Push Pull converter Low output voltage?

[Xerxsea]

Hello, I am messing around with learning the push pull topology in LTSpice and am getting very confused with my results. My attempt is to model a circuit based on the attached document
I'm aware in the proposed circuit he uses a optocoupler to drive the gates but in mine I wanted to keep it as simple as possible and move on from there so I have not included the optocouplers, or any relays.

The probes in the attachment with the circuit are located right between the transformer and diodes. I was expecting to see a more positive voltage but I get a square wave ranging from 0.6v to -5.7v. That's barely enough to pass through the Schottky diode. Yet in all the diagrams I've seen they show it becoming forward biased and passing full voltage to the load/ lc filter.
I've tried so many different things to change it and nothing really seems to be doing the trick. I'm very novice in electronics and even more so in using spice but am trying to learn so any help or tips would be greatly appreciated! I've attached some other probes in different pictures i thought might be useful as well.

 

[d123]

Hi,

Are the connections around the ground symbol, V1, V2, and both NMOS, and in the other direction from the batteries, correct? Are the batteries the right way around or should they be (2.4V-gnd)-(gnd-2.4V)?

 

[Xerxsea]

Based on the PDF circuit diagram all of those are connected in the right way, however that diagram isnt necessarily included the entire circuit and may be simplified. The only exception is gnd as it doesnt show where they connect their ground for their pwm generator gnd to the cells/battery pack, it's my understanding that the logical place for gnd would be on the cell that is providing the charge into the push pull converter (left side)

The cells are connected in series, so yes they should be - + - +. The idea of the circuit is to transfer energy from one cell to the other.

 

[cupoftea]

Hi,
Here is a free SMPS course...in the "little" folders one, there is a folder called pushpull.......there are some working LTspice sims of pushpulls in there should you wish....

SMPS course_Big folders
https://drive.google.com/file/d/17ij_pQLkAzzYeL_TRNci86sXHMirKHmD/view?usp=sharing
SMPS course_little folders
https://drive.google.com/file/d/1WvElhDNfHI0EBei4rouSNEdWWTO1oCtJ/view?usp=sharing

 

[Xerxsea]

Thank you cupoftea, This looks like it has much more detailed information than I could ever find on the small papers I was reading about push pull converters!

 

[Easy peasy]

for your simple sim, as the Vout charges ( i.e. charges the large cap ) the current will fall, we can't see the full detail of the volt sources, the fets, or the R of the diodes, likely these contribute to the limited current even though the Tx is trying to step up....

 

[cupoftea]

Thats an interesting report.....i'll read it fully and get back hopefully....the pushpull used has no output inductor here. (or input inductor)..so presumably they are relying on the leakage inductance of the transformer. The pushpull isnt so good with high leakage inductance in the txfmr. (but maybe if cct voltages are low and fet voltages are high it wont matter too much)
I will read further and get back.
They could of course be relying on stray res of the cct etc.

AYK, the circuit you show is usually done with a flyback....".since flybacks suck in and then blow out current"....this makes them ideal for shunting charge from one battery to another.......a pushpull, as you know, uses the "txfmr action"........as such, the cct you show is a little bit akin to shorting the two batteries together...but doing it a bit more "Politely", , so to speak, via the push pull txfmr.

--- Updated Jun 10, 2022 ---

...Thanks for sending it, Just had a speed read of the report...very thin on detail, no detail about transformer winding or leakage L or even stray R.
Apologies for being cynical, but if there is an effective method here, it looks to me that the writers only wrote that report because their supervisor told them to.......they likely wanted to reveal as little of their "magic" as they could get away with.......it makes you wonder, how a transformer, which is effectively outputting a voltage source, can be used in this way.....certainly i would say an "ideal transformer" (one with zero R and zero leakage) cant really be properly used like this.

BTW, AYK, if the batteries are separate (not in a series stack etc), then no transformer is needed, and just a current source type converter to push current from one cell to the other.

 

[Xerxsea]

you are hardly being cynical, the report is definitely omitting some details. it does list some parts that were used in it though, including the transformer. However one thing I found odd was they list PA6383-AL as their transformer which has a turn ratio of 1:1.5. But in their initial explanation of choosing a transformer they list the transformer turn ratio as 1:1.2.

i've actually never seen this circuit with the flyback topology. I always see the Cell to pack being used with flyback not cell to cell.

I'm going to attempt to learn a lot more about SMPS circuits and transformers in general but I may just be well out of my depth here. Is there something inherently wrong with using the txfmr to "short" the two batteries together? I'm just learning all these concepts still but couldn't I just use an output inductor like a normal push pull converter to limit the current going out into the second cell? does the transformer txfmr not limit the current itself via its inductance?

Yeah unfortunately they have to be in series as they are cells in a battery pack.

 

[cupoftea]

does the transformer txfmr not limit the current itself via its inductance?

....well, in an ideal transformer, with no leakage inductance, and no series resistance, the transformer coils do not offer any current limiting whatsoever.....this is the magic of transformers........only the so called "magnetising current" sees the transformer coil inductance...the "power current" does not see it at all.
Of course , in an ideal txfmr, the magnetising inductance is infinite, so the magnetising current is zero.

So yes, it is a bit odd with transformers....the power current doesnt "see" the transformer coil inductances at all...

So yes, they must be relying on the txfmr leakage inductance and/or the txfmr and circuit stray resistance. That requires various enabling precautions to be taken.

If you use an output inductor, then you can do the current limiting thing yes.......so you would sense the current to the cell and put that in a feedback loop so you get the balancing current that you demand.

Flyback can do cell to cell or cell to pack too.....flyback cheaper on parts count, but if power is high, the pushpull tends to need a smaller transformer.....but the size of transformer and output inductor might be bigger than the flyback transformer...(AYK, the flyback transformer is really called a "coupled inductor"......as there is no "transformer current" in a flyback, only "magnetising current".

Thanks yes , see it now, PA6383-AL

https://www.coilcraft.com/getmedia/9ef973b6-1838-461f-8c3a-c17056e5359f/pa6381.pdf

..so to simulate it, be sure to include the 190nH leakage inductance, and the coil resistances...and use the same switching frequency as them. Also, be wary of needing to snub the leakage inductor spike.....maybe they have relied on the transformer interwindig capacitance to quell that, who knows, they dont say.

--- Updated Jun 10, 2022 ---

After another read, it looks like they are using the transformer and circuit resistances to "limit" the current......and since the cell voltages are pretty much the same anyway...and since the transformer is 1:1...(so theres not much voltage across the stray resistance)..the stray resistance does indeed limit the current......

--- Updated Jun 11, 2022 ---

The diode drop will take away some volts, so you may need to up the txfmr turns ratio to account for this.

 

[Easy peasy]

please don't ignore the listed R in the output wdgs - this will cause less than efficient power transfer and reduce effective Vout at load.

 

[cupoftea]

If you send the LTspice sim, will try and have a look at it.....maybe you need to add gate series resistance or something....i mean.....its a "hacked up" pushpull that appears to rely on cct stray resistance to do current limiting....so maybe the waveforms wont be ideal......but they in their doc , seem to show beautiful flat DC output current, which is a little odd, looking at the schem.

Be wary of the lealage L spike overvoltaging diodes in LTspice...that will mess up the sim and youll need to quench down the leakage L spike.......but i see in your sim you are not using leakage L so ill think again. (but please do add leakage L from the datasheet)

 

[Xerxsea]

Here is the current Sim which i have modified to include the pa-6383-al transformer the best i could.
Just to make sure i didnt do any math wrong here is the formula I used to get secondary coil inductance:

Np/Ns ^ 2 * PL (Primary Inductance)
I'm not sure I understand how to add the leakage inductance in LTSpice. it says in the wiki for asymmetrical transformers to just add a series inductor. but to my understanding but it is symmetrical and I'm not understanding the process the wiki describes for that.

I also changed the diodes to have a higher breakdown voltage (40v now instead of 20v)

Edit wont let me attach Sim file. How did you want me to send it to you?

 

[cupoftea]

Ill try and find time to look at the sim...you can zip and send, or just send as a .txt file.
The attached shows most of the math needed for SMPS design.

Yes to do Leakage L you just put in a primary and sec inductor...or use "K L1 L2 0.999" depending on coupling factor k.

If you put a manual leakage inductor in the pri, then the one in the sec should be the value referred through the transformer.....

Yes L1/L2 = [N1/N2]^2

Eg if you have a 1uH primary leakage inductor and Np/Ns = 2, then add a inductor of 1uH/2^2 in the sec....ie 250nH in the sec

..or just adjust k and do it the coupling factor way.

Leakage L = Lp(1-k^2) where Lp = total primary inductance.