Two LT783 in series?

neazoi

Advanced Member level 6
Joined
Jan 5, 2008
Messages
4,155
Helped
13
Reputation
26
Reaction score
15
Trophy points
1,318
Location
Greece
Activity points
37,164
Hi, I was looking for an easy and cheap HV regulator linear IC and this LT783 https://www.ti.com/lit/ds/symlink/tl783.pdf seems to do the job and provide adequate current for the anodes of small tube projects.
However I would like to experiment wit a bit higher voltages, say 250V.
The datasheet shows a graph of input to output voltage difference in conjunction to the current on page 6.
I wonder, if I can connect two or more of these in series to achieve the goal of regulating higher voltage in "steps" of 125 or less volts per device, so that I can input a higher input voltage in the total chain of these.
 


Because of the price.
A single 150V 50W zener, costs 35-40Euros.
In fact my design used just zeners, actually 8 of them, and no regulator, so the cost was prohibitive.
I have seen active zener circuits, which use a small wattage zener to drive the B junction.
But this LT783 was by far the cheapest option, a "common" component that costs 3-4 Euros max.

In the past I connected linear regulators in series pre-regulate, so that each device is not heated much.
I am thinking of the same thing with two LT783, but can it be done, or the set resistors of the first will "confuse" the second?
Also, I want to verify the max input voltage that one LT783 can tolerate. (I will regulate to more than 100V, as this is a PSU for small vacuum tube circuits)
 

Hi,

and you really want to blow 50+W off heat in the air?
Big heatsink, fan?

I mean: linear is linear. No matter whether you use a resistor, a zener, a BJT, a MOSFET, an IC .... the dissipated power is the same.

Klaus
 

    neazoi

    Points: 2
    Helpful Answer Positive Rating
KlausST said:
Hi,

and you really want to blow 50+W off heat in the air?
Big heatsink, fan?

I mean: linear is linear. No matter whether you use a resistor, a zener, a BJT, a MOSFET, an IC .... the dissipated power is the same.

Klaus
Click to expand...
If you use more than one devices in series and do progressive regulation they should not generate so much heat. As said regulation will not be down to low voltages anyway, but to more than 100V at minimum, so they should not generate that amount of heat.
It is the higher input voltage that I need to tackle.
I could use gas regulator tubes, as they can provide 20 or more mA depended on the tube. I just thought if it could be done with the linear regulators.
 

What’s your output voltage and current? What’s your input voltage? I wouldnt use 125V across a device with an absolute max rating of 125V.

In theory it should work, but I’d worry about both regulators starting simultaneously. What happens if the second regulator starts slowly so that there’s, say,10V across it; now your first regulator will have a much higher voltage across it.
--- Updated ---

Um, wrong. Whether you dissipate power across one device or 100, you’re going to dissipate the exact same amount of heat.
 

    neazoi

    Points: 2
    Helpful Answer Positive Rating
They should not generate that much heat --> simply violates rules of physics.
P = V x I is true .. for any linear circuit.

Klaus
 

    neazoi

    Points: 2
    Helpful Answer Positive Rating
If your source V is not excessive variation one could do it with just a
power R....

Do you care about this ?



The reg spec has no stated limitation of absolute max Vin to ground, maybe contact TI
directly with question.

Then there is the power sequencing problem, how does a cascade power up if input/output
caps are present, will one get out of spec voltage in the transient.....




Regards, Dana.
 

You can use two LT783 regulators in series. You have to take care of the thermal problems.
 

Dissipated power is the voltage drop multiplied by the current, no matter how many devices you use or spread the heat across.

If you want the cheapest solution, use a series resistor and an active clamp. A transistor that can handle the voltage and variation in current with a single Zener diode across collector and base so it always biases itself to Zener voltage plus Vbe.

Brian.
 

with > 90 % efficiency a SMPS can be made for a couple $/W with very low ripple (50mV) at 250V as long as you expect some compromise on load step response time.
 

Anytime you want to design a regulator;

Define input V, power (pk, avg) and output with a tolerance and ripple and any other limitations (THT, SMT, $$, environment)

Without these minimum requirements, it's guesswork or lots of false assumptions.
but with good specs, optimization is easy for those with the tools.
 

You must log in or register to reply here.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…