBySerialNumber() returns a device index to the first such device found.
Unlike GetDevices(), which clears and rebuilds the internal list of devices, GetDeviceBySerialNumber() is "non-destructive" and simply searches the existing internal list.
Linux: also see AIOUSB_GetDeviceByProductID().
All products; Linux, Windows
pSerialNumber - pointer to an 8-byte (64-bit) value containing the serial number to search for
A standard device index if the device was found, or diNone if no device was found with the specified serial number
Gets a "list" of all the ACCES devices found on the USB bus. This "list" is returned in the form of a 32-bit device mask. Each bit set to a "1" indicates an ACCES device was detected at the device index corresponding to the set bit number. For example, if the return value is 0x00000104 then DeviceIndex #2 and #8 correspond to devices that can be controlled by this driver. Returns 0 if no devices were found, which may mean the driver is not installed properly. This function does not return a single device index, but a pattern of bits indicating all the detected devices.
Also see GetDeviceBySerialNumber().
Linux: also see AIOUSB_GetDeviceByProductID().
Windows: A limitation in the current .SYS file prevents detection of more than 32 ACCES devices connected to one computer simultaneously. Let us know if this is of any concern for your application.
All products; Linux, Windows
A 32-bit device mask (0 if no devices found).
Gets the device's unique serial number.
All products; Linux, Windows
DeviceIndex - a standard device index
pSerialNumber - pointer to an 8-byte (64-bit) value to fill with the serial number.
A standard result code
Gets the device's specific properties.
Linux: also see AIOUSB_GetDeviceProperties().
All products; Linux, Windows
DeviceIndex - a standard device index
pPID - pointer to 32-bit variable in which to store the product ID of the device.
pNameSize - pointer to 32-bit value which specifies the size of the pName buffer. It gets set to the number of bytes actually required by the name. If pNameSize is too small to accommodate the entire name, the name will be truncated to fit within your buffer.
pName - pointer to char[] buffer in which to receive the device name. This is an array of characters, not a null-terminated string. Length of string is passed back via pNameSize.
pDIOBytes - pointer to 32-bit value which gets set to how many bytes of DIO the card supports.
pCounters - pointer to 32-bit value which gets set to how many 8254-compatible counters are available.
A standard result code
Converts a possibly abstract device index into an absolute device index.
All products; Linux, Windows
DeviceIndex - the device index you want to resolve. This index may be a standard device index, in which case the same index will be returned, or one of these special constants:
diFirst (0xFFFFFFFEul) - obtains the index of the first ACCES device found on the bus
diOnly (0xFFFFFFFDul) - obtains the index of the only ACCES device found on the bus, meaning that there must be only one ACCES device on the bus
A standard device index or diNone (0xFFFFFFFFul) if a valid device index cannot be returned.
Shuts down the AIOUSB API prior to exiting the program. Must be called if the call to AIOUSB_Init() was successful. No AIOUSB function calls may be made after AIOUSB_Exit() has been called.
All products; Linux
Returns the current timeout setting (in milliseconds) for USB communications.
All products; Linux
DeviceIndex - a standard device index
Current timeout setting (in milliseconds)
Like GetDevices() and GetDeviceBySerialNumber(), AIOUSB_GetDeviceByProductID() searches for devices on the bus. It returns an array of integers containing the list of devices found that match the search criteria. This array consists of <device index>-<product ID> pairs, making it easy to identify all the devices of interest.
Unlike GetDevices(), which clears and rebuilds the internal list of devices, AIOUSB_GetDeviceByProductID() is "non-destructive" and simply searches the existing internal list.
If you are only interested in a single product ID, set minProductID and maxProductID to the same value.
If you want a list of all product IDs, set minProductID to 0 and maxProductID to 0xffff.
All products; Linux
minProductID - the minimum product ID to search for (0 - 0xffff)
maxProductID - the maximum product ID to search for (0 - 0xffff; must be greater than or equal to minProductID)
maxDevices - the maximum number of devices to return in deviceList (1 - 127)
deviceList - pointer to an array of integers in which the found device indexes and product IDs will be returned. The size of this array must be equal to 1 + maxDevices * 2. The array will contain <device index>-<product ID> pairs, which is why it must be twice the size of the number of devices requested. The actual number of devices found is returned in the first integer of the array, index [0]. The <device index>-<product ID> pairs follow, beginning with indexes [1] and [2], respectively.
A standard result code
Gets a richer set of device-specific properties. This function obtains a superset of the information obtained by QueryDeviceInfo().
All products; Linux
DeviceIndex - a standard device index
properties - pointer to a DeviceProperties structure in which the device properties will be returned.
A standard result code
Returns the string representation of an AIOUSB_* result code, useful mainly for debugging purposes.
All products; Linux; debugging
result - a standard result code
Pointer to a null-terminated string representation of result
Returns the AIOUSB module version number as a string with the form, "1.78".
All products; Linux
Pointer to a null-terminated version number string
Returns the AIOUSB module version date as a string with the form, "15 November 2009".
All products; Linux
Pointer to a null-terminated version date string
Initializes AIOUSB API. Must be called before using any other functions in the AIOUSB module. Automatically scans the USB bus and populates the list of ACCES devices found. If this call is successful, then AIOUSB_Exit() must be called before the program exits.
All products; Linux
A standard result code
A crude means to dump a list of all ACCES devices found on the USB bus.
All products; Linux; debugging
Sets the timeout (in milliseconds) for USB communications. The default setting is 5,000 ms.
All products; Linux
DeviceIndex - a standard device index
timeout - the new timeout setting (in milliseconds)
A standard result code
Specifies a clock frequency (in Hertz) for some of the timer-driven functions. The default setting is 1 Hz.
All products; Linux
DeviceIndex - a standard device index
clockHz - the new clock frequency in Hertz
A standard result code
Gets the configuration of the digital I/O ports.
Digital inputs and outputs; Linux, Windows
DeviceIndex - a standard device index
pOutMask - a pointer to the first element of an array of bytes, one byte per 8 ports or fraction. Each bit in the array will be set to "1" if the corresponding port is an output, or "0" if it's an input.
pTristateMask - a pointer to the first element of an array of bytes, one byte per 8 tristate groups or fraction. Each bit in the array will be set to "1" if the corresponding tristate group is in tristate (high-impedance) mode, or a "0" if not.
A standard result code
Configures the digital I/O ports.
Digital inputs and outputs; Linux, Windows
DeviceIndex - a standard device index
bTristate - AIOUSB_TRUE, TRUE causes all bits on the device to enter tristate (high-impedance) mode; AIOUSB_FALSE, FALSE removes the tristate. The tristate is changed after the remainder of the configuration has occurred. All devices with this feature power-on in the "tristate" mode at this time.
pOutMask - a pointer to the first element of an array of bytes, one byte per 8 ports or fraction. Each "1" bit in the array indicates that the corresponding byte of the device is an output. The number of bytes required by the output mask is equal to the number of ports, divided by 8 (one bit controls the direction of an entire port) and rounded up to the next whole value. For example, if the device has 4 ports, then 4 divided by 8 and rounded up is 1; that is, one mask byte is required (actually only the least significant 4 bits of the mask are required).
pData - a pointer to the first element of an array of bytes. Each byte is copied to the digital output ports on the device before the ports are taken out of tristate. Any bytes in the array associated with ports configured as input are ignored. The number of bytes required to store all the data for the device is equal to the number or ports, or the number of channels divided by 8 channels/bits per port/byte. For example, if the device has 96 channels, the number of ports/bytes is 96 divided by 8, or 12.
A standard result code
Configures the digital I/O ports.
Digital inputs and outputs; Linux, Windows
DeviceIndex - a standard device index
pOutMask - a pointer to the first element of an array of bytes, one byte per 8 ports or fraction. Each "1" bit in the array indicates that the corresponding byte of the device is an output. The number of bytes required by the output mask is equal to the number of ports, divided by 8 (one bit controls the direction of an entire port) and rounded up to the next whole value. For example, if the device has 4 ports, then 4 divided by 8 and rounded up is 1; that is, one mask byte is required (actually only the least significant 4 bits of the mask are required).
pData - a pointer to the first element of an array of bytes. Each byte is copied to the digital output ports on the device before the ports are taken out of tristate. Any bytes in the array associated with ports configured as input are ignored. The number of bytes required to store all the data for the device is equal to the number or ports, or the number of channels divided by 8 channels/bits per port/byte. For example, if the device has 96 channels, the number of ports/bytes is 96 divided by 8, or 12.
pTristateMask - a pointer to the first element of an array of bytes, one byte per 8 tristate groups or fraction. Each "1" bit in the array causes the corresponding tristate group to enter tristate (high-impedance) mode. A “0" bit removes the tristate. The tristate is changed after the remainder of the configuration has occurred. All devices with this feature power-on in the "tristate" mode at this time.
A standard result code
Reads a single bit from a digital I/O port.
Digital inputs and outputs; Linux, Windows
DeviceIndex - a standard device index
BitIndex - Number of the bit you wish to read (0-based)
pBuffer - pointer to a byte which will be set to zero or one based on the input bit. Data read from ports configured as output results in a "read-back" of the output.
A standard result code
Reads a single digital I/O port (8-bits).
Digital inputs and outputs; Linux, Windows
DeviceIndex - a standard device index
ByteIndex - Number of the byte you wish to read (0-based)
pBuffer - pointer to a byte in which the input byte will be stored. Data read from ports configured as output results in a "read-back" of the output.
A standard result code
Reads all the digital I/O ports in the device.
Digital inputs; Linux, Windows
DeviceIndex - a standard device index
Buffer - pointer to the first element of an array of bytes. Each port will be read, and the reading stored in the corresponding byte in the array. The number of bytes required to store all the data for the device is equal to the number or ports, or the number of channels divided by 8 channels/bits per port/byte. For example, if the device has 96 channels, the number of ports/bytes is 96 divided by 8, or 12.
A standard result code
Writes a single bit to a digital output port.
Digital outputs; Linux, Windows
DeviceIndex - a standard device index
BitIndex - Number of the bit you wish to change (0-based). Writes to bits configured as inputs are ignored.
bData - AIOUSB_TRUE, TRUE will set the bit to "1"; AIOUSB_FALSE, FALSE will clear the bit to "0".
A standard result code
Writes a single digital output port (8-bits).
Digital outputs; Linux, Windows
DeviceIndex - a standard device index
ByteIndex - Number of the byte you wish to change (0-based). Writes to bytes configured as inputs are ignored.
Data - The byte will be copied to the port outputs. Each set bit will cause the same port bit to be set to "1".
A standard result code
Writes all the digital output ports in the device.
Digital outputs; Linux, Windows
DeviceIndex - a standard device index
pData - pointer to the first element of an array of bytes. Each byte is copied to the corresponding output byte. Bytes written to ports configured as inputs are ignored. The number of bytes required to store all the data for the device is equal to the number or ports, or the number of channels divided by 8 channels/bits per port/byte. For example, if the device has 96 channels, the number of ports/bytes is 96 divided by 8, or 12.
A standard result code
Closes a digital I/O stream opened by a call to DIO_StreamOpen().
Buffered digital inputs and outputs; Linux, Windows
DeviceIndex - a standard device index
A standard result code
Read from, or write to a digital I/O stream opened by a call to DIO_StreamOpen().
Buffered digital inputs and outputs; Linux, Windows
DeviceIndex - a standard device index
FramePoints - number of 16-bit points you wish to stream
pFrameData - pointer to the beginning of the block of data you wish to stream
BytesTransferred - pointer to a variable that will receive the amount of data actually transferred, in bytes
A standard result code
Opens a digital I/O stream. When you are done using the stream, you must close it by calling DIO_StreamClose().
Buffered digital inputs and outputs; Linux, Windows
DeviceIndex - a standard device index
bIsRead - AIOUSB_TRUE, TRUE will open a stream for reading; AIOUSB_FALSE, FALSE will open a stream for writing
A standard result code
Sets the internal read/write clock speed of a digital I/O stream.
Buffered digital inputs and outputs; Linux, Windows
DeviceIndex - a standard device index
ReadClockHz - a pointer to an IEEE double-precision value indicating the desired frequency of an internal read clock. It will be changed to the actual frequency achieved. Use "0" for an external read clock.
WriteClockHz - a pointer to an IEEE double-precision value indicating the desired frequency of an internal write clock. It will be changed to the actual frequency achieved. Use "0" for an external write clock.
A standard result code
The 8254 programming details are not described in this manual. Programming details for the 8254 can be found in the document named CD/ChipDocs/8254.pdf.
Each of these functions is designed to operate in one of two addressing modes. The parameter BlockIndex refers to 8254 chips, each of which contains 3 “Counters”. CounterIndex refers to the counters inside the 8254s. In the primary addressing mode you specify the block and the counter. In the secondary addressing mode, you specify zero (0) for the block, and consider the counters to be addressed sequentially. That is, BlockIndex 3, CounterIndex 1 can also be addressed as BlockIndex 0, CounterIndex 10. The equation to determine the secondary or sequential CounterIndex given the primary or block values is as follows (they simply count consecutively):
CounterIndexsequential = BlockIndex * 3 + CounterIndexprimary
Please note, CounterIndex values associated with BlockIndex 0 are compatible with either addressing mode, there is no need to tell the driver which addressing mode you wish to use.
Loads a count value into a counter.
Counter timers; Linux, Windows
DeviceIndex - a standard device index
BlockIndex - number indicating which 8254 you wish to load
CounterIndex - number from 0-2 indicating which counter on the specified 8254 you wish to load (see note about counter addressing)
LoadValue - a number from 0 to 65535 which you wish loaded into the specified counter
A standard result code
Sets a counter mode.
Counter timers; Linux, Windows
DeviceIndex - a standard device index
BlockIndex - number indicating which 8254 you wish to configure
CounterIndex - number from 0-2 indicating which counter on the specified 8254 you wish to configure (see note about counter addressing)
Mode - a number from 0-5 specifying the 8254 counter mode
A standard result code
Sets a counter mode and loads a count value into the counter.
Counter timers; Linux, Windows
DeviceIndex - a standard device index
BlockIndex - number indicating which 8254 you wish to configure
CounterIndex - number from 0-2 indicating which counter on the specified 8254 you wish to configure (see note about counter addressing)
Mode - a number from 0-5 specifying the 8254 counter mode
LoadValue - a number from 0 to 65535 which you wish loaded into the specified counter
A standard result code
Reads a counter's current count value.
Counter timers; Linux, Windows
DeviceIndex - a standard device index
BlockIndex - number indicating which 8254 you wish to read
CounterIndex - number from 0-2 indicating which counter on the specified 8254 you wish to read (see note about counter addressing)
pReadValue - a pointer to a 16-bit integer in which will be stored the value latched and read from the specified counter
A standard result code
Reads the current count values of all the counters.
USB-CTR-15 only; Linux, Windows
DeviceIndex - a standard device index
pData - a pointer to the first of an array of 16-bit integers in which will be stored the values latched and read from the counters
A standard result code
Reads the current count values of all the counters.
USB-CTR-15 only; Linux, Windows
DeviceIndex - a standard device index
pData - a pointer to the first of an array of 16-bit integers in which will be stored the values latched and read from the counters. After the array in the pointer buffer is an additional 8-bit byte. This byte contains useful information when optimizing polling rates. If the value of the byte is “0", you’re looking at old data, and are reading faster than your gate signal is running.
A standard result code
Reads a counter's current count value, then sets a new mode and loads a new count value into the counter.
Counter timers; Linux, Windows
DeviceIndex - a standard device index
BlockIndex - number indicating which 8254 you wish to read, mode, and load
CounterIndex - number from 0-2 indicating which counter on the specified 8254 you wish to read, mode, and load (see note about counter addressing)
Mode - a number from 0-5 specifying the 8254 counter mode
LoadValue - a number from 0 to 65535 which you wish loaded into the specified counter
pReadValue - a pointer to a 16-bit integer in which will be stored the value latched and read from the specified counter. The reading is taken before the mode and load occur.
A standard result code
Reads a counter's current count value and status.
Counter timers; Linux, Windows
DeviceIndex - a standard device index
BlockIndex - number indicating which 8254 you wish to read
CounterIndex - number from 0-2 indicating which counter on the specified 8254 you wish to read (see note about counter addressing)
pReadValue - a pointer to a 16-bit integer in which will be stored the value latched and read from the specified counter.
pStatus - a pointer to an 8-bit byte in which will be stored the status latched and read from the specified counter.
A standard result code
Selects the counter to use as a gate in frequency measurement on other counters.
USB-CTR-15 only; Linux, Windows
DeviceIndex - a standard device index
GateIndex - number from 0-14 indicating which counter you wish to select as a gate (using sequential counter addressing)
A standard result code
Selects an output frequency for a counter block and starts the counters.
CTR_8254SelectGate() and CTR_8254ReadLatched() are used in measuring frequency. To measure frequency one must count pulses for a known duration. In simplest terms, the number of pulses that occur for 1 second translates directly to Hertz. In the USB-CTR-15 and other supported devices, you can create a known duration by configuring one counter to act as a “gating” signal for any collection of other counters. The other “measurement” counters will only count during the “high” side of the gate signal, which we can control.
So, to measure frequency you:
Create a gate signal of known duration
Connect this gating signal to the gate pins of all the “measurement” counters
Call CTR_8254SelectGate() to tell the board which counter is generating that gate
Call CTR_8254ReadLatched() periodically to read the latched count values from all the “measurement” counters.
In practice, it may not be possible to generate a gating signal of sufficient duration from a single counter. Simply concatenate two or more counters into a series, or daisy-chain, and use the last counter’s output as your gating signal. This last counter in the chain should be reported as the “gate source” using CTR_8254SelectGate().
Once a value has been read from a counter using the CTR_8254ReadLatched() call, it can be translated into actual Hz by dividing the count value returned by the high-side-duration of the gating signal, in seconds. For example, if your gate is configured for 10Hz, the high-side lasts 0.05seconds; if you read 1324 counts via the CTR_8254ReadLatched() call, the frequency would be “1324 / 0.05", or 26.48KHz.
Counter timers; Linux, Windows
DeviceIndex - a standard device index
BlockIndex - number indicating which 8254 you wish to output a output a frequency
pHz - pointer to a double precision IEEE floating point number containing the desired output frequency (in Hertz). This variable is set by the driver to the actual frequency that will be output, as limited by the device’s capabilities.
A standard result code
Sets the A/D trigger and calibration modes. The USB-AI12-16E does not support calibration, and will use CalMode 0.
Linux: there is a parallel set of convenience functions for configuring the A/D. Refer to AIOUSB_SetTriggerMode() and AIOUSB_SetCalMode() for more information.
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
TriggerMode - byte indicating which A/D trigger source to use, see the manual for details. Also sets the clock source for counter 0.
Linux: You may use a bitwise OR of the AD_TRIGGER_* values.
CalMode - byte indicating which A/D source to use. May be one of:
00 hex - actual inputs
01 hex - calibration ground reference
03 hex - calibration high reference
Other values will cause the function to fail and return an error result.
Linux: You may use one of the AD_CAL_MODE_* values.
A standard result code
Starts a large A/D acquisition operation and returns immediately. A result code of AIOUSB_SUCCESS / ERROR_SUCCESS indicates that A/D data is being acquired in the background, and the buffer should not be deallocated or moved. Use ADC_BulkPoll() to query the status of this background operation.
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
BufSize - the size, in bytes, of the buffer to receive the data
pBuf - a pointer to the beginning of the buffer to receive the data
A standard result code
Queries the status of a background A/D acquisition operation initiated by a call to ADC_BulkAcquire(), returning the number of bytes yet to be transferred.
Any data that has been taken is available in the buffer, starting from the beginning. For example, if ADC_BulkAcquire() was called to take 1024 MB of data, and ADC_BulkPoll() indicates 768 MB remains to be taken, the first 256 MB of data is available.
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
BytesLeft - a pointer to a variable which will be set to the number of bytes of A/D data remaining to be taken
A standard result code
Reads a voltage from a single channel, averaging any oversamples, if so configured. This function is easy to use, but relatively slow and often can’t achieve more than 100Hz sampling rates.
Linux: the channel scan range is temporarily changed to include only ChannelIndex. This feature ensures that ADC_GetChannelV() will always return valid data. It also improves performance by reading just the channel requested instead of the entire channel scan range (see ADC_SetScanLimits()).
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
ChannelIndex - number indicating which channel to read
pBuf - a pointer to a double precision IEEE floating point number which will receive the value read (in Volts)
A standard result code
Gets the current A/D configuration from the device. The format of the configuration data returned is described in ADC_SetConfig().
Linux: there is a parallel set of convenience functions for configuring the A/D. Refer to ADConfigBlock for more information.
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
pConfigBuf - a pointer to the first of an array in which configuration bytes will be returned
ConfigBufSize - a pointer to a variable holding the number of configuration bytes to read. Will be set to the number of configuration bytes read.
A standard result code
Takes one scan of A/D data, from start-channel to end-channel, averaging any oversamples for each channel, if so configured. The input array pBuf must contain one entry per A/D channel on the board, even though only entries start-channel through end-channel are altered. This function is easy to use, but relatively slow and often can’t achieve more than several hundred Hertz sampling rates.
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
pBuf - a pointer to the first of an array of 16-bit integers which will each receive the value from one channel (in A/D counts)
A standard result code
Takes one scan of A/D data, from start-channel to end-channel, averaging any oversamples for each channel, if so configured. The input array pBuf must contain one entry per A/D channel on the board, even though only entries start-channel through end-channel are altered. This function is easy to use, but relatively slow and often can’t achieve more than several hundred Hertz sampling rates.
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
pBuf - a pointer to the first of an array of double precision IEEE floating point numbers which will each receive the value read from one channel (in Volts)
A standard result code
Sets the A/D configuration in the device and/or loads a calibration table into the device. If pConfigBuf or ConfigBufSize are NULL (0), the A/D will not be configured. If CalFileName is NULL (0), the calibration table will not be loaded. If all three parameters are NULL (0), no action will be performed.
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
pConfigBuf - a pointer to the first of an array of configuration bytes (see ADC_SetConfig())
ConfigBufSize - a pointer to a variable holding the number of configuration bytes to write (see ADC_SetConfig())
CalFileName - the name of a calibration file to load (see ADC_SetCal())
A standard result code
Queries the device to determine if it supports A/D calibration. A result code of AIOUSB_SUCCESS / ERROR_SUCCESS indicates that the device does support A/D calibration.
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
A standard result code
Sets the range and "ended" mode for an A/D channel.
Linux: there is a parallel set of convenience functions for configuring the A/D. Refer to AIOUSB_SetGainCode() and AIOUSB_SetDifferentialMode() for more information.
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
ADChannel - number from 0-15 indicating an A/D channel on the device
GainCode - the gain code (see ADC_SetConfig())
Linux: You may use one of the AD_GAIN_CODE_* values.
bSingleEnded - for channels 0-7, use AIOUSB_TRUE, TRUE for single-ended mode, or use AIOUSB_FALSE, FALSE to pair it with the respective channel 8-15 in differential mode. For channels 8-15, use AIOUSB_FALSE, FALSE.
A standard result code
Sets the range and "ended" mode for all A/D channels.
Linux: there is a parallel set of convenience functions for configuring the A/D. Refer to AIOUSB_SetAllGainCodeAndDiffMode() for more information.
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
pGainCodes - a pointer to the first of an array of 16 bytes, each of which contains a gain code. This does not include single-ended/differential configuration. To configure single-ended/differential mode on a per-channel basis, use ADC_Range1() or ADC_SetConfig().
Linux: You may use one of the AD_GAIN_CODE_* values for each of the gain codes.
bSingleEnded - Use AIOUSB_TRUE, TRUE for 16-channel single-ended mode, or use AIOUSB_FALSE, FALSE for 8-channel differential mode.
A standard result code
Loads a calibration table into the A/D. The calibration table can either be loaded from a file or generated automaticallly, depending on the file name passed to ADC_SetCal().
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
CalFileName - this can be a full path to a calibration file or simply a file name in the current directory. To generate a calibration table automatically, one of these special command strings (case-sensitive) may be used instead of an actual file name:
":AUTO:" - automatically calibrates the A/D using internal ground and reference measurements and creates an appropriate calibration table, which is loaded into the A/D
":NONE:" - generates a default, uncalibrated table, which is loaded into the A/D
A standard result code
Sets the A/D configuration in the device.
Linux: there is a parallel set of convenience functions for configuring the A/D. Refer to ADConfigBlock for more information.
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
pConfigBuf - a pointer to the first of an array of configuration bytes (described below)
ConfigBufSize - a pointer to a variable holding the number of configuration bytes to write. Will be set to the number of configuration bytes written.
| Configuration bytes for analog input boards (such as USB-AI16-16 family) | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Byte 00 | Bytes 01-0E | 0F | 10 | 11 | 12 | 13 | 14 | ||
| Channel 0 Gain Code | Channels 1-14 Gain Code | Channel 15 Gain Code | Calibration Mode | Trigger & Counter Clock | Scan Start & End Channel | Oversample | MUX Scan Start & End Channel | ||
| A configuration of all zeroes is close to an "ordinary" use; you'll likely want to set external or timer trigger, and start and end channels. | |||||||||
| Channel 0-15 Gain Codes (bytes 00-0F) | |||||||||
| Gain Code | 00 | 01 | 02 | 03 | 04 | 05 | 06 | 07 | |
| Range | 0-10V | ±10V | 0-5V | ±5V | 0-2V | ±2V | 0-1V | ±1V | |
| Add 08 to the gain code for channel 0-7 to pair it with the respective channel 8-15 in differential mode. | |||||||||
| Calibration Mode (byte 10) | |||||||||
| Cal. Mode | 00 | 01 | 03 | ||||||
| Effect | Acquire Normal Data | Acquire Cal. Ground (0.0114V @ 0-10V) | Acquire Cal. Reference (9.9V @ 0-10V) | ||||||
| Trigger & Counter Clock (byte 11) | |||||||||
| Bit | 7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 | |
| Value | Reserved, use 0 | CTR0 EXT | Falling Edge | Scan | External Trigger | Timer Trigger | |||
|
|||||||||
| Scan Start & End Channel (byte 12) | |||||||||
| The start channel (0-F) is specified in bits 0-3, and the end channel (0-F) is specified in bits 4-7. If the end channel is less than the start channel, then the board's behavior is unspecified. | |||||||||
| Oversample (byte 13) | |||||||||
| A number indicating how many extra samples should be acquired from each channel before moving on to the next channel. In a noisy environment, the samples can be averaged together by software to effectively reduce noise. | |||||||||
| MUX Scan Start & End Channel (byte 14) | |||||||||
| In devices with a MUX (i.e. more than 16 channels), the low 4-bits of the start and end channels are stored in byte 12, described above, and the high 4-bits of the start and end channels are stored here in byte 14. The high 4-bits of the start channel are specified in bits 0-3, and the high 4-bits of the end channel are specified in bits 4-7. | |||||||||
A standard result code
Sets the number of over-samples for all A/D channels.
Linux: there is a parallel set of convenience functions for configuring the A/D. Refer to AIOUSB_SetOversample() for more information.
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
Oversample - the number of extra samples to take from each channel in a scan (see ADC_SetConfig())
A standard result code
Sets the start and end channels for A/D scans.
Linux: there is a parallel set of convenience functions for configuring the A/D. Refer to AIOUSB_SetScanRange() for more information.
Analog inputs; Linux, Windows
DeviceIndex - a standard device index
StartChannel - the number of the first channel you want in a scan
EndChannel - the number of the last channel you want in a scan (must be greater than or equal to StartChannel)
A standard result code
Permits the A/D to be calibrated using an external voltage source. The proper way to use this function is to configure the A/D with a default calibration table (such as by calling ADC_SetCal() or AIOUSB_ADC_InternalCal()). Then inject a series of voltages into one of the A/D input channels, recording the count values reported by the A/D (by calling ADC_GetScan()). It's also a good idea to enable oversampling while recording these values in order to obtain the most stable readings. Alternatively, since points is an array of double values, you can obtain individual A/D count measurements and average them yourself, producing a double average, and put that value into the points array.
The points array consists of voltage-count pairs; points[0] is the first input voltage; points[1] is the corresponding count value measured by the A/D; points[2] and points[3] contain the second pair of voltage-count values; and so on. You can provide any number of pairs, although more than a few dozen is probably overkill, not to mention would take a lot of effort to acquire.
This calibration procedure uses the current gain A/D setting for channel 0, so it must be the same as that used to collect the measured A/D counts. It's recommended that all the channels be set to the same gain, the one that will be used during normal operation. The calibration is gain dependent, so switching the gain after calibrating may introduce slight offset or gain changes. So for best results, the A/D should be calibrated on the same gain setting that will be used during normal operation. You can create any number of calibration tables. If your application needs to switch between ranges, you may wish to create a separate calibration table for each range your application will use. Then when switching to a different range, the application can load the appropriate calibration table.
Although calibrating in this manner does take some effort, it produces the best results, eliminating all sources of error from the input pins onward. Furthermore, the calibration table can be saved to a file and reloaded into the A/D, ensuring consistency.
Analog inputs; Linux
DeviceIndex - a standard device index
points - array of voltage-count pairs to calibrate the A/D with
numPoints - number of voltage-count pairs in points (minimum of 2)
returnCalTable - pointer to an array of 65,536 16-bit integers in which the generated calibration table will be returned (may be subsequently loaded into the A/D using AIOUSB_ADC_SetCalTable()); if this parameter is 0, nothing is returned
saveFileName - the name of a file in which the generated calibration table will be saved (may be subsequently loaded into the A/D using ADC_SetCal() or AIOUSB_ADC_LoadCalTable()); if this parameter is 0, the calibration table is not saved to a file
A standard result code
Calibrates the A/D in the same manner as ADC_SetCal(), generating either a default table or using the internal voltage references to generate a calibration table. The advantage of AIOUSB_ADC_InternalCal() over ADC_SetCal() is that AIOUSB_ADC_InternalCal() can return the calibration table or save it in a file.
Analog inputs; Linux
DeviceIndex - a standard device index
autoCal - AIOUSB_TRUE uses the internal voltage references to automatically calibrate the A/D; AIOUSB_FALSE generates a default (uncalibrated) table
returnCalTable - pointer to an array of 65,536 16-bit integers in which the generated calibration table will be returned (may be subsequently loaded into the A/D using AIOUSB_ADC_SetCalTable()); if this parameter is 0, nothing is returned
saveFileName - the name of a file in which the generated calibration table will be saved (may be subsequently loaded into the A/D using ADC_SetCal() or AIOUSB_ADC_LoadCalTable()); if this parameter is 0, the calibration table is not saved to a file
A standard result code
Loads a calibration table from a file into the A/D in the same manner as ADC_SetCal().
Analog inputs; Linux
DeviceIndex - a standard device index
fileName - the name of a file containing the calibration table (such as created by AIOUSB_ADC_InternalCal() or AIOUSB_ADC_ExternalCal()).
A standard result code
Sets the calibration table in the A/D to the contents of calTable.
Analog inputs; Linux
DeviceIndex - a standard device index
calTable - pointer to an array of 65,536 16-bit integers containing the calibration table (such as created by AIOUSB_ADC_InternalCal() or AIOUSB_ADC_ExternalCal()).
A standard result code
Converts a single A/D count value to volts, based on the device’s current gain setting for the specified channel. Be careful to ensure that the count value was actually obtained from the specified channel and that the gain hasn’t changed since the count value was obtained.
Analog inputs; Linux
DeviceIndex - a standard device index
channel - the channel number to use for converting counts to volts
counts - the count value to convert to volts
Volts
Returns the current calibration mode. (Must call ADC_GetConfig() first to get configuration from device.)
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
Current calibration mode (one of the AD_CAL_MODE_* values)
Returns the current end channel for A/D scans. (Must call ADC_GetConfig() first to get configuration from device.)
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
Current end channel for A/D scans
Returns the current gain code for channel. (Must call ADC_GetConfig() first to get configuration from device.)
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
channel - the channel for which to obtain the current gain code
Current gain code for channel (one of the AD_GAIN_CODE_* values)
Returns the current number of over-samples. (Must call ADC_GetConfig() first to get configuration from device.)
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
Current number of over-samples (0-255)
Returns the current start channel for A/D scans. (Must call ADC_GetConfig() first to get configuration from device.)
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
Current start channel for A/D scans
Returns the current trigger mode. (Must call ADC_GetConfig() first to get configuration from device.)
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
Current trigger mode (a bitwise OR of the AD_TRIGGER_* values)
Initializes an ADConfigBlock instance for the specified device index. Must be called before passing config to any other A/D configuration block functions.
Analog inputs; Linux
config - pointer to an uninitialized ADConfigBlock structure
DeviceIndex - a standard device index
defaults - AIOUSB_TRUE will fully initialize the block to proper default values, suitable for passing to ADC_SetConfig(); AIOUSB_FALSE performs a minimal initialization, preparing the configuration block for passing to ADC_GetConfig()
Tells if channel is configured for single-ended or differential mode. (Must call ADC_GetConfig() first to get configuration from device.)
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
channel - the channel for which to obtain the current differential mode
AIOUSB_TRUE indicates differential mode; AIOUSB_FALSE indicates single-ended mode
Tells if the ADC functions will discard the first A/D sample taken. Discarding the first sample may be useful in cases in which voltage "residue" from reading a different channel affects the channel currently being read.
Analog inputs; Linux
DeviceIndex - a standard device index
AIOUSB_FALSE indicates that no samples will be discarded; AIOUSB_TRUE indicates that the first sample will be discarded
Converts an array of A/D count values to volts, based on the device’s current gain settings for the specified channels. Be careful to ensure that the count values were actually obtained from the specified channels and that the gains haven’t changed since the count values were obtained. The arrays must be large enough to accommodate all the channels the device supports, even though only the count values for startChannel through endChannel are converted from counts to volts. An array of count values returned by ADC_GetScan() may be passed to AIOUSB_MultipleCountsToVolts() directly.
Analog inputs; Linux
DeviceIndex - a standard device index
startChannel - start channel to convert
endChannel - end channel to convert (must be greater than or equal to startChannel)
counts - array of A/D counts, such as obtained from ADC_GetScan()
volts - array in which the converted volts will be returned
A standard result code
Converts an array of voltage values to A/D counts, based on the device’s current gain settings for the specified channels. Be careful to ensure that the voltage values were actually obtained from the specified channels and that the gains haven’t changed since the voltage values were obtained. The arrays must be large enough to accommodate all the channels the device supports, even though only the voltage values for startChannel through endChannel are converted from volts to counts. An array of voltage values returned by ADC_GetScanV() may be passed to AIOUSB_MultipleVoltsToCounts() directly.
Analog inputs; Linux
DeviceIndex - a standard device index
startChannel - start channel to convert
endChannel - end channel to convert (must be greater than or equal to startChannel)
volts - array of voltage values, such as obtained from ADC_GetScanV()
counts - array in which the converted A/D counts will be returned
A standard result code
Sets all the A/D channels to the same gain code and differential mode. (Must call ADC_SetConfig() to send configuration to device.)
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
gainCode - one of the AD_GAIN_CODE_* values
differentialMode - AIOUSB_TRUE selects differential mode; AIOUSB_FALSE selects single-ended mode
Sets the A/D calibration mode. (Must call ADC_SetConfig() to send configuration to device.)
Linux: If ground or reference mode is selected, only one A/D sample may be taken at a time. That means, one channel and no oversampling. Attempting to read more than one channel or use an oversample setting of more than zero will result in a timeout error because the device will not send more than one sample. In order to protect users from accidentally falling into this trap, function ADC_GetScan() automatically and temporarily corrects the scan parameters, restoring them when it completes.
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
calMode - the calibration mode. May be one of:
AD_CAL_MODE_NORMAL (0) - normal measurement
AD_CAL_MODE_GROUND (1) - measure ground
AD_CAL_MODE_REFERENCE (3) - measure reference
Sets a single A/D channel to differential or single-ended mode. (Must call ADC_SetConfig() to send configuration to device.)
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
channel - the channel for which to set differential or single-ended mode
differentialMode - AIOUSB_TRUE selects differential mode; AIOUSB_FALSE selects single-ended mode
Specifies whether the ADC functions will discard the first A/D sample taken. Discarding the first sample may be useful in cases in which voltage "residue" from reading a different channel affects the channel currently being read. (Must call ADC_SetConfig() to send configuration to device.)
Analog inputs; Linux
DeviceIndex - a standard device index
discard - AIOUSB_FALSE indicates that no samples will be discarded; AIOUSB_TRUE indicates that the first sample will be discarded
A standard result code
Sets the gain code for a single A/D channel. (Must call ADC_SetConfig() to send configuration to device.)
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
channel - the channel for which to set the gain code
gainCode - the gain code (voltage range) for the channel. May be one of:
AD_GAIN_CODE_0_10V (0) - 0-10V
AD_GAIN_CODE_10V (1) - +/-10V
AD_GAIN_CODE_0_5V (2) - 0-5V
AD_GAIN_CODE_5V (3) - +/-5V
AD_GAIN_CODE_0_2V (4) - 0-2V
AD_GAIN_CODE_2V (5) - +/-2V
AD_GAIN_CODE_0_1V (6) - 0-1V
AD_GAIN_CODE_1V (7) - +/-1V
Sets the number of over-samples for all A/D channels. (Must call ADC_SetConfig() to send configuration to device.)
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
overSample - number of over-samples (0-255)
Sets the start and end channels for A/D scans. (Must call ADC_SetConfig() to send configuration to device.)
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
startChannel - start channel
endChannel - end channel (must be greater than or equal to startChannel)
Sets the trigger mode. (Must call ADC_SetConfig() to send configuration to device.)
Analog inputs; Linux
config - pointer to a properly initialized ADConfigBlock structure
triggerMode - a bitwise OR of these flags:
AD_TRIGGER_CTR0_EXT (0x10) - if set, counter 0 is externally triggered. Otherwise, counter 0 is triggered by the onboard 10MHz clock.
AD_TRIGGER_FALLING_EDGE (0x08) - if set, the A/D is triggered by the falling edge of its trigger source. Otherwise, the A/D is triggered by the rising edge of its trigger source.
AD_TRIGGER_SCAN (0x04) - if set, a single A/D trigger will acquire all channels from start to end, oversampling if so configured, at maximum speed. Otherwise, a single A/D trigger will cause a single acquisition, "walking" through oversamples and channels.
AD_TRIGGER_EXTERNAL (0x02) - if set, the external A/D trigger pin is an A/D trigger source. Otherwise, it's ignored.
AD_TRIGGER_TIMER (0x01) - if set, counter 2 is an A/D trigger source. Otherwise, it's ignored.
Converts a single voltage value to A/D counts, based on the device’s current gain setting for the specified channel. Be careful to ensure that the voltage value was actually obtained from the specified channel and that the gain hasn’t changed since the voltage value was obtained.
Analog inputs; Linux
DeviceIndex - a standard device index
channel - the channel number to use for converting volts to counts
volts - the voltage to convert to A/D counts
A/D counts
Writes a count value to a D/A channel.
Analog outputs; Linux, Windows
DeviceIndex - a standard device index
Channel - the channel number of the D/A you wish to set. The number of channels varies from model to model.
Value - the D/A count value to output. The number of bits of resolution for the D/A outputs varies from model to model, however it's usually 12- or 16-bits. Moreover, some of the 12-bit models actually accept a 16-bit value and simply truncate the least significant 4 bits. Consult the manual for the specific device to determine the range of D/A values the device will accept. In general, 12-bit devices accept a count range of 0-FFFh, and 16-bit devices accept a count range of 0-FFFFh.
A standard result code
Writes a block of count values to one or more D/A channels.
Analog outputs; Linux, Windows
DeviceIndex - a standard device index
pDACData - a pointer to an array of 16-bit integers representing channel/count pairs. The first integer of each pair is the D/A channel number and the second integer is the D/A count value to output to the specified channel. Refer to DACDirect() for an explanation of the channel addressing and count values.
DACDataCount - number indicating how many channel/count pairs are in the array referenced by pDACData
A standard result code
Ends and closes a DAC streaming process. Deprecated: DACOutputCloseNoEnd() is preferred.
Linux: Not yet implemented.
USB-DA12-8A; Linux, Windows
DeviceIndex - a standard device index
bWait - reserved for future expansion; currently, this function always waits for the streaming process to complete before returning
A standard result code
Closes a DAC streaming process without ending it. This is most useful when you've set LOOP or EOM via DACOutputFrameRaw().
Linux: Not yet implemented.
USB-DA12-8A; Linux, Windows
DeviceIndex - a standard device index
bWait - reserved for future expansion; currently, this function always waits for the streaming process to complete before returning
A standard result code
Writes a group of points (a “frame”) into the DAC stream. Deprecated: DACOutputFrameRaw() is preferred.
All points in a frame control the same number of DACs; if, for example, you wish to output one point with all 8 DACs, followed by 99 points with only 2 DACs, set the DAC count to 8, output a frame of just the first point, then set the DAC count to 2, and output a frame of the next 99 points. If the driver’s internal buffer is full, the function will return “ERROR_NOT_READY” (equal to 21 decimal); try again in a moment, as the driver’s buffer should drain some as soon as there’s room in the larger hardware buffer and available time on the USB bus.
Linux: Not yet implemented.
USB-DA12-8A; Linux, Windows
DeviceIndex - a standard device index
FramePoints - the number of points (16-bit integers) in the frame
FrameData - a pointer to the first of an array of 16-bit D/A count values to output
A standard result code
Writes a group of points (a “frame”) into the DAC stream. Similar to DACOutputFrame() except the features are controlled by the upper bits in the data array. This provides the greatest flexibility, at the cost of complexity.
Linux: Not yet implemented.
USB-DA12-8A; Linux, Windows
DeviceIndex - a standard device index
FramePoints - the number of points (16-bit integers) in the frame
FrameData - a pointer to the first of an array of 16-bit D/A count values to output. The DAC count determines the number of samples, even though you can place EOD bits (see below) as you wish.
| FrameData Bit Fields | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Bit | 15 | 14 | 13 | 12 | 11 | 10 | 9 | 8 | 7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
| Meaning | EOM | EOF | EOD | LOOP | DAC Value | |||||||||||
|
||||||||||||||||
A standard result code
Begins a DAC streaming process. The stream is divided into “points”. Each point contains data for one or more DACs, and during the streaming process the onboard counter/timer clocks out points at a steady rate.
Linux: Not yet implemented.
USB-DA12-8A; Linux, Windows
DeviceIndex - a standard device index
pClockHz - a pointer to a double precision IEEE floating point number containing the desired output clock frequency. This value is set by the driver to the actual frequency at which DAC data will be clocked out, as limited by the device’s capabilities.
A standard result code
Sets the number of DACs involved in each DAC streaming point henceforth. When the driver connects to the device, this is initialized to 5 (for ILDA use). You can set this freely between calls to DACOutputFrame() and/or DACOutputFrameRaw() if you wish.
Linux: Not yet implemented.
USB-DA12-8A; Linux, Windows
DeviceIndex - a standard device index
NewCount - number from 1-8 indicating the number of DACs in future points
A standard result code
Enables or disables interlock. While interlock is enabled, DAC streaming is paused unless the interlock pin is grounded, usually through the cable. The interlock pin is pin 12 of the DB25 M connector (or, on the OEM version, pin 7 of the connector named J4).
Linux: Not yet implemented.
USB-DA12-8A; Linux, Windows
DeviceIndex - a standard device index
bInterlock - AIOUSB_TRUE, TRUE enables interlock; AIOUSB_FALSE, FALSE disables interlock
A standard result code
"Manually" starts a DAC streaming process. Normally, DAC streaming will be started automatically by streaming 1¼ SRAMs worth of data (160K bytes, i.e. 81920 samples). It's only if you're using a smaller amount of data that you'd need to "manually" start DAC streaming with this function.
Note that before starting DAC output you must send the lesser of one SRAM worth of data (128K bytes, i.e. 65536 samples) or your entire waveform, due to the use of bank-switched single-ported memory.
Linux: Not yet implemented.
USB-DA12-8A; Linux, Windows
DeviceIndex - a standard device index
A standard result code
Sets the voltage range of the D/A outputs.
Analog outputs; Linux, Windows
DeviceIndex - a standard device index
RangeCode - the voltage range to select. May be one of:
DAC_RANGE_0_5V (0) - 0-5V
DAC_RANGE_5V (1) - +/-5V
DAC_RANGE_0_10V (2) - 0-10V
DAC_RANGE_10V (3) - +/-10V
Linux: Use the named constants listed above.
Windows: Use the numeric values shown in parenthesis.
A standard result code
The ADConfigBlock structure is initialized by a call to AIOUSB_InitConfigBlock() and defined in aiousb.h as:
struct ADConfigBlock {
void *device;
unsigned long size;
unsigned char registers[ AD_MAX_CONFIG_REGISTERS ];
}; // struct ADConfigBlock
This structure is intended to be somewhat opaque, its main purpose being to shield users from having to fill in the bytes and bit fields of an A/D configuration block which is used in calls to ADC_GetConfig() and ADC_SetConfig().
In addition to simplifying the getting and setting of the configuration, using ADConfigBlock and its associated functions permits you to change several configuration settings and then send them all to the device with a single call to ADC_SetConfig(). The other API functions that manipulate individual configuration settings, such as ADC_Range1(), send the modified configuration to the device immediately. There is nothing wrong with sending the configuration to the device after each setting change, but using ADConfigBlock and its associated functions gives users the option of deferring sending the configuration until it's fully set up. Of course, ADC_SetConfig() also lets users send the entire configuration in one step, but it requires them to manipulate the bytes and bit fields of the A/D configuration block.
The functions that manipulate ADConfigBlock are “object-oriented” in the sense that the proper way to use them is to create an ADConfigBlock instance and pass it to all these functions. The ADConfigBlock instance should be initialized immediately after it’s declared using AIOUSB_InitConfigBlock(), which is equivalent to a “constructor” in C++. AIOUSB_InitConfigBlock() makes sure that the configuration block is initialized properly for the device with which it will be used.
The following functions are the "methods" that can operate on ADConfigBlock:
AIOUSB_GetCalMode()
AIOUSB_GetEndChannel()
AIOUSB_GetGainCode()
AIOUSB_GetOversample()
AIOUSB_GetStartChannel()
AIOUSB_GetTriggerMode()
AIOUSB_InitConfigBlock()
AIOUSB_IsDifferentialMode()
AIOUSB_SetAllGainCodeAndDiffMode()
AIOUSB_SetCalMode()
AIOUSB_SetDifferentialMode()
AIOUSB_SetGainCode()
AIOUSB_SetOversample()
AIOUSB_SetScanRange()
AIOUSB_SetTriggerMode()
Below is an example of the proper way to use ADConfigBlock.
/* * procedure for setting A/D configuration */ // create "instance" ADConfigBlock configBlock; // call "constructor" AIOUSB_InitConfigBlock( &configBlock, DeviceIndex, AIOUSB_TRUE ); // ... set properties ... // send configuration block to device ADC_SetConfig( DeviceIndex, configBlock.registers, &configBlock.size ); /* * procedure for reading current A/D configuration */ // create "instance" ADConfigBlock configBlock; // call "constructor" AIOUSB_InitConfigBlock( &configBlock, DeviceIndex, AIOUSB_FALSE ); // read configuration block from device ADC_GetConfig( DeviceIndex, configBlock.registers, &configBlock.size ); // ... use properties ...
A "handle" to a device on the USB bus. Each device on the USB bus is assigned an index number ranging from 0-31. This index number must then be passed to most of the API functions in order to control a specific device.
For convenience when writing quick test programs and such, these two special constants may be used instead of an actual device index:
diFirst (0xFFFFFFFEul) - uses the first ACCES device found on the bus
diOnly (0xFFFFFFFDul) - uses the only ACCES device found on the bus, meaning that there must be only one ACCES device on the bus
The DeviceProperties structure is filled in by a call to AIOUSB_GetDeviceProperties() and defined in aiousb.h as:
struct DeviceProperties {
const char *Name; // null-terminated device name or 0
__uint64_t SerialNumber; // 64-bit serial number or 0
unsigned ProductID; // 16-bit product ID
unsigned DIOPorts; // number of digital I/O ports (bytes)
unsigned Counters; // number of 8254 counter blocks
unsigned Tristates; // number of tristates
long RootClock; // base clock frequency
unsigned DACChannels; // number of D/A channels
unsigned ADCChannels; // number of A/D channels
unsigned ADCMUXChannels; // number of MUXed A/D channels
}; // struct DeviceProperties
This section is a reference for all of the constants used in AIOUSB for Linux which aren't described elsewhere in this document. The values shown in parenthesis next to each constant name are the actual values.
In C, the AIOUSB_BOOL type is declared as an enum. In C++, it's declared as a bool. Other libraries tend to declare the names BOOL, TRUE and FALSE, and worse, they declare them using #defines, which can be surprisingly difficult to work around. So we sidestep such potential conflicts by declaring the same names prefixed with AIOUSB_. It's ugly, but if people want to use the shorter names and they are certain the shorter names won't conflict with anything else, they can define the ENABLE_BOOL_TYPE macro, which defines the constants BOOL, TRUE and FALSE.
AIOUSB_TRUE, TRUE (1, true) - boolean 'true' (in C, has the value '1'; in C++, the value 'true')
AIOUSB_FALSE, FALSE (0, false) - boolean 'false' (in C, has the value '0'; in C++, the value 'false')
Windows: use the names TRUE and FALSE.
The AIOUSB function result codes are a bit confusing. The result codes used in the Windows implementation of the API are defined in a system file named winerror.h. These result codes are generic and can apply to many applications. They are all prefixed with ERROR_, the very first one, ERROR_SUCCESS, sounding like an oxymoron. The result codes used in libusb (Linux), on the other hand, are more appealing: the result code for success is LIBUSB_SUCCESS, while the result codes for errors are LIBUSB_ERROR_xxx. Further complicating matters is that the AIOUSB result codes must be non-negative since all the functions return an unsigned result, whereas the libusb result codes are negative in the case of errors. Fortunately, both schemes use zero to indicate success. Finally, it's desirable to return the original libusb result code in cases where a libusb error causes an AIOUSB API function to fail. So to satisfy all these requirements, we've employed the following scheme:
AIOUSB result codes in Linux are prefixed with AIOUSB_. The result code for success is AIOUSB_SUCCESS, which has a value of zero. The result codes for errors are AIOUSB_ERROR_xxx, which have positive values, starting with one (1).
In order to offer users the option of using result codes whose names are similar to those used in the Windows implementation, we define a second set of result codes with names similar to those used in the Windows implementation but which map to the same values as the AIOUSB_xxx result codes. These alternate result code names can be enabled by defining the macro ENABLE_WINDOWS_RESULT_CODES, which is not defined by default.
In order to preserve the original libusb result codes and pass them back from an AIOUSB API function, we translate the libusb result codes to a format that conforms to the one AIOUSB employs and provide macros for converting the AIOUSB result code back to a libusb result code. Macro LIBUSB_RESULT_TO_AIOUSB_RESULT(code) converts a libusb result code to an AIOUSB result code, and macro AIOUSB_RESULT_TO_LIBUSB_RESULT(code) does the reverse.
In the Linux implementation we provide an extended AIOUSB API function named AIOUSB_GetResultCodeAsString() that returns a string representation of a result code, including those that encapsulate a libusb result code.
Each of the result codes shown has two names: the "native" Linux name, and a Windows-style name, which can be enabled by defining the macro ENABLE_WINDOWS_RESULT_CODES.
AIOUSB_SUCCESS, ERROR_SUCCESS (0) - The function completed successfully.
AIOUSB_ERROR_DEVICE_NOT_CONNECTED, ERROR_DEVICE_NOT_CONNECTED (1) - The internal handle to the device is not valid. This can occur if AIOUSB has not been properly initialized. On Linux this can occur because libusb fails to return a valid handle.
AIOUSB_ERROR_DUP_NAME, ERROR_DUP_NAME (2) - More than one device was found on the bus when diOnly was specified as a device index.
AIOUSB_ERROR_FILE_NOT_FOUND, ERROR_FILE_NOT_FOUND (3) - May mean multiple things. May mean that the attempt to load the A/D calibration table in ADC_SetCal() failed because the specified file does not exist or cannot be read. May also mean that the attempt to save the A/D calibration table to a file in ADC_SetCal() failed because the file could not be created. May also mean that no device was found on the bus when diFirst or diOnly were specified as a device index.
AIOUSB_ERROR_INVALID_DATA, ERROR_INVALID_DATA (4) - Data retrieved from the device is not valid. May mean that the data is corrupt or that an incorrect amount of data was transferred
AIOUSB_ERROR_INVALID_INDEX, ERROR_INVALID_INDEX (5) - The device index passed to an function is invalid.
AIOUSB_ERROR_INVALID_MUTEX, ERROR_INVALID_MUTEX (6) - An internal mutex is invalid. This usually means the library was unable to initialize the mutex.
AIOUSB_ERROR_INVALID_PARAMETER, ERROR_INVALID_PARAMETER (7) - One or more parameters passed to a function are invalid, such as outside the allowed range or null pointers.
AIOUSB_ERROR_INVALID_THREAD, ERROR_INVALID_THREAD (8) - An internal thread is invalid. This usually means the library was unable to initialize the thread.
AIOUSB_ERROR_NOT_ENOUGH_MEMORY, ERROR_NOT_ENOUGH_MEMORY (9) - A function was unable to allocate sufficient dynamic memory.
AIOUSB_ERROR_NOT_SUPPORTED, ERROR_NOT_SUPPORTED (10) - The function is not supported for the device specified, for example, trying to perform A/D operations on a device that doesn't have an A/D.
AIOUSB_ERROR_OPEN_FAILED, ERROR_OPEN_FAILED (11) - Typically means that a streaming device is already open for streaming and cannot be opened again until it's closed.
AIOUSB_ERROR_LIBUSB (100) - If the result code is greater than or equal to this value, then it is a libusb result code indicating an error. Use the LIBUSB_RESULT_TO_AIOUSB_RESULT(code) macro to extract the original libusb result code, or pass the result code to AIOUSB_GetResultCodeAsString() to obtain a string representation of it.
Below is a list (in alphabetical order) of all the result codes returned by functions in the Windows implementation of the AIOUSB API. These result codes are defined in a system file named winerror.h. Please consult the relevant Windows documentation for more information.
ERROR_ALREADY_EXISTS
ERROR_BAD_LENGTH
ERROR_BAD_TOKEN_TYPE
ERROR_DEVICE_NOT_CONNECTED
ERROR_DEVICE_REMOVED
ERROR_DUP_NAME
ERROR_FILE_EXISTS
ERROR_FILE_NOT_FOUND
ERROR_GEN_FAILURE
ERROR_HANDLE_EOF
ERROR_INSUFFICIENT_BUFFER
ERROR_INTERNAL_ERROR
ERROR_INVALID_ADDRESS
ERROR_INVALID_DATA
ERROR_INVALID_INDEX
ERROR_INVALID_PARAMETER
ERROR_NOACCESS
ERROR_NOT_READY
ERROR_NOT_SUPPORTED
ERROR_NO_DATA_DETECTED
ERROR_OPEN_FAILED
ERROR_OUTOFMEMORY
ERROR_SEEK
ERROR_SUCCESS
ERROR_TOO_MANY_OPEN_FILES
The AIOUSB library for Linux is compiled for C and C++, residing within libaiousb.a and libaiousbcpp.a, respectively. However, when using C++, all the public functions, variables, enums and so forth are wrapped up in a namespace AIOUSB declaration to prevent those names from conflicting with anything else. All of the C++ sample programs for Linux contain the statement using namespace AIOUSB; immediately after including the file aiousb.h. Using the namespace in this manner is the simplest technique, but if you prefer to explicitly refer to public elements within this namespace, simply prefix all such names with AIOUSB::. For example, instead of calling GetDevices(), you would explicitly call AIOUSB::GetDevices(). Similarly, constant names must be prefixed with AIOUSB::. For example, AIOUSB_SUCCESS would instead be referred to as AIOUSB::AIOUSB_SUCCESS.
The Linux implementation of AIOUSB is thread-safe to the extent that the internal data structures are protected by a mutex. Furthermore, the lower level libusb module is also thread safe in the same manner. However, neither AIOUSB nor libusb prevent multiple threads from communicating with the same device. A software design that permits multiple threads to communicate with the same device must provide its own mutual exclusion mechanism. AIOUSB for Linux is designed under the premise that one thread will communicate with one device, and optionally, another thread may communicate with a different device, and so on. That form of multithreading is supported by AIOUSB for Linux.
Also beware that certain functions, such as GetDevices() make substantial changes to the internal state of the AIOUSB module. It is not safe to call such functions while other threads are actively communicating with devices! While such internal state changes are made in the context of a locked mutex and will probably not cause the software to fail, in all likelihood any threads which are actively communicating with devices will subsequently experience communication failures with their respective devices. You should only call such functions when no threads are actively communicating with devices. After such changes, threads must reestablish a connection to their devices by obtaining a new device index "handle."