DC to Dc Booster 12 volt to 2300 volt

I want to make DC Booster circuit (12 volt to 2300 volt approximate ) for defibrillator.
i want to know how much size of inductor and capacitor should for DC booster circuit ? ?
so is there any formula or calculation to find circuit parameters values ?
any standard circuit to make out high voltage booster

push pull circuit with step up transformer will be the most successful approach...

Easy peasy said:

push pull circuit with step up transformer will be the most successful approach...

Click to expand...

i want to make without transformer...using inductor and capacitor

i want to make without transformer...using inductor and capacitor

Click to expand...

Possible in principle but inefficient. You get very high primary peak currents, need a high voltage primary switch and get respective high switching losses. A possible compromise could be a moderate primary peak voltage, e.g. 500 V and a cascade rectifier. But flyback transformer topology is clearly preferred.

Hi,

for the single pulse energy stored in an inductor: W = 0.5 * L * I^2

multiply the same formula with your switching frequency and get power.

You don´t give current or power, but you need one of those values.
For a defibrilator safe implementatin I recommend to "limit" the stored energy or limit the power or to limit the time with external "safety" circuit.
Malfunction of the standard circuit should not lead to dangerous situations.


Klaus

my Power rating is input 12 volt battery
output 2300 volt 92 mA that too for only 7 second to charge capacitor

is there any solution to overcome switching Losses
and is there any formula to find inductor value and switching speed ???
i want to charge capacitor for only 8 to 10 sec with 92 mA and 2300 volt

A flyback or push-pull Tx solution will be smaller and more efficient than a straight boost approach, and require smaller mosfets... We have done flybacks to 4500VDC out (10W) from a 30VDC source, knowledge of transformer construction is critical to reliability, as is choice of diodes...

92mA * 2300V refers to a considerable amount of power, makes it even more imperative to achieve high conversion efficiency and compact design.

Its not as simple as it sounds. A straight boost converter will be looking into pretty much a dead short circuit when the high voltage capacitor is fully discharged at the start.

Zero volts x 92mA is zero power (in theory) so the energy storage process is slow to begin. In practice you will need to generate twice the average power to reach your charging goal of 2,300 volts in <10 seconds.

Something like a 400 watt minimum boost converter will be required.
That is 12v x 35 amps average.
Peak current in the inductor will be at least twice that, say 70 Amps.

And the peak flyback voltage 2,300 volts minimum.

What do you plan to use that can quickly switch switch 2,300 volts at 70 amps ?
Its just not practical to use a simple boost inductor circuit for this.

Only sensible way would be a push pull transformer circuit run in current mode.
It will still draw more than 35 amps for ten seconds, possibly a lot more.

And resuscitating someone often needs several shots in rapid succession.
You will also need to control the final charging voltage to suit the patient.
Probably something like 50J to 300J of stored capacitor energy perhaps.

Zapping a five year old child is very different to zapping a morbidly obese giant.

Taking a de-fib cap at 50uF and 2300VDC, this is 132 joules, to charge in 7 secs is 19 watts average over the 7 seconds (ignoring in-inefficiencies) this is about 2A average from your battery over the 7 seconds.

Larger de-fib caps and larger Joule requirements, require more power and hence current from your 12V battery,

500J needs 7.6A from your batt (average) over 7 seconds...

A well designed flyback with say four outputs in series (each 575V) would do the job, a std 500V 20A mosfet would be a suitable choice for the main fet, you could have a fixed duty cycle or peak current limit turn off to control the charging with a terminating volt control.
We did a similar unit for firing explosive charges over very long lines from a low power NiCad battery source.

As per FvM a straight boost is not a viable solution.

Its not that simple Easy P.

Charging up the flyback inductor primary in X microseconds with Y amps is easy.
The problem is discharging the high voltage secondary into a dead short, or at least at a very low voltage.
It simply takes forever for the current to ramp down at a relatively low discharge voltage.

If you try to ramp up the primary before the energy stored in the core has dissipated, the core just saturates on the following cycle.
Reducing the secondary turns helps, but then you run into problems at the high voltage end of the charging cycle. The energy discharge becomes too rapid and the peak currents too high.

The whole process of flyback charging over a very wide voltage range is notoriously inefficient anyway you approach it, and you cannot just assume some average power transfer during the whole charging cycle.

You cannot just assume 100 joules is ten watts for ten seconds drawn from the battery.

My own experience doing this was attempting to charge a high voltage pulse forming network for a very high power pumped laser (a Defence Department project we were contracted to do about thirty five years ago).
Its not as simple as it appears.

The energy delivered per switching cycle is constant for a flyback with peak I control on the primary - yes it takes a little longer to transfer this energy at the start, and more rapidly towards the end - but it still works very well - esp if you use a resonant flyback approach to determine the end of the power delivery pulse, then the freq rises as charging proceeds, thus the (energy delivery) pulse gets shorter and the charge time is minimised. For low power apps like this it works very well. We have built several for well satisfied customers. We do assume ten watts for ten seconds is 100 joules (less losses) and we program the energy delivered per cycle in the flyback to deliver this performance, we also graph the expected charge curves on excel (for rising Vo), and they correlate well to real world units.

- - - Updated - - -

For a programmable controller (such as a PiC) you can do all of this open loop (varying on and off times on the primary mosfet) until you reach the target voltage - no need for any mosfet current sensing... no saturation, efficiency >90%.

All calculated from an excel file, and then turned into a mathematic formula for the pic to generate the correct PWM from 0v out to target Vo.

Yes its definitely possible, but it requires a bit of sophistication to carry it off.

A self oscillating circuit that changes state on peak primary current, and again on reaching zero current in the secondary might work too.
As you say, the frequency would sweep upward as charging progressed.

It would require rather more than a 555 timer, a mosfet and a simple boost inductor.

Easy peasy said:

A flyback or push-pull Tx solution will be smaller and more efficient than a straight boost approach, and require smaller mosfets... We have done flybacks to 4500VDC out (10W) from a 30VDC source, knowledge of transformer construction is critical to reliability, as is choice of diodes...

Click to expand...

yahh right can you provide more about your circuit which you made 30 v dc to 4500 volt dc

Easy peasy said:

Taking a de-fib cap at 50uF and 2300VDC, this is 132 joules, to charge in 7 secs is 19 watts average over the 7 seconds (ignoring in-inefficiencies) this is about 2A average from your battery over the 7 seconds.

Larger de-fib caps and larger Joule requirements, require more power and hence current from your 12V battery,

500J needs 7.6A from your batt (average) over 7 seconds...

A well designed flyback with say four outputs in series (each 575V) would do the job, a std 500V 20A mosfet would be a suitable choice for the main fet, you could have a fixed duty cycle or peak current limit turn off to control the charging with a terminating volt control.
We did a similar unit for firing explosive charges over very long lines from a low power NiCad battery source.

As per FvM a straight boost is not a viable solution.

Click to expand...

can u provide in more detail of your unit's circuit detail ?
how to make this circuit and if i want to simulate it than which software tool is capable for it ?

- - - Updated - - -

hey yahh
can u give more detail for same
basic circuit and component values like timers and inductor

Definitely a self oscillating flyback converter. Yes, the peak currents will be higher than other topologies but the average power is low so you can push the flux density to 280mT.
The secondary winding of the transformer-choke needs to be given some serious consideration. The non-trivial turn to turn capacitance will cause the secondary to resonate with its inductance resulting in a non linear current slope on the primary current ramp. The solution is to space the secondary turns apart and insulate well between layers. This all reduces the secondary turn to turn capacitance.
The beauty of the self oscillating flyback is that it controls its own frequency as the capacitor charges. It only begins a new cycle once the choke has emptied itself of energy.

Don't believe that self-oscillating is first choice for multiple 100 W. Dedicated controller will give better switching performance and higher efficiency. Current mode can achieve suitable adaption to capacitor charging state.

How to make DC Boost circuit up to 2000 volt 0.13 mAh

hi
i want to make DC Boost circuit for up to 2000 volt 130 mAh .
i want to make through fly back converter or else capacitor inductor methods .
so can you provide me basic circuit. calculation and methods.

Re: How to make DC Boost circuit up to 2000 volt 0.13 mAh

Before going further - what voltage are you starting from?

You also quote 0.13mA in the title and 130mA in the message, which do you mean? (I assume the 'h' is a mistake)

If it's 130mA, do you realize the circuit will be producing 260W of power and will probably need a supply capable of providing >300W ?

Brian.