Zener diode regulator?

I am revisiting a problem that was posted a few months ago. I have tried numerous solutions, but nothing on the complex side of things will work for this application.

The problem is that I need to be able to convert an input voltage varying from 20 volts to 80 volts (at up to 100 amps) to 12 volts (at

500ma to 1 amp).

The recommended solutions in the past were not small or cost effective. This needs to be something that can be done with just a few SMT components. Size, weight, and cost are all important.

Using a simple Zener/resistor regulator works fine for a preset voltage (where the resistor value can be calculated correctly). The problem with this arrangement is that if the input voltage is 80 volts and the resistor is chosen so that 500ma of current is available, if the input voltage drops to 20 volts the current becomes way too low to be usable. Setting the current correctly for 20 volts results in extremely excessive current (and fire) when the voltage is 80 volts.

Is it possible to use some sort of transistor with the zener to limit the current based on the voltage? Some type of follower circuit? A schematic of this would be handy. I seem to recall seeing something like this a few years ago, when I didn't need it. :)

Thanks!

Reply to
Jim Drew
Loading thread data ...

Are you able to dissipate 68 watts in this "simple" circuit? If you're supplying 1A through a series element (as opposed to a switcher) you *will* disspate at least 68 watts.

Reply to
John_H

You want a low power zener ( or other reference ) feeding the base of a pass transistor.

Graham

Reply to
Pooh Bear

Are you able to dissipate 68 watts in this "simple" circuit? If you're supplying 1A through a series element (as opposed to a switcher) you *will* disspate at least 68 watts.

Reply to
John_H

This circuit is to power a 5 volt regulator for a CPU, and to provide gate drive power for MOSFETS. The circuit probably needs no more than

100ma, but I wanted to have a reserve for future use of LEDs and such.

I don't think I will need to handle 68 watts of power. Like I said, a

12 zener with the proper resistor for 80 volts works great to provide whatever current level needed. However, as the voltage drops, the current drops, making this solution unusable with a variable power input. I have seen some sort of voltage follower type of circuit that connects to a zener diode to provide a constant current. An op-amp could do something similar I suppose, but I have no idea how this would be wired.
Reply to
Jim Drew

Jim,

you are not understanding what John is telling you....

If you start at 80 Volts and drop to 12 Volts, thats a drop of 68 Voltss. If you deliver 1 Amp then the resistor or regulator that is dropping the voltage will dissipate 68 Watts. The regulator will dissipate 68 Watts, not the load...

The ONLY way around this problem is to use a switching type power supply which is more complex but there are some simple ones out there. They work by "transforming" the power instead of just wasting it. Lookup up "buck switching regulator"

If you really only need 100mA instead of 1 Amp, then your regulator will dissipate 6.8 Watts which might be managable for you but probably not in an SMT part.

Mark

Reply to
Mark

Forget it. You are obviously unqualified to do this even if a solution dropped into your lap from heaven. Look at the evidence:

Exhibit A: You provide no relevant specs:

- how much space/weight is available/acceptable?

- how much power dissipation can be tolerated?

- how much should it cost?

Exhibit B: You already constructed and burned a zener circuit. This is proof that you're out of your depth. Couldn't you have predicted the outcome? Are you going to exhaustively test your future "solutions" under all possible operating conditions? Lacking any theoretical understanding of what's going on, it's your only option.

National Semiconductor makes a completely integrated switching power supply the size of a 24-pin DIP that will do what you want at 80% efficiency - but only up to 40 Volts input, and it's not cheap. If you think you can do better then you should consider applying for a Darwin award <

formatting link
>.

-- Joe Legris

Reply to
Joseph Legris

Its not possible. You havent understood that yet.

There are 2 ways to regulate.

  1. linear, which in your case would require huge power dissipation. No way are you going to do that sm

  1. Switcher, which is what you need here, but no way are you going to do an 80v switcher with a few sm parts. The only route for a switcher is to design your own, I believe I suggested a minimal parts one. But if you havent understood the problem, youre not going to design a switcher successfully.

You might look at Roman Black's switcher page and see if theres one that can be copied direct, if not you're basically out of your depth. But no option will satisfy your above requirements, cant be done.

If I were in your shoes I think I'd look for other ways to get this 5v

1A.

NT

Reply to
meow2222

You got some replies that verged on verbal abuse. You are not ignorant -- you are trying to do something and it means learning.

If you want to take 80 volts and drop it to 12 volts at 1 amp with a simple linear circuit, the circuit will have a voltage drop of 68 volts across it and 1 amp flowing through it. Heat = power = volts X amps = 68 watts. That is too much unless you have a slew of parallel pass transistors on a massive heatsink, with a fan blowing on it. Not impossible, but pointless. A switcher is a reasonable way to achieve your task, however. Switching actually converts voltage by pulsing it through an inductor without much loss, as opposed to the linear technique that just throws away any extra power in the form of heat. It's possible for someone with a technical inclination but limited experience to build a switcher using a pwm chip, a ferrite transformer and other components, but you won't do it in a weekend. You will probably burn up dozens of mosfets and spend weeks (maybe even months?!) getting to where you want to go. But you might get there if you have the time.

Reply to
kell

I am able to do this just fine with SMT parts by using a 12v zener and a SMT (1210) package resistor... it gets mildly warm. The problem is insufficient current when the voltage drops too low.

I have been using a switching design, and I want to get away from it.

Reply to
Jim Drew

LOL! It is by no accident that these were omitted. I wanted ideas not constrained to a specific spec.

I am well aware of what exists, and I have been using a switching version for a few months. The cost is too high, the weight is too much, and the space constraints are too small for this solution. I did not burn anything... I was pointing out what would happen . I am looking for a simplified solution. Thanks for your wonderful help. :)

Reply to
Jim Drew

Yeah.. interesting.

I have a switching unit that I designed (in an afternoon), is all SMT, and works just fine... it's just too big (1.25" x 2.25") and heavy, due to the SMT inductors. I was merely looking for a simple solution, and not a flame fest. I *have* seen a similar circuit in the past, and it was used to supply a 10v gate voltage for a MOSFET drive circuit. The input voltage was 100vdc, and the circuit used a zener and transistor configured as a voltage follower arrangement. I was trying to be as vague as possible to accumulate as many ideas as possible.

Reply to
Jim Drew

The problem is that if it is a linear supply capable of 100mA maximum you *will* have to dissipate 6.8W in the circuit worst case (80V input). Actually, more because you need some safety margin. There is

*no* way around this with a linear regulator-- whatever parts you put in a black box between your 80V supply and your 12V @100mA output will have to dissipate 6.8W unless it's a SMPS. Call it 7.5W with 10mA minimum zener current. You will have additional dissipation in the zener depending on the output current (with no output current, the zener would have to dissipate another 1.3W with no output current-- a total of almost 9W.

If you can redesign your circuit to use less current you may be able to make a simple linear regulator more practical.

Why? If you are having failures, there are ways to protect such circuits.

Best regards, Spehro Pefhany

--
"it\'s the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
 Click to see the full signature
Reply to
Spehro Pefhany

If it's not getting warm, then you're not really using 80v and 100mA, which makes the whole thing a lot easier.

Again, if you don't need much current, you might not need a switcher. MOSFETs don't take much drive. OTOH, if you *need* 100mA, a low-power switcher can be tiny and cheap.

Best, James Arthur

Reply to
dagmargoodboat

Out of curiousity, if 1.25" x 2.25" is too big, what is the regulator going into?

With an input of 100Vdc and an output of 10Vdc, the regulator has to dissipate (90Vdc) * (MOSFET gate current). But back to your original post, what is supplying the large voltage range?

-Dave Pollum

Reply to
vze24h5m

UAV drone.

Lithium Polymer batteries. We are working with a voltage that actually varies between 16v and 65v, but want to be able to support 80v in the future. There are canned switching regulators (2596) that will work at high voltage, but will not reach the voltage that we need.

I guess the total power being used is around 10ma, perhaps up to 25ma briefly when the LED blinks... but we need to be able to power 2.4Ghz camera equipment at 250ma-450ma later on down the road.

Reply to
Jim Drew

Critical question: when your existing circuit is working at 80 volts input, what resistor value is in series with the zener?

John

Reply to
John Larkin

If you go to a >500kHz switching frequency you won't need near as big and heavy an inductor.

Best regards, Spehro Pefhany

--
"it\'s the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
 Click to see the full signature
Reply to
Spehro Pefhany

136 ohm is what is necessary for 500ma @34 watts=fire. I used 6800 ohm for testing (10ma of current @
Reply to
Jim Drew

Your existing solution is warm when supplying 10ma? Imagine it 25-45 times warmer to support the camera. With the "simple" regulator you pursue, the efficiency will be 15% at best when running 12V from an 8V source. For a UAV do you really want efficiency that bad?

1) you need a switcher or 2) you need to reconfigure those batteries so they don't produce a voltage much more than the 12V output you desire.

You can run 5 times longer with a switcher. 5x!

Look at national.com for regulators that supply up to 100V input voltages. If you can get by with 350mA current, they have a chip with an internal FET that will give a teeny-tiny solution without significant complexity (LM5008).

Five times longer run time.

High voltage battery bank with linear regulators? Dude.

Reply to
John_H

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.