Working on push-pull a little help needed

forest · 2021.11.21, 10:06:17

Would is be appropriate to ask a little help with LTSpice ? I'm working on isolation dc-dc converter from 12V battery to 17V 3A. Tested on LTSpice it looks like working too good to be true. I think it would be good to drive mosfets.
If it's offtopic just let me know I would remove this topic.
Waiting for answer

Centraflow · 2021.11.21, 12:48:06

Quote from: forest on 2021.11.21, 10:06:17
Would is be appropriate to ask a little help with LTSpice ? I'm working on isolation dc-dc converter from 12V battery to 17V 3A. Tested on LTSpice it looks like working too good to be true. I think it would be good to drive mosfets.
If it's offtopic just let me know I would remove this topic.
Waiting for answer

It would help to see what you have to start with!

Regards

Mike

forest · 2021.11.22, 09:30:31

I'm working currently on 2 schematics with two models , one simplified and one quite complex. Soon I'll attach them here for corrections and usage by others. I want to create a base schematic for DC-DC push-pull converter which anybody could pick and modify to own needs having reliable simulation.

forest · 2021.11.22, 22:03:32

Attached first simulation. Can you check if all is right especially TL431 pin assignment ?

partzman · 2021.11.22, 22:27:15

Quote from: forest on 2021.11.22, 22:03:32
Attached first simulation. Can you check if all is right especially TL431 pin assignment ?

Forest,

Good work but I see no TL431 in this sim's schematic!

Otherwise, the efficiency is Pout/Pin = 28.548/31.802 = 89.76% with an output voltage of 16.896 from 6ms to 20ms (stabilized). However, your coupling of L1,L2 and L3 at k=1 is not possible to achieve in the real world as you well know plus, you have no inter-winding or inter-turn capacitance specified which will affect the efficiency. There will also be additional resistance of each winding due to skin effect at ~42.44kHz.

There might also be a better mosfet for M1 and m1 as they have a Cgs capacitance of 1.4nf and a Cgd range from 80pf tp 2000pf.

Regards,
Pm

forest · 2021.11.23, 16:40:13

Partzman

tl431.lib is included in zip package, but I'm not sure if I included it correctly. My LTSpice knowledge is limited. Th SG3525 model I took from internet and it seems to has no soft-start simulated so it shows a large spike at start. Feel free to modify schematic to be standalone with all files in the same directory.
I want it to be a simple base for me and other users to create isolated push-pull for experiments. I know the coupling in reality is below 1, that's fine. I'd like to make a simplified method to create push-pull drivers with step included.

Can you help me with fixing schematic ?

forest · 2021.11.23, 18:36:44

btw. I'm working under Kubuntu and wine using LTSpice XVII . Having Core duo2 processor it's a bit slow

forest · 2021.11.27, 13:06:26

Attached fixed simulation. Now PWM is working good

Little help needed only with snubber on primary side. Primary is of course wound in bifilar , then on top wound secondary and then primary is closed . Now I'm preparing to build it. Awaiting your comments. You can use this simulation as a base for other push-pull isolated drivers. Someone with greater knowledge may post here a simulation when the output voltage is above 37V ( the limit of TL431). Please help if you can.

partzman · 2021.11.27, 18:57:57

Forest,

I had to run the sim in ltSpice XVII as it wouldn't load in LtSpice IV for some reason. I personally prefer version IV myself as I've used it for such a long time.

Anyway, the converter looks to regulate well with the switched 10 ohm load every 1ms. With the bifilar wound primary, there will be some capacitance between L1 and L2 which will affect the switching losses with lower primary rise and fall times. Probably will affect the snubber values as well.

Regards,
Pm