UCC1805 error amplifier

Tom,

Per the datasheet internal E/A is highlighted as an Error Amplifier. Note this is a legacy part has been around for a while.
1. Can you email me your schematic.
2. Have you tried running bode plots of the various stages, complete loop, comp to output etc. and see what the response looks like.

Hello Ramesh,

Thank you for replying.  I am not allowed to send out my schematic even though it is just a flyback but I can give you some details about the circuit.  The flyback has two outputs, both are +10V but they are isolated from each other.  One of them is referenced to the same ground that the UCC1805 is referenced to and that is the regulated output.  I am using type II compensation in my feedback.  The voltage divider from the regulated 10V is 20K over 4.99K and the values of the feedback components are; 200k, 8.2nF and 6.8pF.  I have not built the circuit yet but it has been simulated using the model of the UCC1805 from TI.  The simulation shows that the outputs get to about 6V and stall out there.  The reason that the resistances listed are so high is because I thought that I was loading the E/A down but that didn't help.

I hope that this helps.

Thanks again.

Tom 

Tom,

Have you been able to run bode plots to assess the performance. I think running the bode plots will give you good insight into the issue.
Is your schematic similar to Fig. 33 on pg. 25 of datasheet.

In your case, Is one out that is regulated on the primary side - referenced to input side, and second output that is on the secondary side and isolated?

Ramesh,

My schematic is not very similar to figure 33 on page 25 but you are correct in that one output is regulated on the primary side and the other output is isolated.  Whenever I try to run a Bode plot the results don't make any sense at all.  The error amp output has its voltage phase inverted from what it should be.  I think that there is a problem with the model.  When I hook up the E/A in Spice using the TI model and I put in an input with resistive feedback the output goes out of phase with where it should be as the load gets lighter.  Is this real?  I don't know.

Thanks for any help you can give.  I was trying to simulate it to save time and money on a prototype but I think that I have wasted more time with the simulation.

Tom

Tom,

Can you email the simulation that you are running so I can have our modeling team look at it, to see whey there is an issue.
Also have you had a chance to do actual measurements in the lab using a loop analyzer.

Ramesh

Good morning Ramesh,

I do have two simulations that I can send to you.  What email address should I use?  I don't want to post them to the forum.  I have not built the circuit yet and so I don't have any test data to go by.

Thank you.

Tom Patterson

Tom,

If you can email to me at r-khanna@ti.com