Thanks a lot for all your message.
But, excuses me for my ignorance, all theses modification have to be done in the model ?
Sorry but I don't know where.
Maybe can you upload your files ?
Thanks a lot in advance
|
Oh, I forgot, you don¡¯t need parentheses around emax, either
?????????????? V={1/(1/emax+0.5*(1+tanh((v(tptc)-ts)/c)))}
?
toggle quoted message
Show quoted text
From: [email protected] <[email protected]>
On Behalf Of Bell, Dave via groups.io
Sent: Saturday, March 15, 2025 12:05 PM
To: [email protected]
Subject: EXTERNAL: Re: [LTspice] "Physical" PTC model
?
Maybe some help, Pascal:
?
I got this to compile and ¡°run¡± without errors, as below:
First, to fill in some unknown values:
?????????????? .params emax=10 c=5 ts=2
Made a voltage source to create and set a value for node ¡°tptc¡±
?
Then applying your equation to a behavioral voltage source (BV) in the F2 menu:
?????????????? V={1/(1/(emax)+0.5*(1+tanh((v(tptc)-ts)/c)))}
(This syntax didn¡¯t work with a normal V source.)
?
One set of curly braces surrounding everything that needs to be evaluated at run time.
Removing the square brackets? around ¡°v¡± in v(tptc)
?
The BV source put out about 902mV, if that makes any sense in your circuit¡
?
Dave
?
?
?
On Sat, Mar 15, 2025 at 11:07 AM, John Woodgate wrote:
I can tell you that the error messages do not mean that there is a problem with the simulation. For example, a node connected only to a current source, which has infinite impedance, appears?
not to be connected to anything.
Hello,
thanks a lot for your reply,
Ok I understand, it seems logical.
Error: undefined symbol in: "1/(1/(emax)+0.5*(1+tanh(([v](tptc)-(ts))/(c))))"
|
Maybe some help, Pascal:
?
I got this to compile and ¡°run¡± without errors, as below:
First, to fill in some unknown values:
?????????????? .params emax=10 c=5 ts=2
Made a voltage source to create and set a value for node ¡°tptc¡±
?
Then applying your equation to a behavioral voltage source (BV) in the F2 menu:
?????????????? V={1/(1/(emax)+0.5*(1+tanh((v(tptc)-ts)/c)))}
(This syntax didn¡¯t work with a normal V source.)
?
One set of curly braces surrounding everything that needs to be evaluated at run time.
Removing the square brackets? around ¡°v¡± in v(tptc)
?
The BV source put out about 902mV, if that makes any sense in your circuit¡
?
Dave
?
?
toggle quoted message
Show quoted text
From: [email protected] <[email protected]>
On Behalf Of pilou via groups.io
Sent: Saturday, March 15, 2025 11:16 AM
To: [email protected]
Subject: EXTERNAL: Re: [LTspice] "Physical" PTC model
?
On Sat, Mar 15, 2025 at 11:07 AM, John Woodgate wrote:
I can tell you that the error messages do not mean that there is a problem with the simulation. For example, a node connected only to a current source, which has infinite impedance, appears?
not to be connected to anything.
Hello,
thanks a lot for your reply,
Ok I understand, it seems logical.
Error: undefined symbol in: "1/(1/(emax)+0.5*(1+tanh(([v](tptc)-(ts))/(c))))"
|
I can't diagnose that, unfortunately.
On 2025-03-15 18:15, pilou via
groups.io wrote:
On Sat, Mar 15, 2025 at 11:07 AM, John Woodgate wrote:
I can tell you
that the error messages do not mean that there is a problem
with the simulation. For example, a node connected only to a
current source, which has infinite impedance, appears? not
to be connected to anything.
Hello,
thanks a lot for your reply,
Ok I understand, it seems logical.
But how about the
Error: undefined symbol in:
"1/(1/(emax)+0.5*(1+tanh(([v](tptc)-(ts))/(c))))"
?
?
--
OOO - Own Opinions only
If something is true:
* as far as we know - it's science
*for certain - it's mathematics
*unquestionably - it's religion
|
Perhaps it¡¯s the square brackets in ¡°(°Ú±¹±Õ(³Ù±è³Ù³¦)¡±
?
Dave
?
toggle quoted message
Show quoted text
From: [email protected] <[email protected]>
On Behalf Of pilou via groups.io
Sent: Saturday, March 15, 2025 11:16 AM
To: [email protected]
Subject: EXTERNAL: Re: [LTspice] "Physical" PTC model
?
On Sat, Mar 15, 2025 at 11:07 AM, John Woodgate wrote:
I can tell you that the error messages do not mean that there is a problem with the simulation. For example, a node connected only to a current source, which has infinite impedance, appears?
not to be connected to anything.
Hello,
thanks a lot for your reply,
Ok I understand, it seems logical.
Error: undefined symbol in: "1/(1/(emax)+0.5*(1+tanh(([v](tptc)-(ts))/(c))))"
|
Hello,
? ? It's poorly coded. will take time to examine everything.
On Sat, Mar 15, 2025 at 2:15 PM, pilou via groups.io <pilou@...> wrote: On Sat, Mar 15, 2025 at 11:07 AM, John Woodgate wrote:
I can tell you that the error messages do not mean that there is a problem with the simulation. For example, a node connected only to a current source, which has infinite impedance, appears? not to be connected to anything.
Hello,
thanks a lot for your reply,
Ok I understand, it seems logical.
But how about the
Error: undefined symbol in: "1/(1/(emax)+0.5*(1+tanh(([v](tptc)-(ts))/(c))))"
?
?
|
On Sat, Mar 15, 2025 at 11:07 AM, John Woodgate wrote:
I can tell you that the error messages do not mean that there is a problem with the simulation. For example, a node connected only to a current source, which has infinite impedance, appears? not to be connected to anything.
Hello,
thanks a lot for your reply,
Ok I understand, it seems logical.
But how about the
Error: undefined symbol in: "1/(1/(emax)+0.5*(1+tanh(([v](tptc)-(ts))/(c))))"
?
?
|
I can tell you that the error messages do not
mean that there is a problem with the simulation. For example, a
node connected only to a current source, which has infinite
impedance, appears? not to be connected to anything.
On 2025-03-15 17:53, pilou via
groups.io wrote:
Hello to the community,
I'm very interested by the "PTC Thermistors Physical model"
found in
and I have two questions about it.
?
1) When I try to simulate the example ptc_veijola2.asc it
throw me several errors:
LTspice 24.0.12 for Windows
Circuit: * C:\users\pascal\xxx\ptc_veijola2.asc
Start Time: Sat Mar 15 18:34:27 2025
Questionable use of curly braces in "begb 0 egb
?i={1/(1/{emax}+0.5*(1+tanh((v(tptc)-{ts})/{c})))}"
? ? Error: undefined symbol in:
"1/(1/(emax)+0.5*(1+tanh(([v](tptc)-(ts))/(c))))"
solver = Alternate
Maximum thread count: 8
tnom = 27
temp = 27
method = modified trap
ERROR: Node U1:N003 is floating and connected to current
source B:U1:V
Total elapsed time: 0.575 seconds.
Can you help me with that ?
?
2) I would want to simulate this PTC to "measure" a MOSFET
case's temperature. In my case the mosfet have an extra pin
that output a voltage proportional to case's temperature.
The above PTC model take the ambient temperature and have a
third pin to reflect internal temperature.
Is it possible to add an extra pin to be linked to the
mosfet temperature pin ?
?
Thanks a lot in advance,
Pascal
?
--
OOO - Own Opinions only
If something is true:
* as far as we know - it's science
*for certain - it's mathematics
*unquestionably - it's religion
|
Hello to the community,
I'm very interested by the "PTC Thermistors Physical model" found in
and I have two questions about it.
?
1) When I try to simulate the example ptc_veijola2.asc it throw me several errors:
LTspice 24.0.12 for Windows
Circuit: * C:\users\pascal\xxx\ptc_veijola2.asc
Start Time: Sat Mar 15 18:34:27 2025
Questionable use of curly braces in "begb 0 egb ?i={1/(1/{emax}+0.5*(1+tanh((v(tptc)-{ts})/{c})))}"
? ? Error: undefined symbol in: "1/(1/(emax)+0.5*(1+tanh(([v](tptc)-(ts))/(c))))"
solver = Alternate
Maximum thread count: 8
tnom = 27
temp = 27
method = modified trap
ERROR: Node U1:N003 is floating and connected to current source B:U1:V
Total elapsed time: 0.575 seconds.
Can you help me with that ?
?
2) I would want to simulate this PTC to "measure" a MOSFET case's temperature. In my case the mosfet have an extra pin that output a voltage proportional to case's temperature.
The above PTC model take the ambient temperature and have a third pin to reflect internal temperature.
Is it possible to add an extra pin to be linked to the mosfet temperature pin ?
?
Thanks a lot in advance,
Pascal
?
|
Re: proper way to simulate fluctuating load for voltage regulator
On 3/15/25 11:06 AM, Tony Casey wrote: A common way of testing regulators is to use a current source as the load, but a resistor is just as easy. Use a resistor to set the nominal load and put a current source in parallel. Then use the current source as the input for an AC analysis. Load regulation (and line) varies with frequency and that is usually a good thing to know. -- David Schultz "The cheeper the crook, the gaudier the patter." - Sam Spade
|
Re: proper way to simulate fluctuating load for voltage regulator
On 15/03/2025 16:26, john23 via
groups.io wrote:
Hello ,I am trying to test a voltage regulator by re[lacing a
steady resistor with some sort of varying resistor.
Given the circuit in the attached Zip file , what kind of
"resistor" you reccomend?
A common way of testing regulators is to use a current source as the
load, but a resistor is just as easy.
You need to decide what the range of the load current is going to
be. Your circuit presently has a load current of ~2mA, since the
output voltage 2V.
With your circuit, the maximum conceivable output current is about
12mA due to the value of R3. So a sensible nominal current test
range is 1 - 10mA.
Therefore the maximum value of R4 should be: 2V/1m = 2k¦¸, and.
..the minimum value of R4 = 2V/10m = 40¦¸.
So, make the value of R4 equal {RL}, and add the following SPICE
directive:
.step param RL 40 2k 20
Change the analysis type to .OP
When you do this, you will see that the regulation fails beyond
~4.2mA, over which the output voltage varies by 87mV, which isn't
very good. Is this enough for your application?
Only you can say why you're not using a proper LDO device. Honestly,
using a general purpose opamp would be rather better.
--
Regards,
Tony
|
Re: proper way to simulate fluctuating load for voltage regulator
On Sat, Mar 15, 2025 at 11:26 AM, john23 wrote:
BTW -- it might not matter -- but it is slightly odd that you used the higher-power BJTs in the low-power part of the circuit, and a lower-power BJT in the high-power part of the circuit.? With JEDEC part numbers, it's more common to see 2N3904 for the differential pair, and 2N2907 for the output transistor where the current could be larger, depending on load.
?
Andy
?
|
Re: proper way to simulate fluctuating load for voltage regulator
I wrote:
Perhaps a Pi-source?
I meant to write Bi-source.? In other words, a B-source with I=<value>.? You can get fancy where the I depends on its voltage, thus turning it into a controllable resistance.? That's one of many ways to do it.
?
Sorry.? ?Apparently my mind was still on Pi Day.? :-)? (Or :-( since I effectively missed it.)
?
Andy
?
|
Re: proper way to simulate fluctuating load for voltage regulator
On Sat, Mar 15, 2025 at 11:26 AM, john23 wrote:
?
Hello ,I am trying to test a voltage regulator by re[lacing a steady resistor with some sort of varying resistor.
Given the circuit in the attached Zip file , what kind of "resistor" you reccomend?
Your uploaded files have very little information.
?
Is the "regulator" the 3-transistor circuit, with R4 being the load at the regulator's output?
?
Can you? describe how you want the load to vary?? Do you know what you want it to do, or do you just want it to be random?
?
Can you describe it mathematically using a B-source?? Perhaps a Pi Bi-source?? That should lead you to what you should try.
?
Andy
?
|
proper way to simulate fluctuating load for voltage regulator
?
Hello ,I am trying to test a voltage regulator by re[lacing a steady resistor with some sort of varying resistor.
Given the circuit in the attached Zip file , what kind of "resistor" you reccomend?
Thanks.
|
Re: DanTherm model (SOAtherm)
Hello to all,
I just discovered the discussion.
I would be very interested by the SOATherm simulation.
I found it crazy that this is almost unusable as there is almost no documentation on it.
Does anyone have any news ?
Thanks a lot in advance.
|
Re: cannot plot power dissipation from elements in ltspice 24.1.4
On 15/03/2025 07:11, thereitis via
groups.io wrote:
after simulation, i do the following:
?
e.g. on a resistor R1, holding alt button and hovering over
the element R1 shows in the status bar "Left-click to plot R1
dissipation: ...".
but doing so, does not add any plot.
?
other than manually typing and adding the plot, is there a
correct way for the same. or am i doing something wrong here?
You're not, by any chance using Linux+Wine, are you? If so, try
ctrl-alt-leftclick. Apparently, alt-leftclick is assigned to
something else on most Linuxes, even though the alt key alone
produces the thermometer cursor graphic. I guess that's possible on
other platforms, so might be worth trying, if you're on Windows.
Ctrl-alt-leftclick works for me on all versions of LTspice,
including 24.1.4.
--
Regards,
Tony
|
Re: How to create IEC 61000-4-5 surge waveform in time & s behavioral ?
On Sat, Mar 15, 2025 at 02:09 AM, <mhx@...> wrote:
I'm curious as to why you want to (inaccurately) do this in the s-domain?
Your equation is available in the time domain, and it perfectly fits
the way of working of the EXP() voltage source:
Vxxx n+ n- EXP(V1 V2 Td1 Tau1 Td2 Tau2).
-marcel
Hello, Marcel:
?
At first, it's the ideal in sketch the applied circuitry in s-domain on how to attenuate the surge amplitude. Sometimes, in math domain, possibly solutions will be clear to see, though not everytime.
?
Some reasons else maybe:
?
1. For fun to see more ways to implement on simulator.
2. Curiosity the capability of LTspice's Laplace.
3. Puzzle for group ? (puzzle or not , not sure.)
?
Thank you , wish you happy and healthy.
|
cannot plot power dissipation from elements in ltspice 24.1.4
after simulation, i do the following:
?
e.g. on a resistor R1, holding alt button and hovering over the element R1 shows in the status bar "Left-click to plot R1 dissipation: ...".
but doing so, does not add any plot.
?
other than manually typing and adding the plot, is there a correct way for the same. or am i doing something wrong here?
?
thanks.
|
Re: How to create IEC 61000-4-5 surge waveform in time & s behavioral ?
I'm curious as to why you want to (inaccurately) do this in the s-domain?
Your equation is available in the time domain, and it perfectly fits
the way of working of the EXP() voltage source:
Vxxx n+ n- EXP(V1 V2 Td1 Tau1 Td2 Tau2).
-marcel
|
Bell, Dave
John Woodgate
Tony Casey