UCC28070: ERROR(ORPSIM-15090) and WARNING(ORPSIM-15256)

Hi Denis,

I have contacted another resource within TI to check this issue.
But please understand that the responses may be delayed due to holidays.
I will also follow up on this post. Thanks.

Regards,
Teng

Thanks. When do you have holidays? in rusia official holidays and non-working days 30 december-08 january, but I work in this days in home, outside office and I will track information on this post as well. If I find a solution, I will definitely describe the principle of its solution.

Thanks. Errors disappeared after offclude analise Fourier. But my simulation schematic doesn't work. Voltage on Chanel A and Chanel B equals some microvoltes. I'm tested with and without boost_block (parallel on my power parts). If used reference designe voltage on Chanel A and Chanel B have a sinusoidal vaweform whis amplitude 400mV. Because, i'm understand model UCC28070 Trans it's doesn't work on different designe power part, except boost_block?

I'm create new project it's clear (but non use slum097 profiles reference designe). In a channel A and Channel B is a 700mV DC. I'm used Low-side driver UCC27524, unused pins (IRIPA/IRIPB/F_Khz) edit properties FLOAT to RtoGND.
Waveforms simulations in my project and full archive my projectis attached on this message.

HV_PFC_5500W-2019-01-09T15-27.zip

Good day! I'm used in my project Ultra fast IGBT, reference designe on slum097 willn't be abble simulate this event.
I wish to test the work of my design before the release of the pre-production printed circuit boards. Help me please.

Hello Denis

I'll try to run you sim file as soon as possible but it may be Wednesday before I am able to get to it.
In the meanwhile I can only suggest that you try a 'divide and conquer' approach because that is likely to be the approach I will take. For example, the first thing I usually do is to replace the switches (IGBT in your case) with a voltage controlled switch. Use the GDA and GDB outputs to control the switches directly. This eliminates the IGBT and gate driver models from the problem.

Regards
Colin

GDA and GDB it's V_CHA and V_CHB on first screen before posts. The conditions for starting the chip are met.

Hello Denis

I opened your schematic today - I still have to correct some import problems - the file is not finding the model for the UCC28070 and some other parts for example.

However, can you please check the connections for the bridge rectifiers U1 and U2 in your schematic. I simulated this sub-circuit on its own (in PSIM) and HV remains at 0V !. Please check the voltage at the +HV node and let me know if it is OK or not - obviously it should be a rectified version of the input.

I'll wait for your confirmation before proceeding.

Regards
Colin

Model UCC28070 attache in progect and you need re-changed this model in pspice setting on library pages.

Waveforme input signal +HV_IN attache in this message. But waveform input signal doesn't sinusoidal (positive half-wave).

In Orcad 17.2 haven't model FGH60N60, attached new progect whis STGW80V60DF.

hv_pfc_5500w-2019-01-17T06-39.zip

Hello Denis

The +HV_IN waveform looks ok - the reason it does not fall back to zero is that there is too little load on the capacitors on this net so they don't 'follow' the sinusoidal input waveform all the way to zero. This effect will dissapear once the PFC stage is up and running (or if you add a resistor load to the net).
It will probably be tomorrow before I get an opportunity to open your simulation files.

Regards
Colin

Good days. I dosen't have some problems whit the project used BOOST_AVG block. My problemes connects with using own FET (IGBT) and Inductor ( 60KMu cores), this project doesn't work in this model UCC28070. I need modeling transient processe in FET and feedback, and modeling magnetic parametrs in my cores, but doesn't want to search and stydy this process on finished devises.
e2e.ti.com/.../560486
I read this post, but doesn't search solution my problems.
I understand is impossible used this model UCC28070 whis own magnetic core (inductors) and fet, only used boost_avg blocks?

Load in my project it's 5500W, Rload=27.67 Ohm, input rectufier should be hold RMS current 50+A.

Good day! I need simulate magnetics parametrs and transient parametrs my fet(or IGBT) and my core (inductor) materials.

Boost_avg block doesn't allow to consider this parametrs.

I believe that the device will work, if choose some parametrs boost_avg dlock in my circuit, but this study will not be complete.
I have two quetions:

1) My project don't be resolved if is not used boost_avg? (more rhetorical question)
2) How can I get an unencrypted model and is it possible?

Thanks. 

Hi Denis,

Below are the answers to your questions 

1. Yes, we have to use BOOST_AVG block to get correct result.

2. For the device, unfortunately an unencrypted SPICE model is not available for distribution. I apologize for any inconvenience.

Thanks & Regards,

Arpan Gupta

Thank you, but your answers didn't resolve my problems.

Hello Denis

Arpan is the expert for the device model and knows how it behaves much better than I do. However, it seems that the problem is that the Average model isn't going to give you the level of detail of the switching waveforms that you are looking for. Have you considered using the transient model for this device which is available on the UCC28070 page ?. Of course you can expect that the transient model will run more slowly than the average model but that's the tradeoff that has to be made.

The transient model will give you OUTA and OUTB signals which you can use to drive the IGBTs and the power stage. (the simulation may run faster if you turn one of the phases off The exact behaviour of the switching stage, IGBT plus rectifiers and the losses in the magnetics depend on the accuracy of the models for these devices and I'd suggest you discuss this with the respective manufacturers. You should also include an estimate of the parasitic inductances and capacitances due to the layout of course.

If I had to do this myself, I would start with the best models for the inductor that I could find and use simple voltage controlled switches for your IGBT and diode. Then run the system at 50% duty cycle from a simple square wave source (no controller !) and Vin = 50% of Vout. That at least would allow you to judge if the magnetics model was behaving reasonably accurately. Once you were happy with that you could look at 2 or 3 other input/output/Duty cycle conditions and then start to introduce some more complexity. First, an IGBT, then a Diode and then the UCC28070 controller.

This approach allows you to gain confidence in the system model you are building rather than getting bogged down in non-essential details.

Finally, despite what I said earlier about a square wave source, I do think that your choice of the UCC28070 controller is the best one for your application.

Please let us know how you get on.

Regards
Colin