# Interpolate/extrapolate a gate charge curve from two different curves

Gate charge measured data. I have gate-charge measurement plots and Excel data for two operating points for a power MOSFET.

1. Vds = 60 V, Id = 50 A
2. Vds = 600 V, Id = 8 mA

Can I generate a plot with the operating point, 600 V, Id = 50 A from this curve?

• …and yet, no data for us to study. May 22 '15 at 1:35
• Welcome to Mathematica.SE! I hope you will become a regular contributor. To get started, 1) take the introductory Tour now, 2) when you see good questions and answers, vote them up by clicking the gray triangles, because the credibility of the system is based on the reputation gained by users sharing their knowledge, and 3) remember to accept the answer, if any, that solves your problem, by clicking the checkmark sign. May 22 '15 at 1:37
• i've excel data.. how do i attach here?
– Ram
May 22 '15 at 1:50
• You can upload to sendspace, or (more preferable) convert your file to *.csv and put the resulting text on Pastebin. Then, link to it here. May 22 '15 at 1:57
• Thanks a lot. I'll upload that data here from my flash drive in like 10-15 min.
– Ram
May 22 '15 at 2:01

Based on the information you gave, you definitely can NOT. Especially since this is a highly non-linear device (as you mentioned it is a power MOSFET).

You need first to have a mathematical "model" that describes the Id, Vds, and gate charge relation in your device. In fact, this model need to be a predictive model, i.e., one that can be extrapolated to inputs outside your measurement set with sufficient confidence. (May be look at spice models or something similar?)

Once you have such a model, you need to fit its parameters to the measured data (which I suspect only 2 Id-Vds curves might be too few in this case). Then, use the model to predict whatever you need.

Thus the short answer: without a model, you can not extrapolate the behavior of a non-linear device to arbitrary points away from measurements.

• Guys, relax. @ram Bichoy meant no disrespect with that comment. Some community members remove "Thanks" from the body because the Q & A medium doesn't go well with salutations and expressions of gratitude unlike forums/chatrooms/email. I personally don't think it's all that bad to leave a one word "thanks" in the post but I probably have, over the years, removed it during edits and people here generally tend to respect an editor's decision (unless there are compelling reasons to revert). You could always thank the folks who answer the question using comments.
– rm -rf
May 22 '15 at 3:04
• And @Bichoy, I suppose "chit chat" could rub people, especially new users who are unfamiliar with the site, the wrong way. So let's try to avoid that :)
– rm -rf
May 22 '15 at 3:05
• Maybe we can try a bit more help: Bichoy, did you happen to look at the data the OP finally supplied? May 22 '15 at 3:23
• @Ram, I understand your description ... but, sorry, I don't have a better answer than the one above. To accomplish what you are trying to do, it will boil down to predicting the behavior of this transistor for the biasing you looking for. Given it is a non-linear system, this might not be possible, unless you know the physics behind it (=model) or you have lots of testing data that can be applied to system identification techniques. Either way, if this is truly what is asked for, it is an off-topic here on Mathematica. May 22 '15 at 3:50
• I would suggest that you put all the relevant info from the comments into the question. Optimally you could provide a small dataset in the text. Downloading and importing is much less convenient than having working code out of the box. May 22 '15 at 4:23