As a general rule, before doing too much troubleshooting you should do a complete power cycle of the PC, LabJack, and any hubs. Start with no connections to the LabJack except the USB cable and make sure no software is running except for the one LabJack program (LJlogger or LJtest) being used for testing. LJtest only communicates during a test, so it can remain open while LJlogger is run, but LJlogger continually communicates, so it must be completely closed before running LJtest. If other working LabJacks are available, connect one to the cable/port you are using to make sure everything is OK out to the end of the cable. Another general troubleshooting technique is to connect a bad LabJack and then feel each chip to see if any are hot. When trying to measure voltages on screw-terminals (e.g. with a DMM), you must clamp the leads in the screw-terminal to make sure you are getting a good contact.
1. No LED activity at power-up:
2. LED blinks 4 times at power up but then stays off (device does not enumerate over USB). Power up the U12 with IO0 jumpered to STB, disconnect the U12, remove the jumper, and power up again (this turns off the watchdog). Try different computers with different operating systems. If problem is continues:
3. LED blinks repeatedly with various patterns. Power up the U12 with IO0 jumpered to STB, disconnect the U12, remove the jumper, and power up again (this turns off the watchdog). Try different computers with different operating systems. If problem is continues:
4. LJtest fails with only 3 basic tests passing (“Find LabJack”, “OS Version”, “Driver Version”). See item #2 above.
5. LJtest fails with more than the 3 basic tests passing. Before troubleshooting this issue further, run LJtest with “Test Fixture Installed selected”, when it fails let it reset the calibration constants to zero, and then power-cycle the U12.
Often, when this type of failure occurs , it is difficult to determine which chip (IC2=A/D or IC3=RAM) has failed (or both). The problem is that failure of one chip will often disrupt the SPI bus such that communication with the other chip does not work. We have found that the quickest way to repair such a unit is to remove the RAM chip, and then reconnect the LabJack to see if the A/D is communicating. If the A/D starts to communicate, then it is probably just the RAM chip that is bad. If the A/D is still not communicating, it is most likely bad and the status of the RAM chip is unknown (might as well install a new one).
6. LJtest fails "Test Counter (& STB)". Gives a message like "Count error was 2090912.86 percent (596349256)". If you run LJcounter and set the measurement type to Count, you see that the count is stuck at a huge number such as 596377774. Even if you click Reset Counter it does not reset but rather is stuck on that huge number. This problem often means that the watchdog is enabled and thus the Counter is disabled. A common way the watchdog gets accidentally enabled is by using the "Control LJ Watchdog" feature in LJlogger. Some ways to disable the watchdog:
7. Analog Outputs not working correctly when connected only to a DMM or LabJack AI channel. Common symptoms are one or both analog outputs either stuck near zero or will not go over some low voltage (0.5 to 1.0 volts typically). One or both op-amps in IC5 have been damaged. It is also possible for the amps to be damaged to the extent that they work fine when unloaded, but start to voltage limit when loaded at a moderate current within specs.
Customers who sustain hardware damage, are recommended to implement the following suggestions to avoid further failures, since they apparently have a system which causes some sort of fault:
Route all 5 volt and ground connections to a single +5V and GND terminal on the LabJack. Place a 5.6 volt zener (1N4734) on these terminals. If you only have ground connections, connect the zener from that GND to the nearest +5V terminal. A capacitor can also be placed in parallel with the zener. If you must use multiple +5V and/or GND terminals, use a zener at each location. Hopefully, a fault will blow the zener before damaging the LabJack. Particular systems can be evaluated for the possible use of a resistor(s) on either or both side(s) of the zener connection on either or both terminal(s). Often, when connecting a ground on an external supply to the LabJack GND, it is desirable to use a small resistance to prevent the LabJack ground from becoming the “preferred” ground for the power supply.
Analog output failure seems to be due to damage to one or both op-amp buffers. The cause is likely an ESD event or over/under-voltage applied directly to the analog output.
Customers who sustain such a failure, are recommended to implement the following suggestions to avoid further failures, since they apparently have a system which causes some sort of fault:
Place a 5.6 volt zener (1N4734) from the nearest GND to the analog output terminal. If possible (depending on the current drawn from the analog output), a resistor as large as possible is desirable between the AOx terminal and the zener lead. A fault should blow the zener before damaging the LabJack.
CAL failure is very rare (perhaps a couple cases), and is likely due to an ESD event or over/under-voltage applied directly to the CAL pin.
Customers who sustain such a failure, are recommended to implement the following suggestions to avoid further failures, since they apparently have a system which causes some sort of fault:
Place a 4.3 volt zener (1N4731) from the nearest GND to the CAL terminal. As an additional precaution, a small resistor (10 ohms) is desirable between the CAL terminal and the zener lead. A fault should blow the zener before damaging the LabJack.