I need an explanation for this error in ADS
I have got this error many times trying to simulate an oscillator and I would be glad if someone could explain me what it means:
I get the "timesteap is too small" error and then, this list appears:
TRAN Tran1[1] <nxp_lib:circuit complet 2NPN proves 1.8V:schematic> time=(0 s->400 ns)
iter = 1 RHS I= 1.916e+08 V= 8.736e+08 k fail RHS I[Vs]= 2.187e+08 (4.9e+02x) dV= 1.406e+06 dI= 1.070e+01
iter = 2 RHS I= 1.497e+04 V= 4.510e+01 k fail RHS V[Var1.L1.i]= 3.207e+01 (1e+03x) dV= 1.724e+00 dI= 3.240e-04
iter = 3 RHS I= 1.434e+04 V= 4.898e+01 k fail RHS V[Var1.L1.i]= 3.207e+01 (1e+03x) dV= 1.148e+00 dI= 2.212e-04
iter = 4 RHS I= 1.410e+04 V= 4.393e+01 k fail RHS V[BJT1.L3.i]= 3.518e+01 (1e+03x) dV= 5.746e-01 dI= 1.668e-04
iter = 5 RHS I= 1.409e+04 V= 4.898e+01 k fail RHS V[BJT1.L3.i]= 2.882e+01 (1e+03x) dV= 5.855e-02 dI= 1.693e-04
iter = 6 RHS I= 1.400e+04 V= 4.393e+01 k fail RHS V[BJT1.L3.i]= 3.518e+01 (1e+03x) dV= 5.772e-01 dI= 8.929e-05
iter = 7 RHS I= 1.385e+04 V= 4.898e+01 k fail RHS V[BJT1.L3.i]= 2.882e+01 (1e+03x) dV= 5.746e-01 dI= 5.642e-05
iter = 8 RHS I= 1.430e+04 V= 4.393e+01 k fail RHS V[BJT1.L3.i]= 3.518e+01 (1e+03x) dV= 5.741e-01 dI= 1.987e-04
iter = 9 RHS I= 1.397e+04 V= 4.898e+01 k fail RHS V[BJT1.L3.i]= 2.882e+01 (1e+03x) dV= 8.614e-01 dI= 1.092e-04
iter =10 RHS I= 1.425e+04 V= 4.393e+01 k fail RHS V[BJT1.L3.i]= 3.518e+01 (1e+03x) dV= 2.873e-01 dI= 2.015e-04
iter = 1 RHS I= 1.916e+08 V= 8.736e+08 k fail RHS I[Vs]= 2.187e+08 (4.9e+02x) dV= 1.406e+06 dI= 6.686e-01
iter = 2 RHS I= 8.429e+04 V= 5.647e+02 k fail RHS V[Var1.L1.i]= 1.439e+03 (1e+03x) dV= 4.604e+00 dI= 3.090e-04
iter = 3 RHS I= 6.249e+04 V= 4.834e+02 k fail RHS V[Var1.L1.i]= 6.093e+02 (1e+03x) dV= 1.378e+01 dI= 2.232e-04
iter = 4 RHS I= 4.111e+04 V= 4.072e+02 k fail RHS V[Var1.L1.i]= 6.093e+02 (1e+03x) dV= 1.383e+01 dI= 1.468e-04
iter = 5 RHS I= 1.040e+05 V= 4.834e+02 k fail RHS V[Var2.L1.i]= 2.335e+03 (1e+03x) dV= 3.143e-01 dI= 3.700e-04
iter = 6 RHS I= 6.706e+04 V= 4.072e+02 k fail RHS V[Var2.L1.i]= 1.761e+03 (1e+03x) dV= 1.378e+01 dI= 2.233e-04
iter = 7 RHS I= 4.111e+04 V= 4.834e+02 k fail RHS V[Var2.L1.i]= 2.335e+03 (1e+03x) dV= 1.380e+01 dI= 1.468e-04
iter = 8 RHS I= 1.040e+05 V= 4.072e+02 k fail RHS V[Var2.L1.i]= 1.761e+03 (1e+03x) dV= 3.143e-01 dI= 3.700e-04
iter = 9 RHS I= 6.706e+04 V= 4.834e+02 k fail RHS V[Var2.L1.i]= 2.335e+03 (1e+03x) dV= 1.378e+01 dI= 2.233e-04
iter =10 RHS I= 4.111e+04 V= 4.072e+02 k fail RHS V[Var2.L1.i]= 1.761e+03 (1e+03x) dV= 1.380e+01 dI= 1.468e-04
iter = 1 RHS I= 1.916e+08 V= 8.736e+08 k fail RHS I[Vs]= 2.187e+08 (4.9e+02x) dV= 1.409e+06 dI= 4.179e-02
iter = 2 RHS I= 1.026e+06 V= 9.981e+03 k fail RHS I[X4.HPFET1.d]= 4.972e+06 (2.1x) dV= 1.617e+03 dI= 2.276e-04
iter = 3 RHS I= 7.239e+05 V= 1.684e+04 k fail RHS I[X4.HPFET1.d]= 3.253e+06 (1.4x) dV= 2.940e+02 dI= 1.669e-04
iter = 4 RHS I= 5.059e+05 V= 8.591e+03 k fail RHS I[X4.HPFET1.d]= 2.416e+06 (1x) dV= 1.471e+02 dI= 1.133e-04
iter = 5 RHS I= 3.011e+05 V= 1.684e+04 k fail RHS I[X4.HPFET1.g]= 4.561e+05 (4.6x) dV= 2.206e+02 dI= 5.625e-05
iter = 6 RHS I= 1.810e+05 V= 8.591e+03 k fail RHS I[X4.HPFET1.g]= 9.427e+05 (9.6x) dV= 7.309e+01 dI= 4.336e-07
iter = 7 RHS I= 1.607e+04 V= 1.684e+04 k fail RHS V[BJT1.L6.i]= 8.571e+04 (1e+03x) dV= 7.505e+01 dI= 8.549e-07
iter = 8 RHS I= 2.905e+05 V= 8.591e+03 k fail RHS I[X4.HPFET1.g]= 1.392e+06 (14x) dV= 7.726e+01 dI= 5.700e-05
iter = 9 RHS I= 2.285e+05 V= 1.684e+04 k fail RHS I[X4.HPFET1.g]= 4.819e+05 (4.9x) dV= 2.944e+02 dI= 5.462e-05
iter =10 RHS I= 1.744e+05 V= 8.591e+03 k fail RHS I[X4.HPFET1.g]= 9.284e+05 (9.4x) dV= 7.452e+01 dI= 2.815e-06
It seems some convergence problem with this devices, isn't it? Any tips to fix it?
Ty very much!
If you post the circuit, I can say something to overcome the problem..
Hey BigBoss, the circuit is almost the same of an old post where you try to helped me. https://www.edaboard.com/thread305124.html
Now, i'm trying to understand what those kind of error means. The error appears when I change the 1.8V of the bottom for a pulse DC of 1.8V. The rise/fall time and the wide of the pulse doesn't matter.
In oscillator designs, it's sometimes useful to use initial conditions (IC) at sensitive points of your circuit.
But Coust already have Ipulse source for kick start.
Why did you placed pulse source to the gate of current source? Do you want to examine switching on/off behavior?
Mhmm, ok. Is it error from DC-solution before transient simulation and from transient simulation itself? Maybe Coust can post a better pic of his setup, my eyes are not the best anymore, the age .
Looks like it. He has enormous currents and voltages in log :)
I also can't well consider this schematic. I have to hard strain my eyes for it
Maybe better do it not as a screenshot? For example, you can print to file directly from "print' menu ADS.
Maybe he can also annotate the dc-solution if possible, ahh obviously not. But he can try to change the converge criteria in "OPTIONS".
Hello and ty for your answers! I took a print of my circuit in .pdf format:
document1.pdf
I thought that this error appears when one of the variables of the circuit rise too fast, and it's probably what's going here. I don't think that I have a physical problem here but a simulator's one.
This is not happening only when i'm trying to use a pulse source. If I change the value of the bottom resistence from 8 to 7 maintaining the other variables constant I got this kind of error too. Am I doing something wrong? I'm quite new in this oscillator's world and I could make rookie mistakes.
Any tip will be aprreciated.
First question: why you use a E-PHEMT in the bottom current source, any reason, a paper? If you look at the dc-voltages at this transistor, VDS<1.1V, VG=1.8V, mhm? No idea if it's the problem, but maybe replace it with an ideal mos or bjt to examine roughly the dc-voltages.
...And why value of resistor is 8 kOhm? Maybe better to get smaller value of resistor and lowe gate voltage...
Please tell us WHEN exactly the error arises. In DC-simulation or in transient simulation. When it's in DC, try a test simulation with deactivated pulse current source.
Such circuits have always convergence errors.I have explained before in a different post:The simulators don't like uncertainties that include pulse generators,fast steeping square waves,switches,floating nodes,impacted transients etc.
The circuit should be nth order (Non)Linear Time Invariant as much as possible in time and frequency domain.Otherwise, you will see convergence errors again..