Thermal Testing Chapter 3 - FAQ
What is the difference between DVF as measured on a VF40 and on an FEC200?
In an ideal (but unrealistic) DVF test, the diode and the test equipment would both be infinitely fast. So at the end of the heating pulse, the current would go instantly from IH to IM. The heating power would drop to near zero and the junction would begin to cool.
After a delay (TMD) the VF would be measured and compared with the cold reading to get DVF. The junction would of course be a bit cooler than at the end of the IH pulse because of the cooling during TMD.
In a real DVF test however, both the diode and the test equipment take a finite (and variable) time to switch from IH to IM.
The test equipment is limited by the turn-off time of the IH power supply, which, in turn is increased by any inductance in the circuit between the power supply output and the diode under test.
The diode under test may also have a substantial effect on the turn-off time. This is especially true for "standard recovery" rectifiers. One might think that a 2m S Trr would be quite small compared with a typical TMD. However, the 2m S Trr is measured under far different conditions. The Trr in a DVF test situation is MUCH longer. Note that the current being switched from a high Forward Current to a low Forward Current. In this situation the Trr is controlled substantially by the carrier recombination time.
We set the value of TMD accurately, measured from the time we switch off IH to the time we sample the "hot" VF at IM.
The actual effective TMD however is reduced by the switching time of the combined IH supply and the diode under test and circuit inductance.
IMPORTANT NOTE: During the switching time of IH to IM the measured VF is very unpredictable and is substantially meaningless. It is very important to avoid using a TMD in this area.
A very instructive demonstration of this effect is to use the "Cooling Plot" function of the VFS2 software and a VF40 tester. Run this plot on two rectifiers of the same or very similar size. One rectifier should be a "standard recovery" type and the other a "fast recovery". Note the significant difference between the two plots at the low TMD end. Also note what would happen if you made a single measurement of DVF on a slow rectifier at a low TMD that worked well for a fast diode.
In addition to all of this, there is an inductive coupling of the force and sense leads in long cables that effects the stability of the VF sensing circuits to acquire an accurate reading. The cables that we supply with the FEC200 are designed to minimize this effect when connected to our manual test station or to a handler where there is minimal wiring from the test clips to the ends of our cables. At some combination of IH, IM, and TMD the fixturing can be fairly critical.