Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

LLC converter is preferable for step up operation?

Status
Not open for further replies.

devarajguru4

Newbie level 4
Joined
Sep 25, 2013
Messages
5
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
39
I want to convert the 24V to 1000V with good efficiency and high power density. So I have changed the design from push pull to LLC conveter. But when I trying to simulate through the LT spice the calculated output voltage is not coming. Whether this method will work for step up?
 

It might work better if you build the project step-by-step. I guess it has a transformer? Send a square wave (amplitude 24V) directly through the transformer primary. Attach a load at the secondary. Adjust values until you see 1000 V to the load.

I'm running a simulation. Frequency 200 Hz, transformer 40 mH primary, step up ratio 44x. Load 1000 ohms. The 24V supply needs to push 44 A through the primary.

Then add components, one at a time. Inductor, capacitor. Parasitic resistances. Experiment by raising or reducing the value of a component, so that you discover what you need to adjust in order to maintain 1000VAC to the load.

Eventually you can install the half-bridge or H-bridge. Etc.
 

Yes LLC can work - but remember - at high freq and high volts -- you have a lot of current circulating in the winding ( and any other ) capacitances

This is why push pull is used a lot for this application - fixed freq, easy to do, just wind the Tx properly with low capacitances ... fast output diodes with appropriate snubbing ..
 

i remember simulating an llc once, and wasnt getting right vout, and it was because yes like Easy Peasy said, there was a lot of current , and i forgot to account for the drop in the esr of the sim fets.
 

For such high traformer ratio, LLC will request high Cr and very low Lr values somewhat hard to obtain.
What is nominal power requested?
 

I made same calculation and best values for f0=100kHz are: Lp=6micro, Lr=0.8micro, Cr=3.4micro; n=0.012 (about 1:84 or 2: 168).
Fmin (to keep ZVS) = 84kHz; Fmax=115kHz
 

For high o/p volts a sine wave of volts in the Tx can be very helpful, such as a parallel loaded SRC...
 

Hello,
The push pull converter for 24v to 1000v will need a transformer with a turns ratio of around 60 or so. Obviously the higher inductance winding will be the secondary.
It will be difficult to easily get a good coupling between the two primarys and the secondary (I am assuming that you would use a full wave rectified secondary with a single secondary coil)
With a turns ratio that high, it will be difficult to get good coupling, and therefore you will get a relatively high leakage inductance….this will manifest itself mostly in the secondary….so your secondary side diode snubbers will be dissipating quite a bit of power.
You may struggle to find output diodes of the high voltage that you need…and will likely need to series low voltage diodes…you then end up with that conundrum where not all diodes will take up the reverse voltage at the same time, so you will need networks across each diode to kind of ensure that all the diodes go forward biased and reverse biased at pretty much the same time….otherwise , the last one to go forward biased will get an overvoltage for brief intervals
Also, the high secondary leakage inductance will impact on the choice of your output inductor size…because this will need downsizing in its inductance in order to account for the fact that there will be such a high leakage term. So how much do you downsize it?...well it depends on the leakage inductance size…so you will need to wind the push pull transformer in a very repeatable way so that you get tight control of the leakage inductance.

Also, interleave winding is usually used to reduce leakage inductance…but how do you interleave wind two primarys and a secondary?...yes it can be done, but as you can imagine its fiddley…..i suppose the easiest way would be to split the single secondary into two halves and then put the two primarys inside that.
So basically I would use a cascaded boost converter to bump your 24v up to a much higher voltage..and then you can use your transformer isolated converter to do the last bit and get your isolated 1kv.
There is attached a ltspice simulation of a 24v to 1kv push pull, plus the detail on the cascaded boost.
 

Attachments

  • PushPull 24 to 1000v.txt
    11 KB · Views: 43
  • Cascaded Boost.zip
    13.5 KB · Views: 52
  • Like
Reactions: asdf44

    asdf44

    Points: 2
    Helpful Answer Positive Rating
So anyway yes, best boost up, then put your isolated converter after that...

Please avoid the pushpull, as it is virtually bogus, as the following describes...

In fact, in all truth, the pushpull converter is a badboy converter. The leakage inductance is a problem. As you can see in the attached LTspice sim of a 48V 100w pushpull…it has 14.4w of dissipation in the clamp resistors.
This is way too much….and interleave winding of a pushpull converter so as to reduce leakage inductance is more awkward due to the two primaries, which both need interleaving to achieve best interleave results.
The full bridge and two transistor forward don’t burn up the leakage energy in a primary clamp. Thus they are more efficient than the pushpull….and don’t need interleave winding. The only good thing about the pushpull is the low side drives…..but high side bootstrap drivers are cheap and plentiful now anyway.
Surely our “friend” the pushpull is to be considered virtually bogus nowadays.
 

Attachments

  • PushPull 48v to 48v _100w.pdf
    21.5 KB · Views: 62
  • PushPull 48v to 48v _100w.txt
    9.8 KB · Views: 114

It will be difficult to easily get a good coupling between the two primarys and the secondary (I am assuming that you would use a full wave rectified secondary with a single secondary coil)
With a turns ratio that high, it will be difficult to get good coupling, and therefore you will get a relatively high leakage inductance….this will manifest itself mostly in the secondary….so your secondary side diode snubbers will be dissipating quite a bit of power.

Respectfully - I read this as - " I do not know how I would wind a Tx with the necessary good coupling & therefore low leakage"

Also, the high secondary leakage inductance will impact on the choice of your output inductor size…because this will need downsizing in its inductance in order to account for the fact that there will be such a high leakage term. So how much do you downsize it?...well it depends on the leakage inductance size…so you will need to wind the push pull transformer in a very repeatable way so that you get tight control of the leakage inductance.

Not true. Output inductor is a function of Vo ripple and I ripple pri side only.

for your information, push-pulls have been built at up to at over 1kW at up to 200kHz - that work just fine - the tx is critical - yes - but beyond that all else is simple, for 2 or 3 600V high speed diodes in series for the rectifier - a very small snubber only is needed across each - if one avalanches in reverse (for 100nS say) this is no prob as the total heating energy in the device is very low - and 1200V SiC would do the job well - assuming a relatively constant input V.

also for the Tx wind the pri's bi-filar, for 24V you may may 2 sets side by side on the bobbin, then have 5 x layers of pri with 4 x layers of sec in the gaps - connected in series on the pcb for the HV - voila - low Llk - low energy in snubbers both sides - also lower prox loss and skin effect loss - depending on how close the layers are, for TIW pri the layers can be very close...

- - - Updated - - -

One of our standard products contains a Push-Pull, 50 - 160Vin, 240VDC out 3kW, ETD 59, 80kHz - solid copper wire - about 4W of snubbing at full power...
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
well if you are going to wind the transformer like that then I obviously agree with all your points. I thought if the original poster was as clumbsy and non-dexterous as I am, then they will want an easier time winding a transformer and so would pick a full bridge instead…and just put up with the 4 fets and two bootstrap drivers. I agree the output inductor peak to peak current wont be affected if you wind for the low leakage like that. It sounds like you are doing mulitple layer interleaving, with 4 secondaries interleaving. At least the OP knows they should keenly watch for the leakage if they do pick the pushpull.
-----------------------------------
As you know to get low leakage in a transformer wind, its also good if you can get the layers over the full width of the bobbin with windings evenly spaced across the bobbin. So I hope the OP looks out for this aswell.
Sometimes if I only need a small number of turns for a particular coil, then I find they wouldn’t spread across the full bobbin width if not “spiralled”, and so I put windings in parallel and ask for them to be “flat wound” (so as not to use up the bobbin depth)…this way the few windings can nicely fill right across the bobbin width….but then the transformer manufacturers get back and tell that they can’t terminate the multiple turns to a single pin……….or they say that such “flat winding” is expensive.
I hope the OP knows that spec’ing “spiral” winding, with a few number of turns evenly spaced across the bobbin, is risky, because as we found, the winders often don’t bother doing it…they just clump the turns at one end of the bobbin width………peeling back the transformer after manufacture reveals this.

- - - Updated - - -

I am sure the OP knows that if they do pick the full bridge, they do also have to be mindful of the leakage inductance, because as you know, in CCM, the primary current initially slopes up at a di/dt of Vin/Llk….and so with large amounts of leakage inductance, the duty cycle D will have to be that little bit higher.
I still wonder that the OP may find that the easiest of all transformer winds will come if they cheat and shovel a cascaded booster in there first, then with the much increased new vin, they can achieve the vout with the isolated stage with greater ease…..they will have much less current in the primary…and not such a high turns ratio to deal with.
 
Last edited by a moderator:

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top