The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design Languages >> Verilog-AMS >> Output of Kv when my rail-rail is 15v
https://designers-guide.org/forum/YaBB.pl?num=1284716522

Message started by kumar on Sep 17^th, 2010, 2:42am

Title: Output of Kv when my rail-rail is 15v
Post by kumar on Sep 17^th, 2010, 2:42am

Hallo all,

I have implemented a thermistor and ofet using verilog-A and designed a simple temperature sensor. I give a Vdd of -15V but the sensor output is in Kilo volts. I have attached the necessary files. Could somebody tell me what is wrong with it? Is it the simulator or the model?

Title: Re: Output of Kv when my rail-rail is 15v
Post by sheldon on Sep 17^th, 2010, 7:35am

Kumar,

Looked through the code and had one question. Have you tested
the thermistor in isolation? If you bias the thermistor from a current
source and sweep temperature does it behave as expected?
Sometimes it is better to be more methodical when developing a
model. Test the thermistor and then test the thermistor with the
OTFT model.

Best Regards,

Sheldon

Title: Re: Output of Kv when my rail-rail is 15v
Post by kumar on Sep 17^th, 2010, 8:54am

I didn't test the thermistor with current bias. But I used a MOSFET instead of OFET and the output was reasonable. So I guess the problem occurs when the two verilog-A models interact.

Title: Re: Output of Kv when my rail-rail is 15v
Post by Geoffrey_Coram on Sep 22^nd, 2010, 5:45am

Verilog-A is case-sensitive (M=Mega, m=milli), do you really want 2.5e6 Ohms as the resistance?

parameter real res_t0 = 2.5M;

If you have a milliAmp current going through that resistor, you'd get kV.

Title: Re: Output of Kv when my rail-rail is 15v
Post by Geoffrey_Coram on Sep 22^nd, 2010, 5:49am

I also don't like that transistor model -- firstly, that it's completely off for Vds>=0 (I see you expect Vdd=-15, whereas I expected Vdd>0, but still, during Newton's method, you might get a positive Vds in the model).

Have you done some Id/Vd plots at different Vgs? It looks like the current is at least continuous (for Vds<0 but approaching zero, the current does properly go to zero to match the 0 for Vds>=0). But I can't tell by inspection if the conductance (derivative of Id) is well-behaved.

Title: Re: Output of Kv when my rail-rail is 15v
Post by kumar on Sep 22^nd, 2010, 6:47am

Hello Coram,
The resistance is 2.5Gohms. But I used 2.5M.. I actually require very small current in the range of nA.. I have simulated the thermistor in matlab. I got decreasing resistance as temperature increases.(30,40,..70). But in cadence, the resistance follows with that of matlab until temperature=50 and then it behaves strange. I tried to strobe the resistance inside verilog-A model and I mean to say the calculated resistance itself is wrong after temperature=50 .

Title: Re: Output of Kv when my rail-rail is 15v
Post by Geoffrey_Coram on Sep 24^th, 2010, 1:47pm

After 50, eh? I put in this line:
$debug(" temp=", temperature, " res = ", res);

and found that at about 53, the resistance from your polynomial is negative!