Increasing the current capability of this circuit

[liquidzell2]

Any ideas without having to increase the value of R1 or R2 (they are 2M and 1M each)?

D1 , D2, D3 are DIACS that let voltage pass when its 24V or above.. the LEDs are just indicators of High Voltage… the more voltage the more they lit thanks for any tips

 

[RyreInc]

I hope those LEDs aren't on for long--drawing 20mA requires 60kW of power!

Is this homework?

 

[VEBill]

"60kW"? Huh?

The OP's post at top is too vague.

 

[ScottyUK]

VE1BLL - 20mA through a series 2M[Ω] + 1M[Ω] resistor chain...?

 

[VEBill]

Since we're being silly ...

3Mohms * 20mA = 60k Volts

60kV * 20mA = 1200 watts

 

[ScottyUK]

Maths...?! This is getting serious!

 

[RyreInc]

Whoops, I calculated both, but got them mixed up somehow! Still, 1200W is still a bit much just to run some LEDs

 

[Skogsgurra]

The 20 - 30 V typical breakdown voltage of a diac works when you use one diac. But connecting a string of them will break the one with the lowest leakage doen first. Then the others break down. That will produce a highly irregular "warning" level from circuit to circuit and probably also changing with temperature. Parallel resistors are needed to equal out differing and changing leakage currents.

There are much better ways of doing it. Or is this a VERY SPECIAL APPLICATION?


Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[liquidzell2]

Its meant to be a High Voltage Sensor... the lights will go on after X amount of Voltage... then lights wil flash more and more frequently as the voltage goes up

Does anybody have an alternative design / ideas for doing this? Thanks

 

[Skogsgurra]

There is no capacitor in the circuit. So the flashing, if any will be very short.
Add a capacitor to get the flashing and select size of capacitor and series resistor between C and LED to suit.

The string of diacs is not needed. Or do you mean to protect the supervising circuit with them? Why? There are far too few data given for any design engineer to comment seriously on this. Tell us what the application is. Not only that is is supposed to warn for too high voltage. Tell us at what level it shall warn. Tolerances on that level. Temperature range. Constraints on power consumption. Needed visibility - remember that the eye cannot see very short flashes if they are not intense enough.

But, main message is: add C. And use parallell resistors to avoid premature triggering of the series connected diacs. If they are needed at all.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[OperaHouse]

You need a capacitor in place of D1-3 if you want it to flash. And it will still likely have problems. What voltage are you starting with?

 

[VEBill]

High voltage light - a very similar requirement to the neighbouring thread.

 

[sunnysky]

I have no idea where you want to increase current in a HV charging circuit with caps not shown... yet scratched in.... or in the HV indicators LED not shown.
Please define scope of design function and current specs.


For HV indicators, use a HV zener not DIACS, which are negative resistance devices . Then the LV LED's can go in series with the zener to indicate zener limit .. It goes without saying that the LED's need reverse protection diodes built in or added on. All LED's begin to fail at -10V and are rated for -5V. that includes EOS discharge.

 

[OperaHouse]

"All LED's begin to fail at -10V and are rated for -5V. that includes EOS discharge."

Where do you get that? I've tessted large numbers of LED and have never seen any leakage or damage under -200V. I've run tests with these powered at 120V line voltage for months with no damage and even on early technology versions over 30 years old.

 

[MacGyverS2000]

It goes without saying that the LED's need reverse protection diodes built in or added on. All LED's begin to fail at -10V and are rated for -5V. that includes EOS discharge.

Easy to test... hook an LED to a 9V battery backwards, then reverse it to be forward-biased again (briefly). Two shiny dimes and a rusty nickel it still works...

Dan - Owner

http://www.Hi-TecDesigns.com

 

[OperaHouse]

I worked with this guy and at interviews he would always stop a potential employee and say, "I didn't ask where you went to school, I asked where you got your education." Mine increased a little more last week when I repaired a micropressor based wttmeter. The 3.3V linear regulator had shorted and over 12V was on the micro. Thought for sure that chip was toast, but applied 3.3V to it and the instrument functioned perfectly.

 

[IRstuff]

"Thought for sure that chip was toast, but applied 3.3V to it and the instrument functioned perfectly."

It still may be latently toast. Excessive supply voltage is one approach to accelerating a life test. Depending on the duration of the exposure to accelerated conditions, your processor may experience premature failure. OR, it already has a failure, but you just haven't processed anything through the part that is already damaged.

TTFN
faq731-376