suche UC1875

[kamissick]

Looking for an LTSpice compatible model of the UC1875 to create a phase shifted full bridge converter. Anybody have any ideas or recommendations?
Thanks

Translator:
Ich suche nach einem LTSpice-kompatiblen Modell des UC1875, um einen phasenverschobenen Vollbrückenwandler zu erstellen. Hat jemand irgendwelche Ideen oder Empfehlungen?
Danke

 

[Udo]

Hi kamissick,
I'll attach a Lib-file for the UC3875 instead for UC1875. All Functions should be identical for both versions, except for the operating temperature range. Pls check the datasheets for more details.
Regards
Udo

Ich hänge mal die UC3875.lib mit .asy an. Der Unterschied zum UC1875 liegt im Temperaturbereich. Die anderen Eigenschaften sollte dieselben sein.
Gruß
Udo

 

[kamissick]

Hi kamissick,
I'll attach a Lib-file for the UC3875 instead for UC1875. All Functions should be identical for both versions, except for the operating temperature range. Pls check the datasheets for more details.
Regards
Udo

Ich hänge mal die UC3875.lib mit .asy an. Der Unterschied zum UC1875 liegt im Temperaturbereich. Die anderen Eigenschaften sollte dieselben sein.
Gruß
Udo

Do you also know if there is an averaged AC compatible model for UC3875? I am trying to figure out how I can make a Bode plot to examine stability in LTSpice...

Translator:
Wissen Sie auch, ob es ein durchschnittliches AC-kompatibles Modell für UC3875 gibt? Ich versuche herauszufinden, wie ich einen Bode-Plot erstellen kann, um die Stabilität in LTSpice zu untersuchen...

 

[Udo]

I understand what is behind your question. The UC3875 is an older Unitrode type of controller and I could not find an average model of it. To make sure what you are looking for, let me ask some questions as follows:
1. Is your apllication a DC/DC FULL-Bridge Phase Shift Converter ?
2. Operating in Voltage Mode or Current Mode ?
3. Do you have Chris Basso's Averaged Models ? Those can be used independently on a specific IC and are suitable to achieve Bode Plots. You can download them via a Google search

Welcome

For Bode Plot analysis you need to define first the following parameters:
- Switching Frequency Fsw
- Crossover Frequency fcross target. Rule of thumb > 1/10 to 1/20 of Fsw
- Evaluate the Gain of the power Stage only, at fcross as selected. You can simulate it using Basso's Average Models. Alternatively use the Laplace Transfer Function for the topology used (i.e. Full Bridge PC)
- Calculate the compensation Network, (type III amplifier for voltage mode, type II amplifier for Current mode). The compensation network can be calculated by using Venables K-method or others.
- Test the circuit > LTspiceXVII Simulation with Basso's Average Models. Check the Gain and Phase to ensure your compensation network matches with fcrossover target.

Another Possibility is to download the latest Version of LTspice. This software has not an extension, such as LTspiceXVII, the Logo is just LTspice.
With LTspice you can simulate the Bode Plot with your UC3875 circuit directly, like it is.
LTspice offers a component, named "fra", which stands for Frequency Response Analyzer.
Open your feedback Loop, i.e. the Voltage sense divider to Vout, and place the .fra-component between the Divider resistor and Vout (output voltage)
Run the Simulation by using the .fra command. Wait until the process is completed. A window will come up, showing you the results, such as:
> Bode Plot with entries for zero dB at crossover frequency, phase Margin at crossover and gain margin at its frequency.
By doing this, you can test the stability of your circuit, with the compensation network you have used.
What you do not get "free of charge" are the values of the components network itself. That's your turn to do.
------
Regards
Udo

 

[kamissick]

I understand what is behind your question. The UC3875 is an older Unitrode type of controller and I could not find an average model of it. To make sure what you are looking for, let me ask some questions as follows:
1. Is your apllication a DC/DC FULL-Bridge Phase Shift Converter ?
2. Operating in Voltage Mode or Current Mode ?
3. Do you have Chris Basso's Averaged Models ? Those can be used independently on a specific IC and are suitable to achieve Bode Plots. You can download them via a Google search

Welcome

For Bode Plot analysis you need to define first the following parameters:
- Switching Frequency Fsw
- Crossover Frequency fcross target. Rule of thumb > 1/10 to 1/20 of Fsw
- Evaluate the Gain of the power Stage only, at fcross as selected. You can simulate it using Basso's Average Models. Alternatively use the Laplace Transfer Function for the topology used (i.e. Full Bridge PC)
- Calculate the compensation Network, (type III amplifier for voltage mode, type II amplifier for Current mode). The compensation network can be calculated by using Venables K-method or others.
- Test the circuit > LTspiceXVII Simulation with Basso's Average Models. Check the Gain and Phase to ensure your compensation network matches with fcrossover target.

Another Possibility is to download the latest Version of LTspice. This software has not an extension, such as LTspiceXVII, the Logo is just LTspice.
With LTspice you can simulate the Bode Plot with your UC3875 circuit directly, like it is.
LTspice offers a component, named "fra", which stands for Frequency Response Analyzer.
Open your feedback Loop, i.e. the Voltage sense divider to Vout, and place the .fra-component between the Divider resistor and Vout (output voltage)
Run the Simulation by using the .fra command. Wait until the process is completed. A window will come up, showing you the results, such as:
> Bode Plot with entries for zero dB at crossover frequency, phase Margin at crossover and gain margin at its frequency.
By doing this, you can test the stability of your circuit, with the compensation network you have used.
What you do not get "free of charge" are the values of the components network itself. That's your turn to do.
------
Regards
Udo

1. DC-DC Full Bridge Phase Shifted (edit*: with current doubler)
2. Current Mode Control, with primary side current sensing via CS XFMR
3. I do, I am wondering how I can actually apply his averaged switch models for a PSFB and with current mode (PWMCM)


Chris provides a great example on his website of a phase shift full bridge, using voltage mode control, and he modified his PWMVM model to make it compatible with the PSFB topology. I am not sure how to modify the PWMCM model for it to work in my application. Here is his example: http://powersimtof.com/Downloads/Papers/A phase shifted average model.pdf

I also have never had good results using the LTSpice FRA, or even injecting the perturbation signal in the feedback and entering the SPICE commands for performing FRA... Phase and gain response often times is nonsensical and not remotely accurate.

Unless I can figure out how to use Chris's PWMCM model, I was probably going to analytically calculate the various transfer functions on Mathcad and perform an analytical Bode Plot, rather than relying on SPICE.

Thank you!

'

 

[Udo]

Thank's Kamissick for your answers, and for Chris's pdf-article.
I can confirm, the PWMSwitchPS Model is for Voltage mode only. So far, I've no idea myself how to modify it for Current mode usage. I can't offer such a model.

It's interesting to know you never had good results using LTspice FRA. I just opened example circuits and mentioned it as an existing offer by LTspice.

I also use MathCad for my Designs and Bode Plot calculations. Simulations in LTspice using Laplace Expressions show mostly identical results, with acceptable margins.
To verify stability, I use LTspice's .tran-Simulations fror Pulse Load operations. This gives me a "feeling" if compensation looks promising.

I appreciate your words and I suggest we keep the topic on track and post any further news, if there are any worth to mention.

Regards
Udo

 

[Udo]

Have uploaded a Design File, which might be of your interest.
Regards
Udo