Agilent3070 training manual
* Control card control’s relay closures and ASRU’s measurement.
ATL --- analog test language
MOA --- measurement operational amplifier
Rx is the device under test on the PC board. A voltage, Vs, is applied through the S (source) bus to one lead of Rx and the other lead is connected to the I (input) bus of the MOA. Rref is a precision feedback resistor. The output voltage of the MOA, Vmoa, is measured by an internal voltmeter.
All the pieces of the measurement (Vs, Rref, the MOA, and the DVM) are located on the ASRU card.
Rx = -(Vs/Vmoa) * Rref
--- Resistive component tests
Relay connections and test commands;
Relay Commands
Disconnect all opens relay connections on all busses.
Connect to “” connects specified bus to specified node on the PC board
Test commands
Component ,<+tol>,<-tol>,
PerShorts: it is running analog tests to assure that adjustable devices are set in the correct positions before executing subsequent tests.
Analog test source files are written in Analog Test Language (ATL). Initially written by HP IPG, they consist of two sections.
1. Close relays to connect the correct busses to the device under test.
2. Perform the measurement with the specified options.
Remember:
1. Test sources are located under the directory “analog”
2. Each analog test has two sections – relay control statement and test commands.
3. The object file is what the hardware uses to execute the test. if a source file is edited and re-saved, but not re-compiled, the original object is used. Therefore, always remember to re-save and re-compile a source when it is edited.
The heart of this measuring system is the Measurement Operational Amplifier (MOA).
Vs is applied through the S (source) bus to one lead of Rx.
Rref is a precision feedback resistor.
Vmoa is the output of the MOA. It is measured by an internal voltmeter.
All the pieces of the measurement are located on the ASRU card.
The two Fundamental Assumptions regarding Analog In-Circuit test
For the moment we assume the MOA to be ideal, thus its input impedance approaches infinity.
Remember:
This technique depends on the validity of two assumptions:
1. The current, Irx equals the current, Iref
2. These currents can be accurately measured.
Subsequent modules will deal with cases where these two assumptions cannot be guaranteed. You should always keep in mind the basic analog in-circuit equation.
Option list:
reX: Specifies the value of Rref, the feedback resistor. X ranges from 1-6.
arX: Specifies the range of the internal voltmeter. X ranges from 0 to 10 volts for DC measurement. The voltmeter will automatically up-range if set too low but it will not downrange. If the voltmeter does need to autorange, the process can result in rather slow tests, as the voltmeter must make a measurement, detemine that the range is set too low, up-range one level and repeat the process.
amX: specifies the voltage of the DC source. X ranges from –10 to +10 volts for DC sources. If not included in the option list, the source is set to 100mV DC as a default.
Reactive component test options
ico0, ico1: 0 for a source compliance of 35mA, and 1 for a source compliance of 100mA.
frX: 128, 1024, 8192.
pm, sm: the default model for capacitors is “pm” (parallel model), for inductors, it is “sm” (series model).
The resistor’s measurement are noisy, the test is unstable.
Solution: if the test is unstable, use the ed option. Generate a new histogram to confirm the improved stability. You should see a 2x improvement in the stability.
If this did not improve the stability of the test or did not improve it enough, remove the option and widen the test limits.
Look for reactive components near the device under test. Guard any them if they might be causing inconsistent results.
The resistor’s measured value is greater than excepted value.
Solution: if the device is measured high, you are loosing current somewhere. It could be because the Zig is too small. Try reversing the S/I bus connections.
Try adding remote sensing: A bus and B bus
If you cannot get the measurement to be accurate, widen the test limits or comment the test.
If the test is beyond the added 10%, comment the test and don’t test this device.
The resistor’s measured value is less than excepted value.
Solution: First, find the failing component and look at the circuitry around it. Is there a parallel path around the device? If you find, select a ground point and add it to the test. If the measurement was not improved, remove the change. Use the same remote sensing tools (sa, sb) as above to improve the test accuracy.
There are two options that you might try using either in conjunction with the ground or if no ground point is found. These are the ed and wait options.
