« Close

Datasheets and User Guides

App Notes

Software & Driver


Appendix A - Specifications

Specifications at 25 degrees C and Vusb/Vext = 5.0V, except where noted.

Table A-1. Specifications

Parameter Conditions Min Typical Max Units
USB Cable Length       5 meters
Ethernet Cable Length (1)       100 meters
Supply Voltage   3.6 5 5.3 volts
Typical Supply Current (2)          
  Control Low 85   105 mA
  Control High 125   160 mA
Operating Temperature   -40   85 °C
Clock Error ~25°C     ±30 ppm
  -10 to 60 °C     ±50 ppm
  -40 to 85 °C     ±100 ppm
Typ. Command Execution Time (3) Ethernet 1.2      
  USB high-high 1.4      
  USB other 4      
Vs Outputs          
Typical Voltage, USB (4) Self-Powered 4.5 5 5.25 volts
Typical Voltage, Wall-Wart   4.75 5 5.25 volts
Maximum Current (5)     200   mA
Vm+/Vm- Outputs          
Typical Voltage No-load   ±5.8   volts
  @ 1 mA   ±5.6   volts
Maximum Current     1   mA
(1) Expected max Ethernet cable length is at least 100 meters by design, but we have only tested 33 meter cables. Customer feedback on longer cables is welcome.     
(2) Typical current drawn by the UE9 itself, not including any user connections. Minimum value is the typical current when the device is idle. Maximum value is the typical current when the device is very busy.     
(3) Total typical time to execute a single Feedback function with no analog inputs. Measured by timing a Windows application that performs 1000 calls to the Feedback function. See Section 3.1 for more timing information.     
(4) Self-powered would apply to USB hubs with a power supply, all known desktop computer USB hosts, and some notebook computer USB hosts.     
(5) This is the maximum current that should be sourced through the UE9 and out of the Vs terminals. The UE9 has internal overcurrent protection that will turn the UE9 off, if the total current draw exceeds ~490 mA.     
Parameter Conditions Min Typical Max Units
Analog Inputs          
Unipolar Input Range (6) AINx to GND 0   5/G volts
Bipolar Input Range (6) AINx to GND -5   5 volts
Maximum AIN Voltage (7) AINx to GND -15   15 volts
Input Bias Current (8) @ 1 volts   -15   nA
Input Impedance (8)     >10  
Source Impedance (9)       10
Temperature Drift G=1   10   ppm/°C
Absolute Accuracy Res 12-17   ±0.025 ±0.05 % FS
  UE9-Pro, Res=18   ±0.005 ±0.01 % FS
Peak-to-Peak Noise See Appendix B        
Integral Linearity Error G=1   ±0.02   % FS
  G=8   ±0.03   % FS
  UE9-Pro, Res=18   ±0.0001   % FS
Differential Linearity Error 12-bit   ±1   counts
  16-bit   ±4   counts
  UE9-Pro, Res=18   ±1   counts
Stream Data Buffer Size     182361    
C/R Acquisition Time See Section 3.1 1.2   125 ms
Stream Speed (10) 12-bit stream     up to 80k samples/s
  13-bit stream     16000 samples/s
  14-bit stream     4000 samples/s
  15-bit stream     1000 samples/s
  16-bit stream     250 samples/s
Channel-to-Channel Delay (11) 12-bit stream   12   µs
  13-bit stream   44   µs
  14-bit stream   158   µs
  15-bit stream   670   µs
  16-bit stream   2700   µs
(6) For actual nominal input ranges see Section 2.7 of the UE9 User's Guide.     
(7) Maximum voltage to avoid damage to the device. Protection level is the same whether the device is powered or not. When the voltage on any analog input exceeds 6.0 volts, all other analog inputs are also affected, until the overvoltage is removed. At 6.5 volts, there is a ~1 mV offset noticed on other channels, increasing to a ~5 mV offset at 15.0 volts.     
(8) This is the steady state input bias current and impedance. When the analog input multiplexer changes from one channel to another at a different voltage, more current is briefly required to change the charge on the input amplifier. The steady state input bias current is very flat across the common mode voltage range, except for voltages of about 4.5 or higher where the bias current shifts to typically +250 nA     
(9) To meet specifications, the impedance of the source signal should be kept at or below the specified value. With a higher source impedance, noticable static errors can occur due to the bias current flowing through the source impedance. There are also dynamic errors that can become noticable as the source impedance can degrade the ability of the internal multiplexer to settle quickly when changing between channels with different voltages.     
(10) Divide by the number of channels to determine the maximum scan rate. Assumes an Ethernet or USB high-high connection. Other USB connections might not be able to maintain 50 ksamples/s. See Section 3.2 for more information.     
(11) When scanning more than 1 channel in a stream, this is the time between each sample within a scan.     
Parameter Conditions Min Typical Max Units
Analog Outputs          
Nominal Output Range (12) No Load 0.02   4.86 volts
  @ ±2.5 mA 0.225   4.775 volts
Resolution     12   bits
Absolute Accuracy (13) 5% to 95% FS   ±0.1   % FS
Integral Linearity Error     ±2   counts
Differential Linearity Error     ±1   counts
Error Due To Loading @ 100 µA   0.1   %
  @ 1 mA   1   %
Source Impedance     50   Ω
Short Circuit Current Max to GND   100   mA
Slew Rate     0.8   V/µs
Digital I/O          
Low Level Input Voltage   -0.3   1 volts
High Level Input Voltage   2.3   Vs + 0.3 volts
Maximum Input Voltage (14) FIO -10   10 volts
  EIO/CIO/MIO -6   6 volts
Input Leakage Current     10   µA
Output Low Voltage (15) No Load   0   volts
--- FIO Sinking 1 mA   0.55   volts
--- EIO/CIO/MIO Sinking 1 mA   0.18   volts
--- EIO/CIO/MIO Sinking 5 mA   0.9   volts
Output High Voltage (15) No Load   3.3   volts
--- FIO Sourcing 1 mA   2.75   volts
--- EIO/CIO/MIO Sourcing 1 mA   3.12   volts
--- EIO/CIO/MIO Sourcing 5 mA   2.4   volts
Short Circuit Current (15) FIO   6   mA
  EIO/CIO/MIO   18   mA
Output Impedance(15) FIO   550   Ω
  EIO/CIO/MIO   180   Ω
Counter Input Frequency (16)       3 MHz
Input Timer Total Edge Rate (17) No Stream     100000 edges/s
  While Streaming     25000 edges/s
(12) Maximum and minimum analog output voltage is limited by the supply voltages (Vs and GND). The specifications assume Vs is 5.0 volts. Also, the ability of the DAC output buffer to drive voltages close to the power rails, decreases with increasing output current, but in most applications the output is not sinking/sourcing much current as the output voltage approaches GND.     
(13) Analog output accuracy is specified from 5% to 95% of full-scale output.     
(14) Maximum voltage to avoid damage to the device. Protection works whether the device is powered or not, but continuous voltages over 5.8 volts or less than -0.3 volts are not recommended when the UE9 is unpowered, as the voltage will attempt to supply operating power to the UE9 possibly causing poor start-up behavior.     
(15) These specifications provide the answer to the common question: "How much current can the digital I/O sink or source?". For instance, if EIO0 configured as output-high and shorted to ground, the current sourced by EIO0 into ground will be about 18 mA (3.3/180). If connected to a load that draws 5 mA, EIO0 can provide that current but the voltage will droop to about 2.4 volts instead of the nominal 3.3 volts. If connected to a 180 ohm load to ground, the resulting voltage and current will be about 1.65 volts @ 9 mA.     
(16) Hardware counters. 0 to 3.3 volt square wave. Default power level is "High".     
(17) To avoid missing edges, keep the total number of applicable edges on all applicable timers below this limit. See Section 2.10 for more information.     


Looking at the Differential Linearity Error, how does "number of counts" translate into voltage error?  Do we assume the Vref of the ADC is 5V?  If so, for a 12-bit ADC, then one LSB = one count = Vref/2^N =5/4096 = 1.22mV.  So that's the DNL.  The INL error is 0.02% of FS = 0.0002*5=1mV.  Do those two errors simply add up to give 2.22mV of total non-linearity error?

Thanks for the help.

Wiki has general info about DNL and INL.  To translate counts into voltage, it depends on the range.  If you are using the 0-5 volt range, then your calculation for the weight of 1 12-bit count is correct.

You should not have to add DNL and INL.  INL should include DNL.  I think the fact that DNL actually looks slight bigger than INL for a 12-bit G=1 measurement is just the limitation of a 12-bit measurement.  Both errors together are within 1 count, so we can't look closer to separate them.

Absolute Accuracy includes DNL and INL and other error sources.