DC DC converter 3.3V to 100V generation (low current)

[themaccabee]

Hi,

I want to generate up to 100V output, the current will be very less, say less that 100mA or may be even <10mA. Its for a varactor supply.
Im looking into a boost chip implementation.
Is it possible to get an integrated switch version?


Thanks & Regards

 

[betwixt]

You are talking about a serious amount of power for a varactor supply, can you tell us more about exactly what it will be used for.
If the current was 100mA at 100V that is 10W of output power so even at 100% efficiency you would be drawing 3A from the 3.3V supply.

In view of the high in/out ratio I would think a transformer solution would be your best bet.

Brian.

 

[themaccabee]

Yes you are right Brian.The current out put is very less.It will be the typical leakage current only.So assuming a 5mA worst case and as per your calculations ,Output power=100V*5mA=0.5W @100% efficiency it will require 150mA from 3.3V so by putting margin the input supply required will be 3.3V,300mA worst.

So the requirement translates to
Input -3.3V,300mA
Output-100V,5mA.

I was checking different dc dc converter solutions, im supposing that simple boost converter is not easy here since the voltage is very high( correct me if im wrong)

Then remains flyback,push pull,half bridge,full bridge etc..
Can some body give me a suggestion??
Regards

 

[Orson Cart]

First you need to do a web search and find a switchmode controller, or CMOS gates that will work OK at 3V, failing that a self oscillating transistor ckt will work at 3V and provide step up via a 50kHz transformer, if you want tight o/p regulation you could make a discrete boost PWM control from a CMOS-555 (works at 3V) (or an LM339, works to 2V) to drive a 200V NPN xtor and have the feedback pull down pin 5 when the volts exceed 100.
Hope this is of some help, Regards, Orson.

 

[betwixt]

I agree with Orson Cart but the optimum design depends on how stable the 100V has to be. You can stabilize the voltage in two ways, by controlling how much you generate or by regulating it after generating it. I you have a regulator or even a 100V Zener diode to stabilize the voltage, you could do it with a single transistor oscillator and step up-transformer. The principle is the same as used to power mains voltage fluorescent tubes from 12V automotive batteries but in your case the voltages and currents would be far less.

If you have to generate the voltage at exactly 100V I would consider using a small PIC processor to generate a drive signal to a FET and then step the voltage up with a transformer. You can sample the output voltage and read it into the PIC ADC input to maintain a feedback loop. Several small 6 or 8 pin PICs work down to 2V and cost pennies.

Brian.

 

[themaccabee]

Im looking into ICs like MAX5025,MAX15061,MAX15059 etc..

 

[Orson Cart]

also: **broken link removed**

looks good

Regards, Orson Cart

 

[betwixt]

None of those devices will work at 100V.
I still think a single transistor self-oscillating inverter followed by a voltage doubler would be the best solution. I have in mind a circuit with maybe 12 components and no ICs.

Brian.

 

[themaccabee]

None of those devices will work at 100V.

???
MAX1522/24 are rated at 100V but not the ones i had earler listed.

 

[Orson Cart]

None of those devices will work at 100V

I think the idea here is to find a readily avail part that has a 3V or less input capability and use the ouput to drive a mosfet in a boost topology - hence the IC is not subject to the 100V.

Upon mature reflection a step-up ratio of 30.3 is a bit high for a single boost stage, so 2 cascaded boost sections running at 82% or so (max) would do the job (same gate drive each fet), or one of the MAX chips configured as a flyback instead of boost - would give the lowest parts count and footprint,

Regards, Orson Cart.