Post Go back to editing

LT1990 macromodel not being accurate

After struggling a while on my +-25V current sensing circuit which uses an LT1990 for such purpose, I realize that the NMOS proposed on the datasheet for boosting the closed loop gain from x1 to x10 were affecting the behaviour of my circuit.

I found that the drain-source junction of one of the 2N7002 was clamping the negative voltages the output could take. I couldn't reproduce this on LTSpice and my guess is that the output impedance of the gain selector pins is not modelled.


  • Hi Garde707:  from just looking at the circuit, with G=1, it seems to me that as the output goes negative, the diodes will go forward.  That should have the effect of increasing the gain to 10 and also injecting an error due to diode mismatch.  If this is true, then it should have been stated in the text, "Vout positive" or such.

    It should also occur in LTspice, because the opamp+resistors models are indeed modeled as such, provided of course the diode is included in the MOSFET model.

  • Hi GlenBrisebois,

    I agree with your statement and I believe datasheet should state some kind of warning message such as the one you proposed.

    On the other hand, I've been able to reproduce the issue on LTspice. I really don't know why I couldn't do it before, perhaps I didn't dedicate enough time, but now it's clearly happening and as soon as I remove that diodefrom the loop, the issue vanishes into thin air.

    The following picture shows the LT1990 configured in x1 gain with a -1V voltage source on its input. Output should have been -1V.

    Now notice that if I remove the wire that goes to the NMOS that clamps on the feedback, the output goes were we originally thought it should have gone.

    I hope this thread help others and could be added to datasheet's notes.