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.

[SOLVED] Flyback converter design for 250V

Status
Not open for further replies.

kiransbaddi

Member level 2
Joined
Nov 30, 2015
Messages
51
Helped
0
Reputation
0
Reaction score
0
Trophy points
6
Activity points
427
Hello

I m developing on physiotherapy stimulator which works on AC. and output will be high voltage DC

input : 230V 50Hz
Output Voltage : 250V / 80mA . <0.5% of ripples
Switching frequency : 100kHz
SMPS transformer selected : EF20( suitable for 20W and above
To achieve above spec i thought to design a fly-back converter of 20W (250V*0.08mA), i have referred PI application note to design it. here are some problems i have found out

1. Designing snubber circuit. : The reflacted voltage will so high that it can easily destroy the MOSFET
2. Designing SMPS transformer : finding difficulty to select core /bobbin for such a high voltage

can anybody suggest me which is right topology i have to use use

note :
microcontroller PWM with additional driver circuit to drive MOSFET is used
MOSFET VDSS = 600V
 

Designing the converter for 250V*80 mA continuous power is a bad idea, I think. For safety reasons, it shouldn't provide more than the required average power and utilize capacitive energy storage for the pulse energy. Also it doesn't seem appropriate to output more than 80 or 100 V.

A mains operated stimulator must provide reinforced isolation according to medical safety standards. It may be easier to use an off-the-shelf AC/DC converter.
 

you also want to ensure your device will never provide enough current to kill
 

Thanks for the reply

of course the DC is not continuous. instead it is pulse waveform sampled at 8kHz. it is assumed that human skin impedance is in range of 1 to 2kohm. we increment current in the step of 2mA i.d 2m, 4m ,6m .....140mA. hence 2k*140mA =280V but i want to limit it to 250V.



Screenshot at 2019-04-17 12-51-31.png
 

We have designed flybacks up to 6kV DC out - you need to give your self more room on the bobbin to allow for tape between layers and careful layout of wires to ensure high volts do not get close to each other. You need to interleave the windings to reduce leakage - which will lead to less snubbering needed on the pri wdg - also you need to design for DCM only - helps with the o/p diode. For rectified mains in - you want a flyback voltage of 150-200V so the total voltage seen by the fet will be 325VDC ( rectified 230Vac ) plus the flyback volts - this gives you close to 40% ON time at full power and still DCM...
 

Thanks alot.
it will be helpful if you could tell me which bobbin i have to use. i have selected EF20 , literature says above 20W and below 30W we can use EF20. now looks like i need more space
 

i don't have a catalog that shows an EF20

when you pick the next larger bobbin, you want larger minimum winding width and window
it should have enough pins for the required connections
when you wrap the tape, the start and stop should overlap
the overlap will be a larger build up than the rest, so it should be on a side where the core is not.

- - - Updated - - -

compare TDK E20/10/6 (EF20) with TDK E25/13/7 (EF25)

there is also an EF32
 
Last edited:

if you want to reduce your V reflected to pri in a flyback smps, even though your vout is high, then consider having multiple secondaries, and then stack them on top of each other to give an overall high vout.
 

of course the DC is not continuous. instead it is pulse waveform sampled at 8kHz. it is assumed that human skin impedance is in range of 1 to 2kohm. we increment current in the step of 2mA i.d 2m, 4m ,6m .....140mA. hence 2k*140mA =280V but i want to limit it to 250V.

Total power at 140mA and 250V works out to be 35W. You say that the DC is a pulse waveform but do not day the duty cycle (on:eek:ff ratio).

because you are trying to control current, you need to adjustable current source with a compliance voltage of 250V (post #4).

Medical devices have special requirements and you need to pay special attention to these.

I guess EF20 core may be small for 100kHz at 35W (even 20W plus losses).

If the duty cycle is 50% or less, you can still use EF20 and get away with a thinner wire.

In the first post you say 250V at 80mA (that makes it 20W) but that will not work in the high current ranges
 

Thanks for all your replies.
here i want to correct my question. i m designing power supply which gives maximum 250V 80mA( continuous) and 250V 140mA (pulsed 8kHz duty cycle 80%).

as per all your suggestion EF20 looks very small to achieve such high current. i think EE35 suits my design because of larger bobbin size to wind multiple secondaries.

i need your suggestion on selecting EE35 core. what are possible losses or effect i incur when i select bigger bobbin. ( ex. switching frequency, max duty cycle, core loss.. etc..)
note : as per the calculation i have ended up with selecting EE20 core.

here is the link : https://www.microsemi.com/document-...438-lx7309-flyback-and-forward-core-selection
 

Bigger bobbin means lower losses but unduly higher cost. With bigger core, you can actually achieve fewer turns.
 

EE35 max dissipation is likely 2W core and wire - convection cooling. You need to pick a starting point, e.g. 120kHz, fully DCM, Vin min(?) two o/p windings in series at 125Vout each, Bmax = 80mT, core = 3F3, 3C97, 3C96, N87 or similar better.

Then you can work out the turns, then - knowing the rms currents you can work out the wire size for 0.7 watts dissipation on pri, 0.7W on all secondaries - then see if it will fit on the bobbin - if not - go to a bigger bobbin ...
 
Thanks guys

This forum was always been big resource for me. have started the procurement. thanks for all your help
 

Hi Easy peasy Great help from you.
.
One thing i did not get is why i have to use Bmax 80mT, since i m using EE35 @ freq of 100kHz, Vrms =230V then Bmax =0.485mT.

if i
 

hi there, for flyback, B max should be < 160mT, i.e. delta B = 80mT, this is because the ferrite you are using will be a lossy type ...
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top