Zener diode regulator?

You really have no idea what you're talking about, do you?

You'd better get some clarification from your teacher as to what the difference between volts and amps is.

Helpful indeed. Now we see that you really need only 100mA! But what's an order of magnitude between friends?

Simplified solution? How about rounding to the nearest kilowatt, then the power dissipation is zero and you're home free.

-- Joe Legris

You're certainly not going to dissipate that sort of power in a couple of square inches of pc board. Resistor-zener regulators become train wrecks at high power.

Somebody, Linear Tech I think, makes a high-voltage buck switching regulator chip, and LTC may even make a tiny module with magnetics and all. Many others make modules for sure.

I'd think that a drone would be unhappy hauling around and powering a power supply that needs huge heat sinks and is 15% efficient.

John

A zener/resistor regulator is not a good way to go here.

Use a 'cascode' transistor and reference (TL431 or zener diode) to drop the voltage to a manageable level, then use a linear regulator. 3 1/4W resistors, 1 power transistor, 1 TL431 or equivalent, and an LM7812. (A

The cascode transistor should dissipate the brunt of the power by providing a stable 15V or so (ref 15.6V). Select the transistor and heatsink according to load. If you are just using 10mA, you'll only be in for 650mW at 80V, so you can use nearly any transistor that will handle

You tried to be vague, but you succeeded in being misleading. Better luck next time.

well I see we went from 1 Amp down to 10 mA...

yeah at 10 mA you can find a workable solution using a small linear regular....

Mark

Thank you... this is exactly what I was looking for. I did a quick search on the net and have found a couple of sample configurations. Right now, the circuit is using about 37ma max, running at 42v (10 serial Lipo packs), but we would like to add the 2.4Ghz transmitter(s).

Instead of a resistor, how about a constant current diode or equivalent? A DN3545 used with a G-S resistor becomes a settable current source in a rather small space (both parts SMT), with a minimum stand-off near

Jim, You've had a lot of replies to your questions, some good, some not so, some way off beam and through it you have kept your cool. The good replies state the undeniable that W=V*I for *any* linear configuration. I expected someone to challenge the example you gave of driving a mosfet gate from 100V as this was a perfect example of where to use a simple resistor/zener combo as it requires practically zero current. A simple cap across the zener would have supplied any peak currents. As soon as you even consider drawing any real current you run into W=V*I which seems to be for your final configuration (80-12)*0.5 = 34W. Totally, absolutely impractical to consider even for a microsecond unless you have a patented method of converting the heat back into usable power. Your tests have been exactly that, tests, but nothing like or near the real thing.

Now that we have eliminated that and if you are indeed reasonable you would have to concede that point so that we can look at other solutions (that's what you're after, isn't it?).

The switching solution seems to be the best although I am not sure why you are against it. The board size you quoted is huge, who designed it? It sounds like it was designed with antiquated low-frequency parts. For a switcher to be practical in this application you need to consider high-frequency, that is =>400KHz or even 1MHz parts. This means the inductors end up being a lot smaller and lighter, as does the filter capacitor. Now it wouldn't take me long to google up some fast switchers etc but I will leave that to you unless you have something a bit more specific and then I or others should be able to point you in the right direction.

Someone mentioned Roman Blacks switcher page

What other solutions are there? Unfortunately, all the methods I can think of (there are some) will not be as cost-effective as a switcher, nor will they be as light or as small. But why does it need to be so cheap???

my2cents

Jim Drew wrote:

I just had an interesting idea about this circuit:

Normal version: !!-- Vin Vin -----------/\\/\\---+--!! ! !!---+----------- Vout ! \\ ! / ! \\ c \\! / !----- e /! ! /-/ Zener ^ ! GND

We can make a second MOSFET slave its current to the first

!!-- Vin Vin--/\\/\\---+--------!! ! !!- c \\! ! !--/\\/\\---+ e /! \\ !!-- Vin ! / Vin -----------/\\/\\---+--!! ! \\ ! !!---------+-/\\/\\-+---++------ Vout ! ! \\ +-------!

You need a lesson in ASCII drawing.

. Vin --+--------------, . | | . | ||--' . '--/\\/\\-+--|| We can make a second MOSFET slave its current to the first

Your drawing, edited...

. ,------------------, . | |--' . Vin --+------------+--+-/\\/\\-+----||

I couldn't help but follow this thread and figure out all of the parameters of the design. Some questions that still linger:

1. Why does the input voltage range need to be so great? If it is a Lithium-Ion battery pack, or multiple cells, can't a tap be made that always has the same voltage for the MOSFET drive?

1. MOSFETs need a fast high current drive for a few microseconds to charge the gate capacitance, and then a fast discharge. Unless the switching is at a very high frequency, you can trickle charge a big capacitor at about 10 VDC and then use it to drive the gate.

2. The additional components that require such high current either have internal switching regulators that tolerate the wide input voltage ranges mentioned, or else could be driven from appropriate taps on the battery stack. It is unclear why there needs to be such a wide range of inputs, and why a one-size-fits-all regulator is needed.

1. If this is a battery powered application, I would think battery life would be the first consideration, so low power and efficiency would be vital. If weight is a concern, as in a flying apparatus as mentioned, weight would be another absolute necessity. It should be possible to standardize on one battery voltage for all applications, and possibly add more packs in parallel (with sharing diodes or other more efficient supply sharing methods) in order to get more current or longer overall running time.

2. Unless this is a toy, development cost and simplicity are the least considerations. Money spent up front to maximize efficiency and reliability will produce the best possible product at a reasonable production cost. If it is for hobby use, the learning process of fine tuning an elegant design is well worth the time and money spent in research and testing.

Good luck!

Paul E. Schoen

Ooops: yes it will only work in there is a minimum load large enough to ensure that the Rs has at least Vbe on it. Above that point, the two MOSFETs will share the load. Maybe it would be better to use one of those small FETs with V(th)=0 for the second transistor.

Not possible as the battery configuration changes with almost every mission.

The switching frequency of the gate is right at 20Khz.

Things will change frequently, so a setup is necessary that can accomodate the changes.

Weight is a consideration because it affects the runtime of the UAV. Solar power is being used to lengthen the runtime, but realistically, the run time needs to be around 25 minutes. Adding several ounces will reduce that runtime significantly.

This is far from a toy. This is being used for security (homeland type).

Absolutely.

I am not against a switching design. I am against one that weighs 3 ounces vs. a linear version that weighs 1/2 ounce. I think the switching frequency is not nearly high enough on my current design, and that is definitely requires the use of the larger inductors.

Small and light are the key areas of importance. Cheap really doesn't apply too much, other than it usually equates to ease of availability of components.

I appreciate your response. I will look at some other switchers. Roman Black's website is interesting. Even the canned units 5008/5010 are relatively low speed, and require large inductors for high current (1 amp).

You're a godammed pretentious troll and NG "luser" who's too damned dumb to even learn Ohm's law is what you are really seeing, whether you know it or not. Your crap question and discussion doesn't even rise to the level of sci.electronics.basic.

Whoa, if you have access to the battery pack, just take a tap at 3 or 4 cells and regulate from there. The electronics load is trivial compared to the engine motor load.

If the battery pack is changing radically from mission to mission (on the same aircraft) something is being done wrong. If you are using different aircraft on different missions you only need to design one regulator per aircraft. Regardless of the mission it should be acceptable to return from the mission with 90% charge left in the battery pack

There is no access to the battery pack, other than the main input into the main power block. The cells are sealed.