FEC200 IH Test
The IH measures the actual value (as it sees it) as opposed to the GONOGO IH test which is part of the VT test. The IH version is much slower and starts with a (usually) lower current. These two things may effect the correlation between the two methods.
The IH test starts with a current that is full-scale on the scale used to measure IH. The software artificially restricts the limits programming to 25% of the scale. Therefore, the starting current is always at least four times the maximum expected IH.
The forcing voltage entered is the threshold for sensing that the device is turned off.
- TEST(n) 10V >100MA <400MA
In this example the current scale and therefore the initial current would be 4.00A. The current scales are decade multiples of 4000 units and the IH limits are constrained to be equal to or less than 1000 units. The current is slowly ramped down until the DUT voltage reaches 10V (in the example). The point at which this happens is taken to be IH.
DETAILS OF TIMING
The device is first brought up to VBR at 100µA. 7mS is allowed for this. Then the full-scale current is applied. (See details in the VT description). 3mS is allowed for the device to turn on and to settle. Then the current starts to ramp down. The ramp is made up of three segments. Each segment is linear but each has a different rate. The idea is to make each step in the digitally synthesized ramp very roughly the same percentage of the current value.
The first ramp goes from full-scale to 25%. The next from, 25% to 6.25% and the final ramps from 6.25% down to zero.
You can control the rate of these ramps (time per step) by setting a variable called SLOWRAMP in the file CUSTOM.TXT. The value will be effected by the speed of the computer used, so be careful about changing computers.
Going slow results in more heating but going too fast will possibly exceed the response time of the hardware. The following background may help find the most suitable rate.
There are from 250 to 300 steps in each segment. The exact upper speed limit of the hardware is not exactly known as it also depends on the computer. However it is unlikely to operate properly faster than 20-30µS per step.
SLOWRAMP is in units of about 10µS (incrementally) on top of internal overhead. The overhead is in the range of about 10µS to 20µS.
You can see this ramp best by using a scope to monitor the DAC1 output of the B500 board. (The bottom test point). Program SYNC1 to the number of the IH test. Please see description of the SYNC programming elsewhere.