¿ªÔÆÌåÓý

Probing phase


Steve Kale
 

¿ªÔÆÌåÓý

Ok I see that in the log. ?But I don¡¯t get prompted with any error for which I must ¡°press ok to continue¡­¡± which would lead me to look at the log in the first place.

Fine¡­ so (1) there¡¯s an error in the?2SC6144SG sub circuit from On Semi which raises log errors with respect to that sub circuit if all the Qd44vh10 pass transistors are replaced with?2SC6144SG. ?(2) If they are not, then LTspice tells me there¡¯s a missing Qd44vh10 model definition when indeed there isn¡¯t. ?(3) If all pass transistors are Qd44vh10 and the Spice directive in relation to the 2SC6144SG sub circuit is not ¡®commented out' in the schematic then LTspice tells me the model statement for Qd44vh10 is missing when it is not. (4) If all pass transistors are Qd44vh10 and the?2SC6144SG subcircuit statement is commented out of operation then all works just fine.?

For (1) the problem is in the log but with no ¡®flag' that the log should be checked. ?With (2) and (3) LTspice points me to a problem that doesn¡¯t exist and not to where the problem does exist.?

It took a lot of backtracking to identify that commenting out the?2SC6144SG sub circuit statement solved the problem. Of course then an examination of the material that needed to be commented out was readily done. It might well have been more obscure to identify had the?2SC6144SG sub circuit been one of many included via .inc statements. Better that LTspice actually alerts the user to the problem at hand rather than to an area which is just fine.?

Let¡¯s perhaps end this sidebar regarding the?2SC6144SG sub circuit here. The issue is now fixed / no longer needed. I still have outstanding questions with respect to probing phase for which I would be very grateful of assistance.?

Thanks in advance

Steve


EDIT: ?just saw John¡¯s post. ?The Qd44vh10 model is fine with respect to the device name / model name. ?The problem is with the 2SC816144SG sub circuit statement (with the .endl error noted by Philippe) and that LTspice points to an error with a missing Qd44vh10 model definition rather than an error with the 2SC816144SG sub circuit statement.?


On 10 Aug 2016, at 20:44, basier.philippe@... [LTspice] <LTspice@...> wrote:

Hello

With all Qd44vh10 replaced
If I replace .endL? by .ends I get no error or warning.

If I keep .endL I get :


Error on line 1138 : .endl 2sc6144sg
??? ?Unknown control card
Error on line 1138 : .endl 2sc6144sg
??? ?Unknown control card
Error on line 1138 : .endl 2sc6144sg
??? ?Unknown control card
Error on line 1138 : .endl 2sc6144sg
??? ?Unknown control card
Error on line 1138 : .endl 2sc6144sg
??? ?Unknown control card
Error on line 1138 : .endl 2sc6144sg
??? ?Unknown control card

Regards
PhB



 

Hello Steve

You wrote :
"? With (2) and (3) LTspice points me to a problem that doesn¡¯t exist and not to where the problem does exist. "

A/
You are wrong for case (2) [ mixed Qd44 and 2sc ]. There is

Error on line 1137 : .endl 2sc6144sg

B/
" I
t might well have been more obscure to identify had the 2SC6144SG sub circuit been one of many included via .inc statements. "

Definitively not :
If you create a file with the
2SC6144SG subcircuit ( say 2SC6144SG.sub ), even with the uncorrect command .endS , comment the subckt command lines inserted in the schematic and use
.inc 2SC6144SG.sub? ? ? ? ? ? ? ?? (
and that's the right way to work? )
there is no problem in ALL configuration.


Regards
PhB


 

Hello Steve

You wrote :
"? With (2) and (3) LTspice points me to a problem that doesn¡¯t exist and not to where the problem does exist. "

A/
You are wrong for case (2) [ mixed Qd44 and 2sc ]. There is

Error on line 1137 : .endl 2sc6144sg

B/
" I
t might well have been more obscure to identify had the 2SC6144SG sub circuit been one of many included via .inc statements. "

Definitively not :
If you create a file with the
2SC6144SG subcircuit ( say 2SC6144SG.sub ), even with the uncorrect command .endS , comment the subckt command lines inserted in the schematic and use
.inc 2SC6144SG.sub? ? ? ? ? ? ? ?? (
and that's the right way to work? )
there is no problem in ALL configuration.


Regards
PhB


Steve Kale
 

¿ªÔÆÌåÓý

Hi

Any takers for my remaining questions below?

Thanks in advance

Steve

On 9 Aug 2016, at 19:57, Andy wrote:
?
?A differential measurement doesn't show the phase between the two points. It would show you the differential voltage between those two points.??
?
?It can show you the magnitude and the phase of that one voltage. But it definitely would not be the phase between one voltage and the other.
?

Ok. "Phase of that one voltage"...Phase of the differential? (Not so intuitive to a newbie.)

I¡¯ve clearly been getting confused between whether the graph portrays phase (excluding the op amp's 180 degrees) or phase margin directly. Various tutorials are rather?¡°loose¡±?in this respect.

Plus the impact of the minus sign. In the example I posted, if it is not applied then the graph would read +89 degrees (phase margin?). If it is applied then the graph reads -91 degrees. It seems odd to talk about a negative margin but Helmut had some time ago told me to plot with the minus sign. Helmut¡¯s previous advice:

"1. why -V(Vout)/V(Y) rather than V(Vout)/V(Y) ?

You would get a positive phase instead of a negative phase and we know from experience the phase must be negative. :-)"




 

One thing to remember is that in reality, a single signal doesn't have a phase, and phase is only meaningful in frequency domain, or with a single frequency excitation. This is because in reality we don't have an absolute time 0 (or it is so far ago it isn't meaningful to use it). In time domain simulations we have an artificial T0.

Since in reality, all phases are relative to something, changing the something and direction is mostly a matter of convention and what helps the intuition.

On 8/13/16 9:18 AM, Steve Kale stevekale@... [LTspice] wrote:

Hi


Any takers for my remaining questions below?

Thanks in advance

Steve

On 9 Aug 2016, at 19:57, Andy wrote:

A differential measurement doesn't show the phase between the two points. It would show you the differential voltage between those two points.

It can show you the magnitude and the phase of that one voltage. But it definitely would not be the phase between one voltage and the other.
Ok. "Phase of that one voltage"...Phase of the differential? (Not so intuitive to a newbie.)

I¡¯ve clearly been getting confused between whether the graph portrays phase (excluding the op amp's 180 degrees) or phase margin directly. Various tutorials are rather ¡°loose¡± in this respect.

Plus the impact of the minus sign. In the example I posted, if it is not applied then the graph would read +89 degrees (phase margin?). If it is applied then the graph reads -91 degrees. It seems odd to talk about a negative margin but Helmut had some time ago told me to plot with the minus sign. Helmut¡¯s previous advice:

"1. why -V(Vout)/V(Y) rather than V(Vout)/V(Y) ?

You would get a positive phase instead of a negative phase and we know from experience the phase must be negative. :-)"
--
Richard Damon