Two transistor forward converter simulation won't get started

[Flyback]

Hello ,
Do you know why the attached two transistor forward converter simulation will not get started?
I have tried many adjustments but nothing works.
It uses the LTC3705 primary side controller and the LTC3706 secondary side controller
The LTspice simulation is attached.

LTC3705 Datasheet
http://cds.linear.com/docs/en/datasheet/3705fb.pdf

LTC3706 datasheet
http://cds.linear.com/docs/en/datasheet/3706fd.pdf

 

[alec_t]

The sim starts ok for me, but looks like it will take a month of Sundays (if ever) to produce a useful output voltage .

 

[Flyback]

Yes it seems to start, but then it keeps cutting out and hiccupping on and off, I can't work out why.

 

[Dick Cappels]

May be a loop stability problem.

 

[alec_t]

I've often found that sims involving inductors can grind almost to a halt at certain points in their run. Only the status line at the foot of the screen gives a clue that the sim is still running!

 

[crutschow]

Troubleshooting such a complex circuit is a real challenge (and I don't use that term lightly). Can you isolate parts of the circuit and get them to operate by themselves before connecting them together?

 

[eTech]

Try adding a small amount of series resistance to all caps...

eT

 

[ronsimpson]

FlyBack,

I have done too much work on you project.
I have schematics that work.
With yours; anytime the supply voltage on the LT3706 gets to 3.8 volts the thing blows up. This is where the first information comes through the pulse transformer. If I make my own supply for the 3706 and ramp it up the same thing happens.

I think where the two ICs connect together is the problem.
Not error amp problem, not soft start problem, not current limit.......
I do not have it working but.....look at this.

 

[Flyback]

well you certainly did an awful lot better than I did, but as you say, its still not working.
I am confused by it. The switching frequency doesn't correspond to the datasheet, it doesn't say what is the timeout period, also, it sometimes says the series cap at the primary side of the pulse transformer is needed, and sometimes it appears it I not.

 

[eTech]

Hi

I'm not much for forward converters but I played with your circuit a little today and discovered if I make all of the following changes:

1. Reduce the load current by changing the load resistor to, say, 1k.
2. Change the RUN/SS cap to 100n. (longer ramp time)
3. Insert a 37.5k resistor from the 3706 FS/Sync pin to ground (I think this changes the OSC Freq to 50khz)
4. add IC=0 to the spiceline field of C9.
5. Change the upper pulse transformer ratio to .6:1 (360uH|1mH)
6. Change the V1 source to PULSE (0 36 0 1u 1u)

whew...

Then the output eventually rises to about 32v then levels off.

Don't know what to make of this and I could be doing something silly, but I thought I'd share the info..

The simulation takes a heck of a long time..
I stopped it at 959.269 seconds...

See output graph attached.

eT